Upgrading Algae Biocrude for a Low-Nitrogen-Containing Biofuel

May 8, 2017 - and the oxygen contents were between 0.69 and 5.6 wt %. However, a biofuel ... from 0 to 60 min for algae biocrude has not yet been inve...
0 downloads 0 Views 4MB Size
Article pubs.acs.org/IECR

Upgrading Algae Biocrude for a Low-Nitrogen-Containing Biofuel: Compositions, Intermediates, and Reaction Routes Bingwei Zhao, Ze Shi, and Xiaoyi Yang* School of Energy and Power Engineering, Energy and Environment International Centre, Beihang University, 37 Xueyuan Road, Haidian District Beijing 100191, P. R. China ABSTRACT: The upgrading of algae biocrude (obtained by hydrothermal liquefaction) was studied under mild conditions. Here we adopted a guard catalyst to protect the core catalyst and obtained algae biofuel with low heteroatom content at 350 °C. The nitrogen content of biofuel is 0.016 wt %, and the calculated oxygen content is less than 0.51 wt %. Furthermore, the effective yield from biocrude to a low-heteroatomcontaining biofuel was close to the theoretical yield. Through a detailed study of the upgrading process, the content variation trend of biocrude and biofuel compositions and their potential reaction pathways were revealed. The results indicated that hydrogenation reactions almost finished after upgrading at 350 °C for 15−30 min.

1. INTRODUCTION Hydrothermal liquefaction (HTL) technology is suitable for the conversion of algae to biocrude1,2 because it obviates energyintensive drying steps and the reaction temperatures could be below 300 °C. The effect of the HTL reaction time ranging from 1 to 60 min has been studied, and it was found that 5 min was applicable for the conversion of algae, such as Nannochloropsis.3−6 Most studies about HTL or HTL upgrading adopted solvent for the recovery of the biocrude.7−9 However, when HTL biocrude upgrading is conducted, the solvent extraction step of biocrude from aqueous phase may increase some energy consumption (solvent removal). Sometimes air pollution also happens. Nannochloropsis is a kind of promising algal strain10−12 that could be cultivated in wastewater, contain high-value-added products, and grow fast. Thus, the development of a biofuel base on Nannochloropsis biocrude for industrial scale has great potential, especially when the conversion technology, including HTL and catalyst upgrading, are well developed. Different kinds of catalysts, such as Pt/C, Pt/Al2O3, Pd/C, Ru/C, CoMo/γAl2O3, NiMo/γAl2O3, and Raney nickel, have been investigated for biocrude upgrading.3,13−16 The nitrogen contents of the above-mentioned biofuel that was obtained with different catalysts were generally between 1.5 and 4.7 wt %, and the oxygen contents were between 0.69 and 5.6 wt %. However, a biofuel with low heteroatom content and high yield was rarely reported. Elliott et al. obtained a biofuel with low heteroatom content at 405 °C and a space velocity below 0.2 h−1 in a continuous-flow reactor. Its nitrogen and oxygen contents were less than 0.25 and 1.8 wt %, respectively.17 Upgrading temperature ranges from 350 to 500 °C and reaction time ranges from 1 to 6 h for algae biocrude have been investigated.3,13,15,18−20 Many upgrading experiments were conducted above 400 °C, and the biofuel obtained at 350 °C © 2017 American Chemical Society

contains a relatively high ratio of heteroatoms. The yields of algae biocrude upgrading are generally between 40 and 90 wt %, but higher yields are often accompanied by a higher ratio of nitrogen or oxygen in the biofuel, which was not desired. Upgrading at higher temperature (exceeding 400 °C) could result in a decrease of the biofuel yield, and aromatization reactions gradually become dominant reactions.3,18 From the point of the biofuel yield, some attention could be given to upgrading the temperature below 400 °C, but the heteroatom removal effect should also be considered. The whole upgrading process could be classified into two steps based on the temperature variation: the first step is the temperature rising period, and the second step is the temperature holding period. The effect of the reaction time from 0 to 60 min for algae biocrude has not yet been investigated in detail. Longer upgrading time will multiply the energy consumption, which is not desired. Through a study of the two above-mentioned steps of upgrading, it probably could help to reveal the detailed upgrading routes of algae biocrude. Here, considering energy saving, environmental protection, and upgrading performance, we conducted HTL at 280 °C for only 5 min and directly separated the biocrude from water without using a solvent. The upgrading catalyst that was adopted in this study was NiMo/γAl2O3, which is economically friendly and has the potential to be applied to biocrude upgrading for an industrial scale. A guard catalyst was adopted, aiming to improve the upgrading effect of biocrude. In order to obtain biofuel with low heteroatoms, a modified two-stage upgrading was carried out. Control experiments were Received: Revised: Accepted: Published: 6378

