Ubim Fiber (Geonoma baculífera): A Less Known Brazilian Amazon Natural Fiber for Engineering Applications

Abstract

The production of synthetic materials generally uses non-renewable forms of energy, which are highly polluting. This is driving the search for natural materials that offer properties similar to synthetic ones. In particular, the use of natural lignocellulosic fibers (NLFs) has been investigated since the end of 20th century, and is emerging strongly as an alternative to replace synthetic components and reinforce composite materials for engineering applications. NLFs stand out in general as they are biodegradable, non-polluting, have comparatively less CO₂ emission and are more economically viable. Furthermore, they are lighter and cheaper than synthetic fibers, and are a possible replacement as composite reinforcement with similar mechanical properties. In the present work, a less known NLF from the Amazon region, the ubim fiber (Geonoma bacculifera), was for the first time physically characterized by X-ray diffraction (XRD). Fiber density was statistically analyzed by the Weibull method. Using both the geometric method and the Archimedes’ technique, it was found that ubim fiber has one of the lowest densities, 0.70–0.73 g/cm³, for NLFs already reported in the literature. Excluding the porosity, however, the absolute density measured by pycnometry was relatively higher. In addition, the crystallinity index, of 83%, microfibril angle, of 7.42–7.49°, and ubim fiber microstructure of lumen and channel pores were also characterized by scanning electron microscopy. These preliminary results indicate a promising application of ubim fiber as eco-friendly reinforcement of civil construction composite material.

1. Introduction

Composites made of synthetic materials have become a class widely applied in various technological sectors [1]. However, their production is associated with a relatively high consumption of energy and non-degradable wastes. Indeed, synthetic materials not only demand greater manufacturing energy but also produce after service life wastes that are highly polluting and cause serious environmental damage [2]. The search for new materials that are environmentally friendly, sustainable and offer similar mechanical properties at a lower cost, has been arousing the interest of industry and researchers since the end of 20th century, to replace synthetic components by cellulose- based ones [3].

Natural fibers and related fabrics are known to be a renewable resource, and have gained considerable interest in different sectors of the industry, from the production of ropes, handcrafts and textiles, to automobile and construction industries [4,5,6,7]. In recent decades, the search for new natural lignocellulosic fibers (NLFs) as reinforcement in composite materials has increased, as reported in several review papers [8,9,10,11,12]. Consequently, these NLFs are currently applied to reinforce polymeric composites, used in construction materials, food packaging, and automotive parts [13,14,15]. Indeed NLFs such as jute, coir, bamboo, hemp and sisal have been reinforcing polymer composites as eco-friendly civil construction materials for panels, doors, roofing tiles and even low cost buildings [15]. In particular, NLFs are currently being considered as part of multilayered ballistic armor [16] for personnel protection against being shot by a rifle with 7.62 mm ammunition.

NLFs stand out in general as they are biodegradable, non-polluting, have a low energy consumption and are more economically viable. In addition to being lighter than synthetic fibers, they present similar chemical and mechanical properties [7]. The properties of NLFs vary according to their chemical composition, diameter, arrangement of constituents in the fiber, degree of polymerization, crystalline fraction of cellulose, part of the plant organism (stem, leaf, root, seed, among others) and cultivation aspects (age, climatic conditions, degradation processes) [17].

The NLFs extracted from the stem have good quality, mainly due to the presence of crystalline cellulose with a high degree of polymerization [18]. The structure of NLFs is basically composed of two different phases: one crystalline and the other amorphous [3,17,18]. Cellulose is the main chemical constituent of these fibers and is responsible for stiffness. A higher cellulose content in the fiber represents better mechanical properties [19].

Most of the crystalline structure of an NLF is restricted to the cellulose. In fact, cellulose is found not to be uniformly crystalline. However, as indicated by Eichhorn et al. [20], the cellulose ordered regions are extensively distributed throughout the material. According to these authors, the amorphous part of an NLF is composed of hemicellulose, as well as lignin, pectin and waxes. As for the crystalline part, the cellulose consists of a threadlike structure called the microfibril. Although the individual densities of cellulose, hemicellulose and lignin might be ~1.60 g/cm³ [21,22,23], the density of an NLF measured by the geometric method (mass/volume) or the Archimedes’ scale is found to be lower due to the fiber’s porosity. Notably, the cellulose films were reported as having densities of 1.49–1.51 g/cm³ [24].

