Upgrade of bio‐oil produced from the sisal residue composting

The present work studies the composting effects on the chemical characteristics of bio‐oil produced by pyrolysis of sisal residue. Three systems were composted with sisal residue proportions to sisal fiber powder of 100:0, 90:10, and 75:25, respectively. The systems showed reductions of 33%–48% (extractive), 70%–80% (hemicellulose), and 80%–90% (cellulose) after composting. An increase in lignin content was observed in all systems. The pyrolysis of the composted systems was performed at 450°C and 550°C. At both temperatures, this process was selective in producing a large concentration of hydrocarbons (>160% increase), mainly alkanes and alkenes, reducing the concentrations of ketones, aldehydes, and phenolics (>50%) and eliminating esters, furans, and acetic acid to composted biomasses. The higher temperature favored aromatics and cyclic hydrocarbon production from the pyrolysis of composted samples. In addition to these results, composting helped reduce the oxygenated bio‐oil species by approximately 44%–75% at the lowest and ~69% at the highest temperatures. These results indicate that composted sisal residue can produce bio‐oils that are more suitable for biorefineries since they are rich in aliphatic hydrocarbons and non‐oxygenated species.


| INTRODUCTION
Agave sisalana is cultivated in all continents of the earth, whose main producing countries are Brazil, The United Republic of Tanzania, Kenya, China, Cuba, Haiti, Indonesia, Madagascar, Mozambique, Mexico, South Africa, and Thailand (FAOSTAT, 2023).The main product of this crop is sisal fiber, which has Brazil as the world's largest producer, with a production of 100,000 t/ year (FAOSTAT, 2023).However, intense competition from synthetic fibers, prolonged droughts, and poor management of natural resources have made this product less attractive to farmers and field workers (Santos & da Silva, 2017).Consequently, keeping the production of sisal fiber economically viable will contribute to keeping the man in the field and with environmental well-being since the crop absorbs more carbon dioxide than it is produced (Nobel & McDaniel, 1988).
Some initiatives to improve the sisal crop are already being taken, mainly by using residues produced in the defibration process, which reach more than 95% w/w of the leaf (da Silva et al., 2023).Most of the waste is discarded in the field as garbage, harming the environment due to the production of greenhouse gases (Broeren et al., 2017).However, several studies are being carried out to use sisal residues as biogas (Soares et al., 2023), pharmaceutical inputs (hecogenin, inulin, and others) (Apolinário et al., 2017), construction material (Kaushik et al., 2023), bio-oil, and various chemical products (Araujo et al., 2023) These alternative processes for the use of waste will play a fundamental role in the stability of the crop in the producing regions, and they will contribute to the maintenance of man in the countryside.
Among the alternatives presented for the use of sisal residue, the most comprehensive and that can contribute to the solution of global warming on the planet is the production of bio-oil since it can replace petroleum, generating essential biomolecules for industries (Jambeiro et al., 2018).Despite the importance of bio-oil for sisal cultivation, little research has been carried out so far.However, the quality of the bio-oil from sisal residue compared with other biomass makes it a promising raw material for biorefineries.
The sisal residue contains high levels of lignin and extractives, resulting in a bio-oil with an O/C ratio of 0.11 and H/C of 1.48 (Pereira et al., 2022).The value of the O/C ratio of bio-oil from sisal residue is still very high compared to petroleum, but it is one of the lowest compared to others (0.20-0.95).On the other hand, the high H/C ratio is mainly responsible for the high calorific value of the bio-oil from sisal residue (35,331 kJ/kg) and is very close to the value found for oil (46,000 kJ/kg).
Making the bio-oil from the sisal residue with characteristics closer to petroleum implies mainly the increase of the production of hydrocarbons, with consequent reduction of oxygenated compounds.In this case, some methods of bio-oil improvement have been widely studied by several researchers, involving pretreatment of biomass (Chai et al., 2022;Ong et al., 2021) and catalytic transformations of bio-oil (Yuan et al., 2023).The use of catalysts may be necessary to improve the characteristics of the biooil; however, the production of this material involves an expensive technology, which makes the pretreatment of the biomass attractive.
Composting is one of the most used biological processes as a biomass pretreatment, characterized as a natural biological process of aerobic decomposition, which converts organic substances from plant and animal waste into less complex compounds (Xiao et al., 2011;Zhang et al., 2011).Of the applications destined for composting, the production of biofertilizers is the most widely disseminated (Gajalakshmi & Abbasi, 2008;Onwosi et al., 2017).Biological pretreatment for the production of bio-oil has been little explored so far (Alvarez-Chavez et al., 2019;Kumar et al., 2020) despite the possible advantages caused by the breakdown of the lignin and hemicellulose structure, decreased cellulose crystallinity, and increased surface area due to increased porosity structure (Heise et al., 2017;Kumar et al., 2009;Peng et al., 2012).In addition to the intrinsic microorganisms of composting, specific fungi and bacteria can be used to degrade lignocellulose; however, the degradation of specific components from lignocellulose needs selective microorganisms (Alvarez-Chavez et al., 2019), which may be more expensive compared to composting endogenous microorganisms.In some works, the researchers found that composting was responsible for the increase in the lignin concentration, which was associated with the increase in the concentration of hydrogen and carbon monoxide in the pyrolysis (Barneto et al., 2009;Juchelková et al., 2015;Palma et al., 2020).Also, lignin-rich materials can produce less oxygenated bio-oil because of the lower oxygen content on lignin than hemicellulose and cellulose (Hernando et al., 2021).Furthermore, it was observed that composting modifies the biomass, causing an increase in the production of biochar and bio-oil in slow pyrolysis (Parthasarathy et al., 2015).
Influenced by the need to improve the bio-oil of sisal residue and by the results of biomass transformation through composting, a series of composting tests involving mixtures of sisal residue and sisal fiber powder were carried out in this work, with subsequent rapid pyrolysis of the resulting biomass.Analyzes of the materials during composting and their relationships with the species of the bio-oils produced were done for the first time to determine the extent of the upgrade achieved.This innovative work applies composting as a potential biological pretreatment to upgrade bio-oil from sisal residue by reducing the production of oxygenated compounds and increasing the production of hydrocarbons.The present research will benefit the selective production of chemical species aiming at improving the bio-oil from the residues of the sisal culture and subsequent use of this product in biorefineries.

