July 31, 2008
Crystalline Character of Agave Americana L. Fibers
By Oudiani, Asma El Chaabouni, Yassine; Msahli, Slah; Sakli, Faouzi
Abstract The crystalline character of Agave americana L. fibers and its relationship with mechanical properties were studied. Comparative investigations between three different types of fiber extraction were made: raw fibers, distilled water extracted fibers and seawater extracted fibers. A wide angle X-ray diffraction technique was used and cellulosic phases in the three samples were identified. A diffractometric analysis was applied to determine percent crystalline component, microcrystallite sizes, unit cell dimensions and monoclinic angle. The crystallinity of raw fibers was 51.2 % and dropped with extraction in water. The unit cell dimensions a, b and c for raw fibers were 8.9 [Angstrom], 10.3 [Angstrom] and 7.9 [Angstrom], respectively, and the monoclinic angle was 83.6[degrees]. When extracted in distilled water, the values of a and c slightly increased, whereas that of the monoclinic angle decreased. These variations were more pronounced for seawater extracted fibers. Our results demonstrated that differences in the molecular and fine structure caused different mechanical properties. High tenacity and initial modulus values were observed for raw fibers which presented the highest crystallinity. On the other hand, the low tenacity and great extensibility of seawater extracted fibers were attributable to their amorphous character and to the increased unit cell dimensions. Key words Agave americana L., crystallinity, extraction, mechanical properties, microcrystallite size, monoclinic angle, unit cell dimensions, X-rayAgave americana L. is the vegetation of Amaryllidaceae cultivated in North Africa and originated from Central America. Since 1998, the basic fibers of this plant presented the subject of many researches and seemed to be interesting due to their mechanical and physical properties. These natural cellulosic fibers were characterized by a high hydrophility, a low density, a high tenacity and a great extensibility in comparison with other textile fibers .
To explain some of these properties, a previous study on the fine structure of Agave americana L. fibers has been conducted, showing the individual fibers to have a helical structure composed of square shape spirals .
Nevertheless, no work appears to have been done on the crystalline character of these natural fibers using wide angle X- ray diffraction (WAXD) technique. In fact, X-ray diffractometric analysis yields data useful in deducing the arrangement of polymer molecules within cellulosic fibers, whose structure contains highly ordered crystalline domains interposed among amorphous zones, and in varying degree. In addition, crystal size and percent crystallinity have been linked to the physical and mechanical properties of fibers and subsequently fabrics.
In this paper, we report on the crystalline character of Agave amencana L. fibers and its relationship with mechanical properties by measuring the degree of crystallinity, the microcrystallite size and the unit cell dimensions of three different samples.
Materials and Methods
The leaves of Agave amencana L. present a composite structure with an organic matrix and a reinforcement composed of cellulose micro-fibers, which show different kinds of chemical bonding such as covalent, hydrogen or Van Der Waals bonds. The organic matrix is composed of several components including hemicelluloses, pectic matter, lignin and gums, whereas the reinforcing fibers are mainly composed of cellulose.
To extract fibers from the organic matrix, different methods can be used. These methods have a great influence on the fine structure of the obtained fibers. In this work, three different techniques for extracting fibers from the leaves resulting in three different samples were investigated.
i. Raw fibers: these fibers were manually extracted after cutting the leaf in right sections parallel to the fiber's axis. Special care was taken not to subject fibers to any treatment or solicitation. Therefore, the obtained fibers were left in their native state.
ii. Distilled water extracted fibers: leaves were submitted to a hydrolysis treatment in distilled water. Subsequently, fibers were separated from the matrix by calendering the leaf and then they were washed. A temperature of 120 [degrees]C and duration of 90 min appeared to be the best conditions for the treatment, as demonstrated in a previous study .
iii. Seawater extracted fibers: these fibers were obtained by placing leaves on the sea floor for a period of three months [I]. The natural matrix of the leaves was then split up, allowing fibers to be separated.
