Powdered crystal violet (CV, C₂₅H₃₀N₃Cl, ACS reagent, ≥90.0 %) and pellet form sodium hydroxide (NaOH, ACS reagent, ≥97.0 %) were purchased from Sigma-Aldrich and were utilized without pretreatment. Cotton fabric samples procured from Arthur R. Johnson Co., Inc., Brooklyn, N.Y. were provided by the Fashion Institute of Technology (FIT, U.S.A). Deionized water (~20 mΩ/cm, Direct-Q3, Millipore Sigma) was used to make 1.0 × 10⁻⁴ M CV stock solution and 0.1, 0.5, and 1.0 M NaOH stock solutions. To prepare the CV-dyed cotton fabric samples, 1.5 cm × 1.5 cm squares of cotton fabric were placed in separate beakers containing 10 mL of 1.0 × 10⁻⁴ M CV solution. After ensuring that the pieces were fully submerged and covered by CV solution, the samples were left to soak overnight at room temperature. The CV-dyed samples were taken out of solution and placed into an oven overnight at ~60 °C. After removing the samples from the oven, 10 mL of the NaOH solutions were added to each dyed fabric sample. The fabric was kept in the NaOH solution overnight and removed the following day. The samples were placed in separate Petri dishes and again put in the oven to dry overnight at ~60 °C.

The infrared spectra were obtained with a Nicolet iS50 FTIR spectrometer (Thermo Fisher Scientific) equipped with an attenuated total reflectance (ATR) accessory. The samples were contacted with a diamond ATR crystal and the FTIR spectra (32 scans) were recorded between 400~4000 cm⁻¹ with a resolution of 4 cm⁻¹. Raman spectra were obtained using a Horiba-Jobin XploRA™ PLUS Confocal Raman Microscope equipped with a 785 nm laser. Before collecting the Raman spectra of the sample, the spectrometer was calibrated via silicon material using a 100X objective and a 1200 gr/mm grating. The Raman spectra were collected in the 300~1800 cm⁻¹ and 2800~3300 cm⁻¹ Raman shift regions. The acquisition time was 10 seconds and then each spectrum was accumulated from 60 scans. Crystallite structures of the as-received and treated samples were examined using XRD. The XRD patterns were collected with a Philips PW1729/APD3520 diffractometer with copper radiation (λ=1.54184 Å) in the 2θ range of 20~80 ° with a step size of 0.02 °. The surface topography and content were probed by SEM and EDS techniques, respectively. The SEM images were taken with LEO 1550 SFEG SEM (Zeiss) with an EDS detector from EDAX, and software/electronics from iXRFsystems. The samples were coated with nominally 6 nm of gold to reduce charging. Additionally, the samples were imaged using a Robinson backscatter detector (RBSD) and 20 keV electrons. Color measurements of the samples were performed using a ColorReader (Datacolor, U.S.A) and color values were compared by means of CIELAB parameters. The tensile strength, Young’s modulus, and toughness of the woven cotton muslin samples were measured using an Instron 5542 testing station. The diameter of the cotton yarn sample was 250~300 μm and the length of the sample was 150 mm. The extension rate was set at 2 mm/min.

To understand the decolorization of CV-dyed cotton fabric, digital photography and spectroscopic techniques, such as Fourier transform infrared and Raman spectroscopy, were applied. The cotton fabric as-received (Fig. 1(a)) had a yellow-like color. After the interaction between the positively charged CV (carbocation) and the negatively charged cotton fabric (OH−), the color changed from yellow to bluish-violet (Fig. 1(b)) [24, 28]. As shown in Fig. 1(c)–(e), the fading of the violet color increased with increasing NaOH concentration. The decolorization of CV in an alkaline (i.e., NaOH) aqueous media has been well documented. The rate of CV decolorization expressed as −r = k[CV][NaOH] where k is the rate constant [30]. Based on the rate equation, the CV decolorization reaction is first-order with respect to [NaOH]. It is worth noting that a bluish color was still evident in the NaOH-treated samples even at high NaOH concentrations (NaOH/CV = 1.0 × 10³ ~ 5.0 × 10³). Because CV was fully decolorized in the liquid phase reaction under similar experimental conditions (Fig. 1(f)), it is expected that a cellulose-CV interaction decreases the decolorization rate, suggesting that the bond between cellulose and CV is stronger than the interaction between water and CV. This finding has two potential explanations. There may be an attractive force between the positively charged CV and negatively charged surface of the fabric causing adsorption [31]. When treated with NaOH, if attractive forces were predominantly present then it would be expected that the violet color of CV would be present in solution, which was not observed. Alternatively, there may be hydrogen bonding between cellulose and CV, supported by a probable overlapping of C-H and O-H peaks in the FTIR spectra of Fig. 2 [32].

