Sustainable valorization of poultry waste through optimized keratin extraction and its prospective use in dye adsorption

Abstract

Chicken feathers (CFs) are an important by-product of the poultry industry, accounting for 4–6% of the total weight of the chicken. CFs pose serious environmental problems because of traditional disposal methods like incineration and landfilling. CFs are rich in keratin and can be used as a biopolymer for various applications, such as wastewater treatment. The present study optimized keratin extraction by Box-Behnken Design (BBD), evaluating the effects of temperature, time, and reducing agent concentration to maximize the yield. Characterization via Scanning Electron Microscopy (SEM), X-ray Diffraction (XRD), Fourier Transform Infrared Spectroscopy (FTIR), and Thermogravimetric analysis (TGA) confirmed preserved protein structure, semicrystalline nature and thermal stability. Keratin beads (KBs) were fabricated for the removal of Congo Red (CR) from aqueous solutions. Batch adsorption studies investigated pH, adsorbent dosage, contact time, and initial dye concentration, achieving removal efficiencies of ~80%. Kinetics followed the pseudo-second-order model (R² = 0.9974), indicating the primary role of chemisorption. Equilibrium data fitted the Freundlich isotherm better than the Langmuir (R² = 0.9991), which implies that the process is heterogeneous multilayer adsorption, while the Langmuir model estimated a maximal adsorption capacity of 81.3 mg/g. This study supports the United Nations Sustainable Development Goals (SDGs), particularly SDG 6, SDG 12 and SDG 13, by offering a scalable, eco-friendly solution for wastewater treatment while contributing to sustainable waste management practices.

Graphical Abstract

Supplementary information

The online version contains supplementary material available at 10.1186/s13065-026-01766-5.

Introduction

The growing global demand for meat and eggs has resulted in approximately 74 billion chickens being slaughtered annually worldwide [1]. Over the past three decades, the demand for chicken as a main source of protein has increased by nearly 70% [2, 3]. India’s industrial poultry production increased by 9% annually between 2000 and 2019, indicating the rising consumption driven by urbanization and changing food preferences [4]. However, this rapid expansion has led to the generation of vast quantities of poultry waste, with chicken feathers (CFs) being the most prevalent solid byproduct [5, 6]. A single chicken produces about 125 g of feathers, accounting for ~ 4–6% of its total body weight [5]. Despite their abundance, CFs often remain underutilized, with limited applications in animal feed, packaging, and ornamentation [7]. Improper management of CFs can exacerbate environmental issues such as fowl cholera, chlorosis and soil degradation, highlighting the need for sustainable disposal solutions [8].

Conventional disposal methods, such as incineration and landfilling, are harmful to the environment [7]. Incineration is energy-intensive and produces carbon emissions and toxic byproducts, while landfilling poses challenges in storage, odor and handling [9, 10]. Alternative disposal methods, such as composting, are time-consuming and require thorough veterinary inspection [11]. Therefore, valorization of CFs into value-added materials aligns with the principles of the circular bioeconomy and supports the Sustainable Development Goals (SDGs) (SDG 6, 12 and 13).

CFs are composed primarily of keratin, a fibrous structural protein rich in sulfur (S), containing cystine residues that form disulfide bonds. These bonds provide keratin with exceptional chemical, thermal and mechanical stability, but also make it insoluble in most solvents [12–19]. Keratin’s intrinsic biocompatibility, biodegradability, and abundance of reactive functional groups (-NH2, -COOH, -SH) make it ideal for environmental remediation applications, especially as a natural adsorbent [20–28]. However, the extraction of keratin from CFs is still a challenge because of its dense disulfide network, and numerous methods, such as steam explosion, oxidation, alkaline hydrolysis, and enzymatic treatments, often compromise structural integrity and yield [29–32]. Among these, alkaline hydrolysis using sodium sulfide (Na₂S) has been recognized as a cost-effective and efficient method, as it preserves the secondary structure of keratin while avoiding damage to amino acids [29, 33–35]. However, the majority of research relies on empirical, single-factor methods without statistically optimizing extraction parameters like temperature, reaction time, and Na₂S concentration, which results in uneven yields and a lack of insight into factor interactions.

Response Surface Methodology (RSM) provides a strong statistical approach for multivariable optimization, enabling the assessment of both primary and secondary effects of process parameters on extraction efficiency. In addition to increasing extraction yield, a methodical optimization can retain important functional groups, like carboxyl, amide and thiol moieties, which affect the chemical reactivity and adsorption potential of keratin. Keratin’s functional groups make it a promising candidate for the removal of dyes from wastewater [36–38]. Several keratin-based composites have shown effective adsorption of cationic dyes like methylene blue and crystal violet. Conversely, anionic azo dyes such as Congo red (CR) remain challenging to remove because of their complicated aromatic structure, strong solubility, and environmental stability. Industries such as textiles, paper, and leather release dyes into water bodies, causing severe environmental and health issues [39–42]. Conventional wastewater treatment methods have low efficiency in the complete removal of dyes, whereas adsorption has emerged as a desirable method due to its simplicity, cost-effectiveness, and high efficiency in removing a wide range of contaminants [15, 40, 43–48]. Despite the number of studies conducted on keratin-based composites, there is limited information on the application of extracted keratin as the primary adsorbent in the removal of these dyes [49–53].

Recent advances in biopolymer-based adsorbents, including keratin-cellulose, keratin-chitosan, and keratin-graphene composites, show an increasing interest in environmentally benign adsorbent technologies [36, 39, 42]. However, in most studies, keratin extraction and adsorbent development are treated as independent processes. Consequently, limited research has been performed about the simultaneous optimization of keratin recovery and its prompt integration into functional adsorption materials.

This research contributes to SDG 6 (Clean Water), SDG 12 (Responsible Consumption and Production) and SDG 13 (Climate Action) by transforming waste poultry feathers into keratin-based adsorbents. The study exemplifies circular economy strategies by converting an abundant biologically sourced material waste into a value-added biopolymer. Similar strategies have been reported by Majeed and co-workers in 2024, where agricultural by-products were converted into protein-based catalysts [54]. The statistical optimization approaches demonstrate contemporary methods for intelligent process design and sustainable material production [55]. In addition, the use of RSM in this research is indicative of data-driven and intelligent process optimization design, which mirrors the current momentum towards AI-assisted sustainable manufacturing. Recent trends in eco-friendly material design, including sustainable textile processing and biomacromolecule utilization, underscore the importance of applying green chemistry principles with intelligent design methods to mitigate the pollution and environmental impacts of industries [56, 57].

In the present study, the optimisation of Na₂S-mediated keratin extraction using RSM is directly combined with the development of keratin beads (KBs) for CR adsorption. The structural integrity of the keratin extracted under optimal conditions was evaluated using scanning electron microscopy (SEM), X-ray diffraction (XRD), Fourier transform infrared spectroscopy (FTIR), and thermogravimetric analysis (TGA), followed by the adsorption studies in aqueous solutions.

Although there are earlier reports on the use of keratin extraction and its application in composite materials, the systematic synthesis of RSM-optimised keratin extraction and its application in keratin bead-based adsorption systems have not been extensively studied. The present study is based on the hypothesis that keratin extracted via an RSM-designed Na₂S-mediated process, under mild extraction conditions, can preserve the integrity of the keratin structure and maintain sufficient functional activity, leading to enhanced dye adsorption.

Accordingly, the study’s main objectives include optimization and characterization of keratin extraction from waste CFs using a Box-Behnken RSM design and to verify the structural and chemical integrity, respectively. Furthermore, the preliminary dye adsorption performance of KBs in aqueous model dye solutions has been evaluated.

