Comparison study on potential syngas produced by mild thermoconversion of microalgal residues through proton nuclear magnetic resonance and thermogravimetric analysis-fourier transform infrared spectroscopy

This paper presented a mechanism study of syngas production through lipid-extracted microalgal residues by investigating how the structure of these residues affected the pyrolysis characteristics. The results showed that both Chlorella sorokiniana 21 and Monoraphidium 3s35 residues could be pyrolyzed efficiently at low temperatures from 120 to 380 °C, with a final residue of less than 22.75%. Hemicellulose with carboxyl groups was the main component of both microalgal residues, most likely contributing to the production of the C=O compounds and carbon monoxide (CO). A large amount of CHO, CH 2 O, and CH 3 O (60.09% in total) was observed from C. sorokiniana 21, which was 1.58 times more than that from Monoraphidium 3s35. Pyrolysis of the C. sorokiniana 21 residue yielded more C=O and CH n compounds with less activation energy (52.97 – 58.57 kJ mol − 1 ) and a higher reaction rate (0.105% s − 1 ) than that of Monoraphidium 3s35. The different pyrolysis characteristics between the two microalgal residues might be attributed to structural variations.


Introduction
Biomass has been considered one of the most promising renewable energy sources in the world. Compared with fossil fuels, biomass can be processed into liquid transportation fuels, which can substantially decrease the net emissions of carbon into the atmosphere [1]. Biomass can be obtained from land crops as well as ocean habitats. Microalgae, known as sunlight-driven cell factories, can convert carbon dioxide to potential biofuel materials [2] and are an important source of biomass, with the merits of high yields and a short life cycle [3].
There have been major challenges in biochemically converting lignocellulosic biomass into biofuel due to the complex structure of the biomass and the difficulty of separating its components in an economically feasible way [4]. Pyrolysis, however, has provided an alternative solution for converting cellulosic biomass into renewable energy and valuable chemicals [5]. In the last decade, intensive pyrolysis research has been performed on lignocellulosic biomass, ranging from agricultural residues such as corn stover and wheat straw to energy crops such as switchgrass [6]. Forestry wastes such as wood bark have also been the targets for pyrolysis [5]. Research on microalgal biomass pyrolysis using Py-GC/MS and DAEM [7] has been reported. Thermal degradation of the microalgae Chlorella vulgaris ESP-31, Nannochloropsis oceanica CY2, and Chlamydomonas sp. JSC4 has also been analyzed [8]. Recently, microalgae have mainly been used for lipid and biodiesel production [9]. The reutilization of residues after the biofuel conversion process is significant for increasing the economic efficiency of microalgae. However, the mechanism of pyrolysis for lipid-extracted microalgal residues has barely been reported. The online combination of thermogravimetric analysis (TGA) and fourier transform infrared (FTIR) spectroscopy has been successfully applied to the study of the evolution of volatile products with time in the t h e r m a l p y r o l y s i s o f p o l y m e r s [ 1 0 ] as well as Nannochloropsis sp. [11]. It is important to understand the mechanism of microalgal residue pyrolysis, the thermogravimetric decomposition of the volatile products, and the influence of the composition and structure of the microalgae on the pyrolysis kinetics.
In this study, lipid-extracted residues of the microalgae Chlorella sorokiniana and Monoraphidium were investigated with TGA/FTIR to elucidate their dynamic pyrolysis behavior including kinetics and volatile products produced. Furthermore, the chemical structures of the two microalgal residues were analyzed with proton nuclear magnetic resonance ( 1 H NMR) spectroscopy to determine the influence of the structural variation on the pyrolysis.

