Employing nutraceutical industry by-product, cumin seeds spent, for the adsorption treatment of acid blue 113 dye

Cumin seed production is primarily concentrated in India, which accounts for 73% of global production and 80% of its consumption. Other significant producers include Iran (4%), Turkey (3%), and Syria (7%), with the remaining 6% coming from various other countries.¹ Annually, ∼400 000 metric tons of cumin seeds are produced.² The cost of cumin seeds varies based on factors such as quantity, quality, variety, location, season, and demand, typically ranging from $2.2 to $4 per kilogram.³ 

Cumin, the dehydrated seed of the Cuminum cyminum L plant, belongs to the parsley family. The plant features a key stem branching into up to five subordinate divisions, both with 2–3 sections, resulting in a uniform canopy. The stems are branched, glabrous, 3–5 cm in diameter, typically dull green in color, with alternate leaves that are lanceolate, slightly hairy, and bluish-green petioles.⁴ 

The inflorescence of cumin consists of complex umbels of white or pinkish flowers with thread-like leaflets that are pinnate or bipinnate. The small flowers are arranged in umbels with 5–7 umbellates per umbel. The fruit, a schizocarp, is 4–5 mm long and has two mericarps enclosing a single seed. These fruits are oval or fusiform lateral achenes with eight ridges containing oil tubes.⁵ The seeds, similar in arrival to fennel and anise seeds but minor and gloomier, may have noticeable hairs depending on the variety.

The aim of the current study is to subsidize environmental sustainability by utilizing nutraceutical industrial cumin seed waste (NICUS) for bioremediation. Specifically, the study focuses on removing acid blue 113 (AB113) from the effluent of the textile industry. Recent efforts at our research institute have explored creating composites using Nutraceutical Industrial Spent (NIS) as a filler material⁶ and as a bioadsorbent for removing harmful dyes.^7,8 Pilot-scale demonstrations have highlighted the potential of NIS in circular economics.⁹ 

The primary aim was to evaluate whether the cost-effective biosorbent NICUS could effectively remove AB113 from aqueous solutions. This approach highlights the use of nutraceutical industrial spent as an environmentally friendly and affordable solution to mitigate dye toxicity in textile industry wastewater. The study includes modeling studies and a scrutiny of the economics of biosorbent regeneration to support experimental findings.

The Acid Blue 113 (AB113) dye was procured from Sigma-Aldrich in the USA.

NICUS was sourced from a local factory specializing in oleoresin extraction from cumin seeds (Fig. 1). The NICUS underwent a process involving air drying, pulverization, crushing using a ball mill, and sieving rendering to American Society for Testing and Materials (ASTM) standards.

The surface morphology of NICUS was analyzed with a scanning electron microscope (LEO 435 VP model, Japan). The functional groups present in the adsorbent were identified using Fourier Transform Infrared Spectroscopy (FTIR). Infrared spectra were recorded for both control samples (NICUS without AB113 adsorption) and samples loaded with AB113 using an FTIR spectrometer (Inter-spec 2020, Spectrolab, UK). The point of zero charge (pHz) was established as a reference for NICUS.

