Acid and alkali chemical treatments on synthetic and natural cellulosic, fibroin and keratin-based fabrics: study of structural changes

Abstract The characterization of fabrics after applying different degradation conditions appears to be a key factor to understand their behaviour for different applications. Depending on the treatment, morphologic and polymorphic changes may arise in the polymeric chain structure that composes textile fibers. In this paper, a systematic and comparative study between different types of natural and synthetic textile fabrics was carried out in order to shed light on the structural changes occurring under acid and alkali treatments. The natural fabric polymers studied had a cellulose-based composition from plants (cotton, linen and jute), or a protein-based one from animals (silk and wool). The synthetic fabric was polyester. Different treatment times were assessed: 15 min, 1, 2, 4, 8 and 24 h and compared to the initial condition of the fabrics (standard). In the cellulosic fabrics, the alkali provoked the polymorphic transformation of the amorphous CI into CII. However, the acid treatment did not cause any remarkable alteration in the structure. In protein fabrics, the acid treatment increased the amorphicity of the structures, leading to further degradation but not affecting the conformation of proteins, and the alkali dissolved the fabrics. Polyester was not affected by the chemical solutions (neither acid nor alkali).


Introduction
Textile fibers can be classified into two main groups according to the International Organization for Standardization (in ISO/ TR 11827:2012 Textiles-Composition testing-Identification of fibers): natural fibers, and man-made fibers which are divided into artificial and synthetic ones (Treilles et al., 2020). Natural fibers are biopolymers that are either plant-or animalsourced. Artificial fibers are composed of modified biopolymers in a physico-chemical process. Thus, they generally derive from natural materials (Treilles et al., 2020). Synthetic fibers come from petroleum resources and they are obtained through different chemical processes (Treilles et al., 2020). In this article, we dealt with natural and synthetic fibers.
On this basis, textile fibers are composed of different types of polymers: cellulose in the case of the plant-sourced or vegetal fibers (cotton, linen and jute); fibroin and keratin proteins for the animal-sourced ones (silk and wool), within the natural ones; and polyester among the synthetic fibers.
Silk fibers are mainly composed of the fibroin protein polymer being glycine, alanine and serine the main amino acids present (Wei et al., 2012). Silk I and silk II have been reported as polymorphs of silk fibroin. Silk II has an antiparallel ß-sheet structure while silk I is thought to consist of a combination of a-helix, ß-turns and random coils (Wei et al., 2012). Wool has a very complex structure. It is composed of a keratin protein polymer with crystalline structure which may have a-helix or ß-sheet structure (Barani et al., 2018;Morton & Hearle, 2008). a-helix has been reported to be the predominant conformation in wool native state (Barani et al., 2018).
Polyester is the most produced synthetic fiber for textile purposes and the most popular fiber, second to cotton (Chinta & Singh, 2012). It is a polymer formed by the condensation of terephthalic acid and ethylene glycol (Eicchorn et al., 2009).
Industry converts textile fibers into fabrics for widespread uses like fashion, sports, medicine, security, agriculture, etc. Textile fabrics often undergo chemical treatments during their industrial processing to improve or change some of their properties. The characterization of fabrics after applying different conditions is a key factor to understand their behaviour for their different applications (Teli & Terega, 2022;Yazdanbakhsh & Rashidi, 2020;Zhu et al., 2021). Some studies have been carried out to deal with the structural changes occurring during the solubilisation of cellulose in NaOH solution, which entailed swelling (Budtova & Navard, 2016;Isogai & Atalla, 1998;Kamide et al., 1992;Zhu et al., 2021). Morphologic but also polymorphic changes may occur from CI to CII polymorphs if strong alkali treatments are applied (Faruk et al., 2012). On the other hand, different acid hydrolysis conditions have also led to different polymorphs of cellulose (CI, CII and CIII) (Gong et al., 2017;Mahmud et al., 2019). A mixture of CI and CII has been obtained with H 2 SO 4 (50% vol/vol) after 1 h of acid hydrolysis. Furthermore, HCl (37% vol/vol) produced CI, and H 3 PO 4 (85% vol/vol) produced CII (Gong et al., 2017;Mahmud et al., 2019). Cotton fabrics have been described to be submitted to a NaOH treatment under tension called mercerization trying to improve its lustre and dyability (Kafle et al., 2014). Mechanical properties have been improved for jute-based composite materials when treated with alkali NaOH solutions (Kafle et al., 2014). Other authors have also dealt with the effects of chemical treatments on structures and polymorphic changes in silk (Gore et al., 2019) and wool. In this sense, the treatment of silk fibroin with ethanol (EtOH) and methanol (MeOH) has increased its crystallinity and presence of ß-sheets (Puerta et al., 2020). Silk fibroin in formic acid solutions has mainly exhibited ß-sheet structure (Asakura et al., 2001;Gore et al., 2019). The partial break of molecule chains in wool due to alkali solution has also been studied (Chen & Burns, 2006;Zhao et al., 2015). Polyester has been described to have good resistance to acid and oxidizing agents (Chinta & Singh, 2012).
