Continuous fixed-bed column study for removal of Congo red dye from aqueous solutions using Nelumbo nucifera leaf adsorbent

ABSTRACT The adsorption of Congo red (CR) dye from aqueous solutions is conducted in a continuous fixed-bed column by using agricultural biomass of Nelumbo nucifera leaf adsorbent. The column performance is evaluated by varying the adsorbent bed height (2.5‒5 cm), influent dye concentrations (10–50 mg L−1), and inlet flow rate (1‒5 mL min−1). Column experimental data confirmed that the breakthrough characteristics of the adsorption system are dependent on bed height, flow rate, and initial adsorbate concentration. The results show that the decolourisation efficiency and equilibrium uptake (qe) of CR decreases with increasing flow rate and increases with increasing bed height and influent adsorbate concentration. The maximum absorption efficiency (83.12%) was obtained using 5 cm bed height, 15 mg L−1 inlet adsorbate concentration, and 1 mL min−1 flow rate. The increased flow rate and initial adsorbate concentration lead to a shorter column exhaustion time. The length of mass transfer zone increased with increasing bed height. Various mathematical models are applied to column experimental data to predict the breakthrough curve and to evaluate the column capacity and kinetic constants of the models. The correlation between the experimental and theoretical data is quantified by calculating the regression coefficients for the Thomas and Yoon–Nelson models, an R 2 ≥ 0.954 at various operating conditions was obtained, which shows that the trend of experimental data fits well and the overall system kinetics are limited by solid-liquid interphase mass transfer. The N. nucifera leaf fine powder adsorbent is proven to be capable of removing CR efficiently from aqueous solutions in continuous fixed-bed systems at low flow rate and higher bed depth and it can be used effectively in wastewater treatment.


Introduction
For decades, the environment has been suffering from pollution problems, especially the aquatic environment, a source of freshwater, which is responsible for sustaining life on this planet.Water pollution is a major problem faced by the world today [1].Dye-bearing wastewater poses a serious threat to the aquatic environment due to its toxicity and appearance.Synthetic dyes are broadly utilised in a number of different industries, including the dyeing, textile, paper, leather, printing, rubber, cosmetic and plasticsindustries, to colour their products [2].The continuous discharge of dye effluent from these industries imparts color to the receiving natural water bodies, making them unacceptable for human consumption and also causing severe damage to the environment [3].Synthetic dye Congo red (CR) is a water-soluble anionic diazo acid dye, and due to its high affinity for cellulose fibres it is widely used in textile processing industries [4].Azo dyes are highly toxic, carcinogenic and usually constitute a health hazard [5].CR has been known to cause a number of different ailments to the human body including allergic dermatitis, cancer, gastrointestinal, eye, and skin irritation.It may also cause clotting of the blood, and induce drowsiness and breathing problems [6].In addition, even at low concentrations, CR dye can cause diarrhoea, nausea, vomiting, chest and abdominal pain, severe headache, etc. [7].Hence, it is imperative that the effluents are treated to bring down the concentration of dyes to tolerable limits before being released into the waterbodies [8].Conventional treatment methods for the removal of colour from dye wastewater are inefficient and expensive, and hence new treatment methods are desired [9].
Adsorption has been found to be an efficient, insensitive to toxic pollutants and an economical process to treat the dyeing industry effluent [10].Commercially available activated carbon has been widely used as an adsorbent to remove various organic pollutants, because of its excellent adsorption capacity but its use is limited due to the difficulty in the regeneration of spent activated carbon and high cost of the adsorbent [11].Recently, there has been a renewed focus on the development of widely available and highly effective adsorbents that can further decrease the cost of treatment [12].Exploration of effective low-cost adsorbents may contribute to the sustainability of the environment and also offer promising benefits for commercial purposes in the future [13].The performance of an adsorbent in a continuous system is an important factor in assessing the feasibility of the adsorbent in real-world applications.A batch mode analysis alone is not sufficient while designing a treatment system for continuous operation.Column studies are essential to obtain the model parameters required for the design of fixed-bed adsorbers [14].To the best of our knowledge, we believe that there has practically been no work reported for describing the potential of using Nelumbo nucifera (lotus) leaf powder as an adsorbent in a fixed-bed column operated in continuous mode for the removal CR dye from aqueous solutions.Therefore, in this manuscript, we analyse the use of low-cost adsorbent, Nelumbo nucifera leaf (NNL) powder, for the decolourisation of CR from contaminated water in a continuous flow column.Nelumbo nucifera is commonly grown in subtropical and temperate regions.Its leaf as an agricultural waste material contains alkyne, methyl, hydroxyl, and carbonyl functional groups, which are responsible for the adsorption of various pollutants with satisfactory performance at laboratory scale [15][16][17].The NNLs are cheap and widely available in India and can easily be cultivated from seeds or vegetative reproduction.The objectives of this study are to analyse the effect of inlet flow rate, bed height, and influent dye concentration to predict the breakthrough curves (BTCs) and kinetic constants of the various mathematical models using the column experimental data.