FTIR spectroscopy was utilized to examine the structural changes and the interaction of CV with NaOH. The IR spectra of the as-received and treated samples are displayed in Fig. 2. For reference, the IR spectrum of the solid-phase CV is displayed in Fig. S1(I). The full spectra in the range 500~4000 cm⁻¹ can be seen in Fig. 2(I). In the region of 500~1250 cm⁻¹ shown in Fig. 2(II), multiple peak assignments related to the structure of cellulose can be made. The peaks at 1160, 1054, and 900 cm⁻¹ are attributed to C-O-C asymmetric bridge stretching, C-O stretching vibrations, and β-glycosidic linkages found in cellulose I or C₁H deformation, respectively [33–38]. A strong peak is present in the samples treated with 0.5 M and 1.0 M NaOH at 880 cm⁻¹, which is due to the presence of the alkali [39]. Multiple assignments around 1106 cm⁻¹ have been reported related to the C-O-C glycosidic ether and antisymmetric stretching of the pyranose ring [37, 40].

As shown in Fig. 2(III, 1250~1750 cm⁻¹ wavenumber range), a small peak at ~1424 cm⁻¹ is present in samples (a) and (b), which becomes broader and increases in intensity in samples (c–e) as NaOH concentration increases. The peak in samples (a) and (b) is assigned to CH₂ symmetric bending of the ring structures in cellulose, while the same peak for samples (c–e) is due to NaOH (Fig. S2(I)) [39, 41]. The peak at 1588 cm⁻¹ is only present in sample (b) and indicates C=C stretching of an aromatic ring vibration in the aromatic nuclei of CV [42]. Please note that the IR spectrum of 0.1 M CV also shows a weak peak at 1588 cm⁻¹, while the peak is not visible in the 1.0 × 10⁻⁴ M CV IR spectrum due to the strong −OH bending mode peak at 1640 cm⁻¹ as shown in Fig. S1(I). To reduce the impact of −OH on the CV peak intensity, a solid phase CV sample was used, and the spectrum clearly reveals that the peak at 1588 cm⁻¹ is due to the CV. As shown in Fig. 2(III), after NaOH treatment, the 1588 cm⁻¹ peak disappeared due to surface coverage by NaOH.

In Fig. 2(IV), all samples displayed peaks at 2851 cm⁻¹ and 2896 cm⁻¹, suggesting −CH stretching, however sample (e) showed a lower intensity compared to samples (a)–(d) [35, 40]. There are peaks of similar intensities at 3275 cm⁻¹ and 3330 cm⁻¹ in all samples that are attributed to the O-H bond in the O(2)H-O(6) and O(3)H-O(5) inter- and intramolecular hydrogen bonds of cellulose [24, 39, 43, 44]. The inset in Fig. 2(IV) shows the absence of a peak that would have implied a transition from cellulose I to II. This result is encouraging because cellulose crystal transformation from I to II may occur at high NaOH concentrations (>3 M) [33, 39]. Based on the FTIR spectra analysis, it can be concluded that the cellulose structure of the cotton fabric was unaffected by both CV and NaOH under the experimental conditions.

As a complementary technique to the FTIR, Raman spectroscopy is also widely used to acquire structural information on organic/inorganic materials. The Raman spectra of the cotton fabric, CV-dyed fabric (control), and NaOH-treated samples are presented in Fig. 3. The cotton fabric spectrum contains strong bands at ~1098 cm⁻¹ and ~1122 cm⁻¹ which correspond to the asymmetric and symmetric C-O-C stretching band in a glycosidic linkage, respectively (Fig. 3(I)) [45]. These peaks were also shown on the control and NaOH-treated samples’ spectra. In addition to the same peak positions, the intensity ratio of these bands (I_1098cm-1/I_1122cm-1) was not changed much, indicating that the molecular structure of cellulose was not affected by CV dyeing and decolorization processing.

