Technological potential of fibers from 20 Hevea brasiliensis clones for use as pulp , paper , and composite materials

Based on the use of Hevea brasiliensis in latex production, the species has a consolidated role in the Brazilian economy. However, at the end of the production cycle, which lasts from 25 to 30 years, resulting wood, in general, Research, Society and Development, v. 10, n. 10 e549101019102, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i10.19102 2 has no added value and is normally used for firewood, without further exploring its technological properties. Seeking to introduce this species into the pulp and paper industry, we aimed to determine the fiber quality of 20 commercial clones. Wood samples were collected from planted trees (11 to 12 years old) in the municipality of Selvíria MS. We calculated wood quality indexes of H. brasiliensis for cellulose and paper, including Flexibility coefficient, Wall proportion, Runkel ratio, Slenderness ratio (Aspect ratio), and Luce's Shape Factor. The fiber quality indexes that best indicate the potential use of wood from Hevea brasiliensis clones for cellulose and paper production were Wall Fraction, Runkel Ratio and Luce's Shape Factor, indicating that H. brasiliensis fibers, if used for proper industrial purposes, will give origin of a rigid paper with greater bonding contact on the surface. The clones that presented better results for fiber quality indexes were IAC 311, IAC 41, IAN 873, IAC 326, IAC 40 and RRIM 725. Enabling the use of wood after the latex exploratory cycle for pulp and paper will serve as a basis for genetic improvement of these clones and expansion of forest plantations for this purpose.


Introduction
Pulp production in Brazil has grown at rates higher than those seen by other producers in the world. According to Industria Brasileira de Árvores (2020), Brazil is the world's second largest producer of cellulose, behind only the United States, and ranks tenth in the world in paper production.
The success of Brazilian industry in pulp and paper production was the result of forest genetic improvement strategies, mainly for Eucalyptus spp., owing to its rapid growth, short rotation and excellent physical and chemical wood properties (Alves et al. 2021).
In the face of new technologies, common sense would be sufficient to predict a decrease in paper use. However, in practice, the demand for pulp and paper has remained stable or demonstrated an increasing bias. Therefore, it is worthwhile looking for new species that can meet this demand. In this context, we highlight the rubber tree, another species native to Brazil, with long-standing importance in the Brazilian forest industry. Lima et al. (2000) reports that the rubber tree is native to the Amazon region and that Brazil has ten of the eleven known species. Botanically, the rubber tree is a dicotyledon from the Hevea genus, Euphorbiaceae family, all species of which are arboreal. These species have good edaphoclimatic adaptation characteristics and fast growth in tropical and subtropical environments (Santos et al. 2020).
According to Carretero and Mello (1988), in 1972, the first rubber tree production program was created with the goal of reaching 18 thousand planted hectares, in 1977 the second program with a goal of 120 thousand hectares, and in 1981, the third program with a goal of 40 thousand hectares. In 1988, approximately 150 thousand hectares were financed for the planting of rubber trees in Brazil.
Since 1995, the State of São Paulo has been consolidated as the main producer of the species, representing 56% of the country's production (Gonçalves et al. 2001). This can be attributed to genetic improvement research carried out by the Instituto Agronômico de Campinas-IAC. The planted area in Brazil has been increasing significantly from 159,500 ha to 218,307 ha in 2018, and the growth of forest plantations of the species should increase in the next decade, according to projections in the most recent report of the Brazilian Tree Industry -IBA (2020).
The expansion of rubber tree planting in Brazil resulted in an abundant supply of wood from this culture at the end of its rotation (25-30 years), leading to interest in studies on this raw material. In Brazil, wood obtained at the end of the latex production cycle is traditionally used as firewood, despite its good workability characteristics, such as gluing, nailing, and perforation, being easily bent with the use of steam and easily dyed (Eufrade Junior et al. 2015). Research shows that Hevea brasiliensis wood has potential for use as a raw material for bioenergy (Menucelli et al. 2019) and reconstituted panels (Okino et al. 2009;Iwakiri et al. 2017). Furthermore, this wood has good mechanical properties (Leonello et al. 2012), as well as colorimetric characteristics for indoor use (Autran;Gonçalez, 2006).
