Natural mulberry leaves and carboxylic acid-modified mulberry (Morus alba L.) leaves were used for the first time to scrutinize the effects of modification on the retention efficiency of an anionic dye (Remazol Brilliant Blue R (RBBR)) from aqueous solutions to suggest an economical and promising adsorbent for the treatment of dye-contaminated water. The characterization of the adsorbents was accomplished through common techniques including SEM, FTIR, and pHpzc determination. Several parameters studied in batch experiments pointed out that the initial pH of 2.0 and the contact time of 240 min were optimum conditions for all the developed RBBR uptake processes. An artificial neural network (ANN) model was applied to formulate a forecast model for the uptake efficiency of RBBR. The experimental data were assessed by different kinetic and isotherm models to explain the mechanism of the developed processes in more detail. Maximum monolayer adsorption capacities of natural mulberry leaves and acetic acid-, citric acid-, and oxalic acid-modified mulberry leaves were determined as 64.5, 95.2, 84.8, and 91.7 mg g−1, respectively, by the Langmuir isotherm model. These results demonstrated that the modification with carboxylic acids significantly increases the anionic dye adsorption capacity of the mulberry leaves.

  • Natural mulberry leaves and carboxylic acid-modified mulberry leaves were used for the adsorptive removal of an anionic dye for the first time.

  • An economical and promising adsorbent was developed for industrial wastewater treatment applications.

  • Modification with carboxylic acids significantly increases the anionic dye adsorption capacity of the mulberry leaves.

  • ANN model is suitable for the prediction of RBBR adsorption efficiency.

The employment of dyes in industrial branches has become quite popular with the development of industry and technology. However, the production of new dyestuffs and detailing of the color scale actually increased the rate of the usage of synthetic dyes that are harmful to nature. Synthetic dyes that are extremely resistant to biodegradation by microorganisms are commonly employed for product coloration in paper, textile, tanning, leather, plastic, rubber, pharmaceutical, food, and cosmetic industries (Vijayaraghavan & Yun 2008; Ali 2010; Ahmad et al. 2014). Considering the volume and composition of discharged wastewater, the textile industry has become the most polluting industrial sector for the environment and represents a challenge to conventional biological, physical, or chemical treatment methods (Wang et al. 2006). The release of synthetic dye-contaminated industrial wastewater into surface waters induces serious irreversible damage to both living organisms and the environment (Shojaei et al. 2021).

Remazol brilliant blue R (RBBR), an anionic dye derived from anthraquinone, is extensively employed in several industrial activities, especially in the production of polymeric dye, and it causes harmful effects on human metabolism by reaching both aquatic organisms and the food chain even at trace amounts (Mechichi et al. 2006). Removal of RBBR and other hazardous dyes from industrial wastewater – before discharging step – is essential for the safety of the environment and all living organisms. There are several physical, chemical, or biological methods, including adsorption, sedimentation, floatation, flocculation, coagulation, reverse osmosis/ultrafiltration, electrolysis, ion exchange, ozonation, and anaerobic digestion for dye-contaminated wastewater treatment (Ewuzie et al. 2022). Among numerous methods, adsorption has attracted attention due to its fast application, low cost, efficiency, environmental friendliness, as well as its applicability in the treatment of synthetic dyes from large volumes of wastewater generated from worldwide industrial applications (Vakili et al. 2014; Liu et al. 2018). The efficiency and applicability of the adsorption process depend directly on the properties of the adsorbent used. Numerous agricultural and industrial wastes or some other materials including bagasse fly ash (Chumpiboon et al. 2022), shale oil ash (Miyah et al. 2021), cellulose-based activated carbon (Zhang et al. 2022), almond shell (Ozdes et al. 2010; Senturk et al. 2010; Duran et al. 2011), rice bran-based magnetic composite (Ma et al. 2020), metal oxide nanocomposites (Hong & Wang 2017; Reghioua et al. 2021), clay (Ozdes et al. 2014; Jawad et al. 2022), and laccases (Herkommerová et al. 2018) have been employed as an adsorbent for the retention of various synthetic dyestuffs.

Mulberry (Morus alba L.) can be grown in many regions due to its high adaptability to different soil and climate conditions (Karadeniz & Osma 2019). Both natural and modified mulberry leaves have been applied by previous researchers in the adsorption of various dyestuffs. Siraorarnroj et al. (2022) have prepared nanoporous carbon of the mulberry leaves through hydrothermal-carbonization with microwave processing using a chemical reagent (MPC) for the adsorption of methyl orange (MO). Khan & Farooqui (2022) have employed mulberry leaves biochar for the retention of methylene blue (MB) from aqueous solutions. Modification of adsorbents with chemical substances, especially different types of acids, improves their adsorption capacity and enables more effective adsorbents to be obtained. For this purpose, carboxylic acids are widely applied. Tian et al. (2020) have prepared carboxylic acid-modified bamboo powder for Pb(II) ions retention from aqueous media. Nayak & Pal (2021) have modified Abelmoschus esculentus seeds with oxalic, citric, and tartaric acids to enhance the adsorption of gentian violet. Guo et al. (2022) have used acrylic acid-modified walnut shells for the uptake of Rhodamine B.

The present investigation aims to propose a cost-efficient alternative for the adsorption of RBBR from aqueous media by employing carboxylic acid (acetic acid (AA), citric acid (CA), and oxalic acid (OA))-modified mulberry leaves. According to our literature survey, carboxylic acid-modified mulberry leaves were employed for the first time for the retention of a dyestuff in the present research. There are millions of tons of mulberry leaves as agricultural waste all over the world. Mulberry leaves, which have no usage area and are just thrown away, were reprocessed with outstanding environmental practice in the present study. The application of a three-layer artificial neural network (ANN) model was illustrated to model and predict the adsorption efficiency of RBBR. The independent experimental conditions including solution pH, contact time, adsorbent amount, and initial RBBR concentration were applied as input parameters to train the neural network while adsorption percentage, a dependent variable, was considered as the output layer of the neural network. The best experimental conditions were evaluated and optimized for RBBR retention. In addition, the experimental data were analyzed by different kinetics and isotherm models to evaluate the adsorption mechanism.

