Production of Fuel-Like Fractions by Fractional Distillation of Bio-Oil from Açaí (Euterpe oleracea Mart.) Seeds Pyrolysis

This work investigates the effect of production scales (laboratory, bench, and pilot) by pyrolysis of Açaí (Euterpe oleracea Mart.) seeds at 450 °C and 1.0 atmosphere, on the yields of reaction products and acid value of bio-oils. The experiments were carried out in batch mode using a laboratory scale reactor of 143 mL, a bench scale reactor of 1.5 L, and a pilot scale reactor of 143 L (≈1:10:1000). The bio-oil was obtained in pilot scale, fractionated by distillation to produce biofuel-like fractions. The distillation of bio-oil was carried out in a laboratory column. The physical-chemistry properties (density, kinematic viscosity, acid value, and refractive index) of bio-oils and distillation fractions were determined. The qualitative analysis was determined by FT-IR and the chemical composition by GC-MS. The pyrolysis showed bio-oil yields from 4.37 to 13.09 (wt.%), decreasing with reactor volume. The acid value of bio-oils varied from 68.31 to 70.26 mg KOH/g. The distillation of bio-oil produced gasoline, light kerosene, and kerosene-like fuel fractions, and the yields were 16.16, 19.56, and 41.89 (wt.%), respectively. The physical-chemistry properties of distillation fractions increase with temperature. The FT-IR analysis of bio-oils and distillation fractions identified the presence of functional groups characteristic of hydrocarbons (alkenes, alkanes, aromatics, and aromatics rings) and oxygenates (carboxylic acids, ketones, esters, ethers, alcohols, phenols). The GC-MS identified 48.24 (area.%) hydrocarbons and 51.76 (area.%) oxygenates in the bio-oil produced in bench scale and 21.52 (area.%) hydrocarbons and 78.48 (area.%) oxygenates in the bio-oil produced in pilot scale. The gasoline-like fraction was composed by 64.0 (area.%) hydrocarbons and 36.0 (area.%) oxygenates, light kerosene-like fraction by 66.67 (area.%) hydrocarbons and 33.33 (area.%) oxygenates, and kerosene-like fraction by 19.87 (area.%) hydrocarbons and 81.13 (area.%) oxygenates.


