1 INTRODUCTION

In Brazil, the use of wood for energy generation has a long historical association with charcoal production, primarily due to the demand for this product by the steel industry. The country is an absolute leader in steel production using charcoal as a bioreducing agent, resulting in what is known as "green steel". For example, in 2021, 4.25 million tons of charcoal were incorporated into the iron and steel, ferroalloys, and metallic silicon sectors, reaffirming the environmental commitment of the entire production chain (IBÁ, 2022).

In the cultivated tree sector, the percentage of total charcoal production is significant, accounting for 94% (PEVS - IBGE, 2020). Among the raw materials used, Eucalyptus stands out for its high silvicultural productivity, environmental adaptability, uniformity and wood quality. Despite the exponential increase in the establishment of homogeneous forest plantations, they are still not able to fulfill all the demand from companies, resulting in an average annual deficit of approximately 47%. This deficit is filled by natural forests, with the Cerrado biome being one of the main suppliers of charcoal (Cachoeira et al., 2019).

The silviculture of Cerrado species is still underexplored in the establishment of plantations for commercial uses, representing a challenge that needs to be overcome for the diversification of forest biomass. Studies conducted by Mendonça et al. (2008), Pilon and Durigan (2013), and Oliveira et al. (2015) indicate a high survival rate of seedlings from different species of the biome, exceeding 70%, which confirms their plasticity and tolerance to the climatic and soil conditions at the planting sites. Oliveira et al. (2015) discuss some growth strategies of Cerrado species compared to exotic species, highlighting a greater average growth in diameter during the early years, in contrast to traditionally planted species, which initially develop more in height. This would be a physiological adaptation for water and nutrient uptake related to seasonality.

The Cerrado is a tropical savannah region in South America, including a large part of central Brazil, part of northeastern Paraguay and eastern Bolivia, and is Brazil's second largest domain. The Cerrado occupies approximately 24% of Brazil's territory, in a total estimated area of 2,036,448 km² (Santos et al., 2021). The Brazilian Cerrado is recognized as one of the most biodiverse biomes in the world, with more than 12,000 plant species found there, many of which have significant economic and environmental potential (Carvalho, 2022).

Throughout the years, significant transformations in land use and occupation have been observed in this biome (Carvalho, 2022). Despite the appropriate environmental licenses, areas are cleared for agricultural activities, predominantly pastures and reforestation, or for road construction, resulting in a considerable amount of forest waste. Charcoal production has emerged as a solution, not only to add value to this waste, but also to ensure proper environmental disposal. Furthermore, charcoal from native species should also be considered as a promising economic potential, providing income-generating opportunities for local communities and contributing to regional development.

The use of wood from a particular forest species for bioenergy should be based on an analysis of its chemical, anatomical, physical and energy properties (Soleymani et al., 2023). Thus, high-quality biomass must have a high fixed carbon content, thick-walled tissues, few empty spaces in the lumen, low moisture content, high basic density and high calorific value. These variables are influenced by the characteristics of the wood, such as species, age and position in the tree, as well as by silvicultural techniques related to the production of energy forests.

The literature indicates that species native to the Cerrado are suitable for producing heat (Vale et al., 2010; Marques et al., 2020; Siqueira et al., 2020). Notably, genera such as Astronium, Dalbergia and Sclerolobium are highlighted for their high dry mass production, higher specific masses, calorific values above the general average and excellent results for fixed carbon. Costa et al. (2014), evaluating the energy potential of five species native to the Cerrado region of Minas Gerais, highlighted the gravimetric yields (30.88 - 34.39%) and promising charcoal properties of these species, including apparent relative density (0.255 - 0.475 g.cm^-3), higher calorific value (7.135 - 7.730 kcal.kg^-1), fixed carbon (77.2 - 81.0%) and fixed carbon yield (24.3 - 26.56%).

Due to the importance of charcoal in promoting sustainable steelmaking as an agent of environmental decarbonization, in line with the practice of using native forests that would otherwise be cleared for agricultural substitution, this research aimed to evaluate the energy quality of charcoal produced from the wood of seven Cerrado species. The researchers are convinced that, in addition to proper management of forest resources and evaluation of the waste generated, the results obtained can contribute significantly to diversifying the sources of wood used in charcoal production. Furthermore, the research project aims to promote the field of forestry by encouraging studies into the implementation and management of energy plantations involving alternative species. Thus, the objective is not only to diversify raw materials for energy production, but also to promote sustainable development through the responsible management of natural resources.