Due to the occurrence of cellulose and lignocellulose facilities, NICUS exhibits an obviously amorphous and fibrous morphology.¹⁰ Scanning Electron Microscopy (SEM) investigation of the NICUS surface exposed a complex porous structure (Fig. 2). Upon adsorption of the adsorbate (AB113 dye), some of these pores become completely filled, forming a distinct layer on the surface (Fig. 3). The functional groups present in NICUS were identified through IR spectroscopy analysis (Fig. 4). The wideband observed in the IR spectrum of NICUS between 3100 and 3500 cm⁻¹ is due to the hydroxyl groups in cellulose and adsorbed water molecules. Additionally, bands at 1600 cm⁻¹ (C–H stretching) and a weak, sharp band at 3000 cm⁻¹ (C–O stretching) were also detected. Furthermore, bands at 1330, 1300, 1240, 1210, and 1010 cm⁻¹ indicate C–O–C stretching. After the adsorption of AB113 dye onto NICUS, changes in the IR spectra were noted. The broadbands between 3200 and 3550 cm⁻¹, attributed to N–H widening in the –NH₂ group of AB113 dye, disappeared, indicating hydrogen bonding formation between –NH₂ groups of the dye and hydroxyl groups of NICUS. Similarly, the withdrawal of a prominent peak at 1500 cm⁻¹, corresponding to N–N stretching in AB113 dye, further approves resilient adsorption of the dye onto NICUS. Based on the shift in IR absorption frequencies, it can be concluded that AB113 dye has effectively adsorbed onto NICUS. The point of zero charge (pHz) of NICUS was determined to be pH 7.40 (Fig. 5), indicating the pH at which the surface charge of NICUS is neutral.

To achieve maximum adsorption capacity, optimal conditions must be determined. pH significantly influences the surface properties of the adsorbent and the ionization state of the dye molecules in solution, thereby controlling the adsorption process.¹¹ In this study, the highest adsorption volume was observed at pH 2 (Fig. 6).

The number of adsorbent units used directly influences the adsorption process by determining how much adsorbate can be removed under specific operating conditions. In this study, the capacity of NICUS to absorb AB113 dye was investigated across varying amounts of adsorbent (0.02–0.300 g in 50 ml of solution). It was witnessed that increasing the amount of adsorbent resulted in greater removal of AB113 dye.¹² As more NICUS was added, additional AB113 dye molecules were absorbed onto the surface of the adsorbent, beyond what was shown in Fig. 6, where maximum adsorption capacity was observed at pH 2. Eventually, a point of saturation was reached where dye molecules began to bind to the outer layer of the adsorbent once the equilibrium between dye molecules in solution and adsorbed on the adsorbent was achieved. Figure 6 illustrates these findings.

The adsorption of AB113 dye onto NICUS was conducted for different durations—15, 30, 45, 60, 90, 120, 150, and 180 min—to evaluate the effect of contact time. It was observed that rapid adsorption occurred initially, reaching ∼80% within the first 60 min. Subsequently, adsorption continued at a slower rate, reaching equilibrium after about 180 min. These results are depicted in Fig. 6. Dyes molecules gather deeper and more widely inside the adsorbent’s structure as the contact time rises. However, the effectiveness of prolonged contact time diminishes as mesopores become filled, hindering further diffusion of dye molecules into the adsorbent.

The adsorption experiments were conducted over a temperature range of 30–50 °C using three different initial dye concentrations, as illustrated in Fig. 6. It is evident from the results that as the temperature increases, the adsorption volume gradually decreases, representing an endothermic nature of the adsorption process. The dye molecules are more mobile at higher temperatures, characterized by lower activation energy and faster intra-molecular diffusion rates, and may contribute to increased adsorption capacities observed at advanced temperatures.¹³ 

Adsorption isotherm models can be used to describe the adsorption of adsorbate particles onto adsorbent exteriors, each capturing different aspects of the process: Langmuir isotherm model,¹⁴ the dimensionless separation factor R_L,¹⁵ Freundlich isotherm model,¹⁶ and Jovanovic isotherm model.¹⁷ It provides a more flexible approach compared to the classical Langmuir isotherm model. The homogeneity or heterogeneity of the system was not entirely explained by the Langmuir and Freundlich models when it came to the experimental setting of AB113 dye adsorption on NICUS. Therefore, more complex models such as the Jovanovic isotherm (Fig. 7) were explored to better fit the experimental data and understand the adsorption behavior more comprehensively. Table I shows the strategic boundaries of two-structure isotherms.