Materials and methods

Materials

Waste chicken feathers were collected from a local chicken meat shop in Kharar, Mohali, Punjab, India. Sodium sulfide (Na₂S), ethanol, sodium hydroxide (NaOH), sodium alginate, calcium chloride (CaCl₂), Congo red (CR), and HCl were purchased from HiMedia Laboratories (India). Commercial detergent was purchased from the local market in Kharar.

Pre-treatment of chicken feathers (cfs)

Waste CFs were initially washed with warm water after the manual removal of meat residues. The feathers were then cleaned with detergent to get rid of oil, blood stains, and dirt. The feathers were soaked in ethanol overnight to remove any microbial contamination and impurities, followed by rinsing with double-distilled water. After being cleaned, the feathers were dried in the sun, cut into tiny pieces (5–20 mm), and stored for further use [58].

Optimization of keratin extraction procedure

The keratin extraction process was optimized using the BBD model, an example of RSM, with the experiment design carried out using Design Expert software (version 11). RSM is an effective statistical and mathematical tool used to assess the combined effects of multiple process variables and their interactions on a response, with experimental runs determining the optimal conditions. BBD was employed due to its effectiveness in the modelling of quadratic models and its ability to avoid extreme combinations of factor levels that could degrade keratin or raise safety and operational concerns.

The extraction parameters and their respective ranges were chosen based on previously reported literature and preliminary experimental conditions. As shown in Table 1, the three independent variables included temperature (40–60 °C), extraction time (60–180 min), and Na₂S concentration (0.25–0.75 M), while keratin yield was considered the dependent/response variable.

Table 1.

Overview of variables used in experimental runs

Factors	Minimum	Maximum
Temperature (oC)	40	60
Time (minutes)	60	180
Na2S (M)	0.25	0.75

Extraction of keratin

Two grams of pre-treated CFs were immersed in various concentrations of Na₂S ranging from 0.25 to 0.75 M, as detailed in Table 1. The underlying mechanism of keratin extraction using Na₂S is depicted in Fig. 1. During the process, the disulfide bonds (-S-S-) in the cystine residues of keratin are broken by the reducing agent Na₂S. This cleavage disrupts the stiff, cross-linked protein network, transforming the insoluble keratin structure into partially soluble keratin. As a result of the breakdown of the fibrous feather matrix, keratin molecules are released into the solution for further recovery and purification. The parameters generated by the experimental design were used to input and modify the reaction time and temperature. The hydrolysate was centrifuged at 10,000 rpm for 10 min, and the supernatant was filtered. The pH of the solution was neutralized by the addition of HCl. Decomposition of the remaining sulfides was performed slowly under constant stirring in order to safely neutralize the solution and minimize the emission of hydrogen sulfide (H₂S) gas.

Fig. 1.

Illustration of the mechanism of keratin extraction from CFs using Na₂S, showing the breakdown of disulfide bonds and solubilization of keratin.

Modified from [14]

After a certain period, precipitation was observed. The obtained precipitates were thoroughly collected, dispersed in distilled water and centrifuged several times to eliminate the remaining salts and impurities. Lastly, the precipitates were freeze-dried using lyophilization to obtain keratin powder, which was stored at 4 °C until needed.

All experiments were done based on the BBD design matrix, which included replicates at the centre points to determine the error in the experiment and the sufficiency of the model. To confirm the reproducibility of the extraction yield, the optimized conditions predicted by the model were experimentally tested in five replicates (n = 5). All the extraction experiments were conducted in a fume hood that was well ventilated to prevent the accumulation of H₂S gas that could be produced during the reduction of disulfide bonds. The operators were properly dressed with the personal protective equipment (PPE), including lab coats, safety goggles and gloves, and the reaction vessels were tightly covered throughout the experimental procedures.

Quantification of protein

Bradford assay

The protein content of the extracted keratin was measured with the Bradford assay [59]. The calibration curve was based on the standard Bovine Serum Albumin (BSA). The keratin sample was prepared by dissolving 35 mg of keratin in 1 mL of NaOH solution. For the assay, 200 µL of Bradford reagent and 5 µL of sample were combined in microtubes, gently vortexed, and allowed to develop colour for 10 min at room temperature [30]. The combination of Bradford reagent and NaOH were used as a blank under the same conditions. A UV-Vis spectrophotometer was used to measure absorbance at 595 nm, and the protein concentration in the keratin sample was calculated using the BSA calibration curve.

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE)

The molecular weight profile of the extracted keratin was examined using SDS-PAGE, which was slightly modified from Laemmli’s (1970) method [60]. A 12% resolving gel and 4% stacking gel were made. To denature the protein sample, it was combined with LaemmLi sample buffer (5×) containing β-mercaptoethanol and heated at 95 °C for 5 min. Using Tris–glycine–SDS running buffer (pH 8.3), electrophoresis was carried out at 80 V through the stacking gel and then at 120 V through the resolving gel until the dye front reached the bottom. The gel was stained with Coomassie Brilliant Blue R-250 and then destained with a methanol–acetic acid solution until distinct protein bands were evident. The molecular weight of the keratin was estimated using a prestained protein ladder.

Characterization

Following the optimization of the keratin extraction process, the experimental conditions yielding the highest extraction efficiency were identified and selected for further processing. These optimal conditions were then used to extract keratin in significant quantities for thorough characterization. The characterization aimed to investigate the structural, thermal, and chemical properties of the extracted keratin to confirm its integrity and suitability for targeted applications.

Scanning electron microscope (SEM) and energy-dispersive X-ray spectroscopy (EDX)

The surface morphology of keratin was examined using a Jeol JSM-IT 500 SEM operated at an accelerating voltage of 20.0 kV. The sample was carefully prepared by mounting it onto an aluminium stub using a thin carbon film, which provided a stable conductive surface. The sample was sputter-coated with gold under vacuum to increase electron conductivity and avoid surface charge, ensuring a consistent and conductive layer for better imaging resolution.

In addition to morphological analysis, the elemental composition of the keratin sample was determined using EDX, integrated into the SEM equipment. The presence of essential components that are indicative of the protein structure of keratin, such as carbon (C), nitrogen (N), oxygen (O), and sulfur (S), was revealed by EDX. The combined SEM-EDX analysis allowed for a thorough examination of the keratin sample’s surface topography and elemental composition.

Elemental (CHNS) analysis of extracted keratin

Thermo Finnigan CHNS analyzer (USA) was used to perform CHNS analysis on the keratin sample. Approximately 2 mg of the powdered sample was burned at 950 °C in an oxygen environment. Using thermal conductivity detectors (TCDs), the produced gases were separated and identified. The weight percentages (% w/w) of the total sample mass were used to represent the concentrations of CHNS. Protein content was estimated from nitrogen content using the traditional nitrogen-to-protein conversion factor (Protein (%) = N × 6.25) [61, 62]. The collected CHNS data were utilised to determine the purity and proteinaceous nature of the extracted keratin.

Particle size and zeta potential measurement

Particle size distribution and zeta potential measurements were used to assess the stability and dispersion properties of the keratin particles. The average particle size was determined using Dynamic light scattering (DLS), and the surface charge and potential colloidal stability of the suspension were determined by measuring the zeta potential. Measurements were performed using a Lite Sizer 500 device. To avoid multiple scattering, samples were diluted with Dimethyl Sulfoxide (DMSO) and sonicated for five minutes before testing. Three replicate measurements were used to record the intensity-weighted averages of the particle size and PDI data.

X-ray diffraction (XRD)

X-ray diffraction analysis was performed to analyze the crystalline structure and phase characteristics of keratin using a Bruker D8 Advance diffractometer under standard operating conditions.

Fourier transform infrared spectroscopy (FTIR)

The functional groups of the keratin were analyzed using FTIR spectroscopy in transmission mode. The spectra were recorded across a wavenumber range of 4000 cm^− 1 and 400 cm^− 1 to identify characteristic chemical bonds of keratin.