Preparation of NNL powder adsorbent
Mature NNLs were collected from Bastar district in Chhattisgarh State, India.The biomass (leaves) was dried under sunlight to remove the moisture and ground to a fine powder using a pulveriser.The material was thoroughly washed with distiled water several times to remove all the dirt particles from other impurities.Then, the material was dried in a hotair oven at the temperature of 338 K for 8 h, ground, and screened to obtain particles <100 μm in size.The powdered material was then stored in airtight plastic bottles for further use in adsorption experiments [17].

Characterisation of NNL adsorbent
The prepared NNL powder adsorbent was characterised by particle size, pore volume, Brunauer-Emmett-Teller (BET) surface area, field emission scanning electron microscopy/ energy dispersive X-ray spectroscopy (FESEM/EDS), attenuated total reflectance spectroscopy (ATR), and thermogravimetric analysis (TGA).The characterisation of NNL adsorbent is given in our previous batch study published earlier [17].

Chemicals required
The analytical grade anionic diazo acid dye CR with dye content ≥35% and purity 99.8% is obtained from Sigma-Aldrich Chemical Company Limited (India).The other chemicals such as sodium hydroxide, hydrochloric acid, and acetone used are of analytical grade supplied by Merck (India).

Preparation of aqueous CR dye solutions
A stock solution of 1000 mg L -1 was prepared by dissolving 1 g of CR dye powder in 1000 mL of double-distilled water.The experimental solution of different initial dye concentrations ranging between 15 and 50 mg L −1 are made by further diluting the stock solution with pH-adjusted distiled water by adding 0.1 N HCl or 0.1 N NaOH.After dilution, the final pH of the dye solution was measured and was found to be 6.The standard calibration curve is developed through the measurement of the dye solution absorbance at various initial dye concentrations.The CR dye colour remained stable in the pH range of 6-14.Batch experiment results show that maximum adsorption of CR is observed at pH 6 [14,18].At pH 6, a significant electrostatic attraction exists between the protonated binding sites of the NNL adsorbent and anionic dye molecules due to van der Waals forces.Moreover, CR dye colour becomes unstable if the solution pH falls below 6 due to the formation of protonated species, which may lead to a change in the structure of the dye.The CR dye in aqueous solution is black in colour at acidic pH (less than 5), because of the formation of a quinonoid structure [19].Therefore, further adsorption experiments are carried out at pH 6.The structure of the CR dye is shown in Figure S1.

Analytical measurements
The powdered material is weighed using an electronic digital weighing balance (Citizen, India).The pH of the dye solution is measured by a digital pH-meter (Systronics 335, India).After adsorption, the unknown residual dye concentration is determined by measuring the absorbance at 498 nm (λ max ) using a pre-calibrated double beam UVvisible spectrophotometer (Shimadzu UV-1800, Japan).