Microalgal residue preparation
The microalgal strains Chlorella sorokiniana 21 and Monoraphidium 3s35 used in this study were collected from the coastal waters of Shenzhen, Guangdong Province, China. Blue-green medium (BG-11) with 1.5% agar supplement was used to maintain the strains [12]. Each individual cultivation test was carried out with 3 replicates for 10 days. Slots on an incubation shaker platform (Constant Temperature Breeding Shaker, MOMA, Shanghai, China) were randomly assigned. To initiate the cultivation, 80 mL of axenic culture in the exponential phase (dry cell mass density 50 ± 0.08 mg L −1 ) was inoculated into 1 L flasks with 800 mL of sterile BG-11 medium. The flasks were covered with autoclavable foam and cultivated at a room temperature of 26°C with an initial pH of 7.1. Three rows of cool white portable fluorescent light tubes (PM-RGT8-30 W, Mei Optoelectronics Technology Company, Foshan, Guangdong Province, China) were used with a light intensity of 100.5 μmol m −2 s −1 ± 2.0 SE. Aeration of 0.25-0.75 L air per minute was applied to each cultivation flask. To collect the microalgal biomass, the culture s w ere ce ntrifuged (Ce ntrifuge 5 810R, Eppendorf, Germany) at 4000×g for 20 min and resuspended in distilled water. After repeating the above process 3 times, the microalgal biomass was freeze-dried (FreeZone 2.5, LABCONCO, USA) for 3 days. A Soxhlet extractor was used for lipid extraction from the previously obtained microalgal biomass, with 1.5 g extracted with chloroform/methanol (1/2, v/v) at 75°C for 12 h. After lipid extraction, the microalgal residues were then dried at 30°C under vacuum for 12 h. Three biological replicates for each sample of the microalgae biomass residue were prepared for further experiments.

Algal biomass structure analysis with 1 H NMR spectroscopy
1 H NMR spectra of microalgal residues were obtained by a Bruker 500 Ultrashield I100605 400-MHz spectrometer outfitted with a 5-mm broadband probe (Borosilicate Glass Company, China). Firstly, 20 mg of microalgal residue was dissolved in 1 mL of DMSO-d 6 (dimethyl sulfoxide-d6), filtered with a 0.45-μm PTFE (poly tetra fluoroethylene) filter to remove any suspended particulates, and then loaded into 5mm diameter NMR tubes [22]. Three measurement replicates were performed throughout this study.

Thermogravimetric analysis with TG-FTIR (thermogravimetric analysis coupled with fourier transform infrared analysis)
The thermogravimetric characteristics of the microalgal residues were determined by thermogravimetry coupled with Fourier transform infrared (TG-FTIR) analysis (STA449 F3, Naichi Instrument Company, China; TENSOR27 FTIR, Bruker, Germany). High-purity N 2 was used as the protective gas at a flow rate of 20 cm 3 min −1 . Approximately 10 mg of microalgal residue was heated from 25 to 650°C at 10°C min −1 and then held for 1 h at the highest temperature. The volatiles released during pyrolysis of the microalgae were detected online with a Bruker FTIR Tensor Spectrometer (TENSOR27 FTIR, Bruker, Germany). FTIR spectra were recorded from 400 to 4000 cm −1 . Before the microalgal samples were processed in the system, a blank test was carried out first to eliminate the buoyancy effect. Three measurement replicates were performed throughout this study.

Microalgae pyrolysis kinetics
The Arrhenius equation used in previously reported work was employed to describe the pyrolysis reaction rate in this study as follows [7]: where the notation of the Arrhenius equation X (Eq. (1)) was defined as the extent of the pyrolysis reaction. The conversion value X was determined as (w-w f )/(w 0 -w f ), t is the time of reaction, A is the pre-exponential or frequency factor, E is the activation energy, R is the universal gas constant, T is the absolute temperature, and n is the order of reaction. The linearized form of the Arrhenius equation was used to determine the kinetic constants A, E, and n. The simplified form of the linearized Arrhenius equation is as follows: where w 0 is the initial weight at the beginning of the pyrolysis, w f is the final weight at the end of each stage of the pyrolysis (details are explained in a later section), and w is the weight at any time of the pyrolysis process. Eq. (2) can be written in linear form [7]: where y, x, z, B, C, and D (Eq. (3)) are defined as follows (Eqs. (4)-(6)): The constants B, C, and D were estimated with multilinear regression of the TGA data for each stage using Microsoft Excel (Microsoft, Seattle, USA). The kinetic parameters (A, E, and n) were determined according to Eqs. (7), (8), and (9). Three measurement replicates were performed throughout this study.
The data were analyzed with Statistical Package of the Social Sciences (SPSS) software (Windows version 7.0, SPSS, Inc., Chicago, IL). Paired t test was used to analyze continuous data. A p value < 0.05 was considered significant.