Thermogravimetric analysis (TGA)

The sample’s decomposition behaviour and thermal stability were determined using TGA. Approximately two mg of the sample were put in an aluminium crucible. A heating rate of 10 °C/min was used to measure the phase change temperature within the range of 30 °C and 600 °C, under a nitrogen atmosphere.

Nuclear magnetic resonance (NMR)

The sample’s purity and molecular structure were deciphered using NMR spectroscopy. Bruker Avance Neo 500 MHz spectrometer was used to obtain ¹³C NMR spectra of extracted keratin. A 4-mm rotor with a Kel-F end cap was adjusted to spin at 10 kHz. The number of scans was 2048 for each spectrum. The spectra were processed using a size of 32,768 with a 1.00 Hz line broadening and a frequency of 125 MHz.

Application of keratin beads (kbs) for dye adsorption

Preparation of keratin beads (kbs)

Sodium alginate (SA) (2% w/v) was dissolved in distilled water while constantly stirred to prepare the SA solution. The extracted keratin was dispersed in the SA solution at a (1:1) (w/w) ratio to prepare the sodium alginate-keratin (SA-K) mixture. Using a syringe, the resultant mixture was extruded dropwise into a CaCl₂ (4% w/v) solution to prepare the beads. After being cured for 24 h in the CaCl₂ solution, the produced beads were rinsed with distilled water to remove any unbound calcium and left to dry at room temperature. Figure 2A depicts the freshly formed wet KBs, which are translucent, soft and swollen due to the high-water content maintained throughout synthesis. Conversely, Fig. 2B shows the identical beads following the dehydration process. The bead’s size noticeably decreases, and its structure becomes firmer and more compact as a result of the moisture loss, signifying that the beads were successfully dried and structurally stabilized.

Fig. 2.

Comparison of wet and dried KBs (A) Wet beads – translucent and swollen; (B) Dried beads – smaller and firm after dehydration

Batch adsorption studies of congo red (CR)

The removal efficiency of CR using KBs was evaluated by conducting batch adsorption studies. The effect of the adsorbent dose was evaluated by varying KBs dose from 25 mg to 125 mg using 100 mL of CR solution (50 ppm) at neutral pH for 48 h. The effect of pH on adsorption performance was investigated in a pH range of 3.0–11.0 using 100 mL of CR solution (50 ppm) with 125 mg of adsorbent for 48 h. The pH of the dye solutions was adjusted using dilute NaOH or HCl before the addition of KBs. The effect of contact time was investigated for 1,6, 12, 24, 36, 48, 60 and 72 h with 125 mg of adsorbent, pH 5 [, and 100 mL of 50 ppm CR solution. The effect of the initial CR concentration, ranging from 10 to 50 ppm, was studied using 125 mg of adsorbent and 100 mL of solution at pH 5.0 for 48 h. The dye adsorption studies were conducted in triplicate (n = 3) using aqueous dye solutions to provide reproducibility. Initial and final dye concentrations were measured using UV–Vis spectrophotometry, and the adsorption capacity and removal efficiency were expressed as mean ± standard deviation (SD). Furthermore, the low variability in the data from the triplicate samples assures the reliability of the measured adsorption data.

A UV-visible spectrophotometer was used to detect the dye concentrations at λ_max of 498 nm for CR [48]. The removal efficiency (R) (%) and adsorption capacity (q_e) were computed using a standard Eqs. [63, 64]:

where, is the initial dye concentration (mg/L), and is the concentration (mg/L) of dye at any time (min).

where C_o is the initial dye concentration (mg/L), C_e is the final dye concentration (mg/L), V is the volume of the dye solution (L), and m is the mass of the adsorbent used (g).

Adsorption isotherm and kinetic studies

The mechanism and behaviour of CR dye adsorption on KBs were understood by conducting adsorption isotherm and kinetic studies. The adsorption kinetics were analyzed using three models, including pseudo-first-order (PFO), pseudo-second-order (PSO), and Intraparticle diffusion (IPD). Equilibrium adsorption behaviour was investigated using the Langmuir and the Freundlich isotherm models.

Results

Analysis of the extraction of keratin using response surface methodology (RSM)

A total of seventeen experimental runs were conducted to optimize the keratin yield using the BBD design. The magnitude of each coefficient indicated its relative effect on the response compared with other coefficients; the highest coefficient indicated greater significance and was associated with the lowest p-value. The predicted change in response for each unit change in an independent variable was calculated while keeping the other variables constant. In an orthogonal design, a Variance Inflation Factor (VIF) value close to 1 is desirable, with the value of 10 considered an acceptable upper limit. The VIF for each coefficient was within the range of 1 to 1.01, confirming the absence of multicollinearity and supporting the stability of the regression model. The residual plots for the RSM further confirm its adequacy, as shown in figures S1-S5. A second-order quadratic polynomial model was developed to relate the input variables to the percentage keratin yield (Table S1), as shown in the equation below:

where + 63.80 is intercept, + 1.06 A, −0.3125B, −1.50 C are linear terms, + 0.8750AB, + 5.00 AC, −3.00 BC are interaction terms, − 5.59A², – 5.09B², – 3.71C² are quadratic terms, A is the extraction time (minutes), B is the temperature (^oC), and C is the Na₂S concentration (M).

The direction of the variable’s effect (positive or negative) on the response is indicated by the sign (+ or -) of each coefficient. The results of regression analysis usually determine specific signs. The statistical significance and validity of the fitted/developed model was assessed using analysis of variance (ANOVA), as indicated in Table 2.

Table 2.

ANOVA results for statistical evaluation of keratin yield, presenting F-values, p-values, and degrees of freedom

Source	Sum of squares	Df	Mean Square	F-value	p-value
Model	499.12	9	55.46	34.93	< 0.0001	Significant
Lack of fit	0.3125	3	0.1042	0.0386	0.9884	Not significant

The model was observed to be statistically significant, with a model’s F-value of 34.93 and a p-value < 0.0001, suggesting that there is only a 0.01% possibility that an F-value this large may be caused by noise. The lack-of-fit was found to be non-significant (Table 2), further confirming the adequacy of the model. The observed coefficient of variation (C.V.) was 2.21%, showing high experimental precision and reproducibility. Since the difference between the adjusted R² (0.9502) and the predicted R² (0.9571) was less than 0.2, the two values were in reasonable agreement. The signal-to-noise ratio was measured with adequate precision parameter. The measured adequate precision was 17.540 (higher than 4), indicating that this model could be used to navigate the design area. Figure 3 illustrates the 3D response surface plots with contour plots at the base: (A) time and temperature, (B) temperature and the concentration of Na₂S, (C) time and the concentration of Na₂S. These plots are essential for understanding how the factors interact to affect the yield of keratin, representing the relationship between the yield (%) of keratin and different combinations of two independent variables at a time, whereas the third variable remains constant.

Fig. 3.

3D response surface plot graphs illustrate how the yield of keratin is affected by (A) time and temperature, (B) temperature and Na₂S concentration, and (C) time and Na₂S concentration

The response surface plots for keratin yield in relation to time, temperature, and Na₂S concentration indicated significant interactions, highlighting the need to optimize the extraction of keratin from chicken feathers. Figure 3A shows the interaction between time and temperature, showing that the yield of keratin increased with both factors up to an optimal point, beyond which a decline in yield was observed. The curved surface indicates the maximum yield at the centre of the plot when both time and temperature were moderate. The fact that the optimal yield was attained within a particular range of these factors was further confirmed by the concentric contour lines at the base. The yield of the protein was also increased by gradually raising the extraction temperature from 45 to 50 °C over a period of 65 to 120 min. Nevertheless, a decrease in yield was observed with an increase in reaction temperature beyond 50 °C.

