Biofuels from waste fish oil pyrolysis: Chemical composition

Fuel 89 (2010) 563–568 Contents lists available at ScienceDirect Fuel journal homepage: www.elsevier.com/locate/fuel Biofuels from waste fish oil pyrolysis: Chemical composition A. Wisniewski Jr. b,*, V.R. Wiggers c, E.L. Simionatto b, H.F. Meier c, A.A.C. Barros c, L.A.S. Madureira a a Chemistry Department, Federal University of Santa Catarina (UFSC), SC, Brazil Chemistry Department, Regional University of Blumenau (FURB), SC, Brazil c Chemical Engineering Department, Regional University of Blumenau (FURB), SC, Brazil b a r t i c l e i n f o Article history: Received 9 March 2009 Received in revised form 19 June 2009 Accepted 23 July 2009 Available online 6 August 2009 Keywords: Waste fish oil Pyrolysis Biofuels Heavy bio-oil Light bio-oil a b s t r a c t In a previous study, waste fish oil was converted into bio-oil by a fast pyrolysis process at 525 °C in a continuous pilot plant reactor with 72–73% yield. The bio-oil was distilled to obtain light bio-oil and heavy bio-oil and these biofuels were characterized in terms of their physico-chemical properties. In this study, the chemical composition of light bio-oil and heavy bio-oil was determined using GC-FID, GC–MS, 1H and 13 C NMR techniques. The GC–MS analysis of waste fish oil showed the main composition of fatty acids to be the following: C16:0 (15.87%), C18:2 (20.96%), C18:1 (17.29%), C20:5 (5.11%), C20:1 (7.59%), C22:6 (4.53%), C22:1 (10.42%) and others. The GC-FID analysis of the light bio-oil showed 482 compounds that were PIONA classified as paraffins (4.48%), iso-paraffins (8.31%), olefins (26.56%), naphthenes (6.07%) and aromatics (16.86%). The heavy bio-oil had a similar chromatographic profile as diesel oil, with a high content of carboxylic acids and olefins. These results are in good agreement with those for the gasoline and diesel oil fractions of petroleum. Ó 2009 Elsevier Ltd. All rights reserved. 1. Introduction According to the Food and Agriculture Organization (FAO), in 2005 the estimated world fish production was around 142 millions tons. Approximately 75% of this production is used for direct human consumption. The remaining 25% is destined for non-food products, in particular the manufacture of fishmeal and oil. For 2008 the estimated world fish production was 144 millions tons [1,2]. The volume of waste produced by processing plants is calculated to be around 50% of the total processed fish, for which the amount of oil varies from 40% to 65% [3]. The fast pyrolysis of triglycerides has been investigated with a view to biofuel production [4–11]. According to the pyrolysis reaction scheme, presented in Maher and Bressler [12], many different chemical groups can be produced during the pyrolysis reaction. The liquid product (bio-oil) obtained from triglyceride pyrolysis has a very complex composition [13] and requires the use of particular analytical techniques, and a precise determination of the bio-oil composition has not been carried out to date [6,14]. This bio-oil can be used directly as a fuel or can be fractionated to obtain purified hydrocarbons in the range of gasoline and diesel. These biofuels have been compared to fossil fuels and the results show partial agreement with fossil fuel specifications [4,9,10]. The need to improve the quality and the specific regulations of * Corresponding author. Tel.: +55 47 3221 6096; fax: +55 47 3221 6001. E-mail address: [email protected] (A. Wisniewski Jr.). 