| Biomass
Sisal residue (SR) and sisal fiber powder (SFP) were obtained from the semiarid region of Brazil.Sisal residue is a by-product of the process of getting the sisal fiber from the scraping of the Agave Sisalana leaf.After scraping, the SR was stored at 4°C and then dried at 105°C for 4 h (Jambeiro et al., 2018).After drying, the SR was shredded in a mill and sifted in a sieve system (6, 7, 8, 10, 16, 20, and 40 meshes).
The dried sisal fiber powder was obtained from sisal fiber processing warehouses.It consists of small particles of dried fibers and impurities from manipulating long fibers in the storage site.Sisal fiber powder was stored at room temperature, and the material was not treated.

| Composting
The composting processes were performed on a bench system in three thermoplastic cylindrical containers (ethylene poly terephthalate) with a capacity of 5 L, which were kept isolated thermally in a polystyrene box (Figure 1).Each container had two compartments: a superior, where the biomass was placed to be composted, forming a fixed bed, and a lower, used as a receiver of the liquid formed during the process.
The initial moisture content of the composted biomass was 65%, and no water was added to the material during the composting process.The biomass beds were aerated daily from the material's manual mixture during the 44 days of composting.Throughout the process, the bed temperature was measured periodically.During composting, solid material samples were collected weekly and stored below 0°C for later characterization.

| Physicochemical characterization of the samples
Six solid samples of each system initially cooled were dried until humidity was reached below 10%.The moisture content analysis was performed in an OHAUS MB25® analyzer for halogen lamp heating.All tests were made in duplicate.