A similar quadratic trend is shown in Fig. 3B, which illustrates the correlation of temperature and Na₂S concentration. Keratin yield peaked at an intermediate level, which proves that low and high levels of Na₂S and temperatures are not favourable to maximize the yield. The yield rose gradually with the increase in the concentration of Na₂S at 50 °C, as the concentration was raised from 0.25 to 0.5. Moreover, there was a general decline in yield with the rise in temperature and Na₂S concentration. The contours of this plot also show that there is a small range in which the ideal conditions are achieved, with variations resulting in a lower protein yield. The findings show that a good balance between the two variables is needed to achieve optimal extraction efficiency.

Figure 3C shows the relationship between time and Na₂S concentration, showing that increasing both variables increases the yield of keratin, although after reaching a maximum (peak) decrease was observed. When the concentration of Na₂S was increased from 0.25 to 0.5 M, the yield rose steadily at 120 min. Moreover, the overall yield decreased as the concentration of Na₂S rose. Contour plots at the base also supported this observation by showing the optimal yield when both factors were moderate.

The maximum yield was 64 ± 1.6% (n = 5) when the conditions were near the central levels of temperature (50 °C), time (120 min), and Na₂S (0.5 M) concentration. The results were quite close to the yield as was predicted by the RSM model, signifying reproducibility and credibility of the optimization. The quadratic response surfaces emphasize the complex interplay between various variables and the need for precise control of experimental conditions to attain maximum yield of keratin. These findings indicate the efficiency and effectiveness of the keratin extraction process. The comparison between the studies presented on the keratin extraction of the CFs shows that the procedure is efficient and sustainable. In a study, Fagbemi and co-workers [30], with a combination of NaOH and sodium bisulfite (NaHSO₃), optimized using BBD, achieved 68.3% keratin and 65.2% protein yield at a temperature of 87 °C, for 111 min, using 1.78% NaOH and 0.5% NaHSO₃. Another research by Qin and co-workers [65] used an ultrasound-assisted cysteine reduction, showing that the solubility increased with longer treatment hours; however, the thermal stability and amino acid composition were compromised with long treatment. The optimal conditions occurred at 130 W over 2.7 h of treatment, with a maximum 15% cysteine reduction. A comparative research evaluated various reducing agents and concluded that Na₂S yielded the greatest yield (keratin yield of 84.5%, at 80.9 °C, for 9.5 h, using 0.05 M Na₂S), while yields from NaOH and SDS were far less impressive [31]. In another research, the extraction was optimized using 0.43 M Na₂S for 5.43 h, resulting in a yield of 75.39% [66]. Mengistu et al. 2024 performed alkaline hydrolysis with NaOH using a concentration of 0.65 N, at optimal conditions (70 °C, 1 h), yielding 89.71% protein [5]. In comparison, the current research was able to yield 64% keratin using milder and shorter conditions (0.5 M Na₂S, 50 °C, 120 min) and is generally expected to be more energy efficient and scalable. The extraction proposed in the current study reduces the amount of chemicals used, the time of reaction, and the amount of thermal energy introduced and therefore offers a sustainable approach in obtaining keratin using poultry feather waste.

Preliminary scale-up extraction processes, i.e., larger batch sizes and increased amounts of CFs, yielded a high yield of keratin (75–80%), proving the efficiency and scalability of the process under the optimal condition.

Bradford assay

The concentration of the protein in the extracted keratin was quantified by the Bradford assay. The assay was performed in triplicate using BSA as the calibration standard. The standard curve was linear in the range of measurements (0.2–1.0.2.0 mg/mL), and the regression equation was as follows: A₅₉₅ = 1.0965 C + 0.2475 (R² = 0.991). Protein concentration of the extracted keratin was found to be 2.56 ± 0.25 mg/mL. The findings of previous studies have indicated that protein concentrations have differed with respect to the extraction conditions. One study has reported the protein content of hair keratin was ~2.9 mg/mL after 3 days of extraction at 50 °C, this increased to 9.8 mg/mL after 6 days [67]. On the same note, a study established the concentration of ß-keratin solution was 2.76 ± 0.65 mg/mL [68], and another study reported that the concentration of keratin was 0.95 mg/mL [7]. These variances may be related to the variables of extraction time, temperature, solvent content, and source of keratin.

SDS-PAGE

The protein profile of keratin was analyzed using SDS-PAGE to determine the distribution of molecular weight (Fig. 4). The molecular weight marker (protein ladder) was on Lane 1, and Lane 2 was the extracted keratin. The protein profile of Lane 2 displayed distinct bands in the 10–14 kDa region (Fig. 4), indicating that the low-molecular-weight keratin peptides were obtained due to the successful reduction of the feather keratin.

Fig. 4.

SDS-PAGE of extracted keratin, lane [1] pre-stained protein ladder, lane [2] keratin

The occurrence of clear bands and no larger and heavy molecular weight bands demonstrates that the native keratin structure has been effectively disaggregated, resulting in smaller and soluble peptides. The results confirm effective extraction of keratinous protein, and are in line with the reported molecular weight range of 10–22 kDa for β-keratin. The lack of α-keratin with a molecular weight ranging between 40 and 68 kDa, indicates that the process produced mostly keratin fragments of the β type [18]. The difference in peptide size across studies is often attributed to differences in extraction parameters such as reaction time, reagent concentration, and mechanical disruption. For example, Dąbrowska et al. (2021) demonstrated that the size distribution of hydrolyzed keratin peptides strongly depends on the parameters, including the ratio of alkali-to-feather and agitation speed, and obtained fractions of nearly 10 kDa and smaller under optimal conditions [69]. Similarly, soluble protein fractions with molecular weights ranging between ~4 kDa to 14 kDa, depending on the parameters of extraction, were obtained with the help of green extraction of keratin from feathers [19].

Scanning electron microscopy (SEM)

The surface morphology of keratin was examined using SEM analysis to develop detailed information on its structural features. The SEM images of the keratin were observed in Fig. 5A) 3,000x with scale bar = 5 μm, B) 8,500x with scale bar = 2 μm, C) 19,000x with scale bar = 1 μm, D) 22,000x with scale bar = 1 μm revealing that the keratin particles exhibited a globular morphology, with tightly packed and randomly arranged structures. The measurements of the particle sizes were found to range between 1 and 5 μm, which shows a fairly consistent distribution. The globular structure seen in the keratin may indicate a compact structure that is possibly due to the molecular characteristics of the protein, i.e., hydrogen bonding and disulfide bonds. The particles were randomLy oriented and, therefore, did not form highly organized assemblies, which is consistent with their natural state when extracted from feathers. These morphological features are consistent with other past studies that reported that the keratin particles derived from feathers were found to have similar shapes and size distributions [16, 31, 70].

Fig. 5.

SEM images of keratin particles: (A) 3,000 × (5 μm), (B) 8,500 × (2 μm), (C) 19,000 × (1 μm), (D) 22,000 × (1 μm). Globular particles (1–5 μm) show dense, randomLy arranged structures

Energy dispersive X-ray (EDX) analysis

The EDX spectra were used to characterize the elemental composition of the keratin, which showed carbon (C), oxygen (O), sodium (Na), aluminium (Al), sulfur (S), and copper (Cu) [16]. S and Na content in the keratin were 9.46 ± 0.79 and 0.15 ± 0.42, respectively (Table S2). The most prevalent ones were C and O, and the mass percentages were 52.45 ± 2.31% and 35.8 ± 3.58%, respectively [71] (Figure S6). These findings support keratin’s proteinaceous origin, which is mainly made up of organic molecules. The high S content (9.46 ± 0.79%) of keratin confirms the existence of cysteine residues, that assist in forming disulfide bonds, which serve as structural stabilizers of proteins [16]. Trace levels of Na (0.15 ± 0.42%), Al (0.21 ± 0.28%), and Cu (1.88 ± 1.00%) were also found [16]. Na presence may be a result of the usage of Na₂S in an extraction process, Al may result due to the environmental contamination, and Cu may be a result of using Cu components or grids within the EDX equipment. These results are in line with the previous studies, which showed the presence of C, O, and S as the predominant substances in the keratin, with traces of Na, Al and Cu [16, 71].