Column experiments for the removal of colour from synthetic CR dye wastewater
The continuous studies in a laboratory scale fixed-bed adsorption column are carried out at room temperature, T (301 K) using NNL asorbent for the removal of CR from aqueous solution.The experiments are conducted to study the effect of bed height, inlet flow rate and influent dye concentrations on CR dye adsorption.For the column experiments, a Perspex column with an inner diameter of 2.1 cm and a height of 39 cm is used.A schematic explaining the different parts of the experimental fixed-bed column setup is given in Figure 1.A rubber cork of 1.5 cm height is provided at the bottom and top of the column to support the inlet and outlet pipes.The column is packed with 1.8 cm of glass wool followed by glass beads (1.5 mm in diameter) both at the bottom and top [14].The NNL asorbent is packed in the column with varying quantities of 3.72, 5.24 and 7.36 g to produce various bed heights of 2.5, 3.5 and 5 cm respectively.Before starting the experiments, the column was wetted with distiled water in an upward flow direction using a peristaltic pump (Enertech, India) for 2 h to remove the trapped air between the particles [20].The peristaltic pump is provided with an inlet at the bottom of the column.A steady state condition is maintained by measuring the flow rate at the bottom and top of the column.After setting up the column, the water is replaced with an aqueous CR dye solution of known concentration (10, 20 and 50 mg L −1 ) at pH 6, which is continuously fed in the upward direction at the desired flow rate (1, 3 and 5 mL min −1 ) controlled by a peristaltic pump.The treated dye solutions at the outlet of the column are collected at regular time intervals with the flow rate maintained same as the inlet feed stream and the concentration is measured using a UV-visible spectrophotometer.The experiments are continued until the concentration of treated effluent reaches the inlet concentration of adsorbate [21].Continuous decolourisation experiments are carried out by varying the level of one variable while keeping the level of the other variables constant.

Mathematical description of adsorption in a continuous fixed-bed column study
To design a continuous fixed-bed column adsorption process, it is necessary to predict the BTC and adsorbent capacity for the required adsorbate under the given set of operating conditions.Commonly, the BTC is expressed as C t /C o vs. contact time (t) or volume of the effluent (V eff ), where C t and C o represent the effluent and influent adsorbate concentrations, respectively.The effluent volume can be calculated by Equation (1) [22]: where Q and t total are the volumetric flow rate (mL min −1 ) and total flow time (min), respectively.The total amount of dye anion adsorbed in the column is determined by multiplying the area above the BTC by the initial dye concentration and flow rate.It is expressed as [23] where t E is the bed exhaustion time, which is the time at which the dye concentration in the effluent reaches the initial feed concentration.The total amount of adsorbate dispatched to the column (m total ) is obtained from Equation (3) [21]  The column performance can be evaluated based on the total removal efficiency of the adsorbate, which is defined as the ratio of total quantity of solute adsorbed in the column (m ad ) to the total amount of dye sent to the column (m total ) as given in the following equation: Equilibrium adsorption capacity (q e ) in the column is expressed as [14] where W is the mass of dry adsorbent (W) packed in the column.The unadsorbed dye concentration at equilibrium (C e ) in the continuous flow system is evaluated as given in Equation ( 6) [20]

Analysis of column experiments for the removal of CR from synthetic dye wastewater using NNL adsorbent at various operating conditions
The results show that the saturation of binding sites for the removal of CR takes place rapidly in the initial phases of the adsorption process.As the feed solution continues to flow, the amount of solute adsorbed starts decreasing because of the progressive saturation of the adsorbate on the active sites of the NNL adsorbent.The effluent concentration of adsorbate starts to increase until the bed is completely saturated.The shape of the BTC follows a sigmoidal pattern [24,25].

Effect of bed height
The breakthrough curves for the adsorption of CR onto NNL adsorbent at various bed heights from 2.5 to 5 cm by fixing the inlet flow rate (1 mL min −1 ) and dye concentration (15 mg L −1 ) are given in Figure 2. The breakthrough point time (t B ) and bed exhaustion time (t E ) are strongly correlated with the bed height.As the bed height increases, the breakthrough for CR occurs much more slowly, and the time to bed exhaustion increases.For a larger bed height, the exhaustion time is longer, which means that the bed can be operated for much longer durations without the need to change the adsorbent, while for a shorter bed height, the exhaustion time is much shorter [20].Increase in bed height results in a longer diffusion path which in turn leads to an increase in the t B [26].The longer the breakthrough time, the better the intra-particulate diffusion phenomenon and higher will be the decolourisation efficiency and adsorption capacity of the column [22].The results of adsorption as a function of bed height described here are presented in Table 1.It shows that the % adsorption and equilibrium dye uptake of CR increases from 76.43%to 83.12% and 3.236 to 3.986 mg g −1 , respectively, with an increase in bed height from 2.5 to 5 cm.This is due to the fact that more amount of adsorbent at larger bed heights provides more active sites available for adsorption of CR dye molecules which results in a broadened mass transfer zone [14].This phenomenon may be due to an increase in the bed height, decrease in axial dispersion in the mass transfer zone, and as a result, an increase in the diffusion of the dye molecules into the adsorbent.Thus, the adsorbate gets sufficient time to get diffused into the entire particle surface, staying for a longer time in the column and treating more volume of dye solution leading to a decrease in the adsorbate concentration in the effluent [27].