| Ash
For ash content analysis, dry samples were heated at 40°C/min until they reached 800°C, keeping this temperature for 2 h.The ash percentage was determined by Equation (1).The samples' calcination were performed in a muffle oven EDG-3000 3p-S.%Ash = ash percentage, w s = sample weigh and w ash = ash weigh.
All analyses were carried out in duplicates.
The ashes were also characterized to identify metal content using X-ray fluorescence (FRX) on a Bruker S2 Ranger with anode radiation (maximum power of 50 W, maximum voltage of 50 kV, and maximum current of 2 mA) and Xflash® Silicon Drift detector.

| Elemental analysis CHNS
The methodology to identify the C, N, H, and S contents was based on ASTM D5373 e ASTM D4239 standards, and the elements were investigated by Leco Truspec CHN elemental analyzer.S was determined by Leco SC632 elemental analyzer.
Organic carbon (C org ) was determined based on ASTM D5373 and ISO 10694 standards.The organic carbon was analyzed in a Leco Truspec CHN.All analyses were performed in triplicates.

| Extractives
The extractive content analysis was based on Sluiter et al. (2008).This process was carried out in Soxlet extractors initially with ethanol extraction followed by water extraction.The percentage of extractives was made by mass difference, considering the moisture ratio of the respective initial samples (Equation 2).
(1) % Ash = w s − w ash The chemical characterization methodology of the samples used was established by Rocha et al. (2014).
Before determining carbohydrate and lignin contents, the samples were digested through a hydrolysis followed by a filtration process.The filtrate (diluted with ultrapure water) and the cake were used to determine lignin, sugars, organic acids, furfural, and 5-hydroxymethylfurfural (5-HMF).The samples were analyzed in duplicate.

| Lignin
The total lignin content was considered the sum of soluble and insoluble lignin fractions (Equation 3).
%LIG = percentage of total lignin, %LIG i = percentage of insoluble lignin and % LIG s = percentage of soluble lignin.The percentage of insoluble lignin was given by Equation (4).
w c = filter cake weight and % M c = percentage of the filter cake moisture.
To determine the soluble lignin content, the diluted filtrate added to sodium hydroxide (6.5 M) was analyzed in the UV-Vis Ocean Optics USB 2000+ spectrometer, with a 280 nm wavelength.From the absorbance value, the concentration of soluble lignin was calculated (Equations 5 and 6).
[LIG s ] = concentration of the soluble lignin in g/L; A s = absorbance measured at 280 nm; A dps = absorbance of the decomposition sugar products (furfural and 5-HMF) at 280 nm, given by Equation ( 6); DF = dilution factor of the filtrate to spectroscopic analysis; [FUR] = concentration of furfural in g/L analyzed through HPLC and [HMF] = concentration of 5-HMF in g/L analyzed through HPLC.
For the calculations, the following values of the respective absorptions were considered: ( ): fur = 146.85L∕ g. cm and hmf = 114.00L∕ g. cm.
From the value of the concentration, the value of the percentage of soluble lignin was obtained according to Equation ( 7).

| Sugars and organic acids
At this stage of the investigation, the concentrations of glucose, xylose, arabinoses, cellobioses, acetic acid, and formic acid of the samples obtained through composting were analyzed.Such components constitute the lignocellulosic structure of biomass.
The respective diluted filters, obtained from the hydrolysis step, were analyzed in HPLC Hitachi Chromaster® equipment with aminex HPX87H column, with a column temperature of 35°C and analysis time of 22 min.The mobile phase was sulfuric acid (5 mM), with a 0.5 mL/min flow.
2.3.7 | Furfural and 5-HMF For furfural and 5-HMF concentration analysis, the respective diluted filters, obtained from the hydrolysis step, were analyzed in HPLC Hitachi Chromaster® equipment with Merck C18 column, column temperature of 30°C, and analysis time of 18 min.Acetonitrile (1:9) was the mobile phase, with a 0.8 mL/min flow.

| Hemicellulose and cellulose content calculations
From the concentration data of sugars, organic acids, furfural, and 5-HMF, cellulose, and hemicellulose contents were calculated through Equations ( 8) and ( 9).
The identification of the peaks obtained from the chromatographic analysis occurred with a minimum of 80% similarity.