Elemental (CHNS) analysis of extracted keratin

A CHNS combustion analyzer was used to characterize the elemental composition of the keratin, which provided a quantitative determination of the C, H, N, and S components. The elemental percentages determined for keratin were C = 44.39%, H = 5.71%, N = 13.98%, and S = 6.97% (Table S3) (Figure S7) [30, 72–74]. The modest differences could be caused by the presence of residual mineral salts from the extraction process or from the drying. The determined values, however, confirm that the extracted biopolymer has the expected stoichiometry of keratin.

Using the standard nitrogen-to-protein conversion factor (N x 6.25), the nitrogen content (13.98%) corresponds to about 87.4% of the protein content, which is similar to other reported values. An example is in a study where the protein content reported was 87.4% ± 0.6, using the Kjeldahl method [62], and a research which showed protein content of 86.89% [23].

This high number shows that the substance is mainly proteinaceous and contains very minimal contamination of inorganic residues and non-protein organic materials.

The amount of S (6.97%) is especially relevant as the S in the keratin is mainly obtained through the residues of cystine and cysteine that create disulfide bonds (-S-S-). The thermal stability and mechanical stiffness of keratin are due to its covalent bonding. The identified S percentage matched native feathers (4–7%), suggesting that the extraction process did not break the disulfide-containing structure.

Particle size and zeta potential measurement

The DLS analysis was performed to determine the size distribution and homogeneity of keratin particles. The findings showed that the keratin dispersed into relatively large particles, with an average hydrodynamic diameter of 5705 nm (5.705 μm). These findings fall within the range of previous literature, with the hydrodynamic diameters of the keratin particles at pH 3, pH 7 and pH 10 of 5164 nm, 7644 nm and 2227 nm, respectively [19]. The uniformity of the particle sizes was determined through observation of a polydispersity index (PDI) of 0.08, which indicated that the particles of keratin were very homogenous and had a slight disparity in terms of particle size. The given values are intensity-weighted average diameters after mild sonication and appropriate dilution to reduce agglomeration. Figure S8 indicates that the count of the particles is distributed with a strong peak size at 3928 nm. These observations have been consistent because of the aggregation of the keratin particles through intermolecular interactions like hydrophobic, disulfide, and hydrogen bonding. Zeta potential was measured by dispersing keratin in DMSO, and the results indicated that the mean value of the zeta potential is −8.2 mV, implying weak electrostatic stability of the dispersion and moderate aggregation tendency [19]. The keratin particles are likely to form aggregates due to the weak repulsive force, resulting in a negative zeta potential [19].

X-ray diffraction (XRD)

XRD analysis of keratin was carried out to identify its crystal phases. Figure 6A shows the three intense peaks at 2θ = 9.1^o, 19.4^o, and 21.4^o, representing a semi-crystalline nature. The extracted keratin shows α-helix diffraction properties at 2θ = 9.1^o and β-sheet diffraction properties at 2θ = 19.4^o and 21.4^o. Crystalline and amorphous regions were observed in the extracted keratin because of the presence of distinctly diffraction peaks superimposed on a broad and low-intensity background. The observed diffraction peaks are close to the previously reported ranges, where α-helix and β-sheet were given to the prominent peaks at 2θ = 9^o−10^o and 15^o−31^o, respectively. The relatively stronger intensity of the β-sheet peaks denotes a higher prevalence of the β-sheet structures in the extracted keratin. This could be explained by the extraction step, considering the reduced agent and solvent system used. The results of the current study were consistent with the earlier studies, which support that the extracted keratin retains the necessary secondary structures; and confirm the semi-crystalline nature [16, 70, 75].

Fig. 6.

Characterization of keratin by (A) XRD, demonstrating the crystalline structure, (B) FTIR, identifying the functional groups, (C) TGA, showing the thermal stability and degradation behaviour (D) ¹³C NMR spectrum demonstrating the chemical shifts

Fourier transform infrared spectroscopy (FTIR)

FTIR spectra of the keratin samples, which indicate the presence of functional groups associated with the molecular structure of keratin, are demonstrated in Fig. 6B. The characteristic peaks at 546 cm^− 1 indicate disulfide bonds, which are significant for the structural stability of keratin. These linkages are formed due to the oxidation of cysteine residues, which leads to the formation of disulfide bonds that play a crucial role in the mechanical strength and resistance to degradation of the keratin. The C-H stretching vibrations linked to alkyl groups of amino acid residues such as alanine and leucine, are indicated by the 2877 cm⁻¹ and 2959 cm⁻¹ peaks, respectively [8, 17].

The symmetric and asymmetric stretching of the methylene (-CH₂-) and methyl (-CH₃) groups reflects the organic nature of the polypeptide backbone of keratin. The primary reason for the amide I band, observed at 1632 cm⁻¹, is due to the C = O stretching vibrations of the peptide bonds. This band is characteristic of protein secondary structures, especially α-helices and β-sheets, contributing to conformational stability of keratin [71, 76]. The high absorption at this wavenumber indicates that there are structural motifs in the keratin structure.

The amide II band is caused by C-N stretching and N-H bending vibrations, which is determined by the peak at 1522 cm^− 1 [71, 76] providing additional information on the secondary structure of the keratin, as well as its direct dependence on the local folding of the polypeptide chains and the network of hydrogen bonds.

The amide III band is confirmed by the peak at 1239 cm^− 1 and contributes to the C-N stretching and N-H bending [16, 76, 77] and is also an important marker of protein structure. The N-H stretching of the amide A band is present as shown by the peak at 3283 cm^− 1 [6]. The hydrogen-bonded amide groups that make up the hydrogen bonding network of keratin are the major cause of this band and influence its capability to retain water and interact with other molecules. These distinguishing properties (peaks and bands) confirm the integrity and structural components of the keratin, which are consistent with acknowledged protein absorption patterns and previous studies.

Thermogravimetric analysis (TGA)

The thermal stability and decomposition behaviour of keratin were investigated using TGA under a nitrogen atmosphere (20 mL minutes^− 1) at a heating rate of 10 °C minutes^− 1. The TGA-DTG curve denotes four weight-loss stages, each representing a different degree of thermal degradation, as illustrated in Fig. 6C.

The moisture evaporation of keratin has caused the initial weight loss of ~2.6% observed below 150 °C. When heated, biopolymeric materials often release absorbed water. The second and most significant weight loss, accounting for 41% of the total weight, was observed between 230 and 360 °C, with a Tonset of 241.9 °C, a Tmax of 310 °C, and a Tendset of 360 °C, primarily due to the breakdown of the fibrous structural protein. Furthermore, the third degradation (~10%) was observed between 370 and 450 °C, which was linked to the breakdown of keratin fractions that possess intermediate and high thermal stability, like disulfide linkages and closely packed polypeptide chains in the form of β-sheets and larger particle sizes. The final loss of 20.2% between 450 and 600 °C suggests the slow breakdown of thermally stable aromatic structures. Nearly 75% of the total weight loss was attributed to the four physically distinct degradation stages. However, as confirmed by raw instrument data, the TGA curve continued to decline steadily at 550 °C as a result of the slow breakdown of residual organic matter, leading to an 82% overall mass loss and an 18% residual char yield at 600 °C. This distinction implies that there is a modest but continuous degradation tail that goes beyond the primary visible stages. The remaining 18% represents aromatic carbonaceous residues that are thermally stable.