Little research has been performed regarding the structure of textile fabrics instead of raw and/or natural fibers. The influence of acid and alkali treatments on the structure of fabric materials, which can be employed in the same condition as they are acquired, has been scarcely investigated. In this work, a systematic and comparative study between different types of textile fabrics (not individual fibers) was carried out trying to shed light on some unresolved aspects related to the structural changes of these polymeric materials. Chemical treatments in acid and alkali media over natural vegetal (cotton, linen and jute based on cellulose polymers), natural animal (silk and wool based on fibroin and keratin polymers) and synthetic fabrics (based on polyester polymers) were largely studied.

Textile fibers
The samples studied consisted on cotton, linen, jute, silk, wool and polyester pure textile fabrics. They were acquired in a specialised fabric store located in Pamplona (Spain) called Petachos. Fabrics were selected according to minimal industrial process (no dyes and minimal industrial finishing). Before the treatments, they were lightly washed with neutral detergent to remove sizing residues. Finally, they were rinsed with deionised water several times. Fabrics were cut into 4 Â 4 cm squared pieces.

Experimental setup
A set of twelve beakers (0.5 L) were placed on a multi-point magnetic stirrer (Thermo Scientific Cimarec i Poly) settled at 400 rpm ( Figure S1). The acid treatment was carried out in six beakers containing the six different fabrics, and the alkali treatment in the other six. The composition of the beakers was as follows (Table 1): Acid treatment: 15% (wt/wt) H 2 SO 4 solutions containing a 1:10 (wt/wt) fabric pieces to solution ratio (Table S1). Alkali treatment: 15% (wt/wt) NaOH solutions containing a 1:10 (wt/wt) fabric pieces to solution ratio (Table S1).
The fabric pieces were extracted of the solutions at different times: 4, 8 and 24 h, washed with deionised water and dried for 24 h in an oven set at 40 C in order to remove humidity from the solid pieces. In this way, three treated samples (corresponding to the three sampling times) were analysed for each of the six fabrics. Thus, a total of eighteen treated samples were analysed from the acid treatment and another eighteen from the alkali one, along with their corresponding standard untreated fabrics. All the samples were characterized by using the techniques described in the following section.

Characterization techniques
In the case of X-ray diffraction (XRD), the diffractometer used was a Bruker (Massachusetts, USA) D-8 Advance ECO with Cu Ka radiation and a LINXEYE XE-T scintillation detector. The experiments covered a 2h range from 5 to 25 , with a step size of 0.02 and 1 s by step. The crystallinity index (CrI) of the cellulose fabrics was estimated by using (Eq. (1)) (Segal et al., 1959): where I 200 is the maximum intensity of the peak corresponding to the (200) plane at 2h ¼ 22.6 and I am at 2h ¼ 18 represents the amorphous cellulose contribution (Segal et al., 1959). The fabric samples (4 Â 4 cm 2 ) were laid on the holder without any further manipulation. Calculations were carried out with the diffractometer software DIFFRAC.EVA V4.3. The spectrometer Shimadzu (Kyoto, Japan) IRAffinity-1S with a Golden Gate ATR sampling accessory was employed for the Fourier-Transform infrared spectroscopy (FTIR) measurements. The following parameters were established in the LabSolutions IR measurement software: Happ-Genzel apodization, 100 number of scans per sample, 4 cm À1 resolution and a 600-4000 cm À1 wavelength range. The ratio of absorbances at 1335 and 1316 cm À1 (Eq. (2)) was used as an indicator of the amorphous content of CI in lignocellulosic fibers (Colom et al., 2003).