In the case of the control sample’s Raman spectrum (Fig. 3(I, b)), unique CV bands are shown in addition to the cotton fabric Raman band. The peak positions are well matched to band positions in the 0.1 M and solid CV Raman spectra as shown in Fig. S1(II): < 450 cm⁻¹ (C⁺-phenyl stretching and C-N-N mode), 900~1400 cm⁻¹ (ring skeletal vibration and C-H in-plane bending mode), 1500~1650 cm⁻¹ (C-H bending and C-C stretching mode in a ring structure) [46–50]. It is worthwhile to note that Raman spectroscopy has advantages over FTIR spectroscopy in identifying the CV species in aqueous solutions due to Raman spectroscopy’s insensitivity to water molecules (see the 0.1 M CV peaks in Fig. S1(I) and (II)) [51]. Fig. 3(I-c,d,e) shows Raman spectra for the 0.1, 0.5, and 1.0 M NaOH-treated control samples. The CV-related bands disappear after treatment with NaOH. A similar result was shown in the IR spectra (Fig. 2(III)). Furthermore, the Raman spectra of the NaOH-treated samples are very similar to the cotton fabric spectrum (Fig. 3(I-a)), indicating that most surface CV species disappeared or were covered by NaOH. The Raman spectra in the 2800~3300 cm⁻¹ region (Fig. 3(II)) shows a peak at ~2896 cm⁻¹, which is assigned to a typical C-H bond stretching mode and was not affected by CV and NaOH treatment [45, 52]. Note again that Raman spectroscopy is insensitive to the hydroxyl vibration, while IR spectra contains a broad/strong O-H band at ~3275 cm⁻¹ and ~3330 cm⁻¹ (Fig. 2 (IV)).

It has been reported that using surface-enhanced Raman scattering (SERS) technique with metal nanoparticles, such as gold or silver, could improve the quality of the CV Raman spectrum by decreasing the fluorescent background [53, 54]. It is worthwhile to comment that using a 785 nm laser can possibly avoid the fluorescence issues as shown in Fig. 3(I). Since the presence of CV was visually observed and color parameters indicated dye presence on each sample (Fig. 6 in the Colorimetric assays section), it is concluded that Raman spectroscopy has a limitation to detect CV concentrations below 10⁻⁴ M.

The crystalline structures of the as-received and CV/NaOH-treated samples were investigated using X-ray diffraction (XRD) techniques as shown in Fig. 4. For comparison purposes, all spectra were normalized by an aluminum substrate peak (65°). The characteristic peaks (15.1, 16.9, 22.9, and 34.8°), which are assigned to cellulose I, are observed in all samples. It has been reported that the cellulose I structure can transition to cellulose II, shown by 12° and 20° diffraction patterns, with treatment of high concentrations of NaOH [55]. Because the cellulose peak positions were not shifted, it could be concluded that the bulk material had not lost significant crystallinity during the treatment process. In the case of the 0.5 M and 1.0 M NaOH-treated samples, NaOH characteristic peaks are observed at 30.7, 35.6, 40.2, 41.9, 47.0, and 48.6°, and peak intensity of cellulose was not affected by NaOH treatment. To support that the cellulose structure was unaffected by CV/NaOH treatment, the crystallite size of cellulose was calculated using the Scherrer equation, D = 0.9λ/(B·cosθ), where λ is the wavelength (1.54184 Å) of the X-ray used in the experiment, B is the full width at half maximum (FWHM) of the cellulose I peak (22.9 °), and θ is the Bragg’s angle of the cellulose I peak (22.9 °). As shown in Fig. 4 (inset), the FWHM and crystallize size of cellulose were constant, confirming the consistency of the cellulose crystalline structure.

