Fibers were coated with gold to increase conductivity of samples and enable analysis by a Scanning Electron Microscope, JEOL brand, model JSM 6490-LV. ImageJ software was used for processing of digital images and measurement of average diameter and cross-sectional area of fibers.

To investigate the mechanical behavior of natural fibers at low scale through instrumented indentation tests, we used the AFM brand ASYLUM RESEARCH, model MFP-3D-BI. Fibers were cut down to a length of 5mm, and appropriate points for nanoindentation were identified on their surface. This process consists in applying a force to the fiber with the tip of an indenter or cantilever to create indentation marks. The tip is pressed continuously against the fiber for about 2 seconds, and the resulting displacement is measured. Data on force vs. displacement, and on indenter geometry, were used to calculate hardness (H), reduced modulus (Er), rupture modulus (H/Er), resistance to plastic deformation (H³/Er²) and elastic module, after the method proposed in 1992 by Oliver and Pharr. In addition, fibers were swept before and after the indentation at low force using the same indentation tip, thereby generating a topographic record on nanometric scale. The obtained nanoscale topographic images of fiber and indentation traces enabled quantification of the stacking of ductile material around the indenter. For the nanoindentation test, an AC 160 TS Olympus indenter was selected based on the principle that the indenter tip must have an elastic modulus equal to or greater than the sample in order to generate a deformation. Finally, water left over from the NaOH treatment was slowly neutralization (until neutral pH) by diluted acetic acid (vinegar) and disposed by a hazardous waste management company with a valid environmental license.

Fig. 1 shows the cross section of a GAK fiber from the interior of the cluster and treated with 5% NaOH. The fiber has an irregular and elongated shape, while its transversal area is composed of irregularly shaped microfibers. Average value was 13301.806 μm².

Figure 1 Source: Authors.

Fig. 2 shows a comparison between two GAK fibers with and without the 5% sodium hydroxide treatment. The untreated fiber presents irregularities due to the presence of lignin and hemicellulose [38], while the treated fiber exhibits no imperfections, and instead consists of visible microfibrils (observable in the cross-section). Treatment exposes microfibers aligned in the axial direction, which is relevant for applications to composite materials by providing stronger adherence between the polymer matrix and natural fibers.

Figure 2 Source: Authors.

Prior to indentation, we recorded the longitudinal topography of a section of GAK fiber measuring 5um x 5um. Fig. 3 shows that the GAK fiber displays an irregular structure on most of its surface.

Figure 3 Source: Authors.

Fig. 4 is a 3D section along the longitudinal axis of the GAK fiber, revealing surfaces without defined shape that protrude between 0.888μm-0.612μm relative to the general surface. It exhibits superior roughness (Ra) to the areca palm [39], but inferior to flax [40]. This level of roughness may facilitate adhesion between polymer matrices and GAK fibers, confirming our SEM findings.

Figure 4 Source: Authors.

Figs. 5 and 6 show a longitudinal section of GAK fibers before and after nanoindentation. In Fig. 5, cellulose microfibers are observed in lignin within the hemicellulose matrix [34] and aligned with the main axis of the GAK fiber. Due to their orientation, microfibers confer guadua superior resistance to tension along the main axis of the fiber. In addition, the distribution of microfibers confirms that guadua is an orthotropic material [41]. Moreover, Fig. 6 shows 10 nanoimprints with different indentation depths, as shown in Table 2. Nanoimprints 2, 3 and 4 cannot be clearly seen in Fig. 6 since the nanoindentation zone exhibits elastic behavior, recovering by 98.7%, 96.70 and 96.13% respectively after removal of the indentation load, therefore presenting no residual deformation. In contrast, nanoimprints 1, 5-10 presented only minor and non-identifiable recovery.

Figure 5 Source: Authors.

Figure 6 Source: Authors.

Fig. 7 shows the loading and unloading curve of nanoimprint 1. The curve is similar for indentations on other nanoimprints, which is most likely explained by the similar elastic properties of GAK fibers in most of their surface.

Figure 7 Source: Authors.

Following the analysis of curves and the method proposed by Oliver and Pharr, Table 1 presents measures of hardness, reduced modulus, modulus of rupture and resistance to plastic deformation for each of the 10 nanoimprints. Hardness ranges between 10.27 and 13.48MPa, demonstrating the homogeneity of the longitudinal section. However, the relative divergence of hardness in nanoimprint 2 and 9 could be the outcome of specific load distributions. The average hardness of GAK fibers is 11.75MPa, below the reported values for other fibers such as flax, hemp, cotton and pineapple [42]. Reduced modulus ranges between 11.65 and 14.27 MPa, which are low values compared to sisal fibers, and result from the disposition or crystallinity of cellulose microfibers [41]. Average breaking modulus and resistance to plastic deformation were respectively estimated in 0.91 and 9.68MPa.

Table 1

Source: Authors.

The behavior after a full cycle of loading and unloading for the 10 nanoimprints is shown in Fig. 8. Loading and unloading cycles originating from the indentation in the longitudinal section of the fiber exhibit almost no variability. Variation in curves is possibly caused by the distribution of forces during the application and removal of the load. In addition, ridges and ripples may affect the area where load is applied, as they may reduce contact area and therefore also the reaction force.

Figure 8 Source: Authors.

Table 2 presents values of maximum indentation deformation (h_max) obtained from the load vs. displacement curves (Fig. 8), as well as residual depth values (h) for each nanoimprint and the percentage of elastic recovery.

Table 2

Source: Authors.

Residual deformation of nanoimprints is variable; however, nanoimprints 2, 3 and 4 show the highest residual difference, with values of 17.49, 42.31 and 52.15 nm. This confirms that they exhibit the highest values of elastic recovery (Table 2), as shown in Fig. 6 where they do not exhibit an indentation mark. On the other hand, the nanoimprints with the most visible marks in indentation (Fig. 6) show the highest values of residual deformation.

Fig. 9 shows the residual deformation on nanoimprint 2 after application of load. Residual depth on all nanoimprints is obtained from the topographic profile, with a value of 17.49nm in nanoimprint 2, which displays a minimum amount of stacking on the sides but no sinking nonetheless. This is due to the characteristics of plasticity and elasticity of the material.

Figure 9 Source: Authors.

Application of the maximum indentation load (29,91uN) on all nanoimprints resulted in maximum deformations (hmax) of 1290 ± 60um of the GAK fibers (Table 2). When the load is removed, fibers recover from deformation by an amount ranging between 1160 and 1320 nm (Fig. 10). This demonstrates the elastic recovery of GAK fibers, which averaged 94.54%. Recovery occurs since the atoms in the fiber are not permanently displaced, with the force applied in the indentation being stored as a distortion of fiber interatomic bonds [43].

Figure 10 Source: Authors.

Table 3 shows the summary of average mechanical properties of GAK fibers resulting from instrumented nanoindentation.

Table 3

Source: Authors.