0016-2361/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.fuel.2009.07.017 these biofuels has led to an increase in the number of studies related to the determination of their chemical composition [15,16]. Bio-oil has a large variety of compositions as a function of the feedstock [17,18]. It can be produced from biomass based on triglycerides like soybean, palm, castor and canola [5,9,13,19,20], as well as animal fats, lard, poultry fat and fish oil capsules [7,10,21], and the major products are alkanes, alkenes, ketones, aldehyde, aromatics and carboxylic acids. Bio-oils based on lignin-cellulosic biomass contain phenols, benzenediols, furanes and their derivates as major compounds [16,22–25]. Several methods to determine the relative amounts of hydrocarbons in crude oil refining products can be employed [26]. A standardized method using high resolution gas chromatography (GC) with a 100-m capillary column can be used to determine individual components of spark ignition fuels [27]. The gas chromatography method for classification of gasolines into paraffin, isoparaffin, olefin, naphthene and aromatic (PIONA) groups is based on the retention index [28–30]. Proton nuclear magnetic resonance (1H NMR) is appropriate for the measurement of the concentrations of aromatic, olefinic and aliphatic fractions of gasoline samples [31]. Paraffin, aromatic and naphthene contents determined by the same spectroscopy method show a very good agreement when compared with the gas chromatography method [32,33]. In a previous study, some physico-chemical properties of waste fish oil (WFO), bio-oil (BO), light bio-oil (LBO) and heavy bio-oil (HBO) were determined and compared to Brazilian fuel specifications. LBO and HBO were also analyzed to determine the yields of 564 A. Wisniewski Jr. et al. / Fuel 89 (2010) 563–568 compounds according to the carbon numbers in the chain by GC [34]. In this study, the chemical compositions of WFO, LBO and HBO were investigated. The biofuels were analyzed by GC-flame ionization detection (FID), mass spectrometry (MS) and proton (1H) and carbon (13C) nuclear magnetic resonance (NMR) spectroscopy. 2. Experimental 2.1. Chemical composition of waste fish oil The WFO was esterified with methanolic sulfuric acid (90:10, v/ v) [35] before being submitted to gas chromatography/mass spectrometry (Varian CP-3800/Saturn 2000) and quantified by gas chromatography/flame ionization detection analysis. The esterified WFO was compared with standard solutions of Fatty Acid Methyl Esters (FAMEs) and homologous n-alkanes obtained commercially from SupelcoÒ. The FAME standards were utilized to identify the unsaturated fatty acids through the similarity of their retention time (RT). The n-alkane standards were applied to determine the retention index (RI) [36] of all fatty acids in the sample, which can be used as a reference. The analysis was performed using a CP-Sil 8 Cb Low Bleed capillary column (30 m  0.25 mm i.d., 0.25 lm film thickness). Helium (99.999%) was used as the carrier gas with a constant flow rate of 1 mL min 1 for GC–MS and 1.2 mL min 1 for GC-FID, and oven temperature of 80 °C (3 min), 5 °C min 1 to 250 °C (15 min). Injector temperature for both analyses was 250 °C and FID temperature was 280 °C. The MS was conducted with the following operation conditions: transfer line 240 °C, manifold 80 °C, ion trap 175 °C and electron energy 70 eV. 2.2. Waste fish oil pyrolysis The bio-oil was obtained in a previous study in a continuous reactor pilot plant at 525 °C with mass flow rate of 3.2 kg h 1 and underwent simple distillation to produce light bio-oil and heavy bio-oil. The physico-chemical properties of these biofuels have been previously described in the first part of this study [34]. 2.3. 1H and 13 C NMR of biofuels The NMR spectra of BO, LBO and HBO, obtained as described in [34], were recorded at 22 °C using a Bruker AC-300 spectrometer at 300.13 MHz (1H) and 75.47 MHz (13C). Chemical shifts were referenced in parts per million (ppm) relative to the signal of tetramethyl silane (TMS). The sample was dissolved in deuterated chloroform. 