Although it is a tradition in Brazil to use fibers from Eucalyptus grandis, E. saligna, and hybrid E. grandis x E. urophylla for pulp and paper production, the characterization of wood from alternative species, such as Hevea brasiliensis, especially from different genetic materials at an early age, is strategic for the proper disposal of clonal materials in postexploitation stands and for ensuring the supply of fibers to the market. In addition, Hevea brasilienses fibers can be used as reinforcement in composite materials. The application of vegetable fibers as reinforcement in composites depends on the availability, cost and inherent properties of each fiber. Reinforcement capacity depends on the type of fiber, amount used, its geometry, and length and thickness ratio (aspect ratio) (Ardanuy et al. 2015).
The economic importance of Hevea brasiliensis is well established, and productive areas have advanced in Brazil, but available scientific information on the technological characterization of the species is lacking. Therefore, our goal here was to determine the technological potential of 20 clones of Hevea brasiliensis for paper and cellulose production, as well as fiber potential for use in composite materials.

Methodology
Hevea brasiliensis wood samples were collected from 60 trees, three from each clone, in the municipality of Selvíria, State of Mato Grosso do Sul (20 ° 20 ′ S, 51 ° 24 ′ W, elevation 350 m). The planting of Hevea brasiliensis was established in 2006 at a spacing of 3 × 3 m from the seeds of free-pollinated clones ( Table 1). The collection of study material was carried out in 2017 and 2018 at 11 and 12 years of age. The soil of the experimental area was classified as Red Latosol, clayey texture, typical clayey dystrophic, moderate A horizon, hypodystrophic, alic, kaolinitic, ferric, compacted, very deep and moderately acidic (LVd) (Santos et al., 2018) The relief it is characterized as moderately flat and wavy. The local climate is of the Aw type, according to Köppen's classification, with an average annual temperature of 24.5 °C, an average annual Research, Society andDevelopment, v. 10, n. 10 e549101019102, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i10.19102 4 humidity of 64.8%, an average annual precipitation of 1232.2 mm and an average insolation of 7.3 hours / day (Hernandez;Lemos Filho;Buzetti, 1995). The origin of the genetic material of the clones studied is reported in Table 1. The material was split into disks in the forest products laboratory of the São Paulo Forestry Institute. The origin of the genetic material of the studied clones is reported in Table 1. Small wood samples from each clone were used for maceration using Franklin's method (Berlyn and Miksche, 1976). Then they were transformed into small samples for dissociation in hydrogen peroxide following the methodology proposed by Johansen (1940). Then these wood fragments were stained with aqueous safranin and temporarily mounted in a solution of water and glycerin (1:1). Measures followed by the recommendations of the IAWA Committee (1989). Quantitative data is based on at least 30 measurements for each characteristic of each tree, thus meeting the statistical requirements for the minimum number of measurements.
Wood density was calculated according to NBR 11941, Brazilian Association of Technical Standards -ABNT (2003). The samples were saturated in water until the wood presented moisture above the fiber saturation point, then they were dried in ovens until they presented constant mass, we use ten bodies of evidence for each clone. The basic density was obtained according to equation 6.
Cluster analyses were performed with means of clones and their variables. Later, data were standardized so that they all had a common scale, but without distorting their difference, where the mean was equal to zero, and standard deviation was equal to 1. For this standardization, the Standardize tool in the software was used. The Euclidean distance was used as a measure of similarity, and the unweighted pair group method with arithmetic mean (UPGMA) was used to determine the groups. Through the cophenetic correlation coefficient, the dendrogram generated was evaluated. Through the generation of clusters, clone averages in their given groups were used. The analyses were performed based on the correlation between data.
For the number of principal components, it was determined that the sum of two components presented more than half of the data representation (i.e., estimated variation in the ordinate of the auto values). All analyses were performed in the R statistical software environment (R Development core team, 2021).