| Composting
The main characteristics of fresh waste can be seen in Table 1.The high ash content in the sisal fiber powder (35.6%) can be highlighted compared with the sisal residue ash content (11.1%), which may be related to impurities mixed with the fiber powder at the stored site.However, the content of sisal residue extractives (52.7%) is higher than that of sisal fiber powder (29.8%), given the high number of polysaccharides present in the sisal residue (Santos et al., 2015).
Regarding the transformations observed in the system, one of the variables evaluated was the mass loss percentage due to composting.Comparatively, the mass loss observed during composting was similar to the three systems: ~54.3% (SR100-T44) and ~55.4% (SR90-T44 and SR75-T44).This loss is related to the portion of drained liquid biomass (leachate) and the biochemical dynamics established in composting.Generally speaking, the carbonous components of the environment serve as a source of energy for CO 2 , microbial biomass, humic substances, and heat production (Tuomela et al., 2000).
Since self-heating during composting is a process feature, the systems' temperature profiles were investigated and presented in Figure 2. The temperature is considered an essential parameter for composting monitoring since the heat produced in this process is related to the degradation of organic matter (Ryckeboer et al., 2003).
Considering the adaptation limitations of most microorganisms to specific temperature conditions, it is correct to state that there is a relation between temperature and microbial dynamics of the system (Bohacz, 2018;Jurado et al., 2014;Lopez-Gonzalez et al., 2013).
The temperature variations observed in Figure 2 indicate higher microbial count in the early stages of the process and a similar process dynamic in the three systems.It is possible to perceive that, after the 35th day, the temperatures tend to match, reaching room temperature around the 44th day, which is the final stage of composting (Jurado et al., 2014).