2.4. PIONA analysis of light bio-oil The GC-FID and GC–MS analysis were conducted in a CP-Sil PONA (100 m  0.25 mm; film thickness 0.5 lm). Helium (99.999%) was used as the carrier gas with a constant pressure of 50.5 psi for FID and a constant flow of 1.0 mL min 1 for MS; oven temperature of 35 °C (15 min), 1 °C min 1 to 60 °C (20 min), 2 °C min 1 to 200 °C (10 min); injector temperature of 250 °C; FID temperature of 280 °C; and injection volume of 0.3 lL. The GC–MS analysis was conducted with the following operation conditions: transfer line 240 °C, manifold 80 °C, ion trap 175 °C and electron energy 70 eV. The LBO was compared with a Nafta standard solution obtained commercially from SupelcoÒ. The retention indexes of the Nafta standards were determined [36] and the hydrocarbons were identified by retention time similarity. 2.5. Chemical composition of heavy bio-oil The GC-FID and GC–MS analysis of HBO was performed using a CP-Sil 8 Cb Low Bleed capillary column (30 m  0.25 mm i.d., 0.25 lm film thickness). Helium (99.999%) was used as the carrier gas with a constant flow rate of 1.2 mL min 1 for FID and 1.0 mL min 1 for MS; oven temperature of 100 °C (5 min); 5 °C min 1 to 250 °C (20 min); injector temperature of 250 °C; FID temperature of 280 °C; and injection volume of 0.5 lL. The MS was conducted with the following operation conditions: transfer line 240 °C, manifold 80 °C, ion trap 175 °C and electron energy 70 eV. 3. Results and discussion 3.1. The chemical composition of waste fish oil The analysis of methyl esters of WFO revealed 13 peaks and the chemical composition is shown in Table 1. The unsaturated acids C14:0, C16:0 and C18:0 were identified by comparison of their retention times with the FAME standard. Others were identified by comparison of GC–MS spectra with those in the NIST 02 Mass Spectral Database. The major fatty acids found were C16:0, C18:1, C18:2 and C22:1, responsible for 64% of the total composition. Eicosapentaenoic acid (C20:5) and docosahexaenoic acid (C22:6), typical fatty acids of fish oils, were identified and quantified as 5.11% and 4.53%, respectively. 3.2. Light bio-oil fraction of bio-oil The GC-FID analysis of LBO (Fig. 1) shows 482 peaks with relative concentrations higher than 0.01%. The identification of compounds by Detailed Hydrocarbon Analysis (DHA) was carried out using the retention index according to ASTM D 6729-01 and was confirmed by GC–MS through a comparison of the MS spectra with those in the NIST 02 Mass Spectral Database. A correction coefficient related to the detector factor response for specific hydrocarbons was multiplied by the GC-FID relative concentration to obtain the theoretical absolute quantification to express their concentration in percent volume/volume (% DHA). Table 2 shows 31 compounds with a GC-FID relative concentration above 0.5%, which represent almost 44% of the total LBO chemical composition. It can be noted that the major compounds identified are aromatics such as benzene, toluene and ethylbenzene and olefins including 1-pentene, 1-hexene, 1-heptene and 1-octene. The DHA results were classified according to PIONA. Table 3 shows the PIONA classification of light bio-oil in comparison with Table 1 Chemical composition of waste fish oil. Peak RT (min) RI Methyl ester of % Relative GC-FID 1 2 3 4 5 6 7 8 9 10 11 12 13 21.945 25.659 26.102 29.041 29.278 29.393 29.472 29.830 32.218 32.847 35.247 35.481 36.098 1723 1903 1926 2082 2095 2101 2105 2126 2263 2301 2448 2462 2499 C14:0 C16:1 C16:0 C18:4 + C18:0 C18:2 C18:1 C18:1 C18:0 C20:5 C20:1 C22:6 C22:5 C22:1 Total 6.02 4.38 15.87 1.49 20.96 17.29 2.43 3.06 5.11 7.59 4.53 0.85 10.42 100.00 branched 565 A. Wisniewski Jr. et al. / Fuel 89 (2010) 563–568 Fig. 1. GC-FID analysis of light bio-oil. Table 2 Main chemical composition of light bio-oil. Peak RT (min) RI Compound % Relative GC-FID % DHA 12 22 31 54 57 64 72 77 95 101 113 120 144 161 164 170 171 178 200 215 230 252 263 289 315 322 331 360 372 374 407 8.442 10.601 12.517 17.534 18.743 20.278 22.215 23.740 31.619 33.387 37.811 39.847 49.087 55.615 56.914 60.318 60.934 64.072 72.252 76.680 80.918 