Reagents and instrumentation

RBBR, which has a chemical formula of C22H16N2Na2O11S3, a molar mass of 626.53 g mol−1, and a maximum absorbance value of 592 nm, was used as an adsorbate in the present research. All of the chemicals used in the study including RBBR, OA, AA, CA, NaOH, and HNO3 were of analytical purity and obtained from Fluka (Buch, Switzerland) or Merck (Darmstadt, Germany). The working and calibration solutions were obtained by the dilution of the stock RBBR solution, at a concentration of 5,000 mg L−1. Distilled water was used in all stages of the experiments. Diluted NaOH and HNO3 solutions were used for adjusting the initial pH of RBBR solutions. All glassware used in the experimental studies was kept in 5% (v/v) HNO3 solution for 12 h, and then washed with tap water and distilled water, respectively. The RBBR concentration in aqueous solution was detected with Perkin Elmer Lambda 25 model UV–Vis spectrophotometer. Hanna pH-2221 model desktop pH meter to adjust the pH values of the solutions, Edmund Bühler GmbH model mechanical shaker for adsorption experiments, and BOECO S-8 model centrifuge apparatus to separate the adsorbent from the solutions were used. Fourier Transform Infrared Spectrometer (Perkin Elmer 1600 Series) was benefitted to illuminate the functional groups of natural and carboxylic acid-modified mulberry leaves while their morphological structure was elucidated by Scanning Electron Microscope (ZIESS Evo Ls 10).

Preparation of the adsorbents

For the adsorption of RBBR from an aqueous solution, mulberry (M. alba L.) leaves modified with AA, CA, and OA were used. Mulberry leaves were picked up from the countryside of Gümüşhane, a city in the Blacksea region of Türkiye. The leaves were washed with tap water and distilled water several times to remove any environmental dust and then they were dried in an oven at 80 °C for 24 h. A blender was used to grind the leaves and particles smaller than 150 μm (<150 μm) were used in the modification process. The modification of mulberry leaves with carboxylic acids including AA, CA, and OA was implemented according to Tian et al. (2020) with a little change. Briefly, 5.0 g of dried mulberry leaves were mixed with 0.2 M of 100 mL AA, CA, and OA solutions, separately. The stirring of these suspensions was carried out at 120 rpm for 24 h at room temperature. At the end of this process, the mixtures were filtrated through 0.45 μm pore-sized filter paper and the adsorbents on the filter paper were washed with distilled water several times. The modified adsorbents were dried in an oven at 80 °C and then kept in glass bottles until the experimental studies. The natural mulberry leaves and AA-, CA-, and OA-modified mulberry leaves were named NML, AAM, CAM, and OAM, respectively.

Batch adsorption experiments

The batch method was performed to assess the influences of experimental variables such as initial solution pH, equilibrium time, initial RBBR concentration, and adsorbent concentration on the adsorption efficiency of RBBR onto NML, AAM, CAM, and OAM. In a typical procedure, each of the adsorbents was weighed in the range of 10–200 mg, separately in 15 mL of polypropylene centrifuge tubes. Then, 10 mL of RBBR solutions at an initial pH value of 2.0 in the concentration range between 50 and 1,000 mg L−1 were added and the mixtures obtained were mixed in a mechanical shaker at 400 rpm in time intervals of 1–480 min. Based on the results of our previous study, a shaking speed of 400 rpm was found to be sufficient for the adsorption (Ozdes et al. 2010). Therefore, shaking speed was not optimized in the present study and all of the experiments were carried out at a constant shaking speed. At the end of the determined times, the centrifugation was carried out at 1,180 g for 5 min to separate the adsorbent from the solution. The RBBR concentration remaining in the solution was measured with UV–Vis Spectrophotometer at a wavelength of 592 nm. All experiments were performed in triplicate. The amount of RBBR adsorbed by per gram of adsorbent (qe (mg g−1)) was assessed using Equation (1):
formula
(1)
where Co and Ce are the initial and equilibrium concentrations of RBBR (mg L−1), respectively, V is the volume of the aqueous solution (L), and ms is the dry mass of adsorbent (g) used in the experiment.

ANN modeling

In order to get a prediction model for the adsorption of RBBR onto AAM, CAM, and OAM, artificial neural networks (ANNs) were operated by utilizing the Neural Network Toolbox in MATLAB R2017b mathematical software. ANN involves three main layers as input, hidden, and output layers. In the present research, the solution pH, contact time, adsorbent amount, and initial RBBR concentration were preferred as input parameters and the adsorption percentage was designated as the output parameter. In this regard, a three-layer backpropagation (BP) neural network model with a tangent sigmoid transfer function (tansig) at the hidden layer and a linear transfer function (purelin) at the output layer was practiced. In order to train the network, Levenberg–Marquardt back propagation (LMB) algorithm was performed (Khajeh et al. 2015; Ghadirimoghaddam et al. 2021). The performance of the network was evaluated by the correlation coefficient (R2) used as a function of the error. The optimum architecture of the ANN model was constructed as 4-8-1 as given in Supplementary material, Figure S1.