2 MATERIALS AND METHODS

2.1 Biomass characterization

The trees used in this study came from mature native vegetation in the Cerrado Sensu Stricto Biome, harvested under a sustainable forest management plan and donated to the university for scientific research, as shown in Table 1. The disks were collected at breast height (1.3 m above the ground) and then subdivided into four wedges, passing through the pith. Two of these wedges were used to determine physical and anatomical properties, whereas the other two were used for carbonization.

2.2 Physical and anatomical characterization of wood

A moisture analyzer with halogen light was used to determine the moisture content of the wood. The basic density of the wood was determined using the immersion method, in accordance with standard NBR 7190 (ABNT, 2022). The dry mass of the samples was determined on a precision analytical balance (0.1g), after reaching constant weight in an oven at 103 ± 2ºC, and the volumes were determined using the Archimedes principle.

For the anatomical analysis of fiber biometrics, macerated slides were prepared. Small fragments were taken from the heartwood and sapwood portions and immersed in a macerating solution adapted from Franklin (1945), consisting of 50% glacial acetic acid, 58% hydrogen peroxide, and 12% distilled water. The Eppendorf containing wood fragments and the solution were placed in an oven at 50°C for 24 hours to soften the material, followed by maceration through friction. Once the fibers were disaggregated, the macerate was washed in distilled water and filtered using filter paper. Subsequently, the fibers were stained with safranin, and permanent slides were prepared with Entellan. Measurements of length, total diameter, and lumen diameter of 25 fibers were taken under a microscope, and the wall thickness was calculated using the ratio, according to Equation (1):

Wall thickness= ((fiber diameter)-(lumen diameter))/2 (1)

2.3 Wood carbonization and charcoal quality

Previously, charring wood involved extracting a sample from each wedge to determine moisture content using an infrared scale. The laboratory carbonizations were performed in an electric muffle furnace connected to a water-cooled condenser, which is coupled to a condensable gas collector (Figure 1). Heating control was manual, with temperature increments following the specified ramp, corresponding to an average heating rate of 1.25°C/min. The initial temperature was set at 150°C, reaching a final temperature of 450°C, with a total carbonization time of 4 hours.

To evaluate the charcoal, the gravimetric yield was measured as the difference between the mass of charcoal and wood, free of water. In order to evaluate the charcoal, the gravimetric yield was measured by the difference between the mass of the charcoal and that of the wood without water. The moisture content, fixed carbon, volatile material and ash content were determined according to the specifications of standard D1762-84 (ASTM, 2013). The higher heating value was estimated according to Dulong's empirical Equation (2), which considers the fixed carbon content of the sample, as proposed by Green and Perry (2008) and Jenkins (1998).

where: PCS - higher heating value (KJ. Kg^-1); FC - Fixed Carbon.

The determination of apparent relative density was based on the ASTM D-167-93 standard (ASTM, 1999) for mineral coal, adapted for charcoal, in conjunction with the methodology suggested by Oliveira et al. (1982). The charcoal was previously weighed and carefully placed in a wire basket. The basket was immersed in a container of distilled water at room temperature, where it remained for 15 minutes. After this time, the hole in the container was opened to collect the displaced water. After removing the basket from the water, allowing the excess water to drain off, it was left to stand for 1 to 2 minutes and then the moistened charcoal and the displaced water were measured. The apparent density was calculated according to Equation (3):

where: ARD - Apparent Relative Density (Kg/m³); Wcd - Weight of charcoal after drying in an oven (g); Wcb - Weight of water displaced from the container by the immersion of charcoal + basket (g) ; Wcw - Weight of wet charcoal (g).

The pyrolyzed liquor was collected in a suitable recipient during carbonization and, after completion, the recipient was left to cool at room temperature for 15 hours and then weighed to determine the yield of the condensed liquids. The yield was calculated as the ratio between the mass of pyroligneous liquor formed afterwards and the dry mass of the wood before carbonization. The yield of non-condensable gases was calculated by difference, according to Equation (4):

where: NCGY - non-condensable gases yield (%, dry basis); CGY - gravimetric yields of charcoal (%, dry basis); PLY - pyroligneous liquid yield (%, dry basis).

2.4 Data analysis

The experiment was evaluated using a completely randomised design (CRD), considering different species as the variation factor. A total of three discs per species were evaluated in the experiment. To analyze the basic density of the wood and the quality of the charcoal - volatile material, fixed carbon and ash content, apparent relative density and moisture content - the average values of the duplicates are presented. On the other hand, variables such as gravimetric charcoal yield, condensable gases, non-condensable gases (3 repetitions) and fiber biometry (25 repetitions) were assessed using ANOVA. The data were subjected to the Shapiro-Wilk normality test and their means were statistically compared using the Tukey test at a significance level of 5% with the ASSISTAT© statistical software.

