Utilisation of sewage sludge derived adsorbents for the removal of recalcitrant compounds from wastewater: Mechanistic aspects, isotherms, kinetics and thermodynamics

Accepted Manuscript Utilization of Sewage Sludge derived Adsorbents for the Removal of Recalcitrant compounds from Wastewater: Mechanistic Aspects, Isotherms, Kinetics and Thermodynamics Anirudh Gupta, Anurag Garg PII: DOI: Reference:

S0960-8524(15)00958-X http://dx.doi.org/10.1016/j.biortech.2015.07.005 BITE 15238

To appear in:

Bioresource Technology

Received Date: Revised Date: Accepted Date:

19 May 2015 30 June 2015 1 July 2015

Please cite this article as: Gupta, A., Garg, A., Utilization of Sewage Sludge derived Adsorbents for the Removal of Recalcitrant compounds from Wastewater: Mechanistic Aspects, Isotherms, Kinetics and Thermodynamics, Bioresource Technology (2015), doi: http://dx.doi.org/10.1016/j.biortech.2015.07.005

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Utilization of Sewage Sludge derived Adsorbents for the Removal of Recalcitrant compounds from Wastewater: Mechanistic Aspects, Isotherms, Kinetics and Thermodynamics Anirudh Gupta and Anurag Garg* Centre for Environmental Science and Engineering, Indian Institute of Technology Bombay, Powai, Mumbai, 400076, INDIA

ABSTRACT In the present study, the performance of sewage sludge based adsorbents was examined for the removal of two recalcitrant pollutants (i.e. lignin and amoxicillin) from synthetic wastewater solutions (adsorbate concentration = 50 – 250 mg/l). The effect of various reaction parameters such as wastewater pH, adsorbent dosage and temperature was studied. Possible mechanisms for the adsorption process have been proposed which depends upon the behaviour of adsorbent surface and adsorbate molecules under specific reaction conditions. Three-parameter Redlich-Peterson isotherm model was found the best fit to the equilibrium data. Pseudo first and second order models validated the kinetic data for lignin and amoxicillin adsorption systems, respectively and the corresponding activation energy was 3.5 – 4.5 and 12– 22 kJ/mol. The nature of adsorption was elucidated from the thermodynamic parameters. Keywords: Sewage sludge; thermo-chemical process; recalcitrant pollutants; wastewater treatment; adsorption

______________________ *Corresponding author. E-mail: [email protected] (Dr. Anurag Garg), Tel.: +91-2225767861, Fax: +91-2225764650

1. Introduction 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Adsorption and catalytic oxidation are the potential treatment methods for the removal of recalcitrant organic pollutants (e.g. phenols, dyes, lignin and antibiotics) present in industrial effluents. Advanced oxidation processes have been used to eliminate phenol (Lal and Garg, 2015), dyes (Yadav et al., 2013), lignin model compound (Yadav and Garg, 2014) and antibiotics (Giri and Golder, 2014) from the synthetic/ simulated waste streams. The oxidation processes have capability of mineralising the pollutants into less harmful gaseous products. However, the high cost of the oxidant and/or the requirement and severe temperatures hinder their implementation in real life problems. Conversely, the pollutants are transferred from aqueous phase to a solid surface (i.e., adsorbent) in adsorption process. However, the cost of adsorbent and its difficult regeneration after the adsorption process impede the application of this process as well. Therefore, there is a need to develop low cost adsorbents for enabling the process to be economically viable. Being carbonaceous nature, sewage sludge can replace the commercial adsorbents which are produced from natural materials. In this manner, the waste material can be recycled and at the same time, the natural resources required for manufacturing of adsorbent can be conserved. Therefore, the purpose of present work was to determine the adsorption potential of sewage sludge derived adsorbents for the removal of lignin and amoxicillin (AMX) from synthetic wastewater. These compounds are largely found in the effluent generated from paper and pulp mills and pharmaceutical industries, respectively. Lignin imparts colour to the wastewater and resists conventional biological treatment (Andersson et al., 2011a) while AMX has ecotoxicological effects and antimicrobial resistance on the receiving aqueous streams (Putra et al., 2009). So the removal of such pollutants from wastewater is mandatory prior to its final discharge into natural water bodies.