Characterization of the adsorbents

In order to investigate the functional groups on both NBL and carboxylic acid-modified mulberry leaves, the FTIR spectrum of each adsorbent was acquired in the range of 4,000–650 cm−1. FTIR spectra of the modified adsorbents were not significantly different from the spectrum of natural adsorbents, revealing that the modification process did not cause any change in the surface functional groups. A band at 3,290 cm−1 is associated with –OH groups due to the presence of alcoholic, phenolic, and carboxylic groups. The aliphatic C–H stretching vibration is noticed at 2,920 and 2,850 cm−1. The C = C stretching vibration and C = O bond of the carboxylic group are seen at 1,620 and 1,730 cm−1, respectively. The C–O stretching vibration is noticed at 1,027.5 cm−1 (Malook & Khan 2020) (Figure 1). The surface morphology of NML, AAM, OAM, and CAM was scrutinized by SEM analyses. When SEM structures are examined, it is noticed that NML is more porous than the modified adsorbents. After the modification process, the existing porous structures disappeared significantly and the surface of NML was covered with carboxylic acids (Figure 2). In addition to these, the point of zero charge (pHpzc) of NML, AAM, CAM, and OAM was ascertained as 6.0, 5.1, 3.9, and 7.1, respectively.
Figure 1

FTIR spectrum of (a) natural mulberry leaves (NML); (b) acetic acid-modified mulberry leaves (AAM); (c) citric acid-modified mulberry leaves (CAM); and (d) oxalic acid-modified mulberry leaves (OAM).

Figure 1

FTIR spectrum of (a) natural mulberry leaves (NML); (b) acetic acid-modified mulberry leaves (AAM); (c) citric acid-modified mulberry leaves (CAM); and (d) oxalic acid-modified mulberry leaves (OAM).

Close modal
Figure 2

SEM micrographs of (a) NML; (b) AAM; (c) CAM; and (d) OAM.

Figure 2

SEM micrographs of (a) NML; (b) AAM; (c) CAM; and (d) OAM.

Close modal

Influences of initial pH on the adsorption efficiency of RBBR

The initial pH of aqueous solution acts a significant part in the adsorption process since the chemical structure of the adsorbate and the surface charge of the adsorbent are directly affected by the solution pH. Therefore, the uptake efficiency of RBBR on NML, AAM, CAM, and OAM was tested in the initial solution pH range of 2.0–8.0, at initial RBBR concentration of 100 mg L−1 by using 5.0 g L−1 g of each adsorbent. As the initial solution pH is increased from 2.0 to 8.0, the qe values appear to decrease from 18.4 to 2.1 mg g−1 for NML, from 19.1 to 5.7 mg g−1 for AAM, from 19.4 to 7.4 mg g−1 for CAM, and from 19.2 to 6.7 mg g−1 for OAM (Figure 3). At low pH values (pH < pHpzc), adsorbent surfaces are surrounded by H3O+ ions, and their surface functional groups are positively charged (–COOH2+). Therefore, electrostatic interaction occurs between the anionic RBBR molecules and the positively charged adsorbent surfaces, which increases the uptake efficiency. On the contrary, with the increase in the initial solution pH value (pH > pHpzc), the negative charge density on the adsorbent's surface (COO) begins to increase and accordingly, the electrostatic repulsion between RBBR molecules and adsorbent surfaces increases, which decreases the adsorption efficiency (Moharami & Jalali 2013). The pHpzc values of NML, AAM, CAM, and OAM were determined as 6.0, 5.1, 3.9, and 7.1, respectively and it is expected that RBBR, which is an anionic dye, will be removed with the highest efficiency at lower pH values than these values. As a result, further experiments were planned by considering the optimum initial solution pH as 2.0 for all types of adsorbents.
Figure 3

Influences of initial pH on RBBR uptake.

Figure 3

Influences of initial pH on RBBR uptake.

Close modal

Evaluation of equilibrium time and adsorption kinetics

Removal of hazardous dyes from the wastewater should be economical for real industrial applications, since large amounts of industrial wastewater are being discharged into the environment. Therefore, it is critical to ascertain the minimum time at which maximum dye retention occurs. In order to evaluate the optimum equilibrium time for RBBR adsorption; 0.05 g (5.0 g L−1) of AAM, CAM, and OAM were treated separately with 100 mg L−1 of RBBR solution at different contact times in the range of 1–480 min. On completion of each specified period, the adsorbents were separated from the solution through centrifugation and the levels of unadsorbed RBBR in the solution were determined by UV–Vis Spectrophotometer. The RBBR amounts (qt) adsorbed by 1 g of AAM, CAM, and OAM at different time intervals were calculated. In the initial steps of the adsorption process, the adsorption occurred rapidly due to the open adsorption sites on the surface of the adsorbents. Over time, the uptake rate of RBBR decreased as the adsorbent pores were filled, and after 240 min of contact time, the equilibrium was achieved due to the complete saturation of the adsorbent surface. As a result, the sufficient time for the adsorption of RBBR onto all adsorbent surfaces was determined as 240 min for subsequent studies (Supplementary material, Figure S2). Equilibrium time was optimized as 60 min for MO adsorption onto MPC (Siraorarnroj et al. 2022), and as 35 min for MB adsorption with biochar obtained from mulberry leaves (Khan & Farooqui 2022).