Results and Discussion
Material quality for cellulose and paper production can be evaluated by wood properties, e.g., basic density and fiber features (Istikowati et al. 2016). The distribution of cell types determines wood quality attributes of pulp and paper. For example, interconnection, tear resistance, compression, and tension are related to the junction of wood density and fiber features in a manner that defines the properties of the final product (Downes et al. 2002). Mean, maximum and minimum values of each variable are shown in Table 2. The mean density for 20 Hevea brasiliensis clones was 0.46 g.cm³, which is considered medium, according to the classification by Silva et al. (2015). Currently, the industrial standard for pulp and paper derived from Eucalyptus forests prefers density values between 0.40 to 0.55 g.cm³ (Dias;Simonelli, 2013). Woods with a density lower than 0.40 g.cm³ have reduced yield, higher consumption of reagents and a high reject content. Woods with a density above 0.55 g.cm³ present greater difficulties in chipping logs, which causes wear on the mincer knives (Dias;Simonelli, 2013), and pulps with less fiber binding (Ferreira et al. 1997). Thus, wood density of H. brasiliensis clones fits within acceptable limits of basic density for production and yield in the industrial production process. In characterizing wood fiber, flexibility coefficient values above 60% are desired for production of pulp and paper (Foelkel et al. 1987). However, with these FC values, looser and looser mesh papers are produced without much bonding between fibers. Thus, H. brasiliensis clones with an average FC value of 65.50% are likely to have longer and thicker fibers that can be used for purposes requiring mechanical effort, such as packaging and printing (Menegazzo, 2012).
Wall fraction is related to fiber firmness such that the higher its value, the more rigid it can become, affecting paper strength (Gonçalez et al. 2014). Currently, the reference value in the pulp and paper industry for Eucalyptus clones is 40% for production of writing and printing paper, as it results in greater fibrous pulping, greater intertwining of fibers and the potential of fiber-to-fiber contact. These properties improve traction, bursting, tearing, folding, surface resistance and internal resistance of the sheet, according to Baldin et al. (2017). For rubber tree clones, an average value of 34.62% was observed, suggesting that the paper produced will have good opacity, ink absorption and mechanical resistance.
The Runkel ratio measures the degree of fiber collapse during the paper production process such that the lower the RR, the greater the collapse. This allows the fibers to have a larger contact surface, establishing a greater number of bonds, which, in turn, results in paper with greater strength to traction and overflow (Talgatti et al. 2020). RR values can be classified, in general, from 0.56 as good, since Runkel (1952) proposed that fiber with an index up to 0.25 is considered excellent for paper. RR values from 0.25 to 0.50 can be considered very good, from 0.50 to 1.0 good, and 1.0 to 2.0 regular. RR values above 2.0 indicate fibers with thick or very thick walls and easy to tear, resulting in poor paper product. According to Nisgoski Research, Society and Development, v. 10, n. 10 e549101019102, 2021(CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i10.19102 et al. (2012, high values are not desired since the fiber becomes rigid with greater difficulty in accommodation and union on the paper sheet. Therefore, RR values of H. brasiliensis clones in this study can be classified as good for paper production. The slenderness ratio is directly related to tear and breakage strength of fibers. Therefore, values from 40 to 60% present good paper characteristics in tearing and double folding (Mongolon; Aguilera, 2002). The mean value in the present study was 43.81%, which is in line with industrial requirements based on the weakness in paper formation. Slenderness ratio is also related to digestibility of the pulp (Ohshima et al. 2005), meaning that it is easy to obtain water-soluble and absorbable pulp from H. brasiliensis wood.
In architecture, the slenderness ratio is an aspect ratio, i.e., the quotient between building height and width. From this perspective, slenderness is used to calculate the propensity of a column to buckle (Willis, 2016). In this context, aspect ratio (slenderness ratio) is used in many studies on use of wood fibers together with other materials, categorized as composite material, which, according to Hull & Clyne (1996), is composed of two or more phases, which have different physicochemical properties. Composite materials have a matrix and another material that serves as reinforcement (e.g., wood fibers), so the properties are determined by the combination of different materials used. The value of aspect ratio or slenderness ratio can vary, depending on fiber type and design, assuming values from approximately 50 to 500 (Hull;Clyne, 1996). Aspect ratios can affect properties, such as the strain to failure and toughness. A higher aspect ratio will result in lower values of strain at failure and toughness because angular particles induce crack formation (Oréfice et al., 2001).