Whatever the method of extraction used, we obtained technical fibers presented as fibrous bundles, called bundles of ultimate fibers.
Tensile tests were carried out using a Lloyd dynamometer with a constant strain rate. The latter was adjusted to a speed of 250 mm/ min for a 500 mm specimen length to respect the test duration of 20 +- 3 s specified in NFG 07002 French Standard. Prior to testing, fibers were conditioned at standard conditions of 20 [degrees]C +- 2 [degrees]C and 65 % +- 2 % RH for 24 hours. Tests were conducted under the same standard conditions. Due to the high variability found in the tests, we chose a sample size of 50 fibers randomly selected from the same sheaf.
Bundles of technical fibers were aligned manually and lacerated with a scalpel. Short segments of fibers were then obtained and placed into a sliding disk mill. The mill comprised a means of randomly mixing and distributing the small fibers' lengths in various orientations at conditions of surface pressure far below those that would induce crystallite fracture. The sample mass was then placed into a standard holder within the stage of the diffractometer. The spectra were determined using a PANalytical X'Pert Pro MPD-spectrometer set at 45 kV and 40 mA. Samples were scanned from 2theta = 5[degrees] to 2theta = 40[degrees] with increments of 0.02[degrees] using the CuKa radiation (lambda = 1.54 [Angstrom]).
For the air scatter trace, the fibers were removed from the sample holder, the empty holder was reinserted and an X-ray trace was obtained under conditions similar to those for the fiber traces. Corrections for the array of intensities of Agave americana L. fibers were made for air scatter and polarization [3, 4] using the following formulae:
correction for the air scatter:
where (I)^sub 2theta^ is intensity of the fibers obtained by the spectrometer, (I^sub air^)^sub 2theta^ is intensity obtained for the air scatter and mt is optical density of the prepared samples ([mu]t = 1.2);
correction for the polarization effect:
The corrected array of intensities was replotted against the Bragg angles, and a background was simply drawn under the peaks.
In order to obtain maximum use of the X-ray diffraction data, each peak profile was fit to the Pearson VII function  of the form:
where I(x) is intensity, x is 2theta, A is profile amplitude, w is profile width, X^sub c^ is profile center and mu is profile shape factor. In this study, estimated values of X^sub c^, A, w and mu for each peak profile, together with the corrected intensity data points and the Bragg angles (2theta) were used as input values in a software resolution program (Origin 06). The best minimization between observed and calculated intensities occurs when the estimated parameters are well chosen.
Percent Crystalline Component
Measuring amorphous and crystalline components of textile fibers has been approached in many ways. Segal et al.  proposed an empirical method for estimating the degree of crystallinity based on the ratio of peak height measurements of the (002) reflection relative to that attributed to an amorphous component. Wakelin et al.  prepared standards of differing crystalline composition and subsequently calculated a "relative crystallinity index" in which sample crystallinity was compared to that of highly crystalline acid hydrolyzed cellulose and an amorphous ball milled cellulose.
The approach we applied in this work was that proposed by Hermans and Weidinger , who derived an "absolute crystallinity" measure. The method consisted of dividing X-ray scatter into contributions from amorphous and crystalline regions. Intensity minima were located graphically and the area beneath the curves integrated and compared. The percent crystallinity was expressed in the following manner:
where I^sub t^ and I^sub a^ are the integral scattering intensities corresponding to the amounts of total and amorphous parts, respectively.
The lattice spacing d^sub hkl^ was determined for each profile from the resolved position X^sub c^ and by using the Bragg equation :
nlambda = 2dsmtheta (5)
where n is integer, lambda is wavelength of the X-rays, theta is angle of diffraction (2theta = X^sub c^) and d is the distance between the atomic layers (lattice spacing).