Introduction
Açaí (Euterpe oleracea Mart.) is a native palm of natural occurrence in tropical Central and South America [1]. The palm gives a dark-purple, berry-like fruit, clustered into bunches [2]. The fresh fruits are traditionally processed by crushing and/or extracting the pulp and skin with warm water to produce a thick, purple-colored beverage/juice or a paste [3,4]. The fruit is a staple food in rural and urban areas of the Amazon River estuary, particularly in the State Pará (Pará-Brazil), with great economic importance for both rural areas and at regional levels [5]. It has become one of the most important export products of the Amazon River estuary to other parts of Brazil [5], as well as overseas [6].
Of the total 1.228.811 tons/year of fruits produced by the State Pará, between 85% [7] and 83% (wt.) [8] is a residue (Açaí seeds), thus producing between 1.019.913 and 1.044.489 tons/year of a residue. The mechanical processing of Açaí fruits in nature produces around 175.7 tons residue/day in off-season crop and 448.0 tons residue/day in in-season crop in the metropolitan region of Belém (Pará-Brazil), posing a complex environmental problem of solid waste management [9,10]. The Açaí fruit is a small, dark-purple, berry-like fruit, almost spherical, weighing between 2.6 to 3.0 g [11], with a diameter around 10.0 and 20.0 mm [11], containing a large core seed that occupies almost 85% (vol.%) of its volume [3]. Açaí fruit has an oily-fiber seed, rich in lignin-cellulose material [12][13][14][15].
Adjaye et al. [21], investigated the distillation of aspen poplar wood high-pressure liquefaction bio-oil. High-pressure liquefaction was carried out at 5.0 MPa and 360 • C, using 100 g of feed material (∅ Particle~1 .0 mm), 10 g of Na 2 CO 3 , 500 g distilled H 2 O, under CO atmosphere, and 2 h reaction time. The reaction liquid products consist of an organic (bio-oil) and an aqueous phase, showing a bio-oil yield of 30 (wt.%), containing 1.5 (wt.%) H 2 O. The bio-oil distilled within the boiling temperature ranges 85 • C < T Boiling < 250 • C, using 4.0 g of bio-oil. The distillations carried out at 85,115,140,165,175,190,200, 220, and 250 • C. The yields of distillation fractions varied between 21.0 and 62.3 (wt.%), while those of bottom products were between 37.7 and 79 (wt.%). The content of hydrocarbons in the distillation fractions varied within the range 38.5 and 47.4 (area.%), increasing between 85 and 175 • C, reaching a maximum of 47.4 (area.%) at 175 • C, decreasing between 175 and 250 • C. The concentration of oxygenates lies in the range 43 and 56.6 (area.%), showing a tendency to increase between 85 and 250 • C, presenting a maximum of 56.6 (area.%) at 250 • C. In addition, the concentration of phenols in the distillation fractions increases with temperature. The GC-MS analysis of bio-oil identified 85 compounds including carboxylic acids (formic, acetic, and propionic acids), cyclic alcohols, aliphatic alcohols, aldehydes, ketones, aromatic hydrocarbons, aliphatic hydrocarbons, polycyclic hydrocarbons, unsaturated hydrocarbons, substituted furans, substituted phenols, and methoxy phenols (phenol, guaiacol, p-cresol, p-guaiacol, o-guaiacol, iso-eugenol, and catechol).
Carazza et al. [22] investigated the distillation of Eucalyptus tar, recovered from the carbonization retort process. The physical-chemical properties of Eucalyptus tar show a density of 1.180 g/cm 3 and viscosity of 87.17 mm 2 /s, with an acidity of 8.70 (% HAc). The distillation consists of a 3000 L boiler, coupled to a fractionation column of 04 stages, a reflux system, 2 condensers, a homogenization system (centrifuge pump), sample units, a collecting unit, and a vacuum system. The Eucalyptus tar (bio-oil) distilled within the boiling temperature ranges from 110 • C < T Boiling < 300 • C, using a 2000 L feed at 70 mmHg, without reflux. The reaction products consist of an aqueous phase (T Boiling < 110 • C), an organic phase (bio-oil) (110 • C < T Boiling < 300 • C), and a pitch, showing an average bio-oil yield of 26.9 (wt.%), an aqueous phase yield of 17.3 (wt.%), a pitch yield of 50.9 (wt.%), and 4.9 (wt.%) losses. The GC-MS identified 0.63 (area.%) acetic acid, 4.03 (area.%) ketones, and 85.48 (area.%) phenol derivatives, as well as 9.86 (area.%) non-identified chemical compounds. In a secondary fractionation step, the bio-oil re-distilled at 340 • C and 10 mmHg, with reflux ration of 3:1 and 7:1, being the distillates fractionated into 95 samples. The GC-MS identified in all the samples only phenol derivatives (the distribution of phenol derivatives determined for all the samples and/or fractions was phenol, ethyl phenol, cresol, o-cresol, p-cresol, m-cresol, guaiacol, ethyl guaiacol, phenyl guaiacol, syringol, ethyl syringol, methyl syringol, and propyl syringol).