In the recent past, sewage sludge based adsorbents have been produced by different chemical 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

activation methods (Boulaem et al., 2014; Zhuang et al., 2014; Xu et al., 2015) which were then used for the adsorption of phenols, metals and dyes from wastewaters (Smith et al., 2009). In a study, the adsorption capacity of a sewage sludge based adsorbent (surface area = 346.5 m2/g and pore volume = 0.465 cm3/g) was found to be in the range of 96.15 mg/g for phenolic wastewater (phenol concentration = 20 – 150 mg/l) which was comparable to that obtained with the commercial activated carbon (Zou et al., 2013). In the same study, around 68% removal of Rhodamine B dye (initial dye concentration = 20 – 50 mg/l) was also reported by adsorption onto the sewage sludge derived material. In comparison, the sewage sludge derived adsorbents showed the maximum adsorption capacity of only upto 22 mg/g for eleven antibiotics and two anticonvulsants (initial concentration = 0.1 – 200 mg/l) (Ding et al., 2012). Modified agricultural by-products have also been used as adsorbents for the removal of ibuprofen, ketoprofen, naproxen and diclofenac from synthetic wastewater (Baccar et al., 2012). Adsorption capacity of the adsorbents was varied widely from 12 – 57 mg/g. The performance of an adsorbent is a function of its porosity, available specific surface area and the presence of suitable functional groups (such as carbonyl, hydroxyl and ether) on the adsorbent surface. In addition, the nature of pollutants and reaction conditions (e.g., pH, and adsorbent dose) also has profound effect on adsorption process. In earlier studies, activated carbon has been used for the removal of lignin (Andersson et al., 2011a,b) but there is no study reported on lignin removal by adsorption on sewage sludge derived adsorbents up to the best of our knowledge. In our recently published study, the primary sewage sludge was converted into adsorbents by chemical activation process. The performance of these adsorbents was comparable to commercial activated carbon (CAC) for phenol removal from the synthetic wastewater

(Gupta and Garg, 2015). Therefore, the same adsorbents were used for the present research 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

work. In this investigation, the effect of various operating parameters (such pH, shaker speed, contact time and adsorbent dose) on the removal of two model pollutants (i.e., lignin and AMX) was investigated. Adsorption kinetics, isotherms and thermodynamic studies were also carried out to determine various parameters required for the design of an adsorption system. An effort was also made to explain the possible adsorption mechanisms for the two pollutants.

2. Experimental section 2.1. Materials Sewage sludge was collected from the sewage collection chamber located at Indian Institute of Technology (IIT) Bombay, Mumbai, India. The analytical grade chemicals, such as ZnCl2, HCl, NaOH and CAC were purchased from Merck chemicals, Mumbai, India. AMX and alkali lignin (with low sulfonate content) were procured from Sigma Aldrich Private Limited, Mumbai, India. The concentration of both the model compounds in synthetic wastewater was 50–250 mg/l which is similar to those reported in previous studies (Andersson et al., 2011a,b; Putra et al., 2009). The synthetic wastewater solutions were prepared by dissolving appropriate amount of lignin or AMX in distilled water. 2.2. Methods 2.2.1. Activated carbon preparation and characterization Primary sewage sludge was used as precursor to prepare synthetic adsorbents. The oven-dried sludge was activated chemically by mixing with ZnCl2 solution in two different mass ratios (i.e. 1:2 and 1:2.5). The slurry was then subjected to pyrolysis at 600°C for one hour.