| Solid characterization
Table 2 shows the samples' elementary analyses of the composted biomasses.It shows the total and organic carbon content increased by 14% and 23% to SR90, 14% and 34% to SR75, and 3% and 7% to SR100, respectively, from T0 to T44.These results may indicate a more significant increase in the heating value of these materials (Mishra & Mohanty, 2018).The three composting systems' hydrogen contents remained statistically unchanged.More pronounced increases were observed in nitrogen (53% for SR100-T44, 45% for SR90-T44, and 89% for SR75-T44) and sulfur (87% for SR100-T44, 67% for SR90-T44, and 86% for SR75-T44) by composting.The highest nitrogen and sulfur content occurred with SR75-T44 (2.23%) and SR100-T44 (0.28%), respectively, when the final samples were considered.High amounts of nitrogen and sulfur in biomass can cause undesirable environmental impacts through nitrogen and sulfur compound emissions (Protásio et al., 2013).However, some of these final levels were similar to other residual biomasses studied before.For example, the nitrogen content of rice bark (1.85%) reported by Biswas et al. (2017) is comparable to that of the final sample SR90-T44 (1.83%).The sulfur contents in the last samples of the systems had similar values with each other (0.25%-0.28%) and also identical to corn cob (0.24%) and sugarcane bagasse (0.23%), as reported by Danish et al. (2015).As well as the biomass studied by Mishra and Mohanty (2018), whose nitrogen and sulfur contents of their samples (below 6.99% and 0.7%, respectively) indicated minimal generations of NO X and SO X in pyrolysis, the same can be concluded from the values observed in the present work concerning all systems.
The C org /N ratio is an essential factor in evaluating the stabilization of the material in composting.This ratio decreases throughout composting, indicating material stabilization when it reaches values below 20 (Biruntha et al., 2020;Wei et al., 2019).In this work, the C org /N ratio (Table 2) decreased for all studied systems, but only the final sample SR100-T44 reached stability (C org /N = 19.3).Samples SR90-T44 and SR75-T44 presented C org /N > 20, indicating that the process did not reach stability and could show variations in its compositions if kept in composting longer.
We notice a decrease in oxygen content in all systems from T0 to T44: 3% (SR100), 15% (SR90), and 13% (SR75).This result indicates an advantage for the pyrolysis of these biomasses since this suggests fewer oxygenated compounds in bio-oil.This decrease in oxygen content associated with increased carbon content is also a positive factor in increasing the heating value of the material (Ullah et al., 2021).
The ash content before composting (Table 2) was found from the weighted average between the ashes of each biomass (Table 1) and the biomass fractions in each system.With this, the initial ash percentage increased in the following order: SR75 > SR90 > SR100.After composting, the ash content increased by 127% (SR100), 115% (SR90), and 95.9% (SR75), noting that the increase was lower the higher the percentage of fiber powder in each system.Although microorganisms consume some inorganic species that constitute the ashes, the global reduction of material masses by composting is primarily responsible for increased ash content in each system.In addition, the lower variation of ash content after composting indicated that the lower loss of mass caused by composting occurred in the system with the most extensive fiber powder content.
This increase in ash content could also be seen by Bikovens et al. (2012), which reported growth from 3% to 30% in the ash content during composting of lignocellulosic biomass.Although this increase is expected during composting (Juchelková et al., 2015), the most extensive ash content reduces the efficiency of pyrolysis systems as the ashes act as a "heat sink" during pyrolysis (Mishra & Mohanty, 2018).From this point of view, the system SR100 would be better than the systems SR90 and SR75 for pyrolysis.
The mineral composition of the ashes (Table 3) is known for its catalytic effect on pyrolysis (Mishra et al., 2021).In the present study, calcium was a more abundant metal in the ash of the composted and non-composted biomasses (43%-50%).CaO may be responsible for increasing dehydration during pyrolysis and, therefore, for decreased oxygen content in bio-oil by CaCO 3 (Carpenter et al., 2014;Krutof & Hawboldt, 2019).In this work, it is possible to notice a mild increase in the percentage of CaO in SR100 (~3%) and a slight decrease in SR90 (~3%) and SR75 (~5%).The ashes of the systems also had high magnesium (13%-20%), potassium (12%-16%), and phosphorus (8%-10%) content before and after composting.It was possible to notice magnesium content decrease in the three systems through composting, corresponding to 20% (SR100), 25% (SR90), and 20% (SR75).
In composting, part of the minerals in biomasses are consumed as nutrients necessary for microbial development (Bohacz, 2018;Harindintwali et al., 2020;Tuomela et al., 2000), and some components are leached for the liquid phase.

| Chemical characterization
Figure 3 shows the variation of the main components of biomass throughout the composting.The extractive content of the systems dropped throughout the process (Figure 3a).During the composting of lignocellulosic biomass, fungi, and bacteria promote the decomposition of extractives, considered the most available components for this degradation, especially in the initial stage of the process (Barneto et al., 2009;Bohacz, 2018).At the beginning of the composting, microorganisms preferably use soluble and easily degradable carbon sources, such as monosaccharides, starches, and lipids (Tuomela et al., 2000).Jurado et al. (2014) reported that the degradable organic matter of the immature substrate, that is, still in decomposition, is composed of total sugars, phenolic substances, amino acids, and peptides, among other biodegradable components.extractive content is directly related to the production of the liquid phase in pyrolysis and is inversely proportional to the material's thermal stability (Mishra & Mohanty, 2018).Therefore, it is possible to conclude that the degradation of extractives by composting contributes to the increase in the material's thermal stabilization; the high viscosity of sisal residue bio-oil can be attributed to extractives and, with the reduction in these species, pyrolysis of processed material can produce a less viscous fluid (Jambeiro et al., 2018).
The variation in the extractive content for the SR90 was higher than the SR100 and SR75 during the composting.After the composting, 44% (SR100-T44), 48% (SR90-T44), and 34% (SR75-T44) of extractives were degraded.It is also possible to perceive that the degradation of extractives occurred faster at the early stage of composting (first 15 days) for composting systems (SR90 and SR75).Such evidence may be associated with the temperature increase in the same period, indicating the intense oxidative action of microorganisms in these beds and, therefore, greater energy release by breaking chemical bonds than bed SR100 (Gajalakshmi & Abbasi, 2008).
In this work, the hemicellulose and extractives degradation were also more pronounced at the early stage of composting (Figure 3b).About 75% of hemicellulose was degraded in SR100-T44, 74% for SR90-T44, and 80% for SR75-T44.In the system SR100, hemicellulose tended to increase between 5 and 15 days, which can be related to the sharp drop in extractive content in the same period.This result indicates that, during this period, there was no degradation of hemicellulose due to the availability of simpler chemical structures in extractives.
For all systems, the growing tendency in lignin content has been observed throughout the process (Figure 3d).Compared with other components, low lignin degradation has already been reported in some works (Mei et al., 2019;Zhang et al., 2019).