To characterize possible changes in fiber morphology and chemistry due to treatment, SEM images were collected in addition to EDS analysis. Fig. 5 shows low resolution images (a~d), high resolution images (a’~d’), and EDS results (a”~d”) of the cotton fabric, CV-dyed, and NaOH-treated samples. The mesh dimensions of the cotton fabric were ~300 μm × 300 μm and the fabric strands show a flat and smooth surface (Fig. 5(a’)). As expected, carbon (C) and oxygen (O) were dominant chemical components, both observed at ~0.25 and ~0.5 keV, respectively (Fig. 5(a”)). It was also noticed that the cotton fabric contained trace amounts of an impurity, potassium (K), as a weak peak of K shows at ~3.2 keV. Compared to the cotton fabric, the CV-dyed sample (Fig. 5(b’)) shows no distinct differences on neither the fabric surface nor in the chemical composition (Fig. 5(b”)), except for the disappearance of the K peak, indicating that the impurity was removed during the dying process. Although the mesh of the cotton fabric was not changed, the strand surfaces were all covered by CV. In the case of NaOH-treated samples, no significant changes (i.e., mesh dimension and deweaving) on fabric surfaces were noted (Fig. 5 (c–e)). However, it should be highlighted that surface CV had disappeared after interacting with 0.1 M NaOH (Fig. 5 (c’)), and that the strand’s surface became smooth and round. Furthermore, the EDS result provides the presence of sodium (Na), which is shown at ~1.1 keV (Fig. 5 (c”)). Because of the increasing NaOH concentration, the Na peak intensity also increased, and the presence of absorbed Na-containing deposits on strands were clear. Although the presence of NaOH is proved through FTIR (Fig. 2), Raman (Fig. 3), and XRD (Fig. 4), EDS shows the highest sensitivity for NaOH detecting: EDS (0.1 M NaOH) > FTIR and XRD (0.5 M NaOH) > Raman (1.0 M NaOH). Please note that Raman spectroscopy remained as the most sensitive technique for detecting CV. From the spectroscopic and microscopic results, it is concluded that the crystalline structure of cellulose (or cellulose I) was not affected by CV and NaOH, though the chemical compositions on the strands’ surfaces were varied.

In conjunction with vibrational spectroscopic analyses, which provided insight into functional groups present in the samples, a colorimeter was used to measure CIELAB coordinates. Data provided by FTIR and Raman spectroscopies showed that peaks associated with the presence of CV were not detected. Contrary to this data, a violet hue remained on all samples after treatment, indicating the presence of CV. The lack of detection of CV in the treated samples could be explained by a detection limit of both FTIR and Raman spectrometers. This would prevent detection of the reaction products and low CV concentrations, such as that remaining on the fabric. CIELAB coordinates are valuable for characterizing visible changes that may not be distinguishable using the aforementioned methods [56, 57]. These parameters were measured to compare the colorimetry data of CV-dyed and NaOH-treated samples to the fabric as-received (Table 1).

Optical images and the relationship between the L*, a*, and b* values of each sample are displayed in Fig. 6. It is expected that for the sample to be fully decolorized, all values must match those of the fabric as-received. All three parameters for comparing samples must be identical to conclude that they have an identical hue. As the concentration of NaOH was increased, the L* (brightness), a* (red-green coordinate), and b* (yellow-blue coordinate) values shifted towards those of the cotton fabric sample. For instance, it was observed that CV-dyed fabric treated with a higher NaOH concentration had a greater L* shift from the control sample L* value (38.16) to the cotton fabric L* value (84.88). Comparably, since the b* values are negative for the samples dyed with CV, this represents that yellow radiation was absorbed, indicating that these samples appear more blue than yellow [58]. $\mathit{\Delta }{E}_{ab}^{*}$ represents the distance between two color points in the CIELAB space [59]. These values were calculated using the formula $\mathrm{\Delta }{E}_{ab}^{*}=\sqrt{{(L_2^{*}-L_1^{*})}^2+{(a_2^{*}-a_1^{*})}^2+{(b_2^{*}-b_1^{*})}^2}$, with L*, a*, and b* from Table 1. The values decreased as the concentration of NaOH was increased, demonstrating that the violet color faded. However, since the lowest value (19.57) is not 0, it is evident that CV remained on the fabric.

To investigate the influence of the treatment on mechanical properties of the woven cotton muslin samples, tensile testing was conducted on as-received and CV/NaOH-treated cotton fibers. The stress transfer at the fiber interface can be observed by the stress-strain curve as shown in Fig. 7(A). When the load reached the ultimate tensile strength, the dropping of the stress-strain curve differed among the samples. The control and 0.1 M NaOH-treated samples show a moderate slope, indicating that the entangled fiber matrix did not break at the same time. Inversely, the slope of the stress-strain curve of the remaining samples showed is steep, suggesting that the fiber matrix broke immediately.

The calculated ultimate tensile strength is shown in Fig. 7(B). The tensile strength of the samples is similar to the work done by Herrera-Franco and Valadez-González [60]. The Young’s modulus and elongation at break of the samples are shown in Fig. 7(C–D). Evidently, the Young’s modulus of the treated samples increased with increasing NaOH concentration. Consistently, the samples treated with higher NaOH concentration showed lower elongation at break, indicating lower fracture toughness.