Boucher et al. [25], studied the distillation of softwood bark bio-oil produced by vacuum pyrolysis. The pyrolysis was carried out under vacuum in a pilot-scale reactor at 500 • C and 14 kPa, using 92 kg softwood bark residues with a particle diameter ∅ Particle < 14.0 mm, in batch mode. The reaction products yielded 20 (wt.%) bio-oil, 24 (wt.%) aqueous phase, 35 (wt.%) coke (solid phase), and 20 (wt.%) gaseous phase. The aqueous phase composed by 84 (wt.%) H 2 O and 16 (wt.%) soluble/dissolved organic compounds. The density and kinematic viscosity of bio-oil were 1.066 g/cm 3 (20 • C) and 38.0 mm 2 /s (40 • C), respectively, while the flash point was >90 • C. The distillation was carried out in a 250 mL glass flask, coupled to a fractionation column, a condenser, and an electric heater, using 150 g of bio-oil at 140 • C, 1.0 atmosphere. The yield of distillation (aqueous phase + organic phase) was 17.0 (wt.%), while that of organic fraction, with initial boiling point (T IBP = 140 • C), was 11.7 (wt.%). At boiling temperature T Boiling < 100 • C, the distillation curve exhibits the evaporation of H 2 O and low boiling point compounds, reaching a distillation yield of approximately 10.0 (wt.%). After removal of H 2 O, the slope of the distillation curve increases as molecules of high molecular weight evaporate.
Zheng and Wei [45] studied the distillation, under vacuum, of fast pyrolysis rice husk bio-oil at 80 • C, 15 mmHg, using a distillation round bottom glass apparatus of 1000 cm 3 . The pyrolysis products include an aqueous phase (H 2 O, volatile organic acids, and oxygenate compounds), a bio-oil, and a residue. The yields of bio-oil, aqueous phase, and residue were 61 (wt.%), 29 (wt.%), and 10 (wt.%), respectively. The density of biooil was 1.270 g/cm 3 . The content of linear and aromatic carboxylic acids in fast bio-oil (HCOOH, CH 3 COOH, C 6 H 4 (COOH) 2 ) decrease from 14.35 (wt.%) to 1.01 (wt.%) after distillation. The pH increased from 2.8 to 6.8, proving that distillation was effective to de-acidify the bio-oil. In addition, distillation causes a deoxygenation of bio-oil, as the concentration of oxygenates decreased from 50.3 to 9.2 (wt.%). In order to investigate the stability of bio-oil and distilled bio-oil, experiments were carried out to analyze variations on the kinematic viscosity (20 • C) within a period of 30 days. The results show that viscosity (20 • C) of bio-oil is almost constant (~210 mm 2 s −1 ) after distillation, and that of fast pyrolysis bio-oil increases from 130 to 240 mm 2 s −1 , showing that distillation produces a chemical stable distilled bio-oil.
Majhi et al. [50] investigated the distillation of jatropha curcas cake pyrolysis bio-oil. The pyrolysis was carried out in a fixed bed stainless steel tubular reactor (φ id = 40 mm, H = 240 mm, V Reactor = 301.6 mL) in laboratory scale at 550 • C, 5 • C/min heating rate, N 2 flow rate of 50 cm 3 /min, using 250 g biomass with a particle diameter ∅ Particle 0.5 and 0.8 mm. The distillation carried out in a 500 mL glass flask, coupled to a Hempel fractionating packed column and a condenser, using 300 mL of bio-oil. The temperature of gaseous phase was measured at the column outlet using a glass thermometer. The bio-oil was distilled within the boiling temperature ranges T Boiling < 140 • C, 140 • C < T Boiling < 250 • C, and T Boiling > 250 • C. The bio-oil density, kinematic viscosity, and flash point were 1.100 g/cm 3 , 3.96 mm 2 /s, and 180 • C, respectively, containing 15.4 (wt.%) H 2 O. The bio-oil distilled fraction within the boiling temperature range T Boiling < 140 • C had a density of 0.8735 g/cm 3 , kinematic viscosity of 2.35 mm 2 /s, and flash point of 41 • C, containing less than 0.05 (wt.%) H 2 O, and acidity less than 0.05 mg KOH/g. FT-IR analysis identified the presence of alkenes and mono and polycyclic substituted aromatic groups, confirming that T Boiling < 140 • C distillation fraction contains no polar compounds. In addition, GC analysis identified the presence of non-polar compounds including hexane, methyl cyclopentane, 3,3-dimethyl cyclopentane, 3-methylhexane, methyl cyclohexane, and toluene.