The Redlich–Peterson isotherm model,¹⁸ Brouers–Sotolongo isotherm model,¹⁹ and Vieth–Sladek isotherm model²⁰ are related and are depicted in Fig. 7 and Table II. The Sips isotherm model,²¹ Toth isotherm model,²² and Radke–Prausnitz isotherm model²³ projected a Q_m value of 1.80 mg g⁻¹, which differed significantly from the experimentally achieved q_e value, indicating limitations in its applicability to this specific adsorption system. Finally, it is important to note that higher-order models are utilized to better describe complex adsorption mechanisms. The determination coefficient R² applies only to linear models and should not be used to validate data fitting for nonlinear models. Table III summarizes the criteria for the three critical parameters (Q_m, χ², and R²). Discrepancies between model-predicted values and experimental data (q_e) are of interest to researchers, particularly in developing new models that can provide a more accurate understanding of adsorption behaviors in systems involving NIS (Nutraceutical Industrial Spent) and dyes.

Possible phases of the adsorption series that control rate were identified through the analysis of kinetic models. AB113 dye concentrations of 100, 200, and 300 ppm were used in the kinetic investigations. The differences in adsorption rates at 303, 313, and 323 K are demonstrated by kinetic tests. Several kinetic models were applied to the adsorption kinetics data for non-linear analysis, including the pseudo-first order,²⁴ pseudo-second order,²⁵ Weber–Morris intraparticle diffusion model,²⁶ Dumwald–Wagner model,²⁷ and film diffusion model²⁸ (MS Excel 2010). The computed limits are shown in Table IV.

Factors of persistence (R²) and chi-square values (χ²) show that at all preliminary AB113 dye concentrations (100, 200, and 300 ppm) at varied temperatures, the pseudo-second order model suited the experimental data in Fig. 8 restored than the pseudo-first order model. After reaching its peak, the degree of adsorption progressively dropped until it came to a stop. As the temperature rose, the adsorption limit grew (q_e). These results validate that there is no rate limitation in the adsorption operations. Additionally, the results suggest that the solute particles moved from the mass response for the strong surface to the NICUS pores during the initial stages of the adsorption interaction.

The Dumwald–Wagner model [Fig. 9(a)] calculates the true absorption rate constant (K), accounting for practical dispersion parameters (Table V). In the Weber–Morris model [Fig. 9(b)], over the course of the interaction, the solute approval fluctuates with t½. Adsorption energy is often managed by a variety of factors. Our investigational results demonstrate that there are multiple linearity levels at every solute focus. The adsorption rate is strong at lower temperatures and lower initial fixations (100 ppm). The pace eventually equalizes in terms of time after proceeding in a different direct direction. Nevertheless, the amount is more direct at higher temperatures. At higher solute fixations (300 ppm), there is less noticeable variation in the adsorption rate. This is especially clear when the Boyd et al. film dispersion model is used to fit data at higher temperatures. Figure 9(c) shows excellent values for R² and χ², indicating that the data fit the model well, and it gives the liquid film diffusion constant R| (Table V). This demonstrates that at advanced temperatures, diffusional constraints only slightly slowed the rate of adsorption. Distribution is a rate-restricting interaction, as one might expect. The reason why the retention rates change over time is because after it is rapidly ingested, the solute coats the outer layer of the particles.

An important role for energy and entropy is assumed in the connection cycle design. Figures 10 and 11 illustrate how to calculate ∆H° and ∆S°.

Table VI presents the evaluations for the thermodynamic parameters. A positive result for ∆S° demonstrates the acute taste of the AB113 color for the adsorbent as well as the extended randomization at the surface of the strong arrangement. The unusually low upsides of ∆H° imply that the adsorption component is physical, since enthalpy changes of >200 kJ mol⁻¹ are frequently observed during material activities. Table VI shows that, using the dynamic parameters from the pseudo-second-order model and the Arrhenius equation, the activation energy values for the adsorption process increased from ∼100.29 to 1055.99 kJmol⁻¹ at different initial concentrations (100, 200, and 300 ppm).