Chemical composition and structure of C. sorokiniana 21 and Monoraphidium 3s35 residues
The chemical compositions of the microalgal residues of C. sorokiniana 21 and Monoraphidium 3s35 are presented in Table 1. As shown in the table, hemicellulose was the major component (over 40% dry-weight basis) of both microalgal residues. This finding is consistent with the chemical composition research on the marine algae Enteromorpha spp., in which the dominant component of the cellular wall is hemicellulose [23]. Cellulose and lignin were observed as the minor components in both microalgae (less than 3%). Previous research has reported that the polysaccharides of sea algae can be divided into cell wall polysaccharides such as cellulose and hemicellulose [24]. Lignin-like compounds and lignin were also discovered in the cell wall of the red alga Calliarthron cheilosporioides and Coleochaete algae [25]. Statistical analysis indicated that there was no significant difference in the composition between the two microalgal residues [26].
To assess differences in the chemical structure of C. sorokiniana 21 and Monoraphidium 3s35, we acquired 1 H NMR spectra to compare the functional group differences, as shown in Fig. 1 and Table 2. Both microalgal residues contained more than 13% alkanes based on the integrated p e a k s o f C H 3 a n d C H 2 ( 0 -1 . 6 p p m ) . N o t a b l y, C. sorokiniana 21 contains a large amount of CHO, CH 2 O, and CH 3 O (60.09% in total), which is 1.58 times more than that in Monoraphidium 3s35 (only 38.12%). In addition, the functional groups of ArOH, HC=C (nonconjugated), and CHO accounted for 31.51% in Monoraphidium 3s35, which is 4.81 times more than that in C. sorokiniana 21 (6.55%).
Overall, although there is no significant difference between the compositions of these two microalgae, structural variation was observed: C. sorokiniana 21 contained more CHO,  [27]. The presence of amorphous cellulose was indicated by strong C-H stretching vibrations, and the C-O-C stretching vibrations attributed to α-(1-4)-glycosidic linkages were designated as an "amorphous" absorption band, indicating an intensity increase in the amorphous cellulose. The different chemical compositions of cellulose could lead to various thermal decomposition results.
3.2 Pyrolysis of C. sorokiniana 21 and Monoraphidium 3s35 residues 3.2.1 Thermogravimetric characteristics Thermogravimetric analysis Figure 2 shows the thermogravimetry (TG) and derivative thermogravimetry (DTG) results of the microalgal residues of C. sorokiniana 21 and Monoraphidium 3s35 in the pyrolysis process. Both microalgal residues had three distinct weight loss stages during thermal degradation: dehydration, active pyrolysis, and passive pyrolysis, with temperature ranges of 25-120°C, 120-500°C, and 500-650°C, respectively. The active pyrolysis stage contained two substages, which was different from the pyrolysis of plant lignocellulosic biomass such as corn stover [28]. The active pyrolysis temperature for hemicelluloses is usually from 190 to 360°C, so the appearance of active pyrolysis substage 1 (120 to 380°C) is likely attributed to the large amount of hemicelluloses in the microalgal residue. The TG results also revealed that the final residue yields of the microalgal residue (22.20% for C. sorokiniana 21 and 22.75% for Monoraphidium 3s35) were significantly lower than those of plant lignocellulosic biomass, such as corn stover (33.32%) [29] and rice straw (29.5%) [30]. This result indicated that the thermal decomposition of microalgal residue can be more efficient than that of lignocellulosic biomass.  As shown in Fig. 2, the DTG curve for each microalgal residue had two peaks for the reaction rate (−dX/dt) of the pyrolysis. It was also observed that the reaction rate of microalgae pyrolysis increased rapidly from 115 to 310°C and then dropped dramatically from 310 to 504°C with a shoulder from 384 to 445°C. In substage 1, the peak of the reaction rate for C. sorokiniana 21 (0.105% s −1 ) was much higher than that for Monoraphidium 3s35 (0.0799% s −1 ). d 2 X/dt 2 analysis of TG According to Grønli et al. [31], temperatures for the thermal degradation of hemicellulose, cellulose, and lignin could be defined with d 2 X/d 2 t analysis. The d 2 X/d 2 t temperature curves obtained in this study are shown in Fig. 3. As shown in the figure, the characteristics of the thermal decomposition of the microalgal residue can be quantified through the following parameters: T onset(hc) is the extrapolated temperature at the beginning of hemicellulose decomposition; T shoulder1 is the temperature corresponding to the hemicellulose shoulder; T peak is the temperature of the maximum devolatilization rate during the entire pyrolysis; T onset (c) is the extrapolated temperature for the cellulose decomposition; T shoulder2 is the temperature corresponding to the cellulose shoulder and the beginning of the lignin tail in which lignin degradation occurs.
As shown in Fig. 3, the hemicelluloses of the C. sorokiniana 21 and Monoraphidium 3s35 residues started to be thermally decomposed at a T onset(hc) of 193°C and 191°C, respectively and reached the maximum reaction rate at a T peak of 310°C and 311°C. There was a pyrolysis shoulder observed for the microalgae C. sorokiniana 21 and Monoraphidium 3s35 residues at a T shoulder1 of 248°C and 240°C, respectively. Although the amount of cellulose in microalgal residue is much less than that in lignocellulosic biomass, the structure of cellulose is complicated and crystalline [32], compared with that of hemicelluloses. Thus, the pyrolysis stage of cellulose can be observed in Fig. 3. According to this figure, the cellulose of the microalgae C. sorokiniana 21 and Monoraphidium 3s35 residues began thermally decomposing at a T onset (c) of 390°C and 408°C, respectively. It has been reported that the weight loss of cellulose during pyrolysis happens at 315-400°C [33], which was consistent with our results. The lignin of the microalgae C. sorokiniana 21 and Monoraphidium 3s35 residues began pyrolysis at a T shoulder2 of 512°C and 504°C, respectively.