Unit Cell Dimensions and Monoclinic Angle
The crystal structure of native cellulose is known as cellulose I and has a monoclinic unit cell with the dimensions shown in Figure 1 [9, 10]. The cellulose molecule consists of a series of glucose rings linked together by valency bonds and situated in the ab plane (Figure 1). Cellulosic chains are arranged parallel to the b axis, which presents fiber axis. Hydroxyl groups of these macromolecular chains are joined laterally by means of hydrogen bonds, as shown in Figure 2 [9,10]. X-ray spectra of Agave americana L. fibers provided fiber reflections associated with four different orders, only one of which yielded an axial dimension by a single-step calculation. The b- axial dimension was determined directly from:
b = k d^sub 0k0^ (6)
where k is Miller indices (0k0).
The composite product of the a-axial dimension and the monoclinic angle was:
where 1 is Miller indices (001).
The monoclinic angle was found from:
Having found beta, we determined a and c from the respective composite formulae .
The crystallite size normal to the hkl plane was calculated from the integral breadth B of the peak according to Scherrer's equation :
where D^sub [perpendicular]hkl^ is the crystallite dimension in the direction perpendicular to the crystallographic plane hkl, theta is the Bragg angle, lambda is the wavelength of the radiation and K is constant (K = 0.9) (although this factor is sensitive to crystal type) [13, 14].
Results and Discussion
X-ray diffraction traces with corrected intensities are shown in Figure 3. The diffractograms of the three samples showed a pattern quite similar to the four peaks characteristic of native cellulose. These peaks were nearly located at 2theta = 15.5[degrees], 16.5[degrees], 22[degrees] and 34.5[degrees]; which are the positions of the (101), (101), (002) and (040) crystallographic plane reflections, respectively [3, 4, 13,15].
The two peaks at 2theta= 15[degrees] and 30[degrees], occurring on all traces with differences in intensity, were attributable to the inclusion of residual oxidized lignin or pectin in the (101) plane . In the same way, the two peaks at 2theta = 12.5[degrees] and 25[degrees] could be attributed to the same inclusion of noncellulosic matter in the (002) plane. In this study, we excluded crystalline scattering from unidentified phases of non-cellulosic origin from the contribution to the total cellulose crystalline scattering.
All the diffractograms showed that Agave americana L. fibers examined using X-ray diffraction exhibited considerable overlap of the diffraction peaks. Thus, the peak resolution was the only valid method for obtaining reliable parameters to measure crystallinity and apparent crystallite size.
Figures 4-6 represent the profiles resolved into four peaks (101, 101, 002 and 040) and the calculated intensities obtained by summing these peaks. Details of the significant parameters of Pearson VII function are given in Table 1.
In Figures 4(b), 5(b) and 6(b), we can see the great difference between observed and calculated intensities. The best agreement, however, was accomplished when we included additional fifth peak parameters in the peak resolution program. These parameters are presented in Table 2.
The typical output plots after resolving into five peaks are illustrated in Figure 7. From this figure and Table 2, one can note that there was no great difference in the extra peak position among the three samples. However, the difference lay in the intensity of this peak, which increased with extraction in water.
The addition of the extra peak at the order of 2theta = 20[degrees] corroborated the results found by Khalifa et al. , who studied the crystalline character of native and chemically treated Saudi Arabian cotton fibers. These researchers attributed the extra peak to the liquid-like scattering of cellulosic material not in the crystalline registry. They explained that such results may be due to the existence of paracrystalline deformation that occurred during the growth condition, processing and chemical treatment.
Hosemann and Bagchi , in their paracrystalline theory, demonstrated how imperfection in crystal lattice gives a background scatter in addition to the Bragg reflection of the crystal. Herman and Weidinger [18, 19] attributed the background scatter to the amorphous regions.
Considering the theory proposed by Hindeleh and Johnson , who explained the background scatter as a summation of Hosemann and Bagchi's and Herman and Weidinger's background scatters, the fifth peak we added could be taken to represent the scattering by the amorphous material. Therefore, its area was added to the background scatter area. Such an addition caused some drop in the degree of crystallinity of Agave americana L. fiber.