Consistent findings were obtained from triplicate TGA runs (n = 3; Tonset = 241.9 ± 0.6 °C; Tmax = 310.4 ± 0.6 °C; Residue = 18.1 ± 0.2%). These results show the multistage breakdown and thermal stability of the extracted keratin and are in line with the previously published studies, indicating the onset at 240–260 °C and residue of 15–25% [78–81]. The significant weight loss of keratin indicates that a large amount is thermally labile and entirely degrades at high temperatures, implying the breakdown of the majority of the keratin.

Nuclear magnetic resonance (NMR)

The ¹³C NMR spectrum of the keratin displayed definite peaks related to its structural components. The alkyl groups of the side chain corresponded to the peaks at 19 and 24 ppm. The peaks at 39 and 41 ppm represent the β-carbons in the cysteine and leucine residues, and the peak at 61 ppm is associated with the α-carbon. Moreover, the 128 ppm peak indicates the existence of aromatic compounds in keratin. The peak at 174 ppm is reflective of the amide carbonyl carbons in the keratin structure (Fig. 6D). These specific properties support previously reported studies [82–85].

These structural, elemental and functional properties of the keratin were corroborated through the comprehensive characterization. SEM revealed the surface morphology of keratin, and EDX confirms its composition of C, O, S, with traces of other elements. Moreover, the crystallinity, functional groups, and thermal stability of keratin were confirmed with XRD, FTIR and TGA, respectively. In conclusion, the results of this study showed effective extraction of keratin with the anticipated physical and chemical characteristics needed in the intended applications.

Batch adsorption studies

Synthetic dye, such as CR, was used in the dye adsorption studies to evaluate the effectiveness of the keratin-based adsorbent. CR is an anionic azo dye with a strong affinity to natural polymers, and was selected because of its environmental relevance and structural diversity, indicating its presence in industrial wastewater.

The use of KBs from CFs has also successfully removed the anionic azo dye CR and exhibited maximum removal of 79.3 ± 1.0% and adsorption capacity of 31.7 ± 0.9 mg/g (n = 3). This performance can be because of the presence of many functional groups, such as –NH₂, –COOH, and –SH, in keratin, that can interact with dye molecules through electrostatic attraction and hydrogen bonding (Fig. 7).

Fig. 7.

Dye adsorption efficiency of KBs towards CR in aqueous solutions

Effect of pH

The adsorption behaviour of CR onto KBs is significantly affected by the pH of the solution in terms of the surface charge of the adsorbent, as well as the molecular ionization of the dye molecules [86, 87]. The impact of the pH was assessed in the range of 3.0–11.0 with an adsorbent dose of 125 mg, a dye concentration of 50 ppm, a contact time of 48 h, and a temperature of 28 ± 2 °C. The apparent removal efficiency was at pH 3 (84.47 ± 0.25%), followed by pH 5 (78.43 ± 0.45%) and pH 7 (75.5 ± 0.50%) as shown in the Fig. 8A. At pH 3, however, a clear colour change in CR from red to dark blue was observed upon addition of HCl. This is due to protonation of the azo groups, which causes a shift of the π–π* transition to longer wavelengths, which might result in non-adsorptive loss of the dye as well as affect subsequent spectrophotometric quantification [86]. The literature has extensively reported similar protonation-induced colour changes of CR at very strong acidic conditions. Therefore, pH 3 was not chosen for subsequent adsorption studies. Rather, pH 5 was determined as the optimized value since it maintained dye stability and accurate analytical results while providing high adsorption effectiveness. The adsorption was found to decrease significantly at alkaline conditions and removal efficiencies were reduced to 55.27 ± 0.31% at pH 9 and 34.20 ± 0.26% at pH 11 [86, 88]. The stronger adsorption in acidic conditions could be explained by the electrostatic attraction of the negatively charged CR molecules and protonated functional groups (-NH₂, -COOH and -OH) on the surface of the keratin, while surface deprotonation and increased electrostatic repulsion result in lower uptake at higher pH [36, 86].

Fig. 8.

Effect of various parameters on CR adsorption by the KBs: (A) pH, (B) adsorbent dose (mg), (C) Contact time (h), (D) initial dye concentration. Error bars indicate SD (n = 3)

Effect of adsorbent dose

The effect of adsorbent dosage on CR removal was studied by varying the mass of the KBs from 25 to 125 mg while maintaining the dye concentration (50 ppm), solution volume (100 mL), pH [5], contact time (48 h) and temperature of 28 ± 2 °C. The efficiency of removal was found to be increasing in proportion to the increase in adsorbent dose, as shown in Fig. 8B, since removal efficiency was 54 ± 0.78% and 79.23 ± 0.76% at 25 mg and 125 mg, respectively, due to higher availability of active adsorption sites and improved effective surface area at higher doses of adsorbent. On the other hand, the adsorption capacity (q_e) decreased significantly from 109.8 mg/g at 25 mg to 31.69 mg/g at 125 mg. This inverse relationship of q_e and adsorbent dosage can be due to the underutilization of sites, and some of the adsorption sites overlap at higher dosage levels when a constant quantity of dye is divided over a larger mass of adsorbent. Previous studies also reported that increasing the adsorbent dose improved removal efficiency but decreased q_e due to site underutilization and lower effective surface area due to particle clustering [87, 89]. Furthermore, at higher dosages, the rate of removal efficiency is less significant, indicating the partial saturation of adsorption sites and potential particle aggregation, resulting in a lowered effective surface area. Based on these results, 125 mg was selected as the optimal adsorbent dosage for subsequent studies, as it provided the best overall removal efficiency.

Effect of time

The effect of contact time of CR adsorption using KBs was determined at a period of 1–72 h with an adsorbent dose of 125 mg, a dye concentration of 50 ppm, a solution volume of 100 mL, pH 5, and a temperature of 28 ± 2 °C. As shown in Fig. 8C, removability was high in the initial stages, whereby the percentage removal was 48.07 ± 0.95% at 1 h, and it rose steadily to 62.13 ± 0.35% at 12 h. Subsequently, the rate of adsorption slowed down and reached a removal percentage of 69.43 ± 0.45% at 24 h and 79.13 ± 0.40% at 48 h. After this point, only a small increase was noticed, 79.57 ± 0.51% and 80.57 ± 0.40% were recorded at 60 h and 72 h, respectively, implying saturation of adsorption sites, and that adsorption equilibrium was reached. The fast initial adsorption and slower uptake period are possibly due to surface adsorption during the early stages and intraparticle diffusion at the later stage. In recent studies, CR has been shown to exhibit similar two-stage adsorption behaviour, with rapid surface adsorption followed by intraparticle diffusion until equilibrium [86, 90]. As a result, 48 h was chosen as the ideal contact time for the further experiments.

Effect of initial dye concentration

The effect of initial dye concentration of CR on the adsorption performance was explored by varying the concentrations from 10 to 50 ppm with an adsorbent dose of 125 mg, 100 mL of solution volume, pH 5 and temperature of 28 ± 2 °C. As the initial dye concentration was increased, the removal efficiency, initially 87.5 ± 0.40% at 10 ppm, dropped to 79.3 ± 0.36% at 50 ppm (Fig. 8D). This may be due to the saturation of the existing adsorption sites in the instances of high loading of dyes, and this reduces the ratio of dye molecules adsorbed to the adsorbent. Conversely, the adsorption capacity rose with initial concentration, increasing from 8.75 mg/g at 10 ppm to 39.65 mg/g at 50 ppm, due to an increase in the mass transfer driving force caused by the high dye concentration in solution. The concomitant fall in removal efficiency and a rise in qe with initial concentration are typical traits of adsorption systems and reflect competition among dye molecules for a limited number of active sites [91]. Considering these findings, 50 ppm was chosen as the working concentration for subsequent studies to ensure enough adsorption capacity to facilitate reliable adsorption modelling.