Microalgae pyrolysis kinetics
The kinetic parameters for the microalgae pyrolysis in substages 1 and 2 during the active pyrolysis stage are presented in Table 3. As shown in the table, the activation energy of the microalgae pyrolysis in substage 1 (52.97-58.57 kJ mol −1 ) was much lower than that in substage 2 (82.50-83.98 kJ mol −1 ). It was reported that the peaks of the distribution curve of the activation energy for rice straw, rice husk, corncob, and cellulose were 170, 174, 183, and 185 kJ mol −1 [34], respectively. It is known that the activation energy represents the potential barrier needed to be overcome in the   [35]. Compared with lignocellulose biomass, microalgae require less energy to start the thermogravimetric reaction. The results also showed that the activation energy of the residue of C. sorokiniana 21 was less than that of Monoraphidium 3s35, indicating that the former is more easily pyrolyzed and has a lower energy barrier.

Volatile products analysis of TG-FTIR
To identify the major volatile species and continuously measure the volatiles and the devolatilization temperature during microalgae pyrolysis, TG analysis coupled with FTIR spectroscopy was applied in this study. The TG-FTIR spectra of these two microalgae are shown in Fig. 4. The composition of the volatile products was determined with the characteristic wavenumber bands, while the yield history of the products in the time course was identified by the absorbance against time. An integral form of the Lambert-Beer relation was used for the abovementioned determination and identification [36]. The integrated absorbance (IA, measured in cm −1 ) is the integral value of the spectral absorbance over a selected wavenumber interval characteristic for the compound of interest, which can give a quantitative measure of the gaseous species detected.
During the pyrolysis of microalgal residues, light gases (namely, C=O compounds, CH n compounds, CO 2 , and CO) were detected in the specific wavenumber intervals listed in Table 4. The results are shown in Fig. 5. Compounds with characteristic carbonyl bonds, typically organic acids, esters, aldehydes, and ketones, are called C=O compounds [37]. Compounds with characteristic C-H bonds, typically alkyl, aliphatic, and aromatic compounds, can be classified as CH n compounds. Fig. 5a, two peaks were observed for the production of C=O compounds (R-CHO and R-COOH) during the pyrolysis of both C. sorokiniana 21 and Monoraphidium 3s35. According to our results, the temperatures for the maximum production of C=O compounds from the pyrolysis of C. sorokiniana 21 and Monoraphidium 3s35 were 251°C and 333°C, respectively. The active pyrolysis temperature for hemicelluloses is usually from 190 to 360°C. The compositions of hemicellulose in the microalgae C. sorokiniana 21 and Monoraphidium 3s35 were 41.56% and 42.38%, respectively. The production of the C=O compounds is probably due to the pyrolysis of hemicellulose in the microalgae. It was reported that the pyrolysis of Chlorella vulgaris microalgae occurred in the temperature range of  133-537°C, and the main weight loss peak occurred at approximately 296°C, which is consistent with our results. At low temperatures, hemicellulose decomposition contributes to C=O and COOH functional group generation [3].