Effect of flow rate
Flow rate proved to be an important characteristic affecting the performance of the adsorbent in a continuous fixed-bed operation.The effect of flow rate on breakthrough  curves for CR dye adsorption is analysed by varying the flow rate from 1 to 5 mL min −1 while the bed height (2.5 cm) and influent dye concentration (15 mg L −1 ) are kept constant.From Figure 3, we can see that both the t B and t E decrease with increasing flow rate.The experimental results are reported in Table 1.These results indicate that the decolourisation efficiency and equilibrium adsorption capacity of CR decreases from 76.43% to 51.47% and 3.236 to 1.557 mg g −1 , respectively, with an increase in inlet flow rate from 1 to 5 mL min −1 .The shorter t B at higher flow rate is due to a reduction in the contact time between CR dye molecules and NNL adsorbent inside the column and diffusional limitation of the adsorbate into the pores of the adsorbent (improper utilisation of adsorbate in the column) [28].It is observed that a better column performance is achieved at lower flow rates.Hence, the adsorption process remained incomplete, leading to a steep BTC, and concurrently a lower equilibrium dye uptake (q e ) in the column as the flow rate was increased [29].

Effect of the influent adsorbate concentration
The effect of influent dye concentration on CR dye adsorption is investigated by varying the feed concentration from 15 to 50 mg L −1 at a constant bed height of 2.5 cm and flow rate of 1 mL min −1 .The adsorption BTCs are shown in Figure 4 and the experimental parameters obtained from the BTCs are given in Table 1. Figure 4 illustrates that the t B and t E decrease with increasing inlet CR dye concentration.From Table 1 we can see that the adsorption efficiency and equilibrium dye uptake of CR increases from 76.43% to 81.22% and 3.236 to 5.254 mg g −1 , respectively, with an increase in influent adsorbate concentration from 15 to 50 mg L −1 .This is because the driving force for the mass transfer and the CR dye loading rate increases as the influent concentration increases (increase in the concentration gradient between the adsorbate in solution and in the adsorbent) [22,30].At higher concentrations, the availability of dye molecules for the adsorption sites is greater, which leads to a higher uptake of dye resulting in a reduced t B (bed saturates faster when higher amounts of adsorbate are introduced to the adsorbent column) [31].Also, steep BTCs were obtained at high inlet adsorbate concentrations.The increase in the amount of dye adsorbed in the column can be ascribed to the increase in binding sites being occupied as the feed concentration is increased [32].

Analysis of predicted BTCs and estimation of kinetic model parameters in various models
In order to evaluate the relationship of the column parameters, various adsorption models are applied to the data obtained from the experimental studies.Industrial-scale column operation can be designed on the basis of experimental data collected at the laboratory level.Typically, the mathematical models provide insights into the mechanism of the adsorption process, surface properties of adsorbents and degree of affinity of adsorbates and adsorbents.Various mathematical models, such as Adams-Bohart, bed depth service time (BDST), Thomas, Yoon-Nelson and Wolborska models, are used in this study to analyse the breakthrough behaviour of the selected adsorbent-adsorbate system and to evaluate the kinetic model parameters at various bed heights, flow rates, and influent adsorbate concentrations [33].Linear regression analysis is used to evaluate the kinetic constants of the models.