April 5, 2017 May 6, 2017 May 8, 2017 May 8, 2017 DOI: 10.1021/acs.iecr.7b01405 Ind. Eng. Chem. Res. 2017, 56, 6378−6390

Article

Industrial & Engineering Chemistry Research

column detected result. Then, the relative content of each compound was calculated based on the total peak abundance of all compounds. However, it should be pointed out that there may still exist some compounds in the biocrude that could not be detected because the GC−MS testing methods or the solvent that was used to dissolve the biocrude could not guarantee that all compounds in the biocrude were detected. The gas compositions were analyzed by gas chromatography (Agilent 7890A), which contains three detectors: FID (with a Al2O3 packed column), TCD1 (with a 6 ft. × 1/8 in. Porapak Q column), and TCD2 (with a 6 ft. × 1/8 in. 5A molecular sieve column). Higher heating values (HHVs) of the biocrude and biofuel were determined by an oxygen bomb calorimeter (HWR-15E, Shanghai Testing Technology Institute). The biofuel’s moisture was analyzed with the method of Karl Fischer titration (AKF-1, Hogon Scientific Instrument Co., Ltd., Shanghai, China). 2.3. Experiments and Procedure. 2.3.1. Short-Time HTL and Biocrude Direct Separation. The HTL experiment was carried out in a designed 20 L autoclave with 9 L purified water and 1005.1 g of Nannochloropsis oceanica powder. The autoclave was purged with nitrogen for 10 min to eliminate the inner oxygen before heating. The initial pressure was atmospheric. The HTL experiment was conducted at 280 °C for 5 min. However, the real reaction time was probably longer than 5 min because of the preheating time. According to the research results given by Yu et al., the HTL reaction for microalgae (chlorella) generally did not occur until the reaction temperature was improved to 160 °C.21 In this study, it took about 41 min for the inner temperature of the autoclave to rise from 160 to 280 °C. Then, the inner temperature was cooled by flowing cooling water through the inner cooling coil, and it generally took less than 10 min for the temperature to decrease from 280 to 160 °C. The stirring speed was kept at 220 rpm, and the stirring function was not closed until the inner temperature dropped to 65 °C. The flow of cooling water was not stopped until the inner temperature reached 44.5 °C. The autoclave was opened 18 h later, and the inner temperature was 27.6 °C. The biocrude was recovered directly without using a solvent because it was totally isolated from the aqueous phase. The aqueous phase was brownish red with good transparency. In order to remove the potential water in the biocrude, the collected biocrude was kept in a drying oven (110 °C) for about 40 min, and it was stirred several times to accelerate the removal of water vapor. The total weight loss of the biocrude after being dried was 3.4%, including mainly evaporated water and a small amount of potential light end products. The final yield of the biocrude was about 41.8 wt %. 2.3.2. Upgrading Experiments. The upgrading experiments were carried out with a redesigned 100 mL 316L stainless steel autoclave. The distance between the autoclave paddle and the curved inner bottom of the autoclave body was about 5 mm in order to guarantee that the biocrude was efficiently stirred. For each experiment, about 8 g of biocrude was loaded into the autoclave. The catalyst loading was 2.4 g (30 wt % biocrude weight), and the guard catalyst loading was 1 g. They were mixed and loaded into the circular stainless steel wire cloth. The hydrogen pressure was 6 MPa except one control experiment. The upgrading experiment was kept at 350 °C for 1 h with a stirring speed of 600 rpm. Before the experiments, the autoclave was purged with hydrogen. When the experiments were finished, the autoclave was cooled as quickly as possible by water.

conducted to reveal the effects of the guard catalyst. More importantly, in order to clarify the compositions’ variation characteristic during the upgrading process, the compositions of the biocrude and biofuel were studied in detail, including the whole process from the period of the temperature rising to the holding at 350 °C for 60 min. The compositions’ content variation trends and the potential upgrading reaction routes are given in this study.