Discovering and adapting a reasonable model regression to the experimental data obtained through the use of a fractional factorial experimental design (FFED) was the most crucial step in the streamlining process.

The optimal conditions for the statistical optimization experiment included a pH of 2.2, an adsorbent concentration of 6.000 gL⁻¹, and an initial dye concentration of 693.00 mgL⁻¹. The adsorption process was carried out for 132.00 min with orbital shaking at 165 rpm and a temperature of 55 °C. The supreme adsorption consequence was 810.50 mg g⁻¹.

Statistical process optimization allows one to evaluate the effect of the process parameters on adsorption as well as calculate the ideal condition within a predefined range of parameter values. Adsorption capacity improves with time, as shown by 3D graphs that link time with other variables. By optimizing temperature, duration, particle size, and dye concentration, we can accelerate the adsorption process. The maximum adsorption was achieved at 132 minutes. Adsorption capacity benefits from an increase in temperature. Temperature increases and the passage of time both increase adsorption capability. The optimal pH for increased adsorption capacity is about 2.2; above this, increasing the pH has a negative effect on the sorption size. Positive values signify an increasing effect; for example, an increase in temperature (B) causes a substantial increase in sorption capacity. As a result, the surface and contour plots offer a visual depiction of how two parameters together affect biosorption (Fig. 13).

The quadratic ideal designed for method development has been shown to be useful in anticipating the maximum adsorption limit and understanding the interactions between self-regulating components and their effects on the adsorption cycle. Measurable progress resulted in an ∼70.5% rise in adsorption from 475 to 810 mg g⁻¹.

The adsorbed substance can be recovered by recycling dye-loaded NICUS. This solution might not be cost-effective (less than $1 US for 10 kg of NICUS), since the solvent and technique will cost significantly more than the adsorbent employed in the process. Additionally, it will increase the E-factor, which is harmful considering the massive number of toxins in the environment (Sheldon, 1992).

At equilibrium conditions, the maximum experimental adsorption capacity (q_e) achieved for NICUS and AB113 dye was determined to be 96 mg g⁻¹. This capacity was investigated using nine different isotherm models to gain a comprehensive understanding of the adsorption process. Based on the findings, the adsorption kinetics followed the pseudo-second-order model, representing that the interactions between NICUS and AB113 dye were primarily physical in nature. This suggests that the adsorption process involved chemisorption or strong van der Waals forces between the dye particles and the adsorbent external. The laboratory testing validated the efficacy of NICUS in treating textile sector effluents, highlighting its potential for practical applications in environmental clean-up. This ongoing study establishes NICUS derived from nutraceutical industrial cumin seed as a promising material for adsorption processes. Moreover, the study underscores the importance of modeling studies in optimizing adsorption processes across various scales, from laboratory experiments to industrial applications. By employing statistical optimization techniques, the research contributes to advancing the efficiency and effectiveness of using NICUS for environmental remediation purposes.

Authors would like to acknowledge the support provided by Researchers Supporting Project Number RSP2024R424 King Saud University, Riyadh Saudi Arabia.

The authors have no conflicts to disclose.

Syed Noeman Taqui: Conceptualization (equal); Data curation (equal). Usman Taqui Syed: Conceptualization (equal); Data curation (equal). Sameer Algburi: Conceptualization (equal); Formal analysis (equal). Rayees Afzal Mir: Conceptualization (equal); Data curation (equal). Akheel Ahmed Syed: Conceptualization (equal); Formal analysis (equal). Abdullah I. Al-Mansour: Formal analysis (equal); Investigation (equal). Shamshad Alam: Conceptualization (equal); Formal analysis (equal). Mohammad Amir khan: Conceptualization (equal); Data curation (equal). Shareefraza J. Ukkund: Conceptualization (equal); Data curation (equal); Formal analysis (equal).

The data that support the findings of this study are available from the corresponding authors upon reasonable request.