The pseudo-first order (PFO), pseudo-second order (PSO), and intra-particle diffusion (IPD) models were applied to the experimental data to determine the adsorption rate and to explain the adsorption mechanism of RBBR on AAM, CAM, and OAM. The linear rate equations for PFO (Lagergren 1898), PSO (Ho & McKay 1998) and IPD (Weber Jr & Morriss 1963) models are presented in Equations (2), (3), and (4), respectively:
formula
(2)
formula
(3)
formula
(4)
where qe (mg g−1) and qt (mg g−1) are the amounts of the RBBR adsorbed on AAM, CAM and OAM at equilibrium and at any time t (min), respectively. k1 (min−1), k2 (g mg−1min−1), and kid (mg g−1min−1/2) are the rate constants of the PFO, PSO, and IPD models, respectively. The magnitude of C provides some indication about the thickness of the boundary layer. According to the PFO model, k1 and qe values are acquired from the slope and intercept of the plot of ln(qeqt) versus t, respectively. Considering the PSO model, qe and k2 values are calculated from the slope and intercept of the figure drawn between t/qtt, respectively (Supplementary material, Figure S3). On the other hand, IPD model parameters kid and C are observed from the slope and intercept of the plot of qt versus t1/2, respectively. The resulting constants for each model are revealed in Table 1, along with the related correlation coefficients (R2).
Table 1

Kinetic model parameters for RBBR adsorption

Type of adsorbent
AAMCAMOAM
qe,exp 19.4 19.2 19.6 
PFO 
k1 −0.014 −0.014 −0.018 
qe,cal 12.3 9.73 10.1 
R2 0.940 0.940 0.958 
PSO 
k2 0.00302 0.00483 0.00502 
qe,cal 20.1 19.6 20.0 
R2 0.999 0.999 0.999 
IPD 
kid,1 2.24 2.76 2.77 
R2 0.966 0.999 0.994 
kid,2 0.97 0.46 0.79 
R2 0.999 0.964 0.981 
C 5.53 7.31 7.23 
Type of adsorbent
AAMCAMOAM
qe,exp 19.4 19.2 19.6 
PFO 
k1 −0.014 −0.014 −0.018 
qe,cal 12.3 9.73 10.1 
R2 0.940 0.940 0.958 
PSO 
k2 0.00302 0.00483 0.00502 
qe,cal 20.1 19.6 20.0 
R2 0.999 0.999 0.999 
IPD 
kid,1 2.24 2.76 2.77 
R2 0.966 0.999 0.994 
kid,2 0.97 0.46 0.79 
R2 0.999 0.964 0.981 
C 5.53 7.31 7.23 

The qe,exp values for AAM, CAM, and OAM were calculated as 19.4, 19.2, and 19.6 mg g−1, respectively. As a result of applying the PFO model to the experimental data, qe,cal values for AAM, CAM, and OAM were found to be 12.3, 9.73, and 10.1 mg g−1 and R2 values were 0.940, 0.940, and 0.958, respectively. By the application PSO model, qe,cal values were found to be 20.1, 19.6, and 20.0 mg g−1 for AAM, CAM, and OAM, respectively, and R2 values are higher than 0.999 for all three adsorbents (Supplementary material, Figure S3). Considering these data, it is noticed that the PSO model is dominant in the adsorption mechanism of RBBR onto AAM, CAM, and OAM because of the proximity of the qe values computed experimentally with the qe values observed by the application of the model. On the other hand, the higher R2 values were achieved by the PSO model. This suggests that chemisorption is effective in the adsorption mechanism (Amiri et al. 2019). Similar results were obtained by İnal & Erduran (2015) for the adsorption of RBBR on a different type of hydrogel bead and by Zhong et al. (2012) for the application of activated carbon prepared by microwave assisted H3PO4 activation of peanut hull in RBBR removal. In addition to these observations, according to the IPD model, a noticeable multilinear plot (qt versus t1/2) demonstrates that the adsorption process occurs in three main steps. The first step is explained by film diffusion, which corresponds to the transport of dye molecules from the bulk solution to the external surface of the adsorbent. The second step controls the adsorption rate by pore diffusion (intraparticle diffusion) which is attributed to a gradual adsorption occurring from the external surface into the pores of the adsorbent. The third step is explained by the rapid adsorption of dye molecules onto the active sites of the pores' internal surfaces and does not generally determine the adsorption rate (Wang et al. 2005; Hameed & El-Khaiary 2008). In order to decide which of these stages is dominant in the RBBR adsorption mechanism, the rate constant is calculated for each stage. It is known that the stage with the smallest rate constant value is effective on the mechanism. The rate constant (kid) values calculated for each stage are given in Table 1. Since the final stage, which is the equilibrium state, occurs very quickly, the rate constant has high values. Therefore, the effect of this step on the mechanism is neglected. It is seen that the rate constant values obtained from the intraparticle diffusion stage for all three adsorbents are lower than the rate constant values calculated from the film diffusion stage (Table 1). Therefore, it is thought that intraparticle diffusion is effective in the adsorption mechanism of RBBR. However, another parameter to be considered at this point is the C constant. The calculated C constant is nonzero for all adsorbents. Therefore, it can be concluded that film diffusion and intraparticle diffusion are effective together in the adsorption of RBBR on AAM, CAM, and OAM (Bensalah et al. 2017).

Influences of initial dye concentration on RBBR uptake and evaluation of adsorption isotherms

In order to assess the impact of initial dyestuff concentration on the adsorption efficiency of RBBR on NML, AAM, CAM, and OAM and also to obtain the adsorption isotherm data, the RBBR solutions with initial concentrations of 50–1,000 mg L−1 were treated separately with 5.0 g L−1 of adsorbent for 240 min of contact time. Results of the experimental studies indicated that as the RBBR concentration is increased from 50 to 1,000 mg L−1, the uptake amount increases from 9.77, 9.46, 9.67, and 9.90 mg g−1 to 64.8, 91.5, 82.8, and 92.2 mg g−1 for NML, AAM, CAM, and OAM, respectively (Figure 4). The increase in RBBR concentration at constant adsorbent amounts causes a concentration gradient, a driving force for adsorption, and hence the RBBR adsorption amount onto per gram of adsorbent (qe) increases. On the contrary, at high RBBR levels, the percentage of adsorption is expected to decrease due to over-saturated active adsorption surfaces on the adsorbent (Sharma et al. 2010). Namely, as RBBR concentration is increased from 50 to 1,000 mg L−1, the removal percentages decrease from 97.7, 94.6, 96.7, and 99.8% to 32.4, 43.2, 39.4, and 44.3% for NML, AAM, CAM and OAM, respectively (Figure 4).
Figure 4

Impact of initial RBBR concentration on its retention.