2. METHODS 2.1. Materials. The feedstock of Nannochloropsis oceanica was purchased from Shandong Yantai HaiRong Biology Technology Co., Ltd. (Shandong, China). For ultimate analysis and the biochemical composition of Nannochloropsis, refer to our previous study.16 The catalyst adopted in this study is Ni/ Mo/γAl2O3. Except the catalyst Ni/Mo/γAl2O3 (the core catalyst, labeled as Cat-c), the guard catalyst was introduced into this study (labeled as Cat-g). The guard catalyst RG-20B was purchased from QiMao Catalysts Co., Ltd. (Shandong, China). It contains an active component of Ni−Mo and has the same shape of the raschig ring, which is cylinder. The adoption of a guard catalyst is targeted to facilitate the hydrodenitrification effect. The Karl Fischer reagent was purchased from TianJin Sai Fu Rui Science and Technology Co., Ltd. (Tianjin, China). 2.2. Analysis Method. Elemental analysis including carbon, hydrogen, and nitrogen was carried out with an elemental analyzer (Vario EL Cube, Elementar Co. Ltd.). The ratio of an oxygen element was determined by the method of difference. In addition, the content of nitrogen (lower than 0.1 wt %) was also measured by an alkaline potassium persulfate oxidation− ultraviolet spectrometry method developed by the Hach Co. (the standard reagent was purchased from the Hach Co.). The UV spectrophotometer was UV-3300, Shanghai Mapada Instruments Co., Ltd. Composition analysis was conducted with a gas chromatography (GC)−mass spectrometry (MS) instrument (Agilent 7890A-5975C). In order to detect all compositions in the biocrude and biofuel as much as possible, compositions analysis was carried out with two chromatographic columns. One was the polarity capillary column DBWAX (30 m × 0.25 mm × 0.25 μm), and the other was the nonpolar capillary column HP-5MS (30 m × 0.25 mm × 0.25 μm). For polarity column testing, the column oven temperature was kept at 50 °C for 2 min, held at 175 °C for 2 min after ramping at 5 °C/min, and held at 250 °C for 1 min after ramping at 3 °C/min. The highest column oven temperature was 300 °C for a nonpolar column. For other GC−MS parameters, refer to our previous study.16 The results showed that the nonpolar capillary column could detect more compounds, but the polar capillary column could detect some compounds that the nonpolar capillary column could not detect. The two columns may give different peak abundances, even for the same compound. We adopted the same method to solve this problem for all samples. First, the compound result given by the nonpolar column was the basic compounds list of each sample. The compounds that were only detected by the polar column were added to the basic compound list, and as a whole, those compounds together represent the sample’s total composition. The peak abundances of those compounds only detected by the polar column were multiplied by a specific coefficient for each sample, which was determined by the ratio between the total peak abundance of the nonpolar column detected result and that of the polar 6379

DOI: 10.1021/acs.iecr.7b01405 Ind. Eng. Chem. Res. 2017, 56, 6378−6390

Article

Industrial & Engineering Chemistry Research Table 1. Elemental Contents of Upgraded Biofuel at Different Conditions and HHVs elemental content (wt %) biofuel sample biocrude No-Cat Cat-g Cat-c T330 P4 MPa Stage-1 Stage-2