Zhang et al. [51] studied the distillation of fast co-pyrolysis of rice Rusk and ADR (Atmospheric Distillation Residue). The fast pyrolysis was carried out in a downdraft fixedbed reactor, under N 2 atmosphere at flow rate of 0.16 L/min, reaction time between 1-2 s, at 450 • C, using 5.0 g of dried rice Rusk with a particle diameter ∅ Particle < 88.0 µm. The copyrolysis was carried out by mixing dried rice Rusk and ADR powder (∅ Particle < 300.0 mm), and the ADR powder content was set at 15, 20, 25, 30 and 35 (wt.%) of feed. The distillation was carried out in a round-bottom flask placed, using an oil bath heating system, under vigorous magnetic stirring, at 240 • C and 1.0 atmosphere. The distillation products include the distilled bio-oil, a residual solid phase (ADR), non-condensable gases, and residual bio-oil. , the same behavior observed for the content of furfural and phenols derivatives (phenol, o-cresol, p-cresol, and guaiacol). In addition, the H 2 O content present in the distilled fraction FVI was 37.13 (wt.%), higher than that of raw bio-oil, equal to 30.3 (wt.%), showing that dehydration reactions occurred. For the distillation of fast co-pyrolysis bio-oils obtained by mixing rice Rusk and ADR, one can observe that yields of bio-oil remain almost constant (~46-47) (wt.%), while those of char increase from 37 to 41 (wt.%) and those of non-condensable gases decrease from 18 to 14 (wt.%) with increasing ADR content in feedstock. Finally, experimental results based on the elemental analysis and chemical identification by GC-MS show that distillation of bio-oils can be considered a reactive distillation process, as the solid phase (ADR) contains substance of high molecular weight and carbon chain length, not detected and/or identified in the raw bio-oil by GC-MS.
Capunitan and Capareda [52] investigated the distillation of corn Stover bio-oil under atmospheric and vacuum conditions (0.5 bar). The bio-oil produced by pyrolysis of corn Stover using a high-pressure reactor at 400 • C, in batch mode, contains 20.3 (wt.%) H 2 O. The aqueous phase acidity was 24.6 mg KOH/g oil. The distillation set-up consisted of distillation flasks, a fractionating column, a condenser, and collecting flasks, using 10 g bio-oil. For the distillation in atmospheric conditions, bio-oil was distilled within the boiling temperature ranges T Boiling < 100 • C, 100 • C < T Boiling < 180 • C, and 180 • C < T Boiling < 250 • C, while that under vacuum was within the boiling temperature ranges T Boiling < 80 • C, 80 • C < T Boiling < 160 • C, and 160 • C < T Boiling < 230 Elkasabi et al. [57] studied the fractional distillation of tail-gas reactive pyrolysis biooil of horse manure (TGPHM), switch grass (TGPSG), and eucalyptus (TGPE), and the fast pyrolysis bio-oil of horse manure (BHM), switch grass (BSG), and eucalyptus (BE) to fractionate and enrich chemical compounds. TGPHM, TGPSG, and TGPE were produced by pyrolysis of tail-gas with recycle rates of 70, 70, and 50 (vol.%), respectively. A vacuumjacketed distillation apparatus of 100 mL (Vigreux), operating in batch mode, coupled to a Liebig condenser, was used to carry out the distillation experiments under vacuum and 1.0 atm. The distillation apparatus was connected to a 100 mL round-bottomed glass flask, inserted inside a heating mantle with digital temperature control. The vapor temperature was measured at the column top outer joint. The reduced pressure (vacuum) distillation was applied when the bottom flask temperature reached 350 • C. The distillation yields of tail-gas reactive pyrolysis bio-oils TGPHM, TGPSG, and TGPE were 56.3, 55.8, and 54.1 (wt.%), respectively. The distillation yields of fast pyrolysis bio-oils BHM, BSG, and BE were 24.2, 33.1, and 38.9 (wt.%), respectively. The acid values of all distillation fractions of TGPHM, TGPSG, and TGPE decrease with increasing vapor temperature. The acid values of lighter distillation fractions are very high, compared to the initial acid value of TGPSG and TGPE, showing that fractional distillation was not effective in decreasing the acidity of tail-gas reactive pyrolysis bio-oil with initial high acid values.
Huang et al. [70], investigated the fluidized bed pyrolysis with in-line distillation of rice Rusk under atmospheric conditions. The bio-oil was produced by pyrolysis of rice Rusk using a fluidized bed pyrolyzer (φ id = 50 mm, H = 500 mm, V Reactor = 981. The work investigates systematically the influence of production scales (laboratory, bench, and pilot) by pyrolysis of pretreated (drying, milling, and sieving) Açaí seeds at 450 • C and 1.0 atmosphere, on the yields of reaction products (bio-oil, aqueous phase, coke, and gas) and the acid value of bio-oils. The bio-oil obtained in the pilot scale was submitted/subjected to fractional distillation using a laboratory-scale column (Vigreux) of 30 cm to study the feasibility of producing fuel-like fractions (gasoline, light-kerosene, and kerosene), determine the chemical composition of bio-oils and distillation fractions, as well as to perform the physical-chemistry (density, kinematic viscosity, acid value, and refractive index) characterization of distillation fractions.