Adsorption kinetics and isotherm modelling

The kinetics of CR on KBs were studied using PFO, PSO, and IPD models (Fig. 9A, B, and C, respectively). The PSO model was observed to give the highest correlation coefficient (R² = 0.9974) to the experimental data, giving the best fit. In addition, the equilibrium adsorption capacity obtained by the PSO model (Qₑ, calc = 33.33 mg g⁻¹) was quite close to the experimental data (Qₑ, exp = 32.23 mg g⁻¹), which showed that the chemisorption may attribute significantly to the adsorption process.

Fig. 9.

Kinetic and equilibrium modelling of CR adsorption onto the KBs: (A) pseudo-first order, (B) pseudo-second order, (C) intraparticle diffusion, (D) Freundlich isotherm, and (E) Langmuir isotherm, where q_e and q_t represent the adsorption capacity at equilibrium and time t, respectively, and C_e denotes the equilibrium dye concentration

Conversely, the calculated and experimental q_e values were different, and the correlation coefficient (R² = 0.8277) of the PFO model was lower; hence, the physical adsorption process was not sufficient to describe the adsorption process. The intraparticle diffusion (IPD) yielded an R² = 0.9672, and the linear plot was unable to cross the origin (intercept C = 18.34 mg/g). This non-zero intercept indicates that IPD is one of the parameters that affects adsorption; however, it is not the sole rate-limiting step. These findings suggest that adsorption occurs through a multi-step process wherein surface adsorption is followed by diffusion into internal pores, and the chemisorption process plays a significant role. The PSO kinetic model dominance is consistent with reported literature on CR adsorption, indicating that chemisorption predominates in the adsorption mechanism across a variety of adsorbents [86, 92–94].

The Langmuir and the Freundlich isotherm models were used further to evaluate the equilibrium adsorption data (Fig. 9D and E) to understand the nature of the interaction between the adsorbent surface and the adsorbate. The Langmuir model provided a maximum adsorption capacity of 81.3 mg/g with a correlation coefficient of R² = 0.9608, indicating monolayer adsorption on comparatively homogeneous sites. The dimensionless separation factor (R_L) was calculated between 0.10 and 0.53 (0 < R_L < 1), indicating that the adsorption process is favourable.

A better fit was shown by the Freundlich model, R² = 0.9991, suggesting that the adsorption occurs via a multilayer mechanism on a heterogeneous surface. The Freundlich constant K_F with 7.47 mg/g indicates strong adsorption affinity, whereas the intensity constant, n, was observed to be 1.38 (> 1), confirming favourable adsorption. The higher agreement with the Freundlich model is in alignment with the heterogeneous nature of the KBs surface, which is attributed to various functional groups, typically present in keratin-based adsorbents [39, 92, 95].

Overall, the kinetic and isotherm analyses show that the CR adsorption on KBs is dominated by chemisorption on a heterogeneous surface, along with IPD, leading to a favourable adsorption process and a high adsorption capacity [94, 96] (see Table 3).

Table 3.

Kinetic and isotherm parameters for CR adsorption onto KBs based on PFO, PSO, IPD, Langmuir and Freundlich models

Model	Equation used	Parameters
a. Kinetic models
Pseudo-first-order (PFO)		k1 = 0.058 h− 1 qe(cal) = 11.88 mg g− 1 R² = 0.8277
Pseudo-second-order (PSO)		k2 = 9.7 × 10− 3 g mg− 1 h− 1 qe(cal) = 33.33 mg g− 1 R2 = 0.9974
Intraparticle diffusion (IPD)		kid = 1.84 mg g− 1 h− 1/2 C = 18.34 mg g− 1 R2 = 0.9672
b. Isotherm models
Langmuir		qmax = 81.3 mg g− 1 KL = 0.089 L mg− 1 R2 = 0.9608 Homogeneous, monolayer
Freundlich		KF = 7.47 mg g− 1 n = 1.38 R2 = 0.9991 Heterogeneous, multilayer

Where q_t (mg g^− 1) is the amount of dye adsorbed at time t, q_e (mg g^− 1) is the equilibrium adsorption capacity, k₁ (h^− 1) and k₂ (g mg^− 1 h^− 1) are the rate constants of the PFO and PSO models, respectively, K_id (mg g^− 1 h^− 1/2) is the intraparticle diffusion rate constant, and C (mg g^− 1) is the boundary layer thickness, qmax (mg g^− 1) denotes the maximum monolayer adsorption capacity, K_L (L mg^− 1) is the Langmuir adsorption constant, K_F (mg g^− 1) represents the adsorption capacity constant and n indicates adsorption intensity, R² is the coefficient of determination used to evaluate model fitting

The performance of the synthesized KBs in the present study was evaluated by comparing their maximum adsorption capacity (q_max) with various reported adsorbents for CR removal. According to Table 4, the q_{max ˍ}of 81.3 mg g⁻¹ obtained in this work shows a higher adsorption capacity than a few agricultural adsorbents and waste-derived materials, including pineapple plant stems (11.97 mg g⁻¹) [97], banana peels (1.73 mg g⁻¹) [98], and cabbage waste powder (2.31 mg g⁻¹) [99]. While polydopamine-embedded keratin biohydrogels (~2000 + mg g⁻¹) [100] and guar gum/activated carbon (831.82 mg g⁻¹) [101] are examples of engineered nanocomposites that have higher adsorption capacities because of high surface engineering or intensive chemical functionalization, the current study provides a comparatively economical and sustainable balance.

Table 4.

Comparative studies of keratin-based and other polymeric adsorbents for CR dye removal, highlighting their adsorption performance

Source	Composite	Adsorption capacity	Key findings	References
Guar gum/activated carbon (AC) nanocomposite	Guar gum-AC nanocomposite	831.82 mg g⁻¹	High surface area engineered composite	[101]
Vermicompost-derived biochar	Biochar	31.28 mg g⁻¹	Low-cost agricultural waste adsorbent	[102]
Bengal gram fruit shell	Biomass adsorbent	22.22 mg g⁻¹	Natural biosorbent	[104]
Pineapple plant stem	Plant biomass	11.97 mg g⁻¹	Renewable adsorbent	[97]
Cabbage waste	Cabbage waste powder	2.31 mg g⁻¹	Low-cost agricultural waste biosorbent with limited adsorption capacity for Congo Red	[99]
Banana peels	Raw banana peel biosorbent	1.73 mg g⁻¹	Low-cost agricultural waste adsorbent with relatively low Congo Red uptake	[98]
Peanut husk	Tetraethylenepentamine-modified peanut husk composite beads	19.57 mg g⁻¹	Amine functionalization	[86]
Keratin	Polydopamine-embedded keratin biohydrogels	~ 2000 + mg g⁻¹	Increased surface functionality and adsorption capacity with the incorporation of polydopamine	[100]
Chicken feathers	Untreated chicken feathers	97.85 mg g⁻¹	Natural adsorbent	[51]
Hen feathers	Untreated hen feathers	73.84 mg g⁻¹	Moderate Congo Red adsorption using raw feathers	[103]
Chicken feather keratin (RSM optimized)	Sodium alginate–keratin (SA–K) beads	81.3 mg g⁻¹	First integration of RSM-optimized Na₂S extraction with KBs formation; eco-friendly and reproducible	Present study

Furthermore, our KBs demonstrated a competitive performance compared with other modified polymers, including vermicompost biochar (31.28 mg g⁻¹) [102] and amine-functionalized peanut husks (19.57 mg g⁻¹) [86]. Furthermore, the 81.3 mg g⁻¹ capacity is similar to that of untreated hen feather (73.84 mg g⁻¹) [103] and slightly less compared to untreated CFs (97.85 mg g⁻¹) [51] as reported in the literature. Nevertheless, the incorporation of keratin into a bead-based system offers significant benefits in mechanical stability, ease of recovery from aqueous media, and compatibility with continuous-flow applications. Table 4 summarizes the keratin-based and other adsorbents reported in the literature, emphasizing the differences in approach between the present study and the reported studies.