Figure 4

Impact of initial RBBR concentration on its retention.

Close modal

Interactions between the adsorbent and adsorbate can be interpreted by some useful data attained from the adsorption isotherms. Langmuir, Freundlich, and Dubinin–Radushkevich (D–R) isotherm models were utilized for describing the characteristics of RBBR adsorption onto NML, AAM, CAM, and OAM at equilibrium conditions. In the adsorption process, specific homogeneous sites play the role of monolayers which prevents the interactions between the adsorbed species and is the characteristic of Langmuir model (Langmuir 1918). On the other hand, according to the Freundlich model, heterogeneous surfaces which have different energy are available on the adsorbent for multilayer adsorption (Freundlich 1906). The D–R model also provides extra information about the type of adsorption (Dubinin & Radushkevich 1947).

Langmuir model in linear form (Equation (5)) includes qe (mg g−1); the amount of RBBR adsorbed per unit adsorbent mass, Ce (mg L−1); the level of RBBR at equilibrium, b (L mg−1); free energy of the adsorption, and qmax (mg g−1); adsorption capacity. Herein, the intercept and the slope of Ce/qe versus Ce plot were used to obtain b and qmax, respectively:
formula
(5)
The feasibility of an adsorption process is estimated by evaluating the non-dimensional parameter RL, which is calculated with the aid of Equation (6). RL values between 0 and 1 reveal favorable adsorption (Senturk et al. 2010). Here, C0 (mg L−1) is the initial concentration of adsorbate and b (L mg−1) is the Langmuir constant:
formula
(6)
The linear form of the Freundlich isotherm model is expressed by Equation (7):
formula
(7)
where Kf (mg g−1); indicate the adsorption capacity, n; a unitless constant representing the adsorption density. To obtain the numerical values of Kf and n, lnqe versus lnCe a graph is plotted and the constants, n and Kf, are obtained from the slope and intercept of this graph, respectively. The fact that the n constant is in the range of 1–10 indicates the suitability of the developed process. Further information on the adsorption mechanism of RBBR is available from D–R model presented in Equation (8):
formula
(8)
formula
(9)
formula
(10)
where qe is the amount of RBBR adsorbed onto per unit mass of adsorbent (mol g−1), qm is the monolayer adsorption capacity (mol g−1), b is the activity coefficient related to the mean adsorption energy (kJ2 mol−2), ε is the Polanyi potential and Ce is the equilibrium RBBR concentration (mol L−1) in aqueous solution. qm and β are the D–R model constants, observed from the intercept and the slope of the linear plot of lnqe versus ε2, respectively. Mean adsorption energy, E (kJ mol−1), which provides an idea about the adsorption mechanism, can be obtained by using Equation (10). Namely, the E value being in the range of 8–16 kJ mol−1 indicates that the adsorption process mainly proceeds through the ion exchange mechanism. The E value lower than 8 kJ mol−1 indicates physical adsorption, while a value greater than 16 kJ mol−1 demonstrates that chemical adsorption has occurred (Helfferich 1962; Mosai et al. 2020).
Figure 5 provides a comparison of equilibrium isotherms between experimental and theoretical data. The fact that the R2 values based on the Langmuir model for both natural and modified adsorbents are higher than the Freundlich model indicates that RBBR adsorption occurs in the form of a monolayer on the homogeneous adsorbent surface. Similar results were also obtained in the MO adsorption process on the MPC surface (Siraorarnroj et al. 2022). The RBBR adsorption capacity of NML, AAM, CAM, and OAM was achieved as 64.5, 95.2, 84.8, and 91.7 mg g−1, respectively, using the Langmuir equation (Table 2). These results demonstrated that the modification with carboxylic acid significantly increases the anionic dyestuff adsorption capacity of mulberry leaves. The adsorption capacity of the proposed adsorbents is superior to that of most of the other difficult-to-prepare and costly sorbents listed in Table 3 (Iqbal & Saeed 2007; Ada et al. 2009; İnal & Erduran 2015; Li et al. 2015; Chinoune et al. 2016; Abbasi 2017; Saputra et al. 2017; Mate & Mishra 2020). Apart from these adsorbents, the MO and MB adsorption capacities of mulberry leaves modified with different methods were determined as 0.0181 mg g−1 (Siraorarnroj et al. 2022) and 6.55 mg g−1 (Khan & Farooqui 2022), respectively. Between 50 and 1,000 mg L−1 initial RBBR concentrations, the RL values were in the range of 0.180–0.011 for NML, 0.294–0.020 for AAM, 0.217–0.014 for CAM and 0.130–0.007 for OAM, demonstrating the suitability of the developed adsorption method. The fact that the n values obtained from the Freundlich model fall within the range of 1–10 also confirms the suitability of RBBR retention on adsorbents (Le et al. 2019). The E values calculated using the D–R model were in the range of 8–16 kJ mol−1 in all cases, pointing out that the adsorption mechanisms of RBBR probably occur via the ion exchange mechanism (Thamilarasi et al. 2018).
Table 2