Methodology
The process flow sheet shown in Figure 1 summarizes the applied methodology, described in a logical sequence of ideas, chemical methods, and procedure to produce fuel-like fractions (gasoline, light kerosene, and kerosene) by fractional distillation of biooil produced by thermal degradation (pyrolysis) of Açaí seeds at 450 • C and 1.0 atm, in laboratory, bench, and pilot scales. Initially, the Açaí seeds are collected. Afterwards, they are subjected to pretreatments of drying to diminish moisture content, followed by milling and sieving. The pyrolysis carry out in different production scales (laboratory, bench, and pilot) to investigate the influence on reaction products (coke, bio-oil, H 2 O, and gas), particularly the bio-oil, as well as on the acid value of bio-oil. The bio-oil was produced in pilot scale, submitted to fractional distillation to produce gasoline, light kerosene, and kerosene-like fractions. The physical-chemistry properties and chemical composition of distillation fractions were determined.

Materials
By the processing of Açaí fruits in nature with warm water to produce a purplecolored juice or a paste [3,4], a byproduct, the Açaí seeds, is generated [3,7,8]. The seeds of Açaí were obtained in a small store of Açaí commercialization, located in the City of Belém-Pará-Brazil [73].

Pre-Treatment of Açaí Seeds in Nature
The seeds had a high moisture content, not only because the pulping process uses water to extract the Açaí juice, but also due to its incorrect disposal, exposed to the environment. The high moisture content of raw materials favors the generation of undesirable products. In this sense, it was necessary to dry the seeds. Next, 750 kg of Açaí seeds weighed and separated into five loads/charges of 150 kg. Afterwards, the charges were subjected to the drying process in a pilot thermal oven with air recirculation and analog temperature control (FABBE. Ltda, São Paulo-Brazil, Model: 170), at 110 • C for 24 h. The dried seeds were grounded using a pilot knife mill (TRAPP, Model: TRF 600). Bottom sieves with 0.8 mm and 5.0 mm opening diameters were fixed at the cutting mill exit. Two charges with 50.0 kg of dried seeds were sieved, the first using a bottom sieve of 0.80 mm, while the second used a bottom sieve of 5.0 mm. Afterwards, the dried and grounded seeds were sieved using a 0.6 mm sieve to remove the excess fibers from Açaí seeds. Figure 2 shows the material after the drying, milling and sieving process of Açaí seeds to carry out the pyrolysis experiments [Seeds + Fibers (a); Comminuted Seeds + Fibers (b); Fibers (c); and Comminuted Seeds (d)]. The pretreatments of drying, milling, and sieving increases the biomass contact-surface for carbonization and the uniformity (particle size distribution) of the raw material. The pyrolysis experiments were carried out with comminuted seeds, as shown in Figure 2d.