C 66.66 70.62 82.39 83.40 83.33 83.46 84.90 85.32

± ± ± ± ± ± ± ±

H 0.04 0.25 0.37 0.33 0.26 0.03 0.14 0.06

9.61 10.04 11.99 12.64 12.84 12.80 12.94 14.08

± ± ± ± ± ± ± ±

N 0.0.04 0.03 0.17 0.13 0.15 0.04 0.30 0.06

3.73 3.77 2.97 1.71 1.98 2.16 1.27 0.016

± ± ± ± ± ± ±

0.04 0.02 0.11 0.07 0.06 0.02 0.04

O

H/C

O/C

N/C

20.00 15.57 2.65 2.25 1.84 1.57 0.89 ≤0.51

1.7304 1.7056 1.7465 1.8192 1.8493 1.8404 1.8290 1.9803

0.2250 0.1654 0.0241 0.0203 0.0166 0.0141 0.0078 0.0045

0.0480 0.0457 0.0309 0.0175 0.0203 0.0222 0.0128 0.0010

HHV (MJ/kg) 33.0 35.5 43.3 44.5 43.2 44.4 44.4 45.7

± ± ± ± ± ± ± ±

0.0 0.1 0.0 0.2 0.1 0.0 0.2 0.3

They were separately (1) 93.89 wt % n-hexadecane + 6.11 wt % indole (Model-1), (2) 94.13 wt % n-hexadecane + 5.87 wt % pyrrole (Model-2), and (3) 100% pyrrole (Model-3). The option of the model compounds indole and pyrrole was also for the purpose of studying their hydrodenitrification characteristics because the nitrogen in indole and pyrrole was difficult to remove, which was found in our previous study.16 nHexadecane is the most abundant compound in the upgraded biofuel. Upgrading with n-hexadecane could help to deduce the conversion pathways of n-hexadecane and other n-paraffins. As was known, the model compounds were simpler than the biocrude compositions; thus, the real upgrading properties of the model compounds probably have some difference compared with the biocrude. The upgrading research of this paper is not concentrated on small topics, leading to experimental designs that seem a little scattered. For easy understanding of the upgrading conditions for each biocrude upgrading experiment, the experimental symbols are summarized and explained as follows: No-Cat: upgrading without the core and guard catalysts, 6 MPa, 350 °C, 1 h Cat-g: upgrading only the load guard catalyst and no core catalyst, 6 MPa, 350 °C, 1 h Cat-c: upgrading only with the core catalyst and no guard catalyst, 6 MPa, 350 °C, 1 h T330: upgrading at 330 °C with core and guard catalysts, 6 MPa, 1 h P4 MPa: upgrading at 4 MPa hydrogen pressure with core and guard catalysts, 350 °C, 1 h Stage-1: 350 °C, 6 MPa, 1 h, core and guard catalysts; the feedstock was the biocrude Stage-2: 350 °C, 6 MPa, 1 h, core and guard catalysts; the feedstock was the Stage-1 biofuel S260, S290, S310, S330, and S340: five independent experiments, where “S” represents “stop the experiment”, meaning that the experiment should be stopped once the autoclave temperature reaches 260, 290, 310, 330, and 340 °C separately. Upgrading conditions: 6 MPa, core and guard catalysts S350-5, S350-15, S350-30, S350-45, and S350-60: five independent experiments, meaning that the experiments should be stopped once the reaction time at 350 °C reached 5, 15, 30, 45, and 60 min separately. Upgrading conditions: 350 °C, 6 MPa, core and guard catalysts

To reveal the effects of the guard catalyst, several control experiments were conducted. For each control experiment, there was only one variable, and except the variable, the fixed parameters of each control experiment were 6 MPa hydrogen pressure, 350 °C, and 1 h reaction time: (1) upgrading without the core and guard catalyst (the experiment was named NoCat); (2) upgrading only the loading of 1 g of guard catalyst, and no core catalyst was loaded (the experiment and guard catalyst were indicated by Cat-g); (3) upgrading only with the core catalyst Ni/Mo/γAl2O3, and no guard catalyst was loaded (labeled as Cat-c); (4) upgrading at 330 °C with Cat-c and Cat-g (labeled as T330); (5) upgrading at 4 MPa hydrogen pressure with Cat-c and Cat-g (labeled as P4 MPa). In order to obtain the biofuel with low heteroatoms at the optimized conditions, a two-stage upgrading experiment was carried out at the same upgrading conditions, and they were separately named Stage-1 and Stage-2. Stage-1 was conducted at the upgrading conditions of 350 °C, 6 MPa hydrogen pressure, 1 h reaction time, 2.4 g of core catalyst, and 1 g of guard catalyst, and the feedstock was the biocrude. Then, Stage-2 was conducted at the same upgrading conditions as those of Stage-1, but the feedstock was the biofuel that was obtained at Stage-1. It generally took about 50 min for the autoclave to reach 350 °C from ambient temperature. In order to reveal the compositions’ variation characteristics and reaction routes of the biocrude during the whole upgrading process, a series of studies were carried out at the optimized upgrading conditions (6 MPa hydrogen pressure, 2.4 g of core catalyst, and 1 g of guard catalyst). On the basis of the conclusion given by Peng et al., so that stearic acid could be completely upgraded to hydrocarbons at 260 °C in 8 h,22 five runs of experiments were conducted to investigate the upgrading characteristics during the temperature rising period. The experiment was to be stopped immediately when the autoclave temperature reached 260, 290, 310, 330, and 340 °C separately, and they were named S260, S290, S310, S330, and S340 (“S” represents “stop the experiment”). Another five independent runs of experiments were carried out to investigate the effect of the reaction time at 350 °C. The upgrading experiments were to be stopped immediately when the reaction time at 350 °C reached 5, 15, 30, 45, and 60 min separately, and they were labeled as S350-5, S350-15, S350-30, S350-45, and S350-60. A total of 10 upgrading experiments were carried out to reveal the variation characteristic of the biocrude compositions. In addition, another three model compound upgrading experiments were conducted under the same optimized conditions (350 °C, 6 MPa hydrogen pressure, 1 h reaction time, 2.4 g of core catalyst, and 1 g of guard catalyst) to support the analysis of the biocrude compositions’ conversion pathways.