Adams-Bohart model
The Adams-Bohart model is commonly used to describe the initial state of the BTC, and it is based on the theory of surface reaction.The model assumes that the rate of adsorption is proportional to both the residual capacities of solid adsorbent and the concentration of adsorbed species and is expressed by the following linear Equation ( 7) [34]: where t is the flow time (min), K AB is the Adams-Bohart kinetic constant (L mg −1 min −1 ), N o is the maximum saturation concentration (mg L −1 ), Z is the bed height (cm), and U o is the superficial velocity (cm min −1   2, it is observed that as the bed height increases, the values of kinetic constant K AB decreases and maximum saturation concentration N o increases.This may be due to an increase in the number of binding sites available for CR dye adsorption.The same trend is observed as the inlet dye concentration is increased.This may be due to the increase in dye molecules present in the solution [20].The value of K AB increases with increasing flow rate suggesting that the overall rate of dye adsorption process is governed by an external mass transfer [24].The predicted and empirical with respect to various bed heights, flow rates and inlet adsorbate concentrations are shown in Figures S2-S4.They show that the predicted BTCs largely deviate from empirical values.The values of linear regression coefficient, R 2 are in the range between 0.739 and 0.912, indicating that this model does not fit the column experimental data points appropriately.

Bed depth service time (BDST) model
The BDST model is used to predict the relationship between bed height and service time (t) by using different breakthrough values.It assumes that the rate of adsorption is controlled by the interaction among adsorbate molecules and the unutilised capacity of the adsorbent [35].This model ignores the pore diffusion and external film resistance such that the adsorbate is accumulated onto the adsorbent surface directly.Based on these assumptions, the BDST model is given as [36] Table 2. Adams-Bohart model parameters under various operating conditions for CR dye adsorption.The parameters obtained from this model are reported in Table 3, showing that the values of K decreased but N o increased with increasing bed height and inlet adsorbate concentration.This may be due to the available binding sites being saturated with CR dye molecules.It also infers that values of K increase with increasing flow rate.The rate constant, K characterises the rate of solute transfer from liquid to solid phase [35].The predicted and empirical BTCs with respect to various bed heights, flow rates and inlet adsorbate concentrations are shown in Figures S5-S7.They show that there is a small variation among empirical and predicted BTCs.Also, the R 2 values are in the range between 0.929 and 0.982, indicating that this model does not fit the empirical data points very well but still better than the Adams-Bohart model.

Thomas model
The Thomas model is one of the most widely used to describe the column performance, and prediction of BTCs.It is derived based on the following assumptions: (1) external and pore diffusions are not the rate-controlling steps, (2) Langmuir kinetics of adsorption are valid, (3) Thomas model follows pseudo-second-order reversible reaction kinetics with no axial dispersion even when the bed depth is at the minimum, (4) breakthrough occurs immediately once the adsorbent is saturated with the adsorbate and ( 5) constant separation factor [22].The linearised form of the Thomas model can be expressed as follows [14]: where K Th is the Thomas model kinetic rate constant (L min −1 mg −1 ) and q oTh is the maximum solid-phase concentration of the adsorbate (mg g −1 ).The slope K Th and intercept q oTh are estimated from the linear plot of ln C o C t À 1 � � vs. V eff .The adsorption data are fit to this model to evaluate the model parameters at various operating conditions and their values are reported in Table 4. From Table 4, we can see that as the bed height and inlet dye concentration increased, the values of q oTH increased but the values of K Th decreased significantly.The opposite trend is seen when the flow rate is increased.With the increase in flow rate, the values of q oTH decrease but the values of K Th increase indicating that the overall system kinetics is controlled by external film diffusion [37].
The predicted and empirical BTCs with respect to various bed heights, flow rates and inlet adsorbate concentrations are shown in Figures 5-7.The theoretical and empirical data show a similar correlation and breakthrough trend.Also, the calculated bed capacity,q oTh is in good agreement with the real-time data with a high R 2 value (greater than 0.948) at different operating conditions suggesting that this model is valid for CR adsorption behaviour in a longer bed column.However, the adsorption process is typically not controlled/limited by chemical reaction kinetics but is frequently limited by solid-liquid interphase mass transfer, and the impact of axial dispersion might be important at lower flow rates [24].