Pyrolysis Units
The experiments were carried out in laboratory, bench, and pilot scales at 450 • C and 1.0 atmosphere, and all the apparatuses in laboratory, bench, and pilot scales are described in detail elsewhere [77].
The pyrolysis unit in laboratory scale consists of a cylindrical borosilicate glass reactor of 200 cm 3 . The reactor was inserted in a cylindrical oven with a ceramic resistance of 800 W, with a digital temperature and heating rate control system (THERMA, São Paulo-Brazil, Model: TH90DP202-000), and the temperature was measured with the aid of a K-type thermocouple sensor (Ecil, São Paulo-Brazil, Model: QK. 2). A borosilicate glass Liebig condenser was connected to the reactor exit using a Y shaped connection and used to liquefy the gaseous phase, with cooling H 2 O supplied by a thermostatic recirculation bath (VWR) with digital temperature control. The liquid phase products were collected in a 50 mL borosilicate glass flask. The non-condensable gaseous products (CH 4 , CH 3 -CH 3 , CH 3 -CH 2 -CH 3 , O 2 , CO, CO 2 , H 2 , etc.) flow through an opening in the 90 • curve, coupled between the Liebig condenser and the collection flask, to the flare system. Figure 3 illustrates the pyrolysis unit in laboratory scale.
The bench pyrolysis unit, mounted on a movable metallic structure, consists of a cylindrical Reactor (R-1) constructed by AISI 304 Stainless steel of 2.0 L. The experimental apparatus has a control unit with a PLC (Programmable Logic Controller), making it possible to control the reactor temperature, heating rate, and the mechanical stirrer angular velocity. The reactor operating temperature is programmed using a temperature controller (NOVUS, Model: N1100), coupled to a K-type thermocouple inside the reactor (R-1). The reactor connected to a stainless steel double pipe condenser (DN 1 2 ") with a heat exchange area of 0.05 m 2 , coupled to a thermostatic recirculation bath with digital temperature control, using H 2 O as cooling fluid. The liquefied products were withdrawn using a 2.0 L stainless steel collection vessel (C-01). The non-condensable gaseous products flow through an escape valve located between the condenser and the collection vessel, and burned in the flare system. Figure 4 illustrates the pyrolysis unit in bench scale.  The pilot unit is described/presented in detail elsewhere [77]. The unit is composed of seven operational sections (feed/pumping section, pre-heating section, reacting section, cooling section, condensation section, separation & collection, and burning of noncondensable gases, instrumentation and control section). The reacting section consists of the jacketed stainless steel reactor with mechanical stirring system (R-01) of 145 L, operating pressure 1.0 atmosphere, and operating temperature 500 • C. LPG supplies the thermal energy to the reactor (R-01), with the aid of a digital controlled burning system (HOLLAMAQ-MB-20/AB-R, SERIES 05-10), inlet pressure of 0.0-45.0 mbar and energy capacity of 5.000-30,000 Kcal/h, with the mixture LPG-air burning within the annular space concentric to R-01. Figure 5

Experimental Procedures
The pyrolysis of dried Açaí seeds carried out in laboratory, bench, and pilot scale. Pyrolysis in laboratory scale performed using approximately 50.0 g seeds, weighed on a semi-analytical balance (QUIMIS, São Paulo-Brazil, Model: Q-500L210C), placed inside the reactor. Afterwards, the reactor inserted in the jacketed cylindrical furnace. The control system programmed the reaction time, the heating rate, and the process temperature. The heating rate was set equal to 10 • C/min. The reactor was maintained for 10 min at 450 • C after reaching the set point temperature. The liquid product subjected to filtration to separate the aqueous/bio-oil phases and compute the yields of bio-oil and aqueous phases. The coke weighed and the mass of gaseous phase computed by difference. This separation carried out with aid of a glass funnel and a porous filter medium (Qualitative Filter Paper 24.0 Ø; J. PROLAB). Pyrolysis in bench scale performed using approximately 900.0 g seeds, weighed on a semi-analytical balance (QUIMIS, São Paulo-Brazil, Model: Q-500L210C), placed inside the reactor. The control unit programmed the reaction time, the heating rate, and the process temperature. The heating rate was set equal to 10 • C/min. This separation carried out with aid of a glass funnel and a porous filter medium (Qualitative Filter Paper 24.0 Ø; J. PROLAB). The liquid product subjected to filtration to separate the aqueous/bio-oil phases and compute the yields of bio-oil and aqueous phases. The solid phase weighed and the mass of gaseous phase computed by difference. Pyrolysis in pilot scale is performed using approximately 30.0 kg seeds, weighed with the aid of a digital scale (DIGI-TRON, Curitiba-Brazil, Class III). Then, the reactor fed manually by the upper inlet. Afterwards, LPG gas cylinder weighed in order to compute the fuel consumption. The control unit programmed the process temperature. The rector temperature monitored every 10 min to register the formation of liquid and gaseous products. As the temperature rises, the gaseous reaction products condensed inside the collection vessel (VC-02). The non-condensable gases burned in the flare system. The reaction time computed from the moment the reactor reached the operating temperature (set point). The liquid (aqueous fraction + bio-oil) and the solid products weighed to compute the process yields. The liquid (aqueous fraction + bio-oil subjected to filtration using a glass funnel and a porous filter medium (Qualitative Filter Paper 24.0 Ø; J. PROLAB).