3. RESULTS AND DISCUSSION 3.1. High-Quality Biofuel and Effect of the Guard Catalyst. 3.1.1. Biofuel Quality Analysis. The elemental content is one index of the biofuel quality, which is given in Table 1. The lowest heteroatom nitrogen content of the biofuel 6380

DOI: 10.1021/acs.iecr.7b01405 Ind. Eng. Chem. Res. 2017, 56, 6378−6390

Article

Industrial & Engineering Chemistry Research Table 2. Compositions of the Biocrude and Biofuels That Were Obtained at Different Upgrading Conditions .composition contents (%, relative contents based on the peak area) biofuel sample

n-paraffin

cycloparaffin

isoparaffin

aromatics

carboxylic acid

amide

biocrude No-Cat Cat-g Cat-c T330 P4 MPa Stage-1 Stage-2

0.50 43.02 68.49 68.46 68.63 67.88 64.41 60.77

1.12 5.25 6.03 7.49 10.57 8.52 9.85 9.28

7.10 9.63 8.34 6.19 6.82 9.53 11.55

0.76 4.87 6.67 11.15 10.77 8.86 11.85 16.40

40.15 7.45

23.70 7.07

nitrile 17.72 0.18 0.35

0.29

ester

others

5.54 1.57

28.24 5.32 8.71 2.79 3.30 2.53 2.86 0.87

0.63 0.11

0.29 0.16

Table 3. Gas Compositions Obtained under Different Upgrading Conditions gas composition (vol %) biofuel sample No-Cat Cat-g Cat-c T330 P4 MPa Stage-1 Stage-2