Cellulose microfibril, as the most crystalline phase in an NLF, is known as the reinforcement phase incorporated into lignin, which is considered to be an amorphous matrix. The cell wall of a fiber is not a homogeneous membrane: it presents an order of cells composed of distinct walls, with different helicoid angles between the fiber axis and the microfibrils. The strength and ductility of cellulosic fibers depend on the angle, at which these microfibrils are curled in relation to the fiber axis, which is also referred to as the microfibrillar angle (MFA) [25]. The primary wall is the first layer where cellulose fibrils are deposited during cell development. Since fibril orientation is more oblique, MFAs are randomly oriented exhibiting varying degrees of alignment. The primary layer involves a secondary layer, which in turn is composed of three layers S1, S2 and S3. The S1 and S3 layers exhibit microfibrils in a smooth helical orientation, whereas the S2 layer is the thickest on the cell wall and has axial orientation. The orientation of these microfibrils is the main responsible for the fiber’s tensile strength, it can be said that a low MFA implies superior mechanical properties [26].

Among the most diverse NLFs with potential application as polymer composite reinforcement, are those from the Amazon region [27]. A fiber extracted from the ubim plant found in the Amazon, Brazil, is not well known, has been scientifically denominated Geonoma bacculifera, and is a promising reinforcement for composites. In fact, ubim fiber is not mentioned in a study on Brazilian natural fibers with potential for reinforcing polymer composites used in engineering applications [28]. The ubim fiber is extracted from a plant with multiple stems, which is smooth with elongated leaves and unbranched. The plant has a height varying between one to four meters and diameter from one to three centimeters as well as erect or partially crawling stipe, with seven to twelve leaves, little branched inflorescences and globose or ovoid fruits [29]. Ubim fibers have been used in the north of Brazil in hut covers, boat awnings, and rain protection lining. The ubim stem is used in fishing corral walls [30].

Some basic characteristics show that ubim fiber appears as a promising NLF for reinforcing polymeric matrix composites. Thus, the present study aims to investigate the ubim fiber through density, X-ray diffraction (XRD), microfibrillar angle, crystallinity index and scanning electron microscopy. These characteristics may be relevant for composites manufacturing, evaluating their potential for future applications as low cost civil construction materials.

2. Materials and Methods

The ubim fiber (Figure 1a) was extracted from the stem of the plant, scientifically called Geonoma baculifera, purchased in Belém do Pará, Brazil. The stem of the ubim plant was divided into longitudinal splints, which are then manually separated into finer fibers.

Figure 1 illustrates the ubim plant (a), as well as the split splints of the stalk (b) and final split fibers (c).

The characterization of the ubim fiber, in terms of density, microfibrillar angle and degree of crystallinity was preliminarily evaluated. The fiber was shredded and subsequently measured for the dimensions and length of the fiber cross-section.

2.1. Characteristic Density

2.1.1. Geometric Method

To obtain the average diameter, each fiber was divided into five distinct points uniformly separated, and then rotated by 90°. A new reading was carried out at the same five previous points. A total of 3 measurements were taken at each point, thus obtaining a statistical average. The equipment used was an Olympus microscope, model BX53M with 5x magnification.

The length of each fiber was measured using a 0.01 mm precision caliper, and its mass was weighed on a Gehaka electronic scale, model AG-200, with 0.0001 g precision. The fiber cross-section was considered to be ellipsoidal in shape, and its approximate area calculated according to Equation (1), where a and b represent the major and minor axes of the ellipsoid, respectively:

$$A=\frac{\pi ab}{4}$$

(1)

Density results for the five intervals were statistically evaluated using Weibull Analysis (WA) software. This statistical analysis is based on the cumulative Weibull frequency distribution function (F(x)).

$$\mathrm{F}\left(\mathrm{x}\right)=exp\left[-{\left(\frac{x}{\theta }\right)}^{\beta }\right]$$

(2)

where x is measured density, β is the Weibull modulus and θ is the scaling parameter. Weibull modulus (β) is associated with data uniformity, θ has precision adjustment given by R². Equation (2) can be modified to fit a linear graph:

$$\ln \left[\ln \left(\frac{1}{F\left(x\right)}\right)\right]=\beta \ln x-\left(\beta \ln \theta \right)$$