85.399 88.090 93.652 99.380 101.105 103.623 109.384 112.164 113.035 119.841 483.19 541.23 583.56 640.17 651.82 666.62 685.30 700.00 748.64 759.67 787.29 800.00 837.35 863.75 868.99 883.74 886.41 900.00 944.12 968.00 990.86 1022.72 1044.65 1090.00 1146.17 1163.87 1189.70 1248.80 1277.31 1286.24 1356.05 1-Pentene Cyclopentene 1-Hexene Benzene Olefin C7 3-Methylcyclopentene 1-Heptene n-Heptane toluene 2-Methyl-3-ethylpentane 1-Octene n-Octane Ethylbenzene 3-Ethylheptane o-Xylene Unclassified C8 Isobutylcyclopentane n-Nonane n-Propylbenzene 1-Methyl-2-ethylbenzene 1-Decene Unclassified C10 n-Butylbenzene 1-Undecene n-Pentylbenzene Naphtalene 1-Dodecene n-Hexylbenzene 1-Tridecene 2-Methylnaphtalene 1-Tetradecene Total 0.88 0.91 3.69 3.74 0.65 1.36 4.28 0.73 2.95 0.64 3.14 0.69 1.44 0.80 0.75 0.58 2.83 0.64 0.67 0.74 2.54 0.59 0.51 2.24 1.21 1.22 1.94 0.55 1.16 0.66 1.50 46.22 1.04 0.88 4.16 1.79 0.72 1.61 4.66 0.83 2.43 0.73 3.34 0.76 1.19 0.85 0.61 0.64 2.76 0.69 0.57 0.61 2.60 0.62 0.43 2.28 1.03 0.89 1.96 0.47 0.83 0.68 1.53 44.17 petroleum-based fuels (gasolines A and C1). The parameters that differed most between light bio-oil and gasoline A were the contents of paraffins, iso-paraffins and naphthenes compounds, which were lower in the former. On the other hand, the light bio-oil shows a higher content of olefins in comparison with gasolines A and C. Olefins are known to provide more desirable octane ratings than n-paraffins, but are more unstable in the presence of oxygen and can 1 Gasoline A is a petroleum-based fuel and gasoline C is the Brazilian gasoline commercialized with a 20% content of ethanol, as an anti-knocking additive. contribute to the production of gum deposits during long-term storage. This means that fuels with high levels of olefins require the use of antioxidants. The aromatic compounds, like olefins, have good octane numbers, but tend to be more toxic and some countries have specific regulations for certain compounds, for example, benzene. On the other hand, olefins and aromatics are added to gasolines to replace other more toxic compounds used as octane enhancers, for example, organometallic compounds. The results for the PIONA classification also show high contents of C14+ compounds which can be reduced by optimization of the fractionation process to obtain greater similarity between the bio- 566 A. Wisniewski Jr. et al. / Fuel 89 (2010) 563–568 these compounds can be classified according to the PIONA classification system. Table 3 PIONA classification of light bio-oil. Classes % (v/v) Light bio-oil % (v/v) Gasoline C % (v/v) Gasoline A Aromatics Iso-paraffins Naphtenes Olefins Oxygenates Paraffin C14+ Unclassified Total 16.86 8.31 6.07 26.56 0.06 4.48 5.30 32.38 100.00 14.81 19.35 16.26 11.72 17.76 13.65 0.00 6.44 100.00 18.85 23.67 19.89 13.53 0.16 16.55 0.00 7.36 100.00 fuel fraction and fossil fuels. The 32% of unclassified compounds were distributed across a high number of compounds (272) and each one contributes with a relatively low concentration. The difficulty in identifying all of these compounds by GC reflects the complex composition of biofuels obtained from bio-oil derivates, and there is also a lack of studies and standards. The use of complementary techniques like NMR, as described below, shows that 3.3. Heavy bio-oil fraction from bio-oil The results of the GC-FID analysis of HBO are shown in Fig. 2. The chromatogram of HBO was compared with that of diesel oil (DO) and homologous n-alkane standards. The similarity between the chemical compositions of HBO and DO is clear, basically with n-alkanes C10–C26 in the composition. The major peaks for HBO show a small difference in the retention times in relation to those of n-alkane standards. The peaks on the MS spectra were identified as being homologous terminal olefin compounds with a number of carbons lower than C22. The main biofuels obtained from the biomass pyrolysis, as with the heavy bio-oil, had a high acid index as previously reported [34]. In the chromatogram ‘‘a” of Fig. 2, the presence of three peaks (225, 257 and 275) can be noted close to the retention time of n-alkanes C18, C20 and C22 as asymmetric chromatographic bands. The HBO was submitted to the methanolic sulfuric acid procedure to esterify Fig. 2. GC-FID chromatogram of HBO (a) DO (b) and n-alkanes standards (c). Fig. 3. Total ion chromatogram of HBO (a) and esterified HBO (b). 