CH4 0.568 1.103 0.739 0.546 1.405 0.893 0.260

± ± ± ± ± ± ±

0.012 0.005 0.001 0.008 0.014 0.004 0.006

C2H6 0.382 0.752 0.527 0.394 1.063 0.641 0.220

± ± ± ± ± ± ±

0.008 0.010 0.001 0.012 0.025 0.006 0.002

C3H8 0.218 0.359 0.281 0.194 0.544 0.317 0.130

± ± ± ± ± ± ±

CO2

0.004 0.006 0.001 0.005 0.009 0.004 0.002

2.781 0.188 0.077 0.046 1.285 0.047

± ± ± ± ± ±

0.095 0.014 0.000 0.003 0.016 0.003

CO 1.180 1.067 0.658 0.611 0.754 0.697

± ± ± ± ± ±

0.043 0.051 0.001 0.012 0.006 0.007

H2 70.185 67.033 67.922 67.030 68.799 67.729 72.259

± ± ± ± ± ± ±

1.696 1.061 0.146 0.755 0.850 0.348 0.590

ratios of O/C and N/C were decreased from 0.2250 and 0.0480 to 0.0045 and 0.001 separately. The H/C ratio increased from 1.7304 to 1.9803. The composition characteristic of the biofuel is given in Table 2. The compositions were classified into hydrocarbons and heteroatom-containing compounds. Hydrocarbons, including n-paraffin, cycloparaffin, isoparaffin, and aromatics, are the most wanted compounds. The heteroatom-containing compounds, including carboxylic acids, amides, nitriles, esters, and others (which were difficult to sort), are the most wanted as hydrogenated compounds. Nitriles are an indication of the hydrodenitrification effect to some extent, and a detailed discussion is given in section 3.2. Even for upgrading without catalyst, the obtained biofuel still contains 55.37% saturated hydrocarbons and 4.87% aromatics. Among the n-paraffins in No-Cat, 10.32% pentadecane (C15) and 4.20% hexadecane (C16) were included. In addition, the heptadecane (C17) content was higher than that of octadecane (C18) and the tridecane (C13) content was higher than that of tetradecane (C14). However, in most upgrading circumstances, the biofuel’s hexadecane content was about 2 times higher than that of pentadecane; the result was reversed compared with No-Cat. As was known, C13, C15, and C17 were generated separately from fatty acids with even carbon numbers C14, C16, and C18 through decarboxylation or decarbonylation.23 This indicates that the decarboxylation and decarbonylation reactions were superior to the deoxygenation reaction for NoCat. Generally, the decarboxylation reaction leads to the generation of CO2 and decarbonylation for CO.24 From Table 3, it can be obtained that the No-Cat gas contains the highest amount of CO 2 and CO, which also gives proof of decarboxylation and decarbonylation reactions. However, for other upgrading runs, like Cat-g, Cat-c, T330, and so on, the results were reversed: the deoxygenation reaction was superior to the decarboxylation reaction. The saturated hydrocarbon contents for Cat-g and Cat-c were approached, but the remaining heteroatom-containing compositions in Cat-g were higher than those in Cat-c. Especially, the

was below the elemental analyzer’s detection limit, which was less than 0.1 wt %. Also, its oxygen content was less than or equal to 0.51 wt % because it was obtained based on calculations. The nitrogen content of the biofuel Stage-2 was further analyzed for accurate values by an alkaline potassium persulfate oxidation−ultraviolet spectrometry method developed by the Hach Co., and the result was 0.016%. The carbon and hydrogen contents of the Stage-2 biofuel were about 99.4 wt %, which contribute to the HHV of the biofuel of 45.7 MJ/ kg. The biocrude’s HHV was 33.0 MJ/kg. In comparison of Stage-2, the Stage-1 biofuel has a HHV of 44.4 MJ/kg, and it has an improvement of 34.4% compared to the biocrude. The improvement degree of the HHV was higher than that of other literature-reported results;14 most improvement degrees of HHVs were in the range of 7.2−30.8%. The HHV of No-Cat only improved 7.6% compared with that of the biocrude. Biofuels Cat-g and T330 almost have the lowest HHVs compared to those of biofuels obtained under other upgrading conditions (except No-Cat). The nitrogen content of Stage-1 was 1.27 wt %, which was lower than those of Cat-g and Cat-c. Stage-1 was conducted with both core and guard catalysts, and the reduction of nitrogen compared with Cat-c indicated that the adoption of the guard catalyst could facilitate the effect of denitrification to some degree. The oxygen content has also the same characteristic, which states that the adoption of the guard catalyst has a good effect on denitrification and deoxygenation compared with upgrading only with the core catalyst. Because the guard catalyst contains active components Ni− Mo, Cat-g also obtained a biofuel with low nitrogen and oxygen contents. Especially, the oxygen content of Cat-g has a dramatic reduction, but the nitrogen content of Cat-g was still as high as 3 wt %. Table 1 shows that the upgrading effect was far from satisfied without catalyst and guard catalyst. The nitrogen content of Stage-1 was reduced by 35.86% and 41.20% separately compared with T330 and P4 MPa, and the oxygen content was decreased too. This illustrates that the improvement of the temperature and hydrogen pressure has a positive effect on heteroatom removal. From biocrude to Stage-2, the 6381