The ratio of absorbances at 1622 and 1514 cm À1 corresponding to Amide I and Amide II, respectively (Eq. (3)) was used as an indicator of peptide bond hydrolysis in animal fibers (Geba et al., 2014).
The square samples of the fabrics were measured by placing them on the holder of the ATR accessory, assuring a good contact between the samples and the crystal. OMNIC 10 software was employed for the calculations.
The model of the thermogravimetric (TG) and differential thermal (DTA) analyser employed was sDTA/TGA 851-Mettler-Toledo (Greifensee, Switzerland). The temperature program increased from 25 C to 1000 C at a constant rate of 10 C/min in a static air atmosphere. The crucibles were made of Al 2 O 3 and had a 70 lL capacity. To perform these experiments, small pieces of a few milligrams were cut from the square fabric samples. The software employed was STAR SW 9.20.
The morphological changes in fibers were recorded by a PCE-MM200 Microscope. It was set to a 63.2Â magnification.

Cellulosic fiber fabrics
In the raw samples (standard) of cotton, linen and jute fabrics, the characteristic peaks of CI were observed by XRD (Figures 1a-c). The most intense peaks of CI were those at 14.7 , 16.5 and 22.6 corresponding to the (1-10), (110) and (200) planes, respectively, and a shoulder at 20.6 was also observed (Gong et al., 2017;Kafle et al., 2014;Mahmud et al., 2019). Fabric samples immersed in the H 2 SO 4 solution did not seem to be modified regarding the presence of CI; XRD diagrams of the highest value of time considered (24 h) are shown in Figures 1a-c. By contrast, the treatment with NaOH led to a complete polymorphic change from CI to CII in cotton at 24 h of treatment. It followed from the absence of the XRD peaks corresponding to CI in the 24 h treated cotton sample. Instead peaks at 2h ¼ 19.9 and 21.6 , which were assigned to the planes (110) and (020) of the CII polymorph appeared (Gong et al., 2017;Kafle et al., 2014;Mahmud et al., 2019). Both peaks were already observed in 8 h-treated cotton samples. The same result had been previously reported by other authors (Kafle et al., 2014;Mahmud et al., 2019). It is expected that crystallographic properties of these natural polymers change when the hydrolysis media are different. In the case of linen, there was no polymorphic modification, and only CI was Natural Vegetal Cotton -15% (wt/wt) NaOH 8 Linen -15% (wt/wt) NaOH 9 Jute -15% (wt/wt) NaOH 10 Animal Silk -15% (wt/wt) NaOH 11 Wool -15% (wt/wt) NaOH 12 Synthetic Polyester -15% (wt/wt) NaOH observed in alkali-treated samples (Figure 1b). For jute, the peak at 19.9 was also observed but only at 24 h. Thus, in jute, both polymorphs, CI and CII were present at 24 h alkali-treated samples (Figure 1c). Figure 2 shows the tendency of the crystallinity index CrI (see Eq. (1)) over time for the different fabrics when immersed in the acid and alkali solutions. The treatment with NaOH provoked an initial marked decrease in the crystallinity of cotton (from 93.5 to 82.8%), linen (92.5 to 84.3%) and jute (73.5 to 60.4%) after 8 h. The decrease was milder after 8 h and for jute even increasing up to 24 h. CrI for 24 h treated cotton could not be computed due to the absence of the XRD peaks corresponding to the CI polymorph (Eq. (1), Figure 1a). The decrease in CrI was negligible for H 2 SO 4 treatment ( Figure 2).