Yoon-Nelson model
The Yoon and Nelson model was developed based on the theory of adsorption and breakthrough of adsorbate probability.It is based on the following assumptions: (1) rate of decrease in the probability of adsorption for each solute molecule is proportional to the probability of adsorbate adsorption and the probability of adsorbate breakthrough on the adsorbent; (2) effects of axial dispersion are neglected [38].The linearised form of the Yoon-Nelson model is described in Equation ( 10) [14] where k YN and τ are the Yoon-Nelson rate constant (min −1 ) and time required for 50% adsorbate breakthrough (min), respectively.The values of k YN and τ can be obtained from the slope and intercept of the linear plot of ln C t C o À C t � � vs. sampling time t.The adsorption bed should be completely saturated at t = 2τ.Hence, the quantity of solute adsorbed on the adsorbent is half of the total adsorbate fed into the column within a 2τ period.The Yoon-Nelson model adsorption capacity (q oYN ) can be determined as [24]: This model does not require detailed data concerning the characteristics of the adsorbate, the sort of adsorbent and the physical properties of the adsorption bed [39].The Yoon Nelson model is fit to the experimental data to investigate the breakthrough characteristics of adsorption of CR onto NNL adsorbent.The kinetic parameters obtained from this model are reported in Table 5.According to the results, values of k YN decrease and τ increase with increasing bed height.At greater bed heights, the adsorbate molecules have more time to diffuse through the adsorbent, which results in a reduced adsorption rate constant [40].It is known that k YN and τ are inversely related.The opposite trend is observed as the flow rate and influent dye concentration increases.At high flow rate and inlet dye concentration, the number of dye molecules passing through the NNL adsorbent is higher, leading to an increase in the rate constant (k YN ) [41].The column adsorption capacity (q oYN ) calculated based on the results of this model increases with increasing bed height and influent dye concentration but decreases with increasing the inlet flow rate of the CR dye being investigated.When the influent dye concentration increases, the competition between the adsorbate molecules for the adsorption sites is higher, which ultimately results in an increased uptake rate [42].The predicted and empirical BTCs at different bed heights, flow rates and influent adsorbate concentrations are shown in Figures 8-10.The predicted BTCs, Table 5. Yoon-Nelson model parameters at various operating conditions for CR dye adsorption.
K YN (min −1 ) q oYN, expt (mg g −1 ) q oYN, calc (mg g −1 ) τ, expt (min) τ, calc (min) R 2 calculated values of τ and q oYN agree quite well with the real-time data with a standard deviation of τ and q oYN being less than 1.845% under the given set of operating conditions.This can be seen by the high values of R 2 , ranging from 0.949 to 0.987 suggesting that this model is valid for CR adsorption.The advantage of applying the Yoon-Nelson model is that the mathematical application is very direct and it provides the information of 50% column breakthrough, which enables the exhaustion period of the column to be predicted without the need for long experimental time [20].In general, this model is capable of modelling symmetric BTCs and neglects the effect of axial dispersion [24].

Wolborska model
The Wolborska model is used for the description of adsorption dynamics at low adsorbate concentration BTCs.It assumes that the axial dispersion is negligible at high flow rate of solution.The linearised form of this model is expressed as [43] with where β a is the kinetic coefficient of the external mass transfer (min −1 ) and D is the axial dispersion coefficient (cm 2 min −1 ).A plot of ln   slope, respectively [25].The Wolborska model is applied to column real-time data for the description of the BTCs.The model parameters determined under various operating conditions are given in Table S1.From Table S1, we can see that the values of parameters β a and N o are influenced by bed heights, flow rates and influent adsorbate concentrations.It is inferred that as the inlet adsorbate concentration and bed height increases, the values of model constant β a decreases but the values of N o increases.The opposite trend is observed as the flow rate is increased.Increasing the flow rate from 1 to 5 mL min −1 , increases the value of β a which shows that the external film diffusion mainly controls the rate of reaction [43].The predicted and empirical BTCs with respect to different bed heights, flow rates and inlet adsorbate concentrations are shown in Figures S8-S10.As can be seen in Figures S8-S10, there is a large discrepancy between empirical and predicted BTCs.The R 2 values range from 0.739 to 0.912, suggesting that this model does not appropriately fit the column experimental data.Therefore, the Adams-Bohart and Wolborska model validity is limited to the range of the process conditions used [25].
Out of the five mathematical models tested for CR dye adsorption, the Thomas and Yoon-Nelson models show a good fit to the experimental data with good regression coefficients and very low standard deviations.These results signify that higher bed heights, lower flow rates and higher initial dye concentrations are favourable for increased adsorption of CR onto NNL adsorbent in a continuous mode of operation.