Luce's shape factor is particularly important when considering the technological potential of H. brasiliensis wood for use in the pulp and paper industry. This measure is directly related to the final density of the paper sheet and thus a property used in the selection of wood quality in forest improvement programs. The mean value of H. brasiliensis clones was 0.40. This value is lower than the range of values from 0.50 to 0.60 for Eucalyptus spp. used for cellulose and paper production (Baldin et al. 2017;Pirralho et al. 2014), indicating the formation of a thinner, less dense, and lighter weight paper (g.m²).
It was possible to observe heterogeneity in relation to the grouping of variables for 20 clones in the formation of different clusters (Table 3), and four different clusters emerged from the dendrogram (Figure 1). In the cluster analysis, cluster 1, formed by clones IAC 301, GT1, Fx 2261, and PB330, was united because the clone cluster presented lower densities and higher flexibility coefficients (mean of 68.40%) (Table 3), as well as flexibility coefficient close to that of cluster 2. In this group, it is noteworthy that clones Fx 2261 and IAC 301 (Table 1) indicate that this genetic material from Brazil should be used in genetic improvement programs for gains in flexibility coefficients, emphasizing that this index allows better intertwining of fibers in paper formation (Castelo, 2007;Magasha, 2019).  (Table 1). This cluster analysis demonstrates that clones genetically distinct from the Brazilian ones have similar characteristics of a product with greater added value. The characteristics that determined the similarity between the clones were Runkel ratio and Luce's shape factor, indicating that these clones present good interconnection of fibers in paper formation (Magasha, 2019).
Cluster 4 presents the largest number of clones with similar characteristics, totaling nine according to the dendrogram analysis ( Figure 1). The clones in this group were grouped according to their superiority in terms of wood density, fiber diameter and fiber wall thickness. It is expected that wood density presents higher values in the presence of increased diameter and thickness of fiber wall (Paulino;Lima, 2018) and that these features influenced the clustering by similarity among the clones.
The PCA analysis is shown in Table 4. Here, the fiber quality index least related to pulp and paper production is flexibility coefficient showing negative values in two axes. As mentioned in the study, this occurs because H. brasiliensis wood presents values above 60%, which provides for the formation of a loose paper network for bleached pulp, but still has interesting values for production of packaging paper. However, positive values were noted for PC1 and PC2 for wall fraction, Runkel ratio, and Luce's shape factor, the latter being important in the selection of potential species for cellulose and paper based on fiber quality. Intrinsic behavior was observed for slenderness ratio, which has the first negative matrix and the second positive matrix, but with a low relationship, demonstrating that slenderness ratio has little influence on the selection of clones for pulp and paper, considering the influence on tear and fold of the paper.  Figure 2 shows the ordering of cluster groups created based on anatomical fiber features and cellulose and paper indices for 20 H. brasiliensis clones. As proof that forestry is already consolidated in this country, clones grouped into cluster 3 were predominantly Brazilian clones, including IAC 311, IAC 41, IAN 873, IAC 326, IAC 40, and clone RRIM 725 of Malay origin. Cluster 3 shows positive ordering and similarity in terms of fiber quality indexes with Runkel ratio, Luce's shape factor and wall fraction being the most suitable for pulp and paper production. These values highlight the potential of Brazilian forestry of these genetic materials for large-scale production.
The indices that stood out for presenting positive PC1 and PC2 were wall fraction, Runkel ratio and Luce's shape factor. These results indicate that H. brasiliensis fibers, if used by industry, will give rise to a rigid paper with greater contact in the surface union, as well as a paper with greater resistance to traction and bursting. Such properties make it possible to use this material from younger ages for this purpose and provides an economically viable use for the wood after the latex exploratory cycle.

Conclusion
Hevea brasiliensis wood from younger ages presents potential for cellulose and paper production because fiber quality parameters indicate potential for the formation of rigid paper resistant to traction and bursting with greater contact at the surface union. The clones that show similarities in fiber quality indices were IAC 311, IAC 41, IAN 873, IAC 326, IAC 40 and RRIM 725. Most of these clones are from genetic origin developed in Brazil, indicating the capacity and quality of local forestry for the advancement and expansion of plantations of the species for this end purpose.
Additionally, a detailed chemical characterization and pulp yield for the species is recommended in order to determine the percentages that are in accordance with industrial requirements regarding ash content, extractive content, lignin content and holocellulose.