567 A. Wisniewski Jr. et al. / Fuel 89 (2010) 563–568 the possible fatty acid residues. Fig. 3 shows the total ion chromatogram of HBO and an esterified sample of HBO. The total ion chromatogram of the esterified sample of HBO indicates the disappearance of the previously described asymmetric peaks and the appearance of peaks at 225b, 257b and 275b, as Table 4 Main chemical composition of heavy bio-oil (Fig. 2a). Peak RT GC-FID (min) % Relative GCFID Compounds identified by GC– MS 6 10 18 31 50 73 101 129 161 225 249 257 275 276 1.941 2.216 2.768 3.833 5.827 8.479 11.288 14.037 16.811 23.147 25.119 27.239 30.416 30.830 Total 1.32 1.75 1.02 1.33 1.80 1.36 1.34 1.52 1.09 6.46 0.93 11.88 3.13 1.87 36.80 Alkenes + benzene derivatives Decene Undecene Dodecene + naphtalene Tridecene Tetradecene Pentadecene Tetradecanoic acid (C14:0) Pentadecanoic acid (C15:0) Hexadecanoic acid (C16:0) Octadecenoic acid (C18:1) Octadecanoic acid (C18:0) Fig. 4. 1H NMR (a) and shown in the Fig. 3b. These compounds were identified by MS and the relative concentrations were determined by FID (Fig. 2a) and represent 36.8% of the chemical composition of HBO (Table 4). Aromatics, olefins and carboxylic acid residues were the main compounds found in the HBO biofuel. The main characteristics of biofuels such as HBO, a diesel-like fuel, are their high content of olefins and a high temperature value for the cold filter plugging point (+14 °C), as discussed in a previous paper. The presence of carboxylic acid residues is the result of the unreacted carboxylic acids from the feedstock. The acidity of biofuels can reflect in an increased corrosion rating. The advantage is that the origin of this acidity is from weak organic acids, in contrast to fossil fuels where the acidity is from sulfur compounds. Changes in the residence time, the use of a specific catalysts and a post esterification reaction of these biofuels are currently under study, aiming to adapt these biofuels to comply with national regulations. 3.4. 1H and 13 C NMR of bio-oil, light bio-oil and heavy bio-oil The 1H NMR spectrum of the bio-oil is shown in Fig. 4a. As expected, the light bio-oil and heavy bio-oil proton spectra show the same signals with different mol% hydrogen distribution, as listed in 13 C NMR (b) spectral regions for bio-oil. Table 5 Assignment of 1H NMR spectral regions to triglyceride pyrolysis biofuels. Hydrogen type 1 Bio-oil Light bio-oil Heavy bio-oil Aromatics Olefins (–HC@CH–) CH2, adjacent to –CH@CH CH3, adjacent to -Ph CH3, CH2 e CH, adjacent to – (C@O)OR; –(C@O)OH; –(C@O)H CH, adjacent to –CH2–CH@CH CH, adjacent to –CH2–CH2 CH3, adjacent to –CH@CH CH2 e CH, adjacent to –CH2R CH2, adjacent to –CH2–CH2 CH2, adjacent to –CH2–CH@CH CH3, adjacent to –CH2–R CH3, adjacent to –CH2–CH2 CH3, adjacent to –CH2–CH@CH Aliphatics (total) 7.0–9.0 5.0–6.5 2.0–2.5 2.44 6.14 6.31 5.11 10.53 19.84 2.37 4.32 15.15 1.5–2.0 16.23 10.44 9.47 1.0–1.5 55.91 37.63 54.59 0.5–1.0 12.96 16.44 14.08 0.5–3.0 91.41 84.35 93.29 H Chemical shift (ppm) Mol% (% of total Hydrogen) 568 A. Wisniewski Jr. et al. / Fuel 89 (2010) 563–568 Table 5. The simple spectrum can be divided into three distinct bands: aromatics, olefins and aliphatic hydrogens. Resonances between 7 and 9 ppm were assigned to aromatic structures, and between 5 and 6.5 ppm they were attributed to non-conjugated olefins (centered at 5.3 ppm). The region between 0.5 and 2.5 ppm is strongly