In cotton and linen (Figure 3a and b), the H 2 SO 4 treatment over 24 h caused a relative increase of the band at ca. 3331 cm À1 , assigned to intramolecular H-bonds, with respect to that at ca. 3290 cm À1 , which corresponded to inter-and intramolecular H-bonds (Duchemin, 2015). This fact indicated that intermolecular H-bonds were partially destroyed with the acid treatment. This behaviour was less prominent for jute but also occurred (Figure 3c).
The peak clearly appearing at around 1735 cm À1 in untreated (standards) cotton and jute standard samples corresponded to the C ¼ O stretching bond of uronic acid contained in hemicellulose xylan (Mahato et al., 2013). This peak disappeared with NaOH treatment at 4 h (Figure 3d and f). This removal was not achieved with the acid treatment after 24 h (Figure 3d and f) and could be one of the keys to the greater effects of NaOH than H 2 SO 4 on cellulose-based fabrics aforementioned in the XRD results.
NaOH treatment over 24 h caused a strong increase in the peak intensities around 3000-3500 cm À1 for cotton, linen and jute (Figure 3a-c), and also in the signal at 1639 cm À1 (Figure  3d-f). They corresponded to adsorbed water O-H stretching and bending bonds, respectively (Abidi et al., 2010;Duchemin, 2015;Zhu et al., 2021). The amplitude of the signals (clearly wider bands) and the shift towards higher wavenumbers could be assigned to the presence of amorphous material and to the presence of CII (Duchemin, 2015), respectively. This could be explained by the higher effectivity of amorphous cellulose in retaining water because of the greater availability of its hydroxyl groups for establishing hydrogen bonds with water (Abidi et al., 2010). Besides, the hydroxymethyl groups in CI and CII are in different conformations which could possibly  affect the dehydration and depolymerization processes (Wan et al., 2017). The aforementioned shift was mainly observed for cotton and linen at the latter both wavenumbers (Figure 3a-e). Thus, the presence of CII in linen after alkali treatment was suggested thanks to the infrared (IR) study.
The phenomena described in the alkali medium were not observed with H 2 SO 4 treatment (Figure 3), most likely because hemicellulose was not removed (the peak at 1735 cm À1 remained unchanged, Figure 3d and f). Cellulose crystallinity was less affected in this medium ( Figure 2) and the polymorphic transformation of CI into CII did not occur (Figure 1). The results obtained by infrared spectroscopy matched with those found by XRD.
Taking a closer look to the wavenumber region around 900 cm À1 in the spectra of the three lignocellulosic fabrics treated with NaOH solution (Figure 3g-i), the peak at ca. 897 cm À1 was assigned to CI (Yang et al., 2017) and a new peak appeared at 864-878 cm À1 . This had also been found in a previous paper and ascribed to the transformation of CI into CII (Wan et al., 2017). Originally only CI peak (ca. 897 cm À1 ) was observed. For 4 h-and 8 h-treated samples, peaks of both CI and CII were observed. The new peak started to appear at 4 h of treatment and became a very defined peak at 8 h (Figure 3g-i). These observations could arise from a conformational change from CI into CII, as beforementioned. This result indicated that the first steps of the polymorphic transformation of these polymers could happen at very early times (4 h).
The increase in the amorphous content of CI (see Eq. (2)) when NaOH treatment was carried out is clearly shown in Figure 4. The amorphicity results were in agreement with the crystallinity index (CrI) results obtained by XRD (Figure 2). The main increase occurred before 8 h treatment and even the final amorphicity decrease in the 24 h NaOH-treated jute sample agreed with data from XRD ( Figure 2). The results did not clearly change in the amorphous content of CI when fabrics were submerged in acid solution (Figure 4).
Regarding the thermal analyses, mainly water was lost in the 0-120 C temperature interval ( Figure 5). The fabrics treated with the alkali solution suffered a marked initial mass loss in this range (Figure 5d-f) which was minor in their corresponding standard and acid-treated samples (Figure 5a-c). This higher mass loss in alkali-treated samples could have derived from the higher amount of adsorbed water, due to the presence of amorphous material and/or CII polymorph as previously described in the infrared spectra study.