| PY-GC-MS
Biomass pyrolysis is related to the depolymerization of basic biochemical structures, generating monomers that can later be transformed into species representative of all organic functions based on decarboxylation, cyclization, and rearrangement reactions (Kaur et al., 2022).Considering this dynamic, analytical pyrolysis has been essential in studying chemical compounds' production from biomass's thermal decomposition (Kumar et al., 2021).Some typical compounds of bio-oil are produced from the main components of biomass.The amorphous structure of hemicellulose is usually degraded into acids, furans, alcohols, and esters (Kumar et al., 2021;Zhang et al., 2023).Cellulose is responsible for the production of anydro-sugars, for example, levoglucosan, in first degradation that at higher temperatures can be degraded into furans, ketones, aldehydes, and some aliphatics (for example, 5-HMF, acetaldehyde, and acetic acid) (Volli et al., 2021).From the clivage of alkyl chain in lignin units comes phenolics, the main product of lignin degradation (San-Emeterio et al., 2021).The pyrolysis essays were performed at 450°C and 550°C.The gaseous products, analyzed by GC-MS, were grouped according to their organic function and biomass used as a function of the peak area percentage (Figure 4).
The pyrolysis of sisal fiber powder at 450°C produced about 35% of hydrocarbons and 25% of ketones.At 550°C, it produced about 34% of ketones and 22% of aldehydes.
The yield of hydrocarbons is shown in Figure 5. Hydrocarbons were produced from non-composted materials with peak areas ranging from 0% to 35% in the reaction at 450°C and 0% to 9% at 550°C.However, at both temperatures, the peak area substantially creased when produced by composted materials ranging from 56% to 66% at 450°C and 65% to 66% at 550°C.In addition to producing hydrocarbons by wax thermal degradation, they may have been obtained from the degradation of hemicellulose and lignin, mainly in cases of composted biomass (Zhang et al., 2023).The production of aliphatic hydrocarbons occurred from phenolic reactions of decarboxylation resulting from lignin depolymerization (Volli et al., 2021).Since lignin remained at high concentrations after composting, decarboxylation was higher, obtaining peak areas up to 65 times larger than those found in non-composted biomass at 450°C and up to 22 times larger at 550°C.
The pyrolysis at both temperatures of the composted materials showed a marked reduction of oxygenated compounds compared to non-composted biomasses (Figure 6).At 450°C, the area percentages of oxygenated compounds.At 550°C, these percentages ranged from 86% to 98% for the non-composted and 27% to 29% for composted biomasses.This reduction in the oxygenated area may be attributed to the lower elementary oxygen concentration in composted biomass (Table 2) and the existence of methyl and ethyl groups in most of the phenolics produced.The C-O connection of the methoxyl group in lignin was probably easily fractured so that oxygen atoms were removed to a higher degree (Wang et al., 2011).