Isotherm parameters for RBBR adsorption

Type of adsorbent
NMLAAMCAMOAM
Langmuir isotherm 
qmax 64.5 95.2 84.8 91.7 
B 0.091 0.048 0.072 0.134 
R2 0.998 0.999 0.999 0.999 
Freundlich isotherm 
Kf 13.2 9.16 11.6 18.1 
n 3.71 2.45 2.89 3.44 
R2 0.834 0.880 0.876 0.875 
D–R isotherm 
qm 11.9 18.4 15.6 14.8 
β  0.0023 −0.0035 −0.0029 −0.0023 
E 14.7 12.0 13.1 14.7 
R2 0.879 0.922 0.921 0.919 
Type of adsorbent
NMLAAMCAMOAM
Langmuir isotherm 
qmax 64.5 95.2 84.8 91.7 
B 0.091 0.048 0.072 0.134 
R2 0.998 0.999 0.999 0.999 
Freundlich isotherm 
Kf 13.2 9.16 11.6 18.1 
n 3.71 2.45 2.89 3.44 
R2 0.834 0.880 0.876 0.875 
D–R isotherm 
qm 11.9 18.4 15.6 14.8 
β  0.0023 −0.0035 −0.0029 −0.0023 
E 14.7 12.0 13.1 14.7 
R2 0.879 0.922 0.921 0.919 
Table 3

Comparison of the maximum RBBR adsorption capacities of various adsorbents

AdsorbentQmax (mg/g)Reference
ZnO fine powder (Z300) 89.3 Ada et al. (2009)  
ZnO fine powder (Z075) 38.9 Ada et al. (2009)  
Sodium alginate/poly(N-vinyl-2-pyrrolidone) 55.3 İnal & Erduran (2015)  
Bentonite coated with Mg(OH) 66.9 Chinoune et al. (2016)  
Amino-functionalized organosilane 21.3 Saputra et al. (2017)  
Borax cross-linked Jhingan gum hydrogel 9.88 Mate & Mishra (2020)  
Porous Polyurea 78 Li et al. (2015)  
Loofa sponge-immobilized fungal biomass 98.8 Iqbal & Saeed (2007)  
Magnetic nanocomposite of chitosan/SiO2/carbon nanotubes 97.1 Abbasi (2017)  
NML (natural mulberry leaves) 64.5 This study 
AAM (acetic acid-modified mulberry leaves) 95.2 
CAM (citric acid-modified mulberry leaves) 84.8 
OAM (oxalic acid-modified mulberry leaves) 91.7 
AdsorbentQmax (mg/g)Reference
ZnO fine powder (Z300) 89.3 Ada et al. (2009)  
ZnO fine powder (Z075) 38.9 Ada et al. (2009)  
Sodium alginate/poly(N-vinyl-2-pyrrolidone) 55.3 İnal & Erduran (2015)  
Bentonite coated with Mg(OH) 66.9 Chinoune et al. (2016)  
Amino-functionalized organosilane 21.3 Saputra et al. (2017)  
Borax cross-linked Jhingan gum hydrogel 9.88 Mate & Mishra (2020)  
Porous Polyurea 78 Li et al. (2015)  
Loofa sponge-immobilized fungal biomass 98.8 Iqbal & Saeed (2007)  
Magnetic nanocomposite of chitosan/SiO2/carbon nanotubes 97.1 Abbasi (2017)  
NML (natural mulberry leaves) 64.5 This study 
AAM (acetic acid-modified mulberry leaves) 95.2 
CAM (citric acid-modified mulberry leaves) 84.8 
OAM (oxalic acid-modified mulberry leaves) 91.7 
Figure 5

Comparison of equilibrium isotherms between the experimental data and theoretical data for (a) NML; (b) AAM; (c) CAM; and (d) OAM.

Figure 5

Comparison of equilibrium isotherms between the experimental data and theoretical data for (a) NML; (b) AAM; (c) CAM; and (d) OAM.

Close modal

Influences of adsorbent concentration on RBBR removal

The influences of AAM, CAM, and OAM amount on RBBR removal from aqueous media were assessed between 1.0 and 20.0 g L−1 of adsorbent amount, in the experimental conditions of 240 min of equilibrium time and initial solution pH of 2.0 for 200 mg L−1 of initial RBBR concentration. The amount of RBBR adsorption at equilibrium and RBBR removal percentages (Figure 6) plotted against the AAM, CAM, and OAM amounts. The increase in the amount of adsorbent at constant RBBR concentration causes the formation of unsaturated adsorption surfaces and agglomeration of the adsorbent (Crini et al. 2007). Therefore the amount of adsorbed RBBR per gram of adsorbent decreased from 97.1, 101.6 and 102.9 mg g−1 to 9.79, 9.80, and 9.85 mg g−1 for AAM, CAM, and OAM respectively. However, the adsorption percentage increased from 48.6, 50.8, and 51.4% to 97.9, 98.0, and 98.5% for AAM, CAM, and OAM, respectively as increasing in the amount of adsorbent which lead to an increase in the active adsorption sites (Figure 6).
Figure 6

Effect of adsorbent amount on RBBR adsorption.

Figure 6

Effect of adsorbent amount on RBBR adsorption.

Close modal

ANN modeling of the adsorption process of RBBR

The ANN model was applied to generate a prediction model for the adsorption efficiency of RBBR onto AAM, CAM, and OAM by dividing 27 experimental data randomly into three subgroups including training (70%), validation (15%), and testing (15%). Then, three-layer BP neural network model with tansig in the hidden layer and purelin in the output layer was assessed by various neuron numbers. The best performance was acquired when eight neurons were present in the hidden layer. Thereafter to achieve the best possible weights, errors, and correlations various training studies were performed. The ANN regression plot for training, validation, testing and the overall prediction set in the form of network output versus adsorption efficiency is given in Figure 7(a)–7(c) for AAM, CAM, and OAM, respectively. Based on the regression analysis, it is seen that the network output values are close to the experimental adsorption efficiency in all cases. The overall prediction set was obtained as 0.998, 0.994 and 0.999 for AAM, CAM, and OAM, respectively. The results demonstrated that the ANN model is suitable for the prediction of RBBR adsorption efficiency with sensible accuracy.
Figure 7

Outcomes of ANN algorithm for estimation of RBBR recovery percentage of (a) AAM; (b) CAM; and (c) OAM.