(3)

2.1.2. Archimedes’ Principle Method

In addition to the density calculated by geometric method, a density evaluation was also carried out using Archimedes’ principle (buoyancy). In the Archimedes’ test, fiber samples were weighed in air and then subjected to an immersion liquid with a density lower than that of the sample, according to ASTM D3800 [31]. Gasoline was used as the immersion liquid. Density was calculated using the formula indicated in the Equation (4):

$$\rho =\frac{\left(M_3-M_1\right)}{\left(\left(M_3-M_1\right)-\left(M_4-M_2\right)\right)}{\rho }_l$$

(4)

where ρ_l is the density of liquid, as well as M₁ being the weight of the suspension wires in air, M₂ is the weight of the suspension wire in liquid, M₃ is the weight of the suspension wire with fiber in air and M₄ is the weight of the suspension wire with fiber in liquid.

2.1.3. Gas Pycnometry Density

The gas pycnometry was performed by Accupyc 1330 brand Micromeritics, following the procedures established by the ASTM D4892 [32].

2.2. XRD Analysis

X-ray diffraction (XRD) analysis was performed using a PANaytical diffractometer model XPert Pro MRD system, as well as a Pixcel detector and cobald anode with Co K radiation (0.1789 nm). XRD is an effective method of quantifying the cellulose present in the fiber in both its crystalline and amorphous form.

2.2.1. Microfibrillar Angle (MFA)

The microfibril angle (MFA) was determined in association with the cellulose peak (0 0 2), according to the methodology described by Cave [33]. Cave’s methodology is the most accurate method for calculating the MFA. Based on XRD, it suggests that the estimation of MFA occurs from the relationship between three curves obtained from the peak associated with the highest intensity, and namely it is Gaussian along with the first-order and second-order derivatives. Through this analysis, a “T” parameter is obtained and applied in the following polynomial Equation (5).

$$\mathrm{MFA}=-12.19\ast T^3+113.67\ast T^2-348.40\ast T+358.09$$

(5)

2.2.2. Crystallinity Index (CI)

The calculation of the crystallinity index (CI) by XRD was applied using the method described by Segal et al. [34]. The maximum intensity of the peaks (1 0 1) and (0 0 2) was used. These peaks are associated with the amorphous and crystalline phases of the fibers. The CI value was determined using Equation (6) where, I₁ is the intensity of the diffraction minimum (amorphous region) and I₂ is the intensity of the diffraction maximum (crystalline region).

$$\mathrm{CI}=1-\frac{I_1}{I_2}\ast 100$$

(6)

2.3. SEM Analysis

The ubim fiber surface morphology was analyzed by scanning electron microscopy (SEM) in a model Quanta FEG 250 Fei microscope (Hillsboro, OR, USA) operating with secondary electrons at 5 kV.

3. Results

3.1. Density Measurements

3.1.1. Geometric Method

The variation in the cross sections, given by the mean values of a and b, is shown in the histogram of Figure 2, with percentage frequency related to five arbitrary intervals.

The mean frequency of the cross-sectional dimensions for the investigated ubim fibers (Figure 2) reveals an almost normal distribution within the range of 510 to 620 µm. For each interval, the mean density was measured by dividing the mass by the volume before being statistically analyzed by the Weibull method.

Table 1. Density of ubim fibers calculated for different intervals of cross-section dimension.

Diameter Range (µm)	ρ (g/cm3)	Standard Deviation	Β	θ	R2
510–532	0.955	0.06	11.800	0.990	1.000
532–554	0.815	0.02	23.900	0.831	0.960
554–576	0.700	0.03	21.168	0.716	0.990
576–598	0.607	0.03	16.762	0.623	1.000
598–620	0.440	0.02	24.171	0.676	1.000
ρmean	0.703

presents the mean ubim fiber density values for each interval of the mean dimensions of the cross-sections shown in Figure 2. In this table, a trend of greater density for thinner fibers should be noted. Weibull statistical analysis was performed for densities measured by the geometric method. Table 1 presents the Weibull parameters provided by the WA software associated with the shape parameter, scale parameter and precision fit (R²). The quality of these measurements can be evaluated using the R² parameter; the closer to 1, the more reliable the data obtained.