overlapped and contains signals mainly due to cycloalkanes (naphthenes), and normal- and iso-paraffins. The total aliphatic hydrogen intensity (0.5–2.5 ppm) is predominant in both biofuels. The 1H NMR integration results show that the contents of aromatic and olefin compounds in LBO are twice those in HBO. It is possible to consider that in the pyrolysis of triglycerides, the formation of unsaturated compounds occurs with small chains and lower boiling points (BP < 220 °C). The BO, LBO and HBO were also analyzed by 13C NMR (Fig. 4b). The biofuels show the same signals on the 13C NMR spectra. The signal with a chemical shift of 180 ppm confirmed the information obtained from the GC–MS analysis regarding the presence of carboxylic acids in these biofuels with more relative intensity in the HBO sample. The identification of carbons with a chemical shift of 114 ppm as terminal-CH2 and 139 ppm as vinyl confirms the presence of alkenes with terminal unsaturation. Carbons with a chemical shift between 30 and 40 ppm were assigned to naphthenes. Aromatics (127–130 ppm), and methyl and methylene carbons (10–30 ppm) were also identified. 4. Conclusions The full determination of the chemical composition of waste fish oil as described herein shows the carboxylic acids characteristic of fish oil, that is, eicosapentaenoic and docosahexaenoic acids. The carboxylic acids C16:0, C18:1, C18:2 and C22:1 are the main compounds found in the feedstock. The pyrolysis at 525 °C of waste fish oil as an animal source of triglycerides shows that it is possible to obtain biofuels like light bio-oil and heavy bio-oil with a good similarity to petroleum-based fuels. The PIONA analysis identified olefin and aromatic compounds as the main components of light bio-oil. The MS technique was very important to confirm the chemical structure of the PIONA compounds classified. Heavy bio-oil contains basically 1-olefins and carboxylic acid residues of the pyrolysis process. The H1 NMR showed a high content of aliphatic hydrocarbons in all biofuels investigated in this study. The 13C NMR analysis confirmed the presence of carbon in the biofuel composition as aliphatics, olefins and aromatics. Acknowledgments The authors are very grateful to Analytical Instrumentation Laboratory of NIQFAR – UNIVALI for the NMR analysis, National Agency of Fuel, Natural Gas and Biofuels – ANP and Financing of Studies and Projects – FINEP, institutions of the Brazil government for financial support. References [1] FAO – Food and Agriculture Organization. The state of world fisheries and aquaculture. Rome; 2006. [2] FAO – Food and Agriculture Organization. Food outlook, global market analysis. Rome; 2008. [3] Arruda LF, Borghesi R, Oetterer M. Use fish waste as silage – a review. Braz Arch Biol Technol 2007;50(5):879–86. [4] Dandik L, Aksoy HA. Effect of catalyst on the pyrolysis of used oil carried out in a fractionating pyrolysis reactor. Renew Energy 1999;16:1007–10. [5] Lima DG, Soares VCD, Ribeiro EB, Carvalho DA, Cardoso ECV, Rassi FC, et al. Diesel-like fuel obtained by pyrolysis of vegetable oils. J Anal Appl Pyrol 2004;71:987–97. [6] Fortes ICP, Baugh PJ. Pyrolysis – GC/MS studies of vegetable oils from macauba fruit. J Anal Appl Pyrol 2004;72(1):103–11. [7] Adebanjo AO, Dalai AK, Bakhshi NN. Production of diesel-like fuel and other value-added chemicals from pyrolysis of animal fat. Energy Fuels 2005;19:1735–41. [8] Sadrameli SM, Green AES. Systematics of renewable olefins from thermal cracking of canola oil. J Anal Appl Pyrol 2007;78:445–51. [9] Tamunaidu P, Bhatia S. Catalytic cracking of palm oil for the production of biofuels: optimization studies. Bioresource Technol 2007;18:3593–601. [10] Hua T, Chunyi L, Chaohe Y, Honghong S. Alternative processing technology for converting vegetable oils and animal fats to clean fuels and light olefins. Chinese J Chem Eng 2008;16(3):394–400. [11] Alencar JW, Alves