Figure 7

Outcomes of ANN algorithm for estimation of RBBR recovery percentage of (a) AAM; (b) CAM; and (c) OAM.

Close modal

In the present research, the removal of an anionic dye, RBBR, from aqueous media by NML and AAM, OAM, and CAM mulberry leaves via an adsorption process was studied for the first time. This study gets attention from the environmental and industrial perspectives since there is a great need to uptake the harmful RBBR dye by cost-effective and abundant materials from wastewater before the discharge step. The influences of various experimental factors on the adsorption efficiency were investigated. The optimum initial solution pH was determined as 2.0 for RBBR adsorption. It was observed that the adsorption efficiency reached its maximum after 240 min of equilibrium time. RBBR adsorption capacity of carboxylic acid modified adsorbents was also found to be higher than many other adsorbents mentioned in the literature. The results of this study demonstrated that the modification of mulberry leaves with different types of carboxylic acids enhances the adsorption efficiency of anionic dyes and suggests a low-cost and easily available adsorbent for the uptake of harmful RBBR dye from aqueous solutions. In the future, considering the experimental parameters optimized by using aqueous solutions in the laboratory environment, an application will be carried out directly on textile wastewater in order to remove both anionic and cationic dyes. In addition, the usability of the developed adsorbents in the removal of various heavy metal ions from wastewater will also be tested.

The financial support from the Unit of the Scientific Research Projects of Karadeniz Technical University is gratefully acknowledged.

All relevant data are included in the paper or its Supplementary Information.