DOI: 10.1021/acs.iecr.7b01405 Ind. Eng. Chem. Res. 2017, 56, 6378−6390

Article

Industrial & Engineering Chemistry Research

biofuels. Some high-carbon-number hydrocarbons were detected, such as C32; their generation pathways were suggested to have a relationship with carboxylic acids. The carboxylic acids could be upgraded into aliphatic aldehyde or alcohol, followed by a ketonization reaction or aldol condensation and further upgrading to generate high-carbonnumber paraffins.28−31 This was probably because the temperature rising period (from ambient to 350 °C) was a little long, which leads to the generation of high-carbon-number paraffins, and this should be avoided with a fast heating rate. The carbon distribution shift to low carbon numbers is obvious with the rise of the hydrogen pressure and temperature. This indicates that molecules tend to crack into lowcarbon-number compositions at higher temperature and pressure. 3.1.2. Effective Biofuel Yield. The effective biofuel yield has been defined in our previous study,16 which was to multiply the generally defined yield by C% + H% (wt %) content. C% and H% represent the elemental contents of carbon and hydrogen in the biofuel. The effective biofuel yield was defined to reveal the true quality of upgraded biofuel. The ye values of No-Cat, Cat-g, Cat-c, T330, P4 MPa, Stage-1, and Stage-2 are given in Figure 2. The error bar is indicated by the green area.

indole, phenol, and pyrrole contents in Cat-g were much higher than those in Cat-c (4.32%/0.75%). This illustrated that Cat-g has some effect on the upgrading, but its conversion functions were weak. When the composition characteristics of P4 MPa and Stage-1 are compared, the results show that improving the hydrogen pressure helps to increase the contents of cycloparaffins, isoparaffins, and aromatics. In contrast, the n-paraffin content decreased. For aviation kerosene compositions, higher contents of cycloparaffins and isoparaffins are necessary, which benefits some physical and chemical properties of aviation fuel.25 As shown in Table 2, almost the entire composition of the biofuel Stage-2 is hydrocarbons. There are only 0.87% compounds that belong to the others. The n-paraffins are further decreased and the isoparaffins and aromatics are further increased in Stage-2 compared with Stage-1. Not only are the oxygen and nitrogen contents in Stage-2 reduced to an undetectable level, but also the composition distribution is more reasonable. The gases collected during each upgrading experiment were analyzed, and the results are given in Table 3. The gas composition of P4 MPa contained 1.285 and 0.754 vol % CO2 and CO. The content of CH4 was also as high as 1.405 vol %, which was suggested to be generated from CO2 and CO through the methanation reaction. It seems that the decarboxylation and decarbonylation reactions were restricted at Stage-1 compared with P4 MPa. This indicated that higher hydrogen pressure went against the decarboxylation or decarbonylation reaction, and this characteristic was also demonstrated in other related research.26 It has been discussed above that the decarboxylation reaction has priority over the deoxygenation reaction in No-Cat, generating a large amount of CO2 and CO. However, unlike P4 MPa, the content of CH4 was very low in No-Cat gas. The results illustrate that the methanation reaction would not easily happen without participation of the catalyst. Previous studies of CO 2 and CO hydrogenation have revealed that the methanation reaction activity can be improved with a specific catalyst. Also, the one is nickel supported on oxides, such as Al2O3.24,27 Figure 1 shows the biofuel’s carbon number distribution characteristic. The biofuel Stage-2 contains 68.03% C8−C16 compositions, the ratio of which is the highest among the given

Figure 2. Effective biofuel yield of upgraded biofuels.

The ye value of Stage-1 was 83.81 wt %, and that of Stage-2 was 88.90 wt %. Thus, on the basis of the biocrude, the ye value of the final well-upgraded biofuel (Stage-2) could be calculated as 74.51 wt %, which approached the biofuel yield of the continuous-flow upgrading reactor obtained.17 Assuming that the oxygen and nitrogen in the biocrude were completely removed and the carbon was totally transformed into saturated hydrocarbons without any carbon loss, then the maximum theoretical yield could be calculated according to the elemental content of the biocrude (Table 1), which was 79.77 wt %. However, the carbon loss during the upgrading process was inevitable; some of them will transform into gas compositions. Considering the above point, the final effective biofuel yield of 74.51 wt %, from the biocrude to the high-quality biofuel, was relatively high. It should be pointed out that, in several runs of the experiment, a trace amount of water is observed in the autoclave after upgrading, but no obvious water was observed in Stage-1 and Stage-2. The potential existence of a trace amount of water may increase the biofuel yield slightly. However, the analysis results show that the moisture contents of the biofuels Stage-1 and Stage-2 were only 0.073 and 0.046 wt %. Thus, the