Environmental and toxicological considerations

Although the present study showed an effective and optimized method of extracting the keratin and using it for dye adsorption, a comprehensive toxicological and environmental evaluation was beyond the scope of the current study and will be addressed in future studies. Specifically, potential leaching of residual chemicals of keratin-based materials and overall life cycle assessment (LCA) were not investigated and should be considered in future studies. There are underlying safety and environmental issues associated with the use of Na₂S as a reducing agent, though in the current risks were eliminated through the use of mild extraction conditions, controlled neutralization and appropriate laboratory safety precautions. Future studies will be aimed at assessing the leaching behavior, environmental impact, and process sustainability to provide even more in-depth support to large-scale implementation.

Positioning and contribution to the sustainable development goals (SDGS)

Traditional methods of wastewater treatment, like biocoagulation-flocculation and adsorption with agricultural or biological waste derivatives, have been extensively studied in terms of dye removal [105]. These methods are effective, but they rely on chemical additives, produce secondary sludge or use material with a low regeneration potential [106, 107]. Comparatively, the adsorbents developed in this study are derived from a protein-rich poultry waste, are biodegradable and combine valorization of waste with pollution elimination. The present study also makes a direct contribution to various United Nations SDGs by integrating waste valorization with environmental remediation. The transformation of CFs into functional keratin-based adsorbents contributes to SDG 12 (Responsible Consumption and Production) by facilitating the idea of circular bioeconomy and minimising the use of synthetic materials [107–109]. The use of KBs to remove dye is consistent with SDG 6 (Clean Water and Sanitation) since it focuses on wastewater purification with biodegradable and low-cost adsorbents [106, 110]. Moreover, the mild extraction conditions and optimized processing parameters are also indirect contributor of SDG 13 (Climate Action) as the energy use and environmental impact are lower than those of traditional disposal and treatment processes [111]. Overall, this study makes keratin-based materials one of the potential alternatives to conventional adsorbents, supporting their applicability in eco-friendly wastewater management technologies.

Limitations and future perspectives

Even though the present work effectively optimized keratin extraction and showed a significant removal efficiency for CR dye in model aqueous solutions, many challenges must be addressed for the industrial scale. The predominant focus on synthetic mono-pollutant systems does not adequately capture the chemical complexity of real industrial wastewater. In particular, the KB’s reusability, regenerative potential, and long-term operational stability were beyond the scope of the current work. Additionally, because this study was limited to batch-mode studies, further research is required into the adsorbent’s performance in continuous-flow systems, as well as its entire lifecycle environmental impact. Future studies will assess KB’s performance with real commercial textile effluents, emphasizing competitive adsorption in multi-ion systems. To facilitate practical implementation, these future studies will be undertaken in a gradual approach, starting with material optimization and regeneration research to real effluent testing and continuous flow operation, followed by textural characterization and life cycle assessment.

Conclusion

The study successfully optimized the extraction of keratin from CFs, addressing a critical environmental challenge posed by feather waste while highlighting its potential as a valuable polymer. Characterization techniques including SEM, XRD, FTIR, TGA, and NMR, confirmed the high purity and structural integrity of the extracted keratin, validating the effectiveness of the optimized extraction method. The present study demonstrated the effectiveness of KBs in removing CR dye with an efficiency of ~80% under optimized conditions. The Freundlich isotherm model (R² = 0.9991) was a better fit than the Langmuir model (R² = 0.9608) to represent equilibrium data, indicating the heterogeneous features of the surface and multilayer adsorption behaviour. The preferential adsorption was indicated by the Freundlich constant n (> 1), which is consistent with the existence of large numbers of functional groups on the keratin surface. Kinetic studies showed that the adsorption process followed the pseudo-second-order model using a correlation coefficient (R² = 0.997), indicating the key role of chemisorption. The pseudo-first-order model showed a lesser agreement (R² = 0.8277), but the intraparticle diffusion study showed that adsorption takes place through multiple mechanisms, including surface adsorption and pore diffusion, and not only one rate-limiting step. The extraction of the keratin involving Na₂S was performed in a safe and controlled laboratory environment with sufficient ventilation and neutralization of the S residues to ensure the safety of the process, and this also justifies the necessity of the environmentally acceptable scaling up of the industries. Overall, this study establishes keratin as a promising, sustainable, and cost-effective adsorbent material for industrial wastewater treatment, capable of addressing critical challenges associated with dye-laden industries. Future studies focusing on competitive adsorption, regeneration, real effluent treatments and scale-up approaches will further increase its applicability and pave the way for its integration into advanced wastewater technologies.

Commercialization and feasibility perspective

The keratin extraction procedure developed for this study has excellent potential for commercialization as it processes abundant and cheap poultry waste utilizing a mild Na₂S extraction method that can be implemented in continuous processing with potentially minimal environmental impacts. Furthermore, the development of KBs utilizing the optimized extract provides a low-cost and biodegradable adsorbent, thus offering stronger support for a sustainable valorization pathway for poultry waste.

A preliminary SWOT analysis identifies some significant strengths (i.e., low-cost feedstock, environmental process and scale-up potential) as well as opportunities (e.g., possibility of marketable use in wastewater treatment and dye removal). Potential weaknesses were identified in optimization of the process during scale-up, and threats that are mainly focused on competing synthetic adsorbents and the costs of standardizing the process. Overall, the proposed method demonstrates a technical and economic feasibility pathway to valorize feather waste into a high-value product for environmental applications.

Abbreviations

RSM

Response surface methodology

BBD

Box-Behnken design

SEM

Scanning electron microscopy

XRD

X-ray diffraction

FTIR

Fourier transform infrared spectroscopy

TGA

Thermogravimetric analysis

KBs

Keratin beads

CR

Congo red

CFs

Chicken feathers

Na₂S

Sodium sulfide

CaCl₂

Calcium chloride

EDX

Energy-dispersive X-ray spectroscopy

C

Carbon

N

Nitrogen

O

Oxygen

S

Sulfur

DLS

Dynamic light scattering

SA

Sodium alginate

SA-K

Sodium alginate-keratin

VIF

Variation inflation factors

ANOVA

Analysis of variance

Na

Sodium

Al

Aluminium

Cu

Copper

PDI

Polydispersity Index

(-CH2-)

Methylene

(-CH3)

Methyl

NMR

Nuclear magnetic resonance

Sodium bisulfite

NaHSO₃

BSA

Bovine serum albumin

SDGs

Sustainable development goals

H₂S

Hydrogen sulfide

NaOH

Sodium hydroxide

PPE

Personal protective equipment

SDS-PAGE

Sodium dodecyl sulfate–polyacrylamide gel electrophoresis

TCDs

Thermal conductivity detectors

SD

Standard deviation

Author contributions

S.G. - Investigation, Methodology, Data Curation, Formal Analysis, Writing – Original Draft. S.S. - Conceptualization, Supervision, Methodology, Review & Editing.

Funding

The research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.

Data availability

Data is provided within the manuscript.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Use of artificial intelligence (AI) tools

The authors declare that no generative AI tools were used for the creation of figures, data analysis, or scientific conclusions.

Competing interests

The authors declare no competing interests.