Table 3 shows the degree of percent crystallinity and the corresponding background scatter for the three samples of Agave fibers. Note that the percent crystallinity varied with the type of extraction. When extracted in water, the fiber crystalline mass decreased relative to that of raw fiber, which was not immersed. The decrease was more pronounced for seawater extracted fibers (20 %), which had been exposed in water for a long period, in comparison with distilled water extracted fibers.
Such a drop in percent crystallinity for cellulosic fibers immersed in an aqueous medium has also been observed by Foreman and Jakes , who studied crystalline character of cellulosic marine textiles. They noted that this decrease was due to the accessibility of amorphous regions which were more readily attacked by degrading forces, disrupting the overall order. Moreover, as degradation occurred, the polymer chains decreased in length and their mobility increased, once again resulting in greater randomness within the fiber.
Unit Cell Dimensions and Crystallite Sizes
Table 4 exhibits results for interplanar spacings dhk!, unit cell dimensions (a, b, c and beta) and crystallite sizes of the three samples of Agave fibers. It is clear from this table that crystallographic parameters were greatly affected by the type of extraction. The unit cell parameters of raw fibers (a = 8.9 [Angstrom], b = 10.3 [Angstrom], c = 7.9 [Angstrom] and beta = 83.6[degrees]) closely matched the values reported in the literature [9, 10]. For distilled water extracted fibers, the value of a increased to 9.1 [Angstrom] and that of the monoclinic angle beta decreased to 80.6[degrees]. However, when extracted in seawater, unit cell parameters were more affected, and we noted an increase of a and c values which reached 9.5 [Angstrom] and 8.2 [Angstrom], respectively, and subsequently a significant decrease of the monoclinic angle beta to 71.3[degrees].
From Table 4, it is also clear that the length of repeating unit in cellulose I had not been altered with the type of extraction (b = 10.3 [Angstrom]).
The microcrystallite sizes D^sub [perpendicular]hkl^ of the samples that were immersed in water were larger than those of raw fibers not immersed. This increase in the crystallite size may be attributed to the conditions of extraction in aqueous solution. As water penetrates less ordered areas of the fiber, these areas are damaged by the action of degrading forces. Therefore, rearrangement of associated chains can occur, resulting in an increased crystallite size [13, 21]. On the other hand, the hydrolytic degradation results in a decrease in polymer chains' length, and consequently, an increase in their mobility; encouraging the formation of more ordered crystallite regions .
Perel  noted an increase in crystallite size with increasing moisture content in the sample. He stated that this growth could be explained as "amorphous regions at the surface of or near the crystallites can then move so as to adjust their orientation and location to fit those of the crystallites".
The inter-reticular distance dhu indicates the spacing between diffraction planes of the crystallite unit cell. Therefore, the number of these planes making up the crystallites can be deduced by the ratios between crystallite sizes D^sub [perpendicular]hkl^ and the interplanar spacings d^sub hkl^ .
where N^sub hkl^ is the number of Miller planes hkl making up the crystallite.
Table 5 gives the number of hkl diffraction planes of the crystallites for the three samples of Agave americana L. fibers. One can note from this table that N^sub 101^ did not change with the type of extraction, whereas other plane numbers N^sub 101^, N^sub 002^ and N^sub 040^ were affected with different proportions. These numbers increased with extraction in water. A possible explanation is that 101, (002) and (040) planes were more or less easily exfoliated under the action of water, and the forces between these kinds of planes were affected.
The tensile mechanical properties (stress, strain and initial modulus) of the three samples of Agave americana L. fibers are given in Table 6. The values obtained for the mechanical properties of distilled water extracted fibers were close to those of raw fibers, which exhibited the highest tenacity and the lowest extensibility. The tenacity decreased from raw fibers (sigma = 34.5 cN/tex) to distilled water extracted fibers (sigma = 30.2 cN/tex), and then to seawater extracted fibers (sigma = 28.2 cN/tex). The extensibility increased significantly by 250 % from raw fibers (epsilon = 14 %) to seawater extracted fibers (epsilon = 49 %).