| DISCUSSION
The temperature profile and degradation profiles (Figures 2 and 3) indicate that the system SR90 has higher microbial activity at the beginning of the process, higher extractive degradation, lower hemicellulose and cellulose degradation, and a higher percentage increase of lignin content compared to other systems.The SR90 was also the system with the highest oxygen content reduction, implying that this system is more advantageous as a pretreatment for pyrolysis.
The results presented show that the composting SR90 provided higher degradation of extractives (Figure 3a); however, the SR75 presented higher hemicellulose and cellulose degradation during the process (Figure 3b,c), and for both systems, such degradations were more pronounced in the initial stage.increases in the lignin content in all tems have occurred, it can be observed that the SR90 and SR75 dropped in the lignin content between the 20th and 30th days, followed by an increase, which can be attributed to a degradation followed by lignin repolymerization (Kausar et al., 2010).On the other hand, the lignin of SR100 and SR90 tended to fall next to the 44th day, indicating the highest count of ligninolytic microorganisms in the final stage of the process (López-González et al., 2015).The SR75 presented a lower increase in lignin content, which can be associated with the lower degradation of extractives through this process.These results showed that no lengthy processing is required to raise lignin concentration through composting as a pretreatment for pyrolysis (Jurado et al., 2015;Ventorino et al., 2016).
Compared to other biological pretreatments, composting showed similar changes in lignocellulosic biomass to anaerobic fermentation and anaerobic digestion shown, for example, by Wang et al. (2018) and by Song et al. (2021).In both works, an increase in lignin content was reported after biological pretreatment.Hence, in these cases, the changes led to higher phenol and aromatic content in bio-oil.Overall, biological pretreatments applied to lignocellulosic biomass reduce the recalcitrance of some of its compounds, such as lignin and cellulose.Nevertheless, the changes depend on the kind of enzyme secreted by microorganisms in the biological process, the time, and the kind of biomass.These can be seen in the work of Martínez-Patiño et al. (2018), which shows seven different strains of white rot fungus acting in the different degradation rates of lignocellulose of olive tree biomass, and the work of Wang et al. (2018) showed how time influenced in lignocellulose degradation of poplar wood through fermentation.
Biological pretreatment has been reported as effective in increasing the selectivity of pyrolysis production, reducing activation energy (Hamieh et al., 2014;Wang, Ye, Yin, Jin, et al., 2014;Wang, Ye, Yin, Lu, et al., 2014).In this work, the products from composted material pyrolysis stand out due to the high production of hydrocarbons at both temperatures (Figure 4).Compared with the non-composted material products, it observed a reduction in ketone content, of esters and furans and partial elimination of acetic acids, phenolics, and aldehydes.
Acetic acid production was not found or was low in all composted systems pyrolysis products at both temperatures, indicating that composting biomass treatment has improved bio-oil by raising its stability.Acetic acid production from sisal residues and sisal fiber powder may be related mainly to hemicellulose degradation during composting (Kumar et al., 2021).Furans and ketones' low production in the bio-oil of the composted samples is related to the low concentrations of cellulose and hemicellulose in these materials (Zhang et al., 2023).Non-composted materials had low lignin concentrations (9.5%-11.4%),which generated small areas of peaks corresponding to phenolic compounds (2%-5% at 450°C and 6%-11% at 550°C) once lignin was the primary precursor of these compounds (Kaur et al., 2022).On the other hand, composted biomasses did not follow the same behavior as other biomass, with even lower phenolic areas (0%-1% at 450°C and 2%-4% at 550°C), despite the high lignin concentration (28%-35%).In prior works, the increase of lignin through biological pretreatment led to bio-oil with higher phenolic content (Wang, Ye, Yin, Jin, et al., 2014;Wang, Ye, Yin, Lu, et al., 2014), which can indicate a higher degradation rate through composting.
Hydrocarbons produced from composted and noncomposted biomass pyrolysis are a mixture of alkanes, alkenes, aromatic, and cyclic molecules (Figure 5).Alkanes and alkenes were the main products formed from composted biomass, with peak areas between 38%-40% and 16%-28% (at 450°C), and 26%-27% and 27%-30% (550°C), respectively.However, at 550°C, non-composted materials produced few alkanes (0%-1%) and alkenes (0%-8%).This result may be directly linked to higher hydrogen production due to the higher lignin content of composted materials (Juchelková et al., 2015).The increase in lignin content after biomass composting was in the 195%-268% range for sisal residue (Table 1 and Figure 3), which is much larger than the interval found by Barneto et al. (2009) (20%-30%).Higher hydrogen production may have occurred due to C-connections breaking in lignin (Juchelková et al., 2015).alternative for the production of alkanes and alkenes would be the total hydrogenation of the benzenic ring and cyclohexane production (Jiang et al., 2020).Cyclohexane can be broken through the Haag-Dessau mechanism, where silicon catalyzes the reaction (Slagtern et al., 2010).In this mechanism, the cyclohexane is protonated, forming the carbon ion (CH 3+ ), which forms olefins and alkanes when broken.It also explains the low phenolic production observed in bio-oils from composted biomasses.
Aromatic hydrocarbons remained in a narrow peak area (0%-4%) for all biomass studied at both temperatures.This result exposes the difficulty of producing aromatic from biomass, even after composting.According to Ke et al. (2022), lignin, cellulose, and hemicellulose are combined by covalent reticulation, making the material very resistant to aromatic production and promoting the production of oxygenated species.In addition, high metal levels (K, Na, Ca, and Mg) reduce the yield of aromatic hydrocarbons.
At 450°C, no peaks of cyclic hydrocarbons were identified for any of the samples.At 550°C, the peaks of cyclic hydrocarbons were identified only in composted biomasses, with pick area ranging from 9% to 13%.The large production of cyclic hydrocarbons occurred due to the extensive production of alkanes and alkenes, which should have occurred by cycling reactions.
In the composted samples, temperature rise favored the production of hydrocarbons, especially aromatic and cyclic.These indicate that alkane and alkene cycling processes are favored at the highest temperature (550°C).Shi et al. (2017) reported that the rise in temperature (above 500°C) favors the breaking of aliphatics and the formation of cyclic hydrocarbons followed by dehydrogenation for aromatic formation through the Diels-Alder reaction.
Although the results showed few differences among the products from pyrolysis of the composting systems, it can be highlighted that SR75 generated a biomass capable of producing the lowest oxygenated bio-oil at the mildest temperature without affecting hydrocarbon production.
Furthermore, these hydrocarbons are concentrated mainly in alkanes and alkenes, favorable to biorefinery applications.