Subsequently, the pyrolysed material was washed with acid and water in the same sequence. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

The detailed preparation method of the adsorbent has been reported in our earlier publication (Gupta and Garg, 2015). The adsorbents were represented as ACZn2 and ACZn2.5 in accordance to the mass ratio of the activating agent to sludge material (i.e., 2 and 2.5, respectively). The samples were subjected to elemental analysis, surface area and porosity determination, scanning electron microscopy– energy dispersive X-ray spectroscopy (SEMEDX), X-ray diffraction (XRD) and Fourier Transform Infra-red spectroscopy (FTIR). 2.2.2. Determination of lignin and AMX concentration in liquid samples The concentration of an individual compound in aqueous solution was determined on a UV/vis spectrophotometer (model: UV 210 A, Schimadzu, Japan). The solutions containing individual compound were scanned in the wavelength range of 200 – 700 nm. The wavelengths of the maximum absorbance for lignin and AMX were found to be 280 nm and 230 nm, respectively at which the calibration curves were drawn for the two compounds to determine their concentration in unknown liquid samples (Fig. S1). 2.2.3. Batch adsorption experiments The batch adsorption study was initiated with the determination of equilibrium time, i.e., the time needed to achieve the maximum adsorption level. For each experimental run, 50 ml of the wastewater sample with known amount of the adsorbent was taken in 100 ml conical flask and shaken in an orbital shaker. For kinetic study, the initial concentration of the adsorbate was 100 mg/l. The conical flasks were removed at regular time intervals and the residual adsorbate concentration was determined after filtering the treated sample through 0.45 µm pore size Nylon membrane filter paper. The amount of adsorbate lost on the walls of conical flasks and filter paper was found insignificant (less than 2%). All the experiments were performed in triplicates and the average of three values was used for the calculation of design parameters. For isotherm and thermodynamic studies, the changes in the equilibrium

adsorbate concentrations were recorded for different initial concentrations of the adsorbate 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

and wastewater temperatures.

3. Results and discussion 3.1. Characterization of adsorbents The major characteristics of the synthesized adsorbents and CAC are presented in Table 1. The detailed characteristics of various adsorbents have been discussed in our recent publication (Gupta and Garg, 2015). According to the elemental analysis results, carbon in the synthesized adsorbents was 62 – 70% compared to ~92% in CAC. BET surface area and total pore volume of the synthesized adsorbents and CAC were ~500 m2/g and 0.3 cm3/g, respectively. The porous morphology in all the adsorbents was confirmed from SEM images while the presence of the favourable functional groups (such as carbonyl, ether and hydroxyl) was confirmed from FTIR spectra of the adsorbents. The point of zero charge (pHpzc) for the various adsorbents was slightly above neutral pH (i.e. 7.2 – 7.4). Insert Table 1 here 3.2. Batch adsorption studies The performance of sewage sludge derived adsorbents and CAC under varying reaction conditions was compared and discussed in subsequent sub-sections. 3.2.1. Equilibrium time determination To determine the equilibrium time for lignin adsorption on activated carbon samples, the initial runs were conducted for a time period of 12 h at ambient temperature. An adsorbent concentration of 2.5 g/l was added to 50 ml of the synthetic wastewater solution (lignin concentration = 100 mg/l). The wastewater pH and shaker speed were 7 and 100 rpm, respectively. The lignin removal curve was levelled off after 6 h duration thereby indicating

the attainment of equilibrium condition. Hence, all the succeeding runs with lignin 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