C=O compounds As shown in
Hemicelluloses are heterogeneous polymers of pentoses (xylose, arabinose), hexoses (mannose, glucose, galactose), and sugar acids, with a random, amorphous structure. The backbone of hemicellulose consists of O-acetyl, α-larabinofuranosyl, α-1,2-linked glucuronic, or 4-Omethylglucuronic acid substituents. The release of carbonyl groups can be caused by the cracking of xylan and glucomannan from hemicellulose [4]. Acidic products, especially acetic acid and formic acid, with C=O groups, are commonly detected in the bio-oil from the pyrolysis of hemicellulose [5]. The prevalent mechanism of the formation of acetic acid involves a primary elimination reaction of the active Oacetyl groups linked to the main xylan chain at the C2 position [6].
Carbon monoxide Carbon monoxide (CO) released from pyrolysis of the microalgae is shown in Fig. 5b. The peaks of CO release were observed at 230°C and 323°C for C. sorokiniana 21 and Monoraphidium 3s35, respectively. C. sorokiniana 21 reached a higher maximum production of CO at a lower temperature than did Monoraphidium 3s35.
Since the release of CO was mainly caused by the cracking of ether (C-O-C) and carboxyl (C=O) groups [38], the superiority of C. sorokiniana 21 in producing CO might be attributed to the chemical structure difference between these two microalgae. On the other hand, it has been reported that the release of CO from cellulose was very small, with only a small peak, and that almost no CO was evolved from lignin pyrolysis at low temperature (< 600°C) [33]. The authors also reported that there was a peak for CO release from hemicelluloses at 280°C, which was within the temperature range for CO release in our observation. Thus, we believe that the CO production of microalgal residues is mainly from hemicelluloses.
CH n compounds The CH n compounds released from pyrolysis of the microalgal residues are shown in Fig. 5c. As shown in the figure, the CH n compounds showed two main release peaks at 278°C and 378°C for C. sorokiniana 21 and a shoulder from 322 to 385°C for Monoraphidium 3s35. The primary pyrolysis may give rise to the first release peak, while the secondary pyrolysis at a higher temperature might be the driving force for the second release peak of gases, which is in agreement with a report by Yang et al. (2007). The first release peak of CH n compounds from C. sorokiniana 21 might be caused by the cracking of the high content of CH 3 O compounds as shown in Table 2. The cracking procedure mainly happened during primary pyrolysis at low temperature. For Monoraphidium 3s35, the production of CH n compounds might be due to secondary pyrolysis. It was reported that 1,4-anhydro-d-xylopyranose is one of the main condensable products from hemicellulose pyrolysis, which is attributed to cleavage of the glycosidic linkage of the xylan chain, followed by rearrangement of the depolymerized molecules. Studies have inferred that most of the produced 1,4-anhydro-dxylopyranose would be instantly consumed to produce twocarbon and three-carbon fragments as well as gases including CH 4 , acting as an intermediate product from the pyrolysis of hemicellulose [5].
Carbon dioxide Carbon dioxide (CO 2 ) released from pyrolysis of the microalgae is shown in Fig. 5d. As shown in the figure, CO 2 generation reached a peak at 250°C and 313°C for C. sorokiniana 21 and Monoraphidium 3s35, respectively. Based on the results, Monoraphidium 3s35 produced less CO 2 than C. sorokiniana 21 during pyrolysis. The production of CO 2 is due to the cracking and scission of C-C and C-O bonds during pyrolysis [39].

Conclusion
The present study showed that microalgal residues have clear advantages over plant lignocellulosic materials in terms of pyrolysis, with a lower activation energy (52.97-83.98 kJ mol −1 ) and lower final residue yields (22.20-22.75%). Due to the high content of hemicellulose (41.56-42.38%) and the predominant structure of carboxyl and carbonyl functional groups, the main evolution products were C=O and CH n compounds, presenting great potential as syngas. Moreover, the chemical structure differences between C. sorokiniana 21 and Monoraphidium 3s35 probably resulted in the observed variation in the pyrolysis characteristics and production of volatiles.
Authors' contributions Details of the individual author's contribution towards the manuscript have been illustrated below: Professor Shuangfei Li, as the first author, collected, processed and analyzed the data. He was also responsible for all drafts of the manuscript.
Professor Xuewei Yang, as the corresponding author, was responsible for the conceptualization of the study, design and planning of the manuscript, as well as the analysis and interpretation of the results.
Dr. Liangxu Liu, as the coauthor, was responsible for analyzing the data and improving the figure quality, English writing and discussion section of the manuscript.
Professor Jay Jiayang Cheng, as the coauthor, was responsible for critical reading and finalization of the manuscript.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.

Statement of informed consent Not applicable.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.