The mass loss occurring at ca. 360 C for cotton, linen and jute standard samples ( Figure 5) was related to cellulose (C) decomposition as it occurred over a narrow temperature range and corresponded to over a 90% of the sample's mass. The shoulder to the left occurring along with this loss in most H 2 SO 4 treated samples (Figure 5a-c) corresponded to the loss of hemicellulose (Hc) (Subhedar & Gogate, 2014). This shoulder was absent in the NaOH-treated samples because hemicellulose had already been removed at early times (4 h) by the solution (Figure 5d-f) as concluded by the infrared studies.
As the hours of treatment increased, temperature was shifted towards lower values which meant lower thermal stability, even though these values of temperature were not directly dependent of the time of treatment due to the heterogeneity of the fabric samples ( Figure 5). By comparing the temperature of decomposition of cellulose (C) in the standard and also after 24 h treatment (Table 2), the highest shift was produced in jute for both acid and alkali treatments, also showing the minor temperature of decomposition for C. This result could be related to the more amorphous character of this fabric, as concluded by XRD and IR. The mass loss of hemicellulose seemed to be linked to the mass loss of cellulose as they were both shifted together (Figure 5a-c). This result suggested a closely related cellulose and hemicellulose structure and supported the idea of the removal of hemicellulose being the triggering event of the action of NaOH on cellulose as aforementioned.
The micrographs of cotton, linen and jute fabrics after 24 h of treatment clearly show the higher degradation degree  of the polymeric fibers with NaOH treatment ( Figure S2). The cotton fabric was almost completely frayed. In the case of jute, the weft and warp of the fabrics appeared separated and the fibers were observed isolated.

Animal fiber fabrics
The standard silk fabric showed XRD reflections (Figure 6a) because it is mostly made up of fibroin fibers which contain ordered polymeric peptide chains (Asakura et al., 2001;Gore et al., 2019;Puerta et al., 2020). The diffraction peak at 2h ¼ 20.2 corresponded to the amorphous phase of the silk I structure and those at 16.7 and 22.7 corresponded to the antiparallel ß-sheet structure of silk II (Rowell et al., 2005;Subhedar & Gogate, 2014). Thus, both conformations (I and II) were visible for standard samples and also for H 2 SO 4 treated samples after 4, 8 and 24 h. This showed the absence of structural modifications upon acid treatment, even though the signal of the amorphous phase was higher at 24 h (Figure 6a). Silk fabrics were completely dissolved after 4 h of alkali treatment and the tendency of conformational change could not be properly followed. In other media, such as MeOH and EtOH (in saturated vapor), the enrichment of crystalline structure has been assigned to the transformation of amorphous random coil to ß-sheet structure (Puerta et al., 2020).
Wool showed the most complex structure with a-helix and b-sheet conformations. The standard sample presented a defined peak at 2h ¼ 22.4 which could be related to the ß-sheet structure (Figure 6b) (Abdelghaffar et al., 2018;Gore et al., 2019;Puerta et al., 2020;Wang et al., 2016). Its intensity with respect to the amorphous background decreased with acid treatment time, indicating lower crystallinity along time. The broadband starting at 2h ¼ 15 indicated the presence of a disordered region. The very weak peak at ca. 2h ¼ 10 may arise from an a-helix structure (Wang et al., 2016). a-helix has been reported to be the predominant conformation in the wool native state (Barani et al., 2018). Wool fabrics after 4 h of alkali treatment could not be studied due to the almost complete dissolution of the samples. In this sense, both animal fiber fabrics were destroyed by the alkali solution employed in our experiments.   Amide I, II and III bonds corresponding to the characteristic peptide bonds of proteins were found in the animal fiber fabrics in the wavelength region within 1700-1200 cm À1 : fibroin in silk ( Figure 7a) and keratin in wool (Figure 7b). Amide I signal appeared in the range 1700-1600 cm À1 , amide II in 1600-1500 cm À1 and amide III in 1350-1200 cm À1 (Puerta et al., 2020). The peak appearing at 1625 cm À1 for silk and at 1633 cm À1 for wool was indicative of ß-sheet structure (Abdelghaffar et al., 2018;Jaramillo-Quiceno et al., 2017). Similarly to the IR study, the b-sheet conformation was also observed by XRD for all the samples studied ( Figure 6). The very weak shoulder at around 1730 cm À1 appearing for both protein fabrics after 24 h of acid treatment could be related to amide bonds belonging to disordered structures that appeared after long treatment times.