In Table 1 it is important to note that thinner ubim fibers, with lower diameters, display significantly higher density, while the thicker ones are much less dense. As compared to the cellulose, hemicellulose and lignin densities, ~1.6 g/cm³ [21,23], the ubim fibers revealed an amount of porosity that increases markedly with the fiber diameter.

3.1.2. Archimedes’ Technique

The density value was also evaluated by the Archimedes’ technique and found as 0.73 ± 0.02 g/cm³. In both cases, ubim fiber density values (0.70 g/cm³ for geometric and 0.73 g/cm³ for Archimedes’ technique) are among the lowest reported for NLFs and most commonly applied in reinforcing polymer composites such as flax, jute, sisal and cotton [7].

Using the Archimedes’ technique it was not possible to precisely identify the difference in densities between thinner and thicker ubim fibers. This was due to the eventual penetration of gasoline (immersion liquid) into the fiber’s open porosity. Nevertheless, the obtained value of 0.73 g/cm³ is close to that found by the geometric method, 0.70 g/cm³, which in practice attests the mean value of ubim fiber density of 0.72 g/cm³.

3.1.3. Gas Pycnometry Results

The absolute density obtained by gas pycnometry of the ubim fibers was 1.86 ± 0.26 g/cm³. This relatively higher density found by the pycnometer is due to the ability of the helium gas to fill all the open pores of the material. The gaseous pycnometry provides the value of the absolute density, including all empty parts of the NLF, and as it is very porous, this higher value is justified. This higher density result by pycnometry compared to other methods was also found by Oliveira et al. [35], as 1.61 g/cm³.

Higher values for the absolute density, such as that of ubim fiber obtained by pycnometry of the order of ~1.6 g/cm³, are justified by the density of plain cellulose [21,22,23]. Most reported densities of NLFs [36,37], including cellulose films [24], are ~1.50 and used for technological application such as polymer composite reinforcement.

The significant amount of porosity in the ubim fiber is responsible for a lower apparent density, 0.72 g/cm³, which is good for a lighter composite reinforcement in spite of its relatively higher absolute density of 1.86 ± 0.26 g/cm³.

3.2. XRD Results

Figure 3 shows XRD patterns of samples associated with different related ubim fiber conditions: (a) vertical stalk (VS); (b) vertical fiber (VF); (c) horizontal stalk (HS); and (d) horizontal fiber (HF). It should be noted that in both conditions (a) and (b), the peak associated with (2 0 0) is the prominent one. This peak is related to the cellulose crystalline part of the fiber and was the peak used to assess MFA [38]. As for condition (c) and (d), the same pattern is not repeated and it was not possible to calculate the corresponding MFA.

A preliminary evaluation of the percentage of crystalline cellulose by Rietveld revealed values of around 66% for conditions (a) and (b).

3.2.1. MFA Results

The calculated MFA for the VS and VF corresponded to 7.49° and 7.42°, respectively. An average value of 7.46° can be considered similar to some MFA values reported for other NLFs such as ramie and hemp [3,28,39,40].

3.2.2. CI Results

The degree of crystallinity of cellulose is an important structural parameter of NLFs. The stiffness of plant fibers is mainly due to the relatively large amount of crystalline regions in the cellulose. Figure 4 shows the convoluted peaks obtained from background free XRD patterns to calculate the crystallinity index (CI) according to Equation (6). In this figure the crystallographic planes (2 0 0) and (1 1 0) are associated with the crystalline and amorphous phases of the ubim stalk (VS) and fiber (VF), respectively. As for VS in Figure 4a, the crystalline peak at 2θ = 26.3° has an intensity I₂ = 402.1 a.u, while the amorphous peak at 2θ = 11.8° with I₁ = 149.7 a.u allowing to obtain a CI = 63% by Equation (6). Similarly for VF, Figure 4b revealed a crystalline peak at 2θ = 25.6° with I₂ = 694.5 a.u and an amorphous peak at 2θ = 10.4° with I₁ = 116.6 a.u, giving CI = 83%, which will be considered the proper one for ubim fiber.

These CIs are within the range of values reported for several common NLFs such as sisal [41], hemp [40], jute [42,43], flax [44], and coir [28], used as reinforcement in polymer composites, which are shown in

Table 2. Density, cellulose content, crystallinity index and microfibrillar angle of ubim fibers as compared to other NLFs applied as polymer composite reinforcement.