The crystalline character of Agave americana L. fibers revealed their mechanical properties, as seen in Figures 8 and 9. The most important factors influencing the tenacity and the elongation were the degree of crystallinity and the unit cell dimensions. The crystallite sizes had little influence on the mechanical properties. A high degree of crystallinity and low unit cell dimensions enabled raw fibers to achieve high tenacity and low extensibility values, though they presented short crystallites.
In fact, when a force is applied to a fiber, extension occurs for two reasons: i) a slight stretching of the chain molecules themselves and ii) a straightening of the molecules in the non- crystalline regions, with a resultant straining of the hydrogen- bond cross-links between them . Thus, the magnitude of distortion of cellulose molecules and cross-links depends on the spacing between neighboring chains represented by the unit cell parameters a and c. When these parameters have high values, cellulose molecules are more distant and the fiber becomes more extensible and less strong (seawater extracted fibers). On the other hand, the proportion of ordered regions (degree of crystallinity) is also responsible for the fiber mechanical properties. In crystalline regions, the molecules are firmly fixed and stable, holding the structure in place and allowing the fiber to be stronger and less extensible (raw fibers). Conclusions
The work presented in this paper reported on the crystalline character of Agave americana L. fibers. Three different methods for extracting fibers from the leaf were investigated: raw fibers manually extracted, seawater extracted fibers and distilled water extracted fibers. The main differences in structure, and consequently in fiber mechanical properties, originated from variations in extracting methods.
Extracting in water decreased the crystallinity of fibers and increased both the crystallite sizes and the unit cell parameters a and c. Thus, although the crystallites of fibers immersed in water grew in size relative to raw fibers not immersed, the overall crystalline component decreased. A few crystallites reordered as a consequence of extraction in water, but these same causes created greater disorder overall.
The first difference that had a definite influence on mechanical properties was the degree of crystallinity. This parameter was 51.2 % for raw fibers and decreased to 50.1 % for distilled water extracted fibers, then to 41.2 % for seawater extracted fibers.
Moreover, the mechanical properties were influenced by the unit cell parameters. Values of a and c were 8.9 [Angstrom] and 7.9 [Angstrom], respectively, for raw fibers, and increased with extraction in water to reach 9.5 [Angstrom] and 8.2 [Angstrom], respectively, for seawater extracted fibers. The high tenacity and low extensibility values of raw fibers could be explained by a high degree of crystallinity and low unit cell dimensions. On the opposite side, the decreased tenacity and increased extensibility of seawater extracted fibers were attributable to their low degree of crystallinity and high unit cell parameters. The crystallite sizes had negligible influence on the mechanical properties.
1. Msahli, S., Etude du Potentiel Textile des Fibres d'Agave americana L., PhD Thesis, No. [degrees]02Mulh0691, Universite de Haute Alsace, Mulhouse, France (2002).
2. Chaabouni, Y., Drean, J. Y., Msahli, S., and Sakli, F., Morphological Characterization of Individual Fiber of Agave americana L., Textile Res. J. 76(5), 367-374 (2006).
3. Khalifa, B. A., Abdel-Zaher, N., and Shoukr, F. S., Crystalline Character of Native and Chemically Treated Saudi Arabian Cotton Fibers, Textile Res. J. 61(10), 602-608 (1991).
4. Sotton, M., Arniaud, A. M., and Rabourdin, C., Etude par Diffractomerie des Rayons X des Parametres de Cristallinite et de Desordre dans les Fibres Textiles, Bull. Sci. ITF 7(27), 265-290 (1978).
5. Kreze, T., and Malej, S., Structural Characteristics of New and Conventional Regenerated Cellulosic Fibers, Textile Res. J. 73(8), 675-684 (2003).
6. Segal, L., Creely, J. J., Martin, A. E., and Conrad, C. M., An Empirical Method for Estimating the Degree of Crystallinity of Native Cellulose Using the X-ray Diffractometer, Textile Res. J. 29, 786-794 (1959).