Distillation Unit
The fractional distillation of bio-oil was performed using an experimental apparatus similar to those described in the literature [73,[77][78][79]. The distillation apparatus, illustrated in Figure 6, is described in detail by Ferreira et al. [79]. The distillation fractions were subjected to the pretreatment of decantation to separate the aqueous and organic phases, and the organic phase submitted to filtration to remove small solid particles.

Chemical Composition of Bio-Oils and Distillation Fractions
The composition bio-oils and distillation fractions determined by CG-MS and the equipment and operational procedures described in details elsewhere [73]. The identification of compounds in bio-oils and distillation fractions performed by CG-MS, and the equipment details described in the literature [73]. The GC-MS operating conditions illustrated in Table 1. The intensity, retention time, and compound identification were analyzed according to the NIST mass spectra library. The concentration of all oxygenates and hydrocarbons in each sample was expressed in each area, as no internal standard was injected to compare the peak areas.

Pre-Treatments and Centesimal Characterization of Açaí Seeds
The yields of drying, comminution and sieving pre-treatments were 58 Table 2 summarizes the material balance, process conditions, and yields of reaction products by pyrolysis of Açaí seeds at 450 • C and 1.0 atmosphere, in laboratory, bench, and pilot scales. Figure 7 shows the yields of reaction products by pyrolysis of Açaí seeds carried out in different production scales (laboratory, bench, and pilot), using reactors of 140, 1500, and 143,000 mL, that is, volume scales of approximately ≈ 1:10:1000 [77]. One observes that the yield of bio-oil lies between 4.37 and 13.09 (wt.%), decreasing with increasing reactor volume, while the yields of coke and gas increase with increasing reactor volume. Since the pyrolysis experiments in different production scales used no mechanical stirring system, the higher the reactor volume, the worse the energy transport by conduction in the solid phase (seeds) and convection in the fluid phase (air), and hence the lower the carbonization grad of biomass. The lower the carbonization grade of biomass, the lower the bio-oil yield. This explains a decrease of bio-oil yield with increasing reactor volume. The yield of gas ranged from 24.14 to 27.0 (wt.%), showing no differences between laboratory and bench scales and an increment of~10% compared to the pilot scale.

Physico-Chemical Characterization of Distillation Fractions
The physical-chemical properties of distillation fractions (gasoline, 80-175 • C; light kerosene, 175-200 • C; and kerosene-like fraction, 200-215 • C) of bio-oil are illustrated in Table 6. Table 6. Physico-chemical properties of distillation fractions of bio-oil produced by pyrolysis of Açaí seeds at 450 • C and 1.0 atmosphere, in pilot scale.

Physico-Chemical
Properties It can be observed that acidity of distillation fractions increases with boiling temperature. However, the acidity of gasoline-like fraction is much lower than that of raw bio-oil (70.26 mg KOH/g), as described in Table 4. The high acid value of bio-oil is due to the presence of 78.48 (area.%) oxygenates [35]. The same behavior was observed for the densities, kinematic viscosities, and refractive indexes of gasoline, light kerosene, and kerosene-like like fractions with increasing boiling temperature. This is probably due to the high concentration of higher-boiling-point compounds in the distillate fractions, such as phenols, cresols (p-cresol, o-cresol), and furans [35], as the concentration of those compounds in the distillation fractions increases with increasing boiling temperature as reported elsewhere [35,66,70,72].
The results reported by Elkasabi et al. [53], show that fractional distillation was not effective to diminish the acid values of TGRP bio-oil with initial high acid values, what does not agree with the results reported by Capunitan and Capareda [48], as well as those presented in Table 6, showing that the acid values of distillation fractions are lower than that of raw bio-oil, proving that distillation was effective.