PB, Craveiro AA. Pyrolysis of tropical vegetable oils. J Agric Food Chem 1983;31:1268–70. [12] Maher KD, Bressler DC. Pyrolysis of triglyceride materials for the production of renewable fuels and chemicals. Bioresource Technol 2007;98:2351–68. [13] Zhang Q, Chang J, Wang T, Xu Y. Review of biomass pyrolysis oil properties and upgrading research. Energy Convers Manage 2007;48:87–92. [14] Garcia-Perez M, Chaala A, Pakdel H, Kretschmer D, Roy C. Characterization of bio-oils in chemical families. Biomass Bioenergy 2007;31(4):222–42. [15] Pousa GPAG, Santos ALF, Suarez PAZ. History and policy of biodiesel in Brazil. Energy Policy 2007;35:5393–8. [16] Qiang L, Xu-lai Y, Xi-Feng Z. Analysis on chemical and physical properties of bio-oil pyrolyzed from rice husk. J Anal Appl Pyrol 2008;82:191–8. [17] Oasmaa A, Meier D. Norms and standards for fast pyrolysis liquids. J Anal Appl Pyrol 2005;73:323–34. [18] Bridgwater AV. Biomass fast pyrolysis. Therm Sci 2004;8(2):21–49. [19] Idem RO, Katikaneni SPR, Bakhshi NN. Thermal cracking of canola oil: reaction products in the presence and absence of steam. Energy Fuels 1996;10(6):1150–62. [20] Schwab AW, Dykstra GJ, Sekke E, Sorenson SC, Pryde EH. Diesel fuel from thermal-decomposition of soybean oil. JAOSC 1988;65(11):1781–6. [21] Santos ALF, Quirino RL, Ribeiro RAM, Martins DU, Iha OK, Rubim JC, et al. Physico-chemistry properties of poultry fat bio-oil. <http:// www.biodiesel.gov.br/docs/congresso2007/producao/59.pdf> [accessed 15.04.08]. [22] Luo Z, Wang S, Liao Y, Zhou J, Gu Y, Cen K. Research on biomass fast pyrolysis for liquid fuel. Biomass Bioenergy 2004;26:455–62. [23] Wang C, Du Z, Pan J, Li J, Yang Z. Direct conversion of biomass to bio-petroleum at low temperature. J Anal Appl Pyrol 2007;78:438–44. [24] Lede J, Broust F, Ndiaye F, Ferrer M. Properties of bio-oils produced by biomass fast pyrolysis in a cyclone reactor. Fuel 2007;86(12–13):1800–10. [25] Martins AF, Diniz J, Stahl JA, Cardoso AL. Characterization of liquid products and char from the pyrolysis of eucalyptus sawdust. Quím Nova 2007;30(4):873–8. [26] Blomberg J, Schoenmakers PJ, Brinkman UATh. Gas chromatographic methods for oil analysis. J Chromatogr A 2002;972:137–73. [27] ASTM D 6729-01. Standard test method for determination of individual components in spark ignition engine fuels by 100 meter capillary high resolution gas chromatography. Annual book of ASTM standards, vol. 05.01; 2002. [28] Zhu X, Zhang L, Che X, Wang L. The classification of hydrocarbons with factor analysis and the PONA analysis of gasoline. Chemometr Intell Lab 1999;45:147–55. [29] Durand JP, Boscher Y, Petroff N, Berthelin M. Automatic gas chromatographic determination of gasoline components. J Chromatogr 1987;395:229–40. [30] Ré-Poppi N, Almeida FFP, Cardoso CAL, Raposo Jr JL, Viana LH, Silva TQ, et al. Screening analysis of type C Brazilian gasoline by gas chromatography – Flame ionization detector. Fuel 2009;88:418–23. [31] Burri J, Crockett R, Hany R, Rentsch D. Gasoline composition determined by 1H NMR spectroscopy. Fuel 2004;83:187–93. [32] Kapur GS, Singh AP, Sarpal AS. Determination of aromatics and naphthenes in straight run gasoline by 1H NMR spectroscopy. Part I. Fuel 2000;79:1023–9. [33] Sarpal AS, Kapur GS, Mukherjee S, Tiwari AK. PONA analyses of cracked gasoline by 1H NMR spectroscopy. Part II. Fuel 2001;80:521–8. [34] Wiggers VR, Wisniewski Jr A, Madureira LAS, Barros AAC, Meier HF. Biofuels from waste fish oil pyrolysis: continuous production in a pilot plant. Fuel 2009;88:2135–41. [35] Saify ZS, Akhtar S, Khan KM, Perveen S, Ayattollahi SAM, Hassan S, et al. A Study on the fatty acid composition of fish liver oil from two marine fish, Eusphyra blochii and Carcharhinus bleekeri. Turk J Chem 2003;27:251–8. [36] Adams RP. Identification of essential oil components by gas chromatography, mass spectroscopy. 4th ed. vol. viii. Carol Stream: Allured; 2007. 804 p.