Ada
K.
,
Ergene
A.
,
Tan
S.
&
Yalcın
E.
2009
Adsorption of Remazol Brilliant Blue R using ZnO fine powder: Equilibrium, kinetic and thermodynamic modeling studies
.
Journal of Hazardous Materials
165
(
1–3
),
637
644
.
Ali
H.
2010
Biodegradation of synthetic dyes – A review
.
Water, Air, & Soil Pollution
213
,
251
273
.
Bensalah
H.
,
Bekheet
M. F.
,
Younssi
S. A.
,
Ouammou
M.
&
Gurlo
A.
2017
Removal of cationic and anionic textile dyes with Moroccan natural phosphate
.
Journal of Environmental Chemical Engineering
5
(
3
),
2189
2199
.
Chinoune
K.
,
Bentaleb
K.
,
Bouberka
Z.
,
Nadim
A.
&
Maschke
U.
2016
Adsorption of reactive dyes from aqueous solution by dirty bentonite
.
Applied Clay Science
123
,
64
75
.
Chumpiboon
A.
,
Thongsubsai
K.
,
Pongsiri
T.
,
Knijnenburg
J. T. N.
&
Ngernyen
Y.
2022
Removal of cationic dye from textile wastewater using treated bagasse fly ash: An industrial waste
.
Engineering and Applied Science Research
49
(
3
),
381
394
.
Dubinin
M. M.
&
Radushkevich
L. V.
1947
Equation of the characteristic curve of activated charcoal
.
Proceedings of the Academy of Sciences of the USSR, Physical Chemistry Section
55
,
331
333
.
Duran
C.
,
Ozdes
D.
,
Gundogdu
A.
&
Senturk
H. B.
2011
Kinetics and isotherm analysis of basic dyes adsorption onto almond shell (Prunus dulcis) as a low cost adsorbent
.
Journal of Chemical & Engineering Data
56
(
5
),
2136
2147
.
Ewuzie
U.
,
Saliu
O. D.
,
Dulta
K.
,
Ogunniyi
S.
,
Bajeh
A. O.
,
Iwuozori
K. O.
&
Ighalo
J. O.
2022
A review on treatment technologies for printing and dyeing wastewater (PDW)
.
Journal of Water Process Engineering
50
,
103273
.
Freundlich
H. M. F.
1906
Uber die adsorption in Losungen
.
Zeitschrift für Physikalische Chemie
57
,
385
470
.
Ghadirimoghaddam
D.
,
Gheibi
M.
&
Eftekhari
M.
2021
Graphene oxide-cyanuric acid nanocomposite as a novel adsorbent for highly efficient solid phase extraction of Pb2+ followed by electrothermal atomic absorption spectrometry; statistical, soft computing and mechanistic efforts
.
International Journal of Environmental Analytical Chemistry
https://doi.org/10.1080/03067319.2020.1861260
.
Guo
X.
,
Liu
Z.
,
Tong
Z.
,
Jiang
N.
&
Chen
W.
2022
Adsorption of Rhodamine B from an aqueous solution by acrylic-acid-modified walnut shells: characterization, kinetics, and thermodynamics
.
Environmental Technology
https://doi.org/10.1080/09593330.2021.2011430
.
Hameed
B. H.
&
El-Khaiary
M. I.
2008
Kinetics and equilibrium studies of malachite green adsorption on rice straw-derived char
.
Journal of Hazardous Materials
153
(
1–2
),
701
708
.
Helfferich
F.
1962
Ion Exchange
.
McGraw-Hill
,
New York
.
Herkommerová
K.
,
Dostál
J.
&
Pichová
I.
2018
Decolorization and detoxification of textile wastewaters by recombinant Myceliophthora thermophila and Trametes trogii laccases
.
Biotech
8
,
505
.
Ho
Y. S.
&
McKay
G.
1998
Kinetic models for the sorption of dye from aqueous solution by wood
.
Process Safety and Environmental Protection
76
(
2
),
183
191
.
Jawad
A. H.
,
Abdulhameed
A. S.
,
Kashi
E.
,
Yaseen
Z. M.
,
ALOthman
Z. A.
&
Khan
M. R.
2022
Cross-linked chitosan-glyoxal/kaolin clay composite: Parametric optimization for color removal and COD reduction of remazol brilliant blue R Dye
.
Journal of Polymers and the Environment
30
,
164
178
.
Karadeniz
M.
&
Osma
E.
2019
Determination of heavy metal accumulation in mulberries (Morus alba L.) growing in malatya
.
Erzincan University Journal of Science and Technology
12
(
2
),
893
901
.
Khan
F. A.
&
Farooqui
M.
2022
Removal of methylene blue dye from aqueous solutions onto Morus nigra L. (mulberry tree) leaves powder and its biochar— equilibrium, kinetic and thermodynamic study
.
International Journal of Environmental Analytical Chemistry
https://doi.org/10.1080/03067319.2022.2103691
.
Lagergren
S.
1898
About the theory of so-called adsorption of soluble substance, Kungliga Svenska Vetenskapsakademiens
.
Handlingar
24
,
1
39
.
Langmuir
I.
1918
The adsorption of gases on plane surfaces of glass, mica and platinum
.
Journal of American Chemical Society
40
(
9
),
1361
1403
.
Le
V. T.
,
Tran
T. K. N.
,
Tran
D. L.
,
Le
H. S.
,
Doan
V. D.
,
Bui
Q. D.
&
Nguyen
H. T.
2019
One-pot synthesis of a novel magnetic activated carbon/clay composite for removal of heavy metals from aqueous solution
.
Journal of Dispersion Science and Technology
40
(
12
),
1761
1776
.
Miyah
Y.
,
Benjelloun
M.
,
Lahrichi
A.
,
Mejbar
F.
,
Iaich
S.
,
El Mouhri
G.
,
Kachkoul
R.
&
Zerrouq
F.
2021
Highly-efficient treated oil shale ash adsorbent for toxic dyes removal: Kinetics, isotherms, regeneration, cost analysis and optimization by experimental design
.
Journal of Environmental Chemical Engineering
9
(
6
),
106694
.
Moharami
S.
&
Jalali
M.
2013
Removal of phosphorus from aqueous solution by Iranian natural adsorbents
.
Chemical Engineering Journal
223
,
328
339
.
Mosai
A. K.
,
Chimuka
L.
,
Cukrowska
E. M.
,
Kotze
I. A.
&
Tutu
H.
2020
Removal of platinum (IV) from aqueous solutions with yeast-functionalized bentonite
.
Chemosphere
239
,
124768
.
Ozdes
D.
,
Duran
C.
,
Senturk
H. B.
,
Avan
H.
&
Bicer
B.
2014
Kinetics, thermodynamics, and equilibrium evaluation of adsorptive removal of methylene blue onto natural illitic clay mineral
.
Desalination and Water Treatment
52
(
1–3
),
208
218
.
Ozdes
D.
,
Gundogdu
A.
,
Duran
C.
&
Senturk
H. B.
2010
Evaluation of adsorption characteristics of malachite green onto almond shell (Prunus dulcis)
.
Separation Science and Technology
45
,
2076
2085
.
Reghioua
A.
,
Barkat
D.
,
Jawad
A. H.
,
Abdulhameed
A. S.
,
Al-Kahtani
A. A.
&
ALOthman
Z. A.
2021
Parametric optimization by Box–Behnken design for synthesis of magnetic chitosan-benzil/ZnO/Fe3O4 nanocomposite and textile dye removal
.
Journal of Environmental Chemical Engineering
9
(
3
),
105166
.
Saputra
O. A.
,
Rachma
A. H.
&
Handayani
D. S.
2017
Adsorption of remazol brilliant blue R using amino-functionalized organosilane in aqueous solution
.
Indonesian Journal of Chemistry
17
(
3
),
343
350
.
Sharma
P.
,
Kaur
R.
,
Baskar
C.
&
Chung
W.-J.
2010
Removal of methylene blue from aqueous waste using rice husk and rice husk ash
.
Desalination
259
(
1–3
),
249
257
.
Thamilarasi
M. J. V.
,
Anilkumar
P.
,
Theivarasu
C.
&
Sureshkumar
M. V.
2018
Removal of vanadium from wastewater using surface-modified lignocellulosic material
.
Environmental Science and Pollution Research
25
(
26
),
26182
26191
.
Vakili
M.
,
Rafatullah
M.
,
Salamatinia
B.
,
Abdullah
A. Z.
,
Ibrahim
M. H.
,
Tan
K. B.
,
Gholami
Z.
&
Amouzgar
P.
2014
Application of chitosan and its derivatives as adsorbents for dye removal from water and wastewater: A review
.
Carbohydrate Polymers
113
,
115
130
.
Wang
S.
,
Li
L.
,
Wu
H.
&
Zhu
Z. H.
2005
Unburned carbon as a low-cost adsorbent for treatment of methylene blue-containing wastewater
.
Journal of Colloid and Interface Science
292
(
2
),
336
343
.
Wang
S.
,
Soudi
M.
,
Li
L.
&
Zhu
Z. H.
2006
Coal ash conversion into effective adsorbents for removal of heavy metals and dyes from wastewater
.
Journal of Hazardous Materials
133
(
1–3
),
243
251
.
Weber
W. J.
Jr
&
Morriss
J. C.
1963
Kinetics of adsorption on carbon from solution
.
Journal of the Sanitary Engineering Division
89
,
31
60
.
Zhang
H.
,
Xing
L.
,
Liang
H.
,
Ren
J.
,
Ding
W.
,
Wang
Q.
,
Geng
Z.
&
Xu
C.
2022
Efficient removal of Remazol Brilliant Blue R from water by a cellulose-based activated carbon
.
International Journal of Biological Macromolecules
207
,
254
262
.
This is an Open Access article distributed under the terms of the Creative Commons Attribution Licence (CC BY 4.0), which permits copying, adaptation and redistribution, provided the original work is properly cited (http://creativecommons.org/licenses/by/4.0/).

Supplementary data