A group analysis was conducted using a dendrogram constructed using the hierarchical grouping method proposed by Ward (1963). The Mahalanobis distance was adopted as a measure of dissimilarity, as suggested by Cruz and Carneiro (2006), indicating characteristics measured for the species in this study, along with information on other species taken from the literature (Table 2), for the attributes of basic density, gravimetric charcoal yields, apparent relative density, volatile material, fixed carbon and ash content. The Elbow method (Thorndike, 1953) was used to determine the ideal number of groups in the dendrogram. By means of multivariate data analysis, the authors sought to observe the proximity or distance of the native species in this study in relation to those already described in the literature and even some used commercially. The analysis was processed using R Studio software version 4.0.2 (R Core Team 2020).

3 RESULTS AND DISCUSSION

3.1 Physical and anatomical characterization of wood

The average moisture content of the wood was 10.30 per cent, below the equilibrium moisture content of 12.4 per cent for the city of Montes Claros, where the material was stored and the experiment was conducted (Baraúna et al., 2022). These values are in line with the specifications described in DN 227 (COPAM, 2018) for charcoal production, which recommends carbonizing wood with a moisture content below 40%.

Among the evaluated woods, only two are classified as having low basic density: Strychnos pseudoquina and Persea willdenovii, while the others were categorized as having medium density (Table 2), according to the classification by Melo et al. (1990); Vale et al. (2002); Silveira et al. (2013); and Silva et al. (2015). The fibers exhibit medium length and thin to thick walls in all species, except for Casearia mariquitensis, which has short fibers, and Strychnos pseudoquina, which has fibers with thin walls (IAWA, 1989).

The specimen of Bowdichia virgilioides proved to be the densest, a factor that can be attributed to the length of the wood fibers. Conversely, fibers with thin walls confer the lowest basic density value to Strychnos pseudoquina. The results corroborate the statement that wood basic density is highly dependent on the dimensions of the cells composing it (Panshin; Zeeuw, 1980).

In order to provide a broader explanation of basic density values through the wood's anatomical constitution, other equally important characteristics should be validated, such as pore frequency and diameter. These assessments could elucidate, for example, how Persea willdenovii had the second lowest basic density value (0.527 g/cm³), despite having long fibers (1193.05 µm) and thin to thick walls (3.37 µm).

Dense woods with long, thick fiber walls, such as A. brasiliensis and B. virgilioides, generally have a higher lignin content. Lignin is the main chemical component that influences the production and quality of charcoal. In addition, trees with higher densities generally have a lower S/G ratio in the lignin, which is considered an ideal scenario for pyrolysis because the guaiacyl unit is more stable and preserved during the process (Massuque et al., 2023).

According to Pereira et al. (2016), the presence of positive correlations between high density, fiber dimensions with thick cell walls and reduced lumens indicates the potential for producing denser charcoal with a smaller pore volume. According to the authors, the selection of tree species which provide wood for the production of high-quality charcoal is based on the assumption of high basic density, representing mass gains per volume allocated within the carbonization furnace. In addition, the higher density of the wood will result in charcoal with a higher density and greater mechanical strength, desirable properties for most applications of this input.

The basic density of Eucalyptus wood, which is used commercially to produce charcoal, must exceed 0.500 g/cm³ (Santos et al., 2011; Pereira et al., 2016; Protásio et al., 2021). Therefore, considering this characteristic in isolation, all the native species evaluated show high potential as biomass for carbonization.

3.2 Wood carbonization and charcoal quality

The average moisture content of the charcoal produced was 3.31 %, suitable for use as an energy input for combustion. Humidity levels below 6 per cent in the dry season and 12 per cent in the rainy season are desirable because higher values significantly reduce the heat of combustion, the temperature of the combustion chamber and the temperature of the exhaust gas. In addition, it is worth noting that moisture content has an inverse relation with calorific value (Fortaleza et al., 2019).

The average gravimetric coal yield with the method used was 32.70%, with values ranging from 29.85% to 36.84%. There was no significant effect of the species factor (Figure 2). Notably, Qualea grandiflora showed the highest conversion of wood into charcoal, consequently producing lower quantities of pyroligneous liquor.

In order to meet the demands of the consumer market, it is desirable to obtain a high gravimetric yield in charcoal, as this maximizes the use of wood in carbonization furnaces, leading to greater energy production and lower yields in liquid and non-condensable gases, two by-products of the pyrolysis process. However, there are currently changes in the evaluation of carbonization products, in which pyroligneous liquor is also gaining commercial importance, ceasing to be a discarded waste product and becoming an additional source of income for the producer.