Possible interactions between CR dye and NNL adsorbent
In order to understand the adsorption process of CR dye onto the NNL adsorbent, a mechanism of adsorption is important.In fact, the adsorption of dye occurs through physisorption or chemisorption at pH 6, depending on the nature of mutual interaction, such as (1) Π-Π interaction between aromatic rings of NNL adsorbent and CR dye molecules, (2) electrostatic interaction between negatively charged SO 3 − groups of CR and the positively charged adsorbent surface that results from protonation of O atoms at the adsorbent surface (3) hydrogen bonding between adsorbent surface and adsorbate [44].CR is an anionic dye that contains a sulphonic group in its structure, which ionises in aqueous solution, forming coloured anions, together with aromatic rings.The principal constituents of Nelumbo nucifera leaves are abundant floristic fibre, protein, cellulose, hemicellulose, flavonoids, alkaloids and all of them together show potential adsorption capacity for various toxic pollutants [15,16].The mechanism of adsorption of anionic dye on the adsorbent is proposed to be chemical adsorption through the strong Π-Π stacking and anion-cation interaction between the anionic dye and the aromatic rings of the NNL adsorbent.The proposed mechanism for adsorption of CR dye onto NNL adsorbent is shown in Figure 11.ATR studies revealed that the surface of the adsorbent contains abundant hydroxyl, methyl, alkyne, and carbonyl groups and this has been reported in our previous paper published elsewhere [17].These groups may interact with the Π electron of the aromatic ring of the CR dye molecules.The adsorption of CR dye can take place at the functional groups or binding sites on the surface of the adsorbent in a monolayer manner [17].Figure 11 shows that the hydroxyl species getting involved in the binding of adsorbate molecules is on the surface of the adsorbent.A similar type of mechanism is also reported in the literature [45].In addition, film and pore diffusion models have been used for examining the diffusion mechanism and it has been elaborately discussed in our previous paper published elsewhere [17].

Conclusion
The NNL powder is an effective adsorbent for the removal of CR from aqueous solutions in continuous mode experiments.The removal efficiency of the dye strongly depends on bed height, flow rate and the influent adsorbate concentration.The higher the bed height and influent adsorbate concentration, better the column performance.The decolourisation efficiency decreased with an increase in inlet flow rate.The equilibrium dye uptake is linearly correlated with bed height and the influent adsorbate concentration.The opposite trend was observed as the flow rate increased.The NNL adsorbent is capable of adsorbing a maximum of 5.254 mg g −1 of CR at 50 mg L −1 influent dye concentration, 2.5 cm bed height, and 1 mL min −1 flow rate.The breakthrough point time and bed exhaustion time increased with an increase in bed height and decreased with an increase in inlet adsorbate concentration and flow rate.A steep breakthrough curve was observed at high inlet flow rates and high inlet adsorbate concentrations.Various models were applied to column experimental data to estimate the breakthrough curves and evaluate the model parameters.While considering the regression coefficient, standard deviation, predicted breakthrough curves, and calculated bed capacity, it is clear that the Thomas and Yoon-Nelson models fit well with the real-time column experimental data as compared to the other models tested.The overall system kinetics are controlled by solid-liquid inter-phase mass transfer.The experimental results revealed that NNL biomass can be used as an effective adsorbent for the removal of colour from synthetic CR dye wastewater in a continuous mode of operation using a fixed-bed adsorption column.The higher adsorption efficiency of synthetic dye effluent suggests that the NNL powder may be used effectively to decolourise anionic dyes from industrial effluents.

Figure 1 .
Figure 1.Schematic diagram of the fixed-bed adsorption column experimental setup.

Table 1 .
Effect of bed height, flow rate and influent adsorbate concentration on CR dye adsorption using LLP adsorbent.
).The values of the model parameters, K AB and N o can be determined from the slope and intercept of the straight-line plot of ln C t Bohart model is applied to the column empirical data for the analysis of the BTCs.The kinetic model parameters are reported in Table2at various operating conditions.From Table has been used to determine the kinetic model parameters, N o and K from the intercept and slope.The BDST model is fit to the empirical data to investigate the breakthrough characteristics of the adsorption of CR onto NNL adsorbent at various process conditions.

Table 3 .
Bed Depth Service Time model parameters at various operating conditions for CR dye adsorption.

Table 4 .
Thomas model parameters at various operating conditions for CR dye adsorption.