Chemical Compositional of Bio-Oil and Distillation Fractions by GC-MS Chemical Compositional of Bio-Oils by GC-MS
The chromatograms of bio-oils obtained by pyrolysis of Açaí seeds pyrolysis at 450 • C and 1.0 atmosphere, in bench and pilot scales is shown in Figure 12 Tables 8 and 9. The high acidity of both bio-oils, described in Table 4, is probably due to the presence of carboxylic acids, ketones, aldehydes, phenols, and cresols confer the high acidity of bio-oil.
The composition of bio-oil shows similarity to those reported in the literature [27,34,41,47,48,53,61], showing the presence of hydrocarbons, phenols, cresols, furans, carboxylic acids, and esters, among other classes of compounds [73]. The identification of hydrocarbons with carbon chain length between C 11 and C 15 , shows the presence of heavy gasoline compounds with C 11 (C 5 -C 11 ), light kerosene-like fractions (C 11 -C 12 ), and light diesel-like fractions (C 13 -C 15 ), according to Tables 8 and 9.

Chemical Compositional of Distillation Fractions by GC-MS
The chromatograms of bio-oil obtained in pilot scales and distillation fractions is shown in Figure 13. One observes that the spectrum of peaks is moving to the right, showing that distillation was effective to fractionate the bio-oil.       Table 12. The presence of ketones, furans, ethers, esters, aldehydes, and phenols confer the high acidity of kerosene-like fraction, as summarized in Table 12. Finally, the content of hydrocarbons within gasoline (40 • C < T Boiling < 175 • C), light kerosene (175 • C < T Boiling < 200 • C), and kerosene-like fraction (200 • C < T Boiling < 215 • C) are higher than those reported in the literature [17][18][19][46][47][48]66,70,72], proving that distillation was not only effective to diminish the acidity, but also to concentrate hydrocarbons. The classes of compounds, summation of peak areas, CAS number, and retention times of chemical compounds identified by CG-MS in bio-oil obtained by pyrolysis of Açaí seeds at 450 • C and 1.0 atmosphere, in pilot scale, as well as in distillation fractions as summarized in details in Supplementary Tables S1-S4.

Conclusions
The yields of bio-oil decrease with increasing reactor volume, while that of coke and gas increase. The yield of distillation fractions (gasoline, light kerosene, and kerosene-like like fractions), 77.61 (wt.%), is higher but according than those reported in the literature for both atmospheric and vacuum conditions [17][18][19]21,41,[46][47][48]53,66,70,72]. The acid values of distillation fractions increase with increasing boiling temperature. However, the acidity of gasoline-like fraction is much lower than that of raw bio-oil (70.26 mg KOH/g). The same behavior was observed for the densities, kinematic viscosities, and refractive indexes of gasoline, light kerosene, and kerosene-like like fractions with increasing boiling temperature.
The FT-IR analysis of bio-oil and distillation fraction identify the presence of hydrocarbons (alkanes, alkenes, and aromatic hydrocarbons) and oxygenates (phenols, cresols, carboxylic acids, alcohols, ethers, ketones, and furans). The bio-oil is composed of 21.52 (area.%) hydrocarbons and 78.48 (area.%) oxygenates. The presence of carboxylic acids, as well as phenols and cresols, is associated to the high acidity of bio-oil.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/en14133713/s1, Table S1: Classes of compounds, summation of peak areas, CAS number, and retention times of chemical compounds identified by CG-MS in bio-oil obtained by pyrolysis of Açaí seeds at 450 • C and 1.0 atmosphere, in pilot scale., Table S2: Classes of compounds, summation of peak areas, CAS number, and retention times of chemical compounds identified by CG-MS in gasoline-like fraction (40-175 • C) after distillation of bio-oil obtained by pyrolysis of Açaí seeds at 450 • C and 1.0 atmosphere, in pilot scale., Table S3: Classes of compounds, summation of peak areas, CAS number, and retention times of chemical compounds identified by CG-MS in light kerosenelike fraction (175-200 • C) after distillation of bio-oil obtained by pyrolysis of Açaí seeds at 450 • C and 1.0 atmosphere, in pilot scale., Table S4: Classes of compounds, summation of peak areas, CAS number, and retention times of chemical compounds identified by CG-MS in kerosene-like fraction (200-215 • C) after distillation of bio-oil obtained by pyrolysis of Açaí seeds at 450 • C and 1.0 atmosphere, in pilot scale.