The average charcoal yield in Brazil is approximately 35 % (Rosillo-Calle; Bezzon, 2005; Ramos et al., 2023), which is very similar to the performance of Q. grandiflora and B. virgilioides. Similar results are reported in the literature for carbonizations at final temperatures similar to those of this research, 450 degrees Celsius. Costa et al. (2014) for five Cerrado species (Luehea divaricata, Casearia sylvestris, Guazuma ulmifolia, Rapanea ferruginea and Trema micranta), with an average of 32.29 %. Santos et al. (2016) and Protásio et al. (2021) found average values ranging from 31.25% to 35.80%, respectively, for commercial Eucalyptus clones. Fortaleza et al. (2019), during the carbonization of residues from the native species Ceiba pentandra, Guatteria sp. and Brosimum sp. in final carbonizations at 500 ºC, found values ranging from 30.03% to 32.35%.

The gravimetric yield is affected by the carbonization process and the quality of the wood. The increase in temperature results in lower yields, with a higher fixed carbon content and, consequently, higher calorific value. The density of wood results in denser charcoal, but the high density value must be a reflection of the increased length and thickness of the fibers (Silva et al., 2020). This relation may be observed in B. virgilioides, with the highest basic density (0.722 g/cm³) and longest fibers (1495.42 µm), crucial factors for the high gravimetric yield in charcoal (34.83%). In agreement with this fact, Santos et al. (2011) mention that 60% of the wood's mass is degraded during the carbonization process in the production of charcoal; therefore, the greater the specific mass of the woody material, the greater the mass of charcoal by volume, resulting in a product with superior mechanical strength.

Regarding wood quality, it is important to note that, in addition to basic density, moisture content and chemical composition play significant roles in influencing the outcome of carbonization. Lower moisture content wood, a higher proportion of lignin and a lower proportion of holocellulose have been shown to be factors that contribute positively to obtaining higher yields in charcoal production (Massuque et al., 2021).

The results were good for the apparent density of the charcoal, with a maximum value of 0.688 g/cm³ (B. virgilioides) and a minimum of 0.273 g/cm³ (P. willdenovii). These specimens also have the highest and lowest basic wood density, respectively. Therefore, it is possible to state that dense wood generates dense charcoal, as reported by Numazawa (1986), Pastore, Okino and Pastore Junior (1989), Santos et al. (2011), Fortaleza et al. (2019) and Delatorre et al. (2020). The other species in the study produced charcoal at decreasing densities: A. brasiliensis (0.575 g/cm³), Q. grandiflora (0.450 g/cm³), C. mariquitensis (0.446 g/cm³), S. pseudoquina (0.440 g/cm³) and T. aurea (0.381 g/cm³).

The denser charcoal tends to increase its mechanical strength and energy density, which are crucial characteristics, especially when the product is used in the steel industry. In the blast furnace, charcoal is stacked with iron ores and melting materials, which makes its mechanical strength essential to support the load on it. If the charcoal is not strong enough, it can crumble and compromise the operation. In addition, charcoal with good mechanical strength is less likely to break during transportation, reducing losses and improving the energy efficiency of the steelmaking operation (Massuque et al., 2023). Thus, according to the results of the analyses evaluated in this research, the B. virgilioides species is an excellent option for this industrial segment. The authors recommend specific analyses for the mechanical strength of charcoal in the future to complement these results.

A charcoal with a density higher than 0.380 g/cm³ is preferable for industrial steel production, as it increases energy productivity (Massuque et al., 2023). The authors mention that bulk density also reflects the stock of fixed carbon per volume of the bioreductor and therefore higher values are recommended. Among the species studied, P. willdenovii (0.273 g/cm³) does not meet this prerequisite.

It is important to highlight the results obtained with regard to the immediate chemical composition of the A. brasiliensis and Q. grandiflora species (Figure 3) in this scientific article. The species in question had the highest concentrations of fixed carbon and volatile material, respectively. In terms of ash, the values were remarkably similar, although numerically higher for Q. grandiflora and lower for P. willdenovii. However, it is important to emphasize that, assuming a uniform charcoal yield for all species, the variability in fixed carbon, volatile material and ash can be attributed to the use of different genetic materials.

The increase in fixed carbon content results from a rise in the final carbonization temperature, followed by a simultaneous reduction in volatile material, yield and loss of strength. The increase in temperature and fixed carbon content also implies a decrease in the yield of the carbonization process. However, a significant reduction in carbon, associated with a higher volatile material content, compromises the efficiency of charcoal in the blast furnace, as it generates a high fines content (Picancio et al., 2018).