For both silk and wool, the ratio between Amide I and Amide II bands (Eq. (3)) corresponding to H 2 SO 4 treated samples increased with time, as shown in Figure 8. An increase in this ratio was the result of the higher degree of hydrolysis of peptide bonds (Geba et al., 2014), showing the effect of the acid treatment.
The breakage of the peptide bonds could be related to a higher degree of amorphousness, as previously described by XRD.
The mass losses occurring between 280 and 500 C ( Figure 9a) and between 220 and 550 C (Figure 9b) corresponded to the denaturation of silk and wool proteins, respectively (Cheng et al., 2018;Subhedar & Gogate, 2014). Thermal curve profiles, similarly to XRD patterns, suggested that the hydrolysis of polymeric peptide chains due to the acid treatment did not significantly affect the conformation of proteins. Nevertheless, treatment over 24 h with H 2 SO 4 led to the appearance of a mass loss around 210 C in silk and to the shift of the main peak towards lower temperature (the shift was ca. 18 ). For wool, the treatment provoked the appearance of a shoulder around 270 C. These effects could be related to the higher hydrolysis occurring at the longest treatment times and the presence of less bonded protein chains.
The micrographs of silk and wool after 24 h of treatment did not show appreciable degradation of the polymeric fibers with the acid treatment ( Figure S3). For both protein fabrics, the samples disappeared when they were introduced in the alkali solutions.

Synthetic fiber fabrics
The main XRD peaks of polyester were observed at 17.5 and 23.1 . The crystalline structure of the polyester fabric was neither affected by the acid (Figure 10a) nor by the alkali (Figure 10b) solution. IR spectra suggested the same as no modifications were observed upon treatments. The peak at 1715 cm À1 corresponded to the stretching vibration of C ¼ O in the ester group, and those at 1338, 1239, 1093 and 1015 cm À1 to C-O stretching in ester or carboxylic acid (Figure 10c   due to depolymerization which consists in the cleavage of bonds. DTG curves agreed with XRD and IR results. The micrographs of the polyester fabrics after 24 h of treatment did not show any signs of degradation of the polymeric fibers after the acid and alkali treatments ( Figure S4).

Conclusions
Based on the proposed objectives and the experimental setup, the following conclusions could be drawn regarding the structure of the textile fibers made of polymeric chains: The cellulose-based fabrics, cotton, linen and jute did not suffer remarkable changes upon the acid treatment, probably, because the acid solution was unable of removing hemicellulose, which is closely packed with cellulose. The alkali treatment removed hemicellulose and caused a significant initial increase in the cellulose I amorphous content and eventually the transformation of the amorphous cellulose I into cellulose II. The alkali treatment also led to increased water retention because of the increase in the amorphous content. This effect was more marked after 24 h of treatment, probably because the cellulose II structure had already been established with a higher capacity of establishing hydrogen bonds with water molecules than cellulose I. The protein-based fabrics -from animals-, silk and wool, were insoluble in an acid medium and completely soluble in alkali at 24 h. The acid treatment did not much affect the conformation of proteins. However, amide bond signals and diffraction reflections belonging to disordered structures Figure 10. XRD patterns (a, b), infrared spectra (c, d) and differential thermogravimetry curves (e, f) corresponding to polyester for the acid (a, c, e) and alkali (b, d, f) treatments at different times (4, 8, 24 h).
increased with time in solution for both. Regarding the synthetic fabric, the structure of polyester was not affected by the acid nor by the alkali solution.

Disclosure statement
No potential conflict of interest was reported by the authors.