Figure 1. Carbon number distribution of upgraded biofuels. 6382

DOI: 10.1021/acs.iecr.7b01405 Ind. Eng. Chem. Res. 2017, 56, 6378−6390

Article

Industrial & Engineering Chemistry Research

Table 4. Elemental Content Variation Trend of the Biofuel during the Temperature Rising and Temperature Holding Processes of Upgrading elemental content (wt %) biofuel sample biocrude S260 S290 S310 S330 S340 S350-5 S350-15 S350-30 S350-45 S350-60

C 66.66 68.80 68.90 72.43 74.50 83.73 84.31 83.40 82.18 83.58 84.90

± ± ± ± ± ± ± ± ± ± ±

H 0.04 0.52 0.09 0.03 0.05 0.29 0.17 0.15 0.71 0.08 0.14

9.61 9.85 10.31 11.14 11.29 12.67 12.81 12.45 12.42 12.96 12.94

± ± ± ± ± ± ± ± ± ± ±

N 0.04 0.03 0.05 0.10 0.07 0.19 0.11 0.38 0.27 0.11 0.30

3.73 3.88 3.88 3.32 3.39 2.68 2.23 1.11 1.13 1.56 1.27

± ± ± ± ± ± ± ± ± ± ±

0.04 0.06 0.05 0.05 0.05 0.04 0.04 0.05 0.02 0.04 0.04

O

H/C

O/C

N/C

20.00 17.48 16.91 13.11 10.81 1.44 0.64 3.05 4.27 1.90 0.89

1.7304 1.7175 1.7956 1.8463 1.8182 1.8154 1.8238 1.7911 1.8143 1.8601 1.8290

0.2250 0.1906 0.1840 0.1357 0.1089 0.0129 0.0057 0.0274 0.0390 0.0171 0.0078

0.0480 0.0483 0.0483 0.0393 0.0390 0.0274 0.0227 0.0114 0.0118 0.0160 0.0128

Figure 3. Biocrude yield variation trend during the upgrading process.

potential water in the biofuel probably has little effect on the biofuel yield. Upgrading at 330 °C or 4 MPa obtained similar effective yields, both of them below the ye value of Stage-1. This illustrated that improving the temperature from 330 to 350 °C and the hydrogen pressure from 4 to 6 MPa could increase the effective yield to some degree. The ye value of Cat-c was slightly higher than that of Stage-1. However, considering the margin of error, it could be assumed that Cat-c had a ye value similar to that of Stage-1. In other upgrading conditions, such as Cat-g, No-Cat, the ye values were below 80 wt %. Also, the ye value of No-Cat was 63 wt %, which was the lowest. The gas composition of No-Cat contains 3.96 vol % CO2 and CO, which is shown in Table 3. This revealed that much more carbon was transferred to gas compositions and led to a decrease of the No-Cat biofuel yield. 3.2. From Biocrude to Biofuel: Compositions’ Content Variation Trend and Reaction Routes. To reveal the content variation trend and reaction routes of the compositions, a total of 10 experiments were investigated from the temperature rising period to the temperature holding period. Model compound upgrading was conducted to assist in the analysis of the biocrude conversion pathways. 3.2.1. Variation Trends of the Elemental Content, Yield, and Gas Composition. First of all, the elemental content variation trend of the biofuels during the upgrading process is given in Table 4. This indicates that the carbon content increased with rising temperature. From ambient temperature to 260 °C, the carbon content just improved from 66.66 to 68.80 wt %. Even when the temperature reached 290 °C, the biofuel’s carbon content still remained about 68.90 wt %,

without any remarkable improvement. In comparison, 310 and 330 °C brought a relatively obvious change for the biocrude elemental content. The remarkable elemental change of the biocrude occurred when the temperature reached 340 °C. The carbon content improved to 83.73 wt % rapidly, and the nitrogen content decreased to 2.68 wt %. When the temperature reached 350 °C, along with the reaction time, the elemental contents became relatively stable with slight fluctuation. The carbon content was in the range of 83−85 wt %. The nitrogen content of the biofuel decreased to a satisfied level (