7. Wakelin, J. H., Virgin, H. S., and Crystal, E., Development and Comparison of Two X-ray Methods for Determining the Crystallinity of Cotton Cellulose, J. Appl. Phys. 30,1654-1662 (1959).
8. Hermans, P. H., and Weidinger, A., Quantitative X-ray Investigations on the Crystallinity of Cellulose Fibers, a Background Analysis, J. Appl. Phys. 19,491-506 (1948).
9. Morton, W. E., and Hearle, J. W. S., "Physical Properties of Textile Fibers," 2nd Edn., The Textile Institute, Heinemann, London, England (1975).
10. Champetier, G., and Monnerie, L., "Introduction a la Chimie Macromoleculaire," Masson et Cie, Paris, France (1969).
11. Foreman, D. W., and Jakes, K. A., X-ray Diffractometric Measurement of Microcrystallite Size, Unit Cell Dimensions, and Crystallinity: Application to Cellulosic Marine Textiles, Textile Res. J. 63, 455-464 (1993).
12. Scherrer, P., Bestimmung der Grosse und inneren Struktur von Kolloidteilchen mittels Rontgenstrahlen, Nach. Ges. Wiss. Gottingen 2, 8-100 (1918).
13. Davidson, T. C., Newman, R. H., and Ryan, M. J., Variations in the Fiber Repeat Between Samples of Cellulose I from Different Sources, Carbohydr. Res. 339, 2889-2893 (2004).
14. Murdock, C. C., The Form of the X-ray Diffraction Bands for Regular Crystals of Colloidal Size, Phys. Rev. 35, 8-23 (1930).
15. Ouajai, S., and Shanks, R. A., Composition, Structure and Thermal Degradation of Hemp Cellulose After Chemical Treatments, Polym. Degrad. Stab. 89(2), 327-335 (2005).
16. Keshk, S., Suwinarti, W., and Sameshima, K., Physicochemical Characterization of Different Treatment Sequences on Kenaf Bast Fiber, Carbohydr. Polym. 65(2), 202-206 (2006).
17. Hosemann, R., and Bagchi, S. N., "Direct Analysis of Diffraction by Matter," North Holland, Amsterdam, Netherlands (1962).
18. Hermans, P. H., and Weidinger, A., X-ray Studies on the Crystallinity of Cellulose, J. Polym. Sci. 4, 135-164 (1949).
19. Hermans, P. H., and Weidinger, A., Quantitative X-ray Investigation on the Crystallinity of Cellulose Fibers, J. Appl. Phys. 13, 491-509 (1948).
20. Hindeleh, A. M., and Johnson, D. J., Crystallinity and Crystallite Size Measurement in Polyamide and Polyester Fibers, Polymer 19, 27-32 (1978).
21. Scallan, A. M., A Quantitative Picture of the Fringed Micellar Model of Cellulose, Textile Res. J. 41, 41-47 (1971).
22. Atalla, R. H., The Crystallinity of Cellulosic Fibers, in "Preservation of Paper and Textiles of Historic and Artistic Value II," Advances in Chemistry Series, No. 193, J. C. Williams, Ed., American Chemical Society, Washington DC, USA, pp. 69-176 (1981).
23. Perel, J., An X-ray Study of Regain-dependent Deformations in Cotton Crystallites, J. TextileInst. 81, 241-244 (1990).
Asma El Oudiani1, Yassine Chaabouni, Slah Msahli and Faouzi Sakli
Unite de Recherches Textiles, ISET, Avenue Ali Soua,
5070 Ksar Hellal, Tunisia
1 Corresponding author: Av Haj AIi Soua, BP 68. Ksar-Hellal, Ksar Hlel, 5070, Tunisia. Tel: 00 216 73 475; fax: 00 216 73 475 163; e- mail: [email protected]
Copyright Textile Research Institute Jul 2008
(c) 2008 Textile Research Journal. Provided by ProQuest Information and Learning. All rights Reserved.