| CONCLUSIONS
This work sought to contribute to improving bio-oil from the pretreatment of biomass by composting.Composting changed the original biomasses, decreasing the fraction of extractives, cellulose, hemicellulose, and elemental oxygen.However, lignin and ash content increased its fractions mainly due to the mass reduction of the materials and the difficulty of degradation.
The transformation of biomass by composting meant that the main pyrolysis products were the alcohol and the hydrocarbons, not producing esters and furans.There was also a significant reduction in carboxylic acids, aldehydes, ketones, and phenols.Lignin was the major precursor of aliphatic hydrocarbons due to hydrogen production and transformation of phenols into cycloalkanes.Composting also helped reduce the oxygenated species of bio-oils at both temperatures studied, probably due to the hydrogen in the reaction medium.However, increased ash content can hamper the application in fluidized bed pyrolysis due to corrosion potential.Although there are slight differences between the results of the composted biomasses, they allow us to conclude that the inclusion of sisal fiber powder did not cause significant effects on the produced bio-oils, besides the lowest oxygenated components production at the lowest temperature studied in the system with higher sisal fiber powder content (SR75).
The results presented in this work show the benefits of biomass composting for bio-oil quality.This work is an initial study, but it brings significant elements that could serve as a basis for the evolution of the theme.

ACKNOWLEDGMENTS
Thanks to the Federal University of Bahia (UFBA) and the Federal Institute of Education, Science, and Technology of Bahia (IFBA).Association of Sustainable and Solidarity Development of the Sisaleira Region (APAEB).This study was financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior-Brasil (CAPES)-Finance Code 001.
Percentage of metallic compounds present in the ash from the initial and final samples.

F
Production by functional groups (a) at 450°C and (b) at 550°C.

F
Percentage of hydrocarbons (a) at 450°C and (b) at 550°C.

F
Percentage of oxygenated and non-oxygenated (a) at 450°C and (b) at 550°C.