NLFs	Density (g/cm3)	Cellulose (%)	Crystallinity Index (%)	Microfibrillar Angle (◦)	Reference
Ubim	0.70	66	83.0	7.46	PW
Ramie	1.50	85	75.0	7.50	[17,28,39]
Hemp	1.50	72	87.8	7.50	[7,40]
Sisal	1.50	75	72.2	20.0	[7,17,28,41]
Coir (coconut)	1.40	53	44.0	51.0	[7,28]
Curaua	0.92	71	75.6	18.8	[43,45]
Jute	1.45	60	71.3	8.00	[7,40,42,43]
Flax	1.38	71	63.1	10.0	[7,44]
Banana	1.50	65	39	11.0	[7,37,46]
Pineapple	1.60	83	38	14.0	[7,37,38]
Bamboo	1.21	45	60	6.0	[7,37,38]
Sugarcane Bagasse	0.49	69	45	14.5	[7,37,38]
Kenaf	1.50	72	72	4.1	[37,46]
Cotton	1.60	51	65	25	[7,36,46]

PW: Present Work.

. In addition, this table also presents the density, cellulose content, crystallinity index and microfibrillar angle of less known NLFs such as curauá [43,45], banana [7,37,46], pineapple [7,37,38], bamboo [7,37,38] and sugarcane bagasse [7,37,38]. Also included in Table 2 were cotton [7,36,46] and kenaf [37,46], with relatively higher densities and CI.

A close look at the values of density, cellulose content, IC and MFA in Table 2 fails to reveal any direct relationship between the density with either the cellulose content or the CI for the presented NLFs. One possible reason is that the reported densities, like the case of ubim fiber in Table 2, are the apparent ones measured by geometric or Archimedes’ methods. Indeed, the apparent density considers the amount of porosity, which is strongly influenced by the internal fiber microstructure (lumen and channels), and the characteristics of the fiber species. On the other hand, cellulose content, IC and MFA correspond to another structural hierarchy, which is more associated with the fiber strength. As such, in several cases a lower cellulose content might be related to higher MFA and weaker fiber like in the coir, banana and sugarcane bagasse fibers. To the knowledge of the authors these relationships have never been fully investigated and are the subject of future research.

3.2.3. SEM Results

The SEM analysis disclosed the morphologies of the cross-section and surface of the ubim fiber as shown in Figure 5. The analyzed characteristics reveal the existence of an irregular rough surface, which can contribute to the mechanical interlock between the fiber and the matrix. In Figure 5a,b, one should notice this heterogeneous surface.

The measurement of the diameter of a random fiber is also shown in Figure 5a at a magnification (300×). Figure 5c–e show, with higher magnification, details of the cross section of the ubim fiber.

One should see in Figure 5f a polygonal lumen channel, which is associated with a greater amount of porosity. These characteristics are similar to those observed by other authors [47,48], such as rough surface, defects, flaws and irregularities along the length of the fiber.

4. Summary and Conclusions

A preliminary investigation on the physical properties of the ubim fibers extracted from the stem of the Brazilian Amazon plant, scientifically known as Geonoma baculifera, revealed characteristics supporting a possible reinforcement of novel epoxy composites;

• Densities measured by the geometric method from 0.955 to 0.440 g/cm³ and analyzed by the Weibull statistical method revealed a tendency to decrease with increasing fiber cross-section dimensions, from 510 to 620 µm. After using the Archimedes’ technique, the density showed a mean value of 0.73 g/cm³, quite similar to the 0.70 g/cm³ that was found after using the geometric method, which are among the lowest reported so far for natural lignocellulosic fibers. By the gas pycnometry method the density was found as 1.86 ± 0.26 g/cm³ due to the exclusion of the porosity by the helium gas, but consistent with other NLF and pure cellulose with ~1.60 g/cm³.
• Microfibril angles from 7.42 to 7.49° were found by X-ray diffraction for different conditions of ubim stalk and fiber. A crystallinity index of 83% was evaluated by XRD in ubim fiber powder. This CI is within the range of reported values for several NLFs.
• These preliminary physical and microstructural characterizations performed for the ubim fiber, in association with a lower cost, indicate a promising application as reinforcement of polymer composites to be used as eco-friendly civil construction materials.