The performance levels of more than half of the charred wood were less than ideal, and some were significantly below the limit recommended by the steel industry (below 4%). However, the ash content of charcoal from Bowdichia virgilioides (4.2%) and Strychnos pseudoquina (4.2%) could meet the industry's desired limit. According to the literature, the ideal percentage of ash for charcoal from the Eucalyptus genus is lower, specifically below 1.5% (Pincelli; Mineiro, 2022).

The presence of ash in charcoal should be reduced to a minimum because, as well as reducing its calorific value, it causes wear in the blast furnace and can compromise the quality of pig iron (Santos, 2008). In the steelmaking process, ash acts as a catalyst in the C - CO2 reaction due to the presence of metal oxides such as CaO, Fe₂O₃, K₂O, and MgO (Matos; Rios, 1982). It should be noted that the ash content is related to the wood and is not influenced by the carbonization process at the same final temperature, but rather by the species, the amount of bark and, above all, contamination with soil and/or sand in the wood (Donato et al., 2020).

The relative proportions of fixed carbon, volatile material, and ashes are crucial for recommending biomasses for heat generation, as they influence combustibility and facilitate fuel ignition. Therefore, the essential characteristics of charcoal highlighted by metallurgical industries are: 70 to 80% fixed carbon, 25 to 35% volatile material, 1 to 6% moisture, and 0.5 to 4% ashes (Delatorre et al., 2020; Dias Júnior et al., 2021).

Charcoal properties varied among species, higher heating value from 20.584 to 24.268 kJ.kg^-1. The highest averages were for the species Agonandra brasiliensis, Bowdichia virgilioides, and Persea willdenovii. However, Q. grandiflora showed the lowest value. Similar results were reported by Nisgoski et al. (2014) for five native Amazonian species (26.878 to 31.117 kJ.kg^-1). The higher the content of fixed carbon and the lower the levels of volatiles and ash, the greater the higher heating value of the wood or charcoal will be (Vale et al., 2010).

3.3 Multivariate analysis

The Agglomerative Coefficient (AC) was adopted as a metric in this study to analyze the quality and accuracy of the dendrogram produced, as suggested by Roth et al. (2016). The AC obtained was 0.94, indicating a strong grouping arrangement in the data sets, as values close to 1 denote a satisfactory grouping structure (Crispim et al., 2020). Thus, the dendrogram obtained can be considered consistent, and inferences of interest can be made based on its visual assessment (Figure 4).

The use of the Elbow Method allowed the number of clusters to be determined, and the appropriate number for the set of forest species based on attributes such as basic wood density, charcoal yield, apparent relative charcoal density, volatile material, fixed carbon content and ash content was determined in 4 groups, with a breakpoint equivalent to 18% of the total distance. It was observed that, for the species evaluated in this study, A. brasilliensis, B. virgilioides and Q. grandiflora were classified as group I, while C. mariquitensis, S. pseudoquina, P. willdenovii and T. aurea were classified as group II. As for the data on Eucalyptus species obtained from the literature, most were classified in groups III and IV.

The presence of species characterized as suitable for charcoal production for use in the steel industry, such as E. benthamii (Nunes et al., 2015) and the hybrids E. urophylla x E. grandis and (E. camaldulensis x E. grandis) x E. urophylla x E. grandis (Andrade et al., 2018) in group I, alongside A. brasilliensis and B. virgilioides, reinforces the performance and quality of these species for charcoal production.

However, for the other species under study, Q. grandiflora, C. mariquitensis, S. pseudoquina, P. willdenovii and T. aurea, caution should be exercised when comparing them with the Eucalyptus genus. The average data from groups I and II for volatile materials, fixed carbon and ash content diverged from the average values obtained for groups III and IV, which are mainly made up of species normally used to produce charcoal in the steel industry (Table 4).

5 CONCLUSIONS

Among the seven species studied, the best results were achieved by Bowdichia virgilioides, which presented the highest basic wood density and the longest fibers, important characteristics for its high performance in charcoal production and quality. The multivariate analysis reinforces that this species offers performance and quality for charcoal production, approaching that of some commercial Eucalyptus species. The medium-density wood of Persea willdenovii produced low-density charcoal, while Qualea grandiflora and Tabebuia aurea are not recommended for use in metallurgical industries as they do not meet desirable fixed carbon standards.

This study aims to contribute to the production of charcoal by diversifying the sources of wood currently used. Furthermore, it serves to stimulate the field of forest silviculture in the development and implementation of energy forests involving alternative species to Eucalyptus.

REFERENCES

Publication Dates

History