contaminated wastewater were performed for 6 h duration. For the treatment of AMX (initial concentration = 100 mg/l) contaminated synthetic wastewater, the equilibrium study was conducted for 72 h duration at ambient conditions. The aforementioned duration was selected based on previously reported study (Carabineiro et al., 2012). The wastewater pH, adsorbent dosage and shaker speed were 6, 3 g/l, and 100 rpm, respectively. During first 8 hours, the adsorption rate of AMX was high which was followed by slow removal rate until the equilibrium was reached after 60 h. Therefore, all further experiments with AMX laden synthetic wastewater were performed for 60 h duration. The stability of the two synthetic solutions containing lignin and AMX was also checked by conducting control tests (without addition of adsorbent) for the respective equilibrium duration (i.e., 6 and 60 h, respectively). No change in the solution absorbance (at 280 and 230 nm, respectively) was observed. So, it can be confirmed that the lignin and AMX solutions were stable for the duration of adsorption process. 3.2.2. Effect of adsorbent dosage The adsorbent dose was varied from 0.5 – 5 g/l to find their effect on the removal of lignin and AMX (initial pollutant concentration = 100 mg/l). With increase in adsorbent dose up to 2.5 and 3.0 g/l for lignin and AMX respectively, the adsorbate removal was increased (Fig. S2). The maximum corresponding lignin and AMX removal of 86% and 81% was obtained with CAC. Further Increase in the adsorbent concentration caused no additional reduction of pollutants from the synthetic solutions. It can be suggested that a dynamic equilibrium might have established between the adsorbate molecules adhere on adsorbent surface and in synthetic aqueous solution. As a result, no further adsorption of the adsorbate molecules occurred despite the availability of more active adsorption sites at the higher adsorbent doses. One of the waste derived adsorbents, i.e., ACZn2.5, showed comparable lignin removal

(~85%) at the same optimum dose of 2.5 g/l. But both waste derived adsorbents showed 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

inferior performance for AMX removal from the wastewater. For instance, AMX removal was only 72% in the presence of ACZn2.5. 3.2.3. Effect of initial pH and other possible adsorption mechanisms To investigate the effect of initial pH on the adsorption process, the wastewater pH was varied from 3 – 11 (Fig. 1). The corresponding adsorption capacity of CAC for lignin and AMX solutions was found to be 18.4 – 72 and 14.2 – 59.2 mg/g, respectively. The corresponding adsorption capacity of sewage sludge derived activated carbons, ACZn2 and ACZn2.5, for AMX was 12 – 48.3 and 12.7 – 50.8 mg/g which is ~15 – 20% less than that obtained with CAC. On the other hand, the waste derived adsorbents ACZn2 and ACZn2.5 showed better affinity for lignin molecules compared to AMX showing adsorption capacity of 17.2 – 60 and 18 – 68 mg/g, respectively. It should also be noted that the performance of ACZn2.5 was comparable to CAC for lignin adsorption. The adsorption of lignin onto the waste derived materials and commercial adsorbent was found to decrease on either side of the neutral pH (Fig. 1a). Lignin polymer is mainly composed of three alcohol units: coniferyl, sinapyl and p-coumaryl with pKa values ranging between 9.9 and 10.3 (Ragnar et al., 2000). Therefore, the adsorbing species dissociate in anionic form above a pH of ~10 resulting in the repulsion from the negatively charged adsorbent surface layer which would cause the reduced adsorption. In the acidic range, the additional protons may compete with lignin molecules for the carbonyl sites. As a consequence, the removal of lignin molecules would have diminished. The highest adsorption capacity was 72 mg/g obtained with CAC compared to 60 and 68 mg/g with ACZn2 and ACZn2.5, respectively. The other probable mechanisms of lignin adsorption on activated carbon are discussed in the subsequent paragraph.

The lignin adsorption on CAC and synthesized activated carbon may be attributed to the 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

formation of hemiacetal after the reaction between hydroxyl groups present in lignin and carbonyl group on activated carbon (Fig. S3). In addition, the charge transfer mechanism can also lead to the adsorption of lignin molecules (Rivera-Urtilla et al., 2013). The carbonyl surface oxygen groups and the aromatic ring of phenolic group may act as electron donor and acceptor, respectively which would result in the formation of ‘donor-acceptor’ complex. As another possibility, the formation of hydrogen bonds between the carbonyl (or ether) surface oxygen atoms and hydrogen atoms in primary and secondary alcohols (present in the adsorbate molecule) may also be responsible for the removal of lignin from the synthetic wastewater. Since lignin exists as a neutral molecule in a wide pH range, their adsorption most likely involves hydrogen bonding and/or van der Waals interaction (Baccar et al., 2012). Similarly, AMX adsorption can also be discussed based on pH of the wastewater and other mechanisms. The dissociation constants (i.e. pKa values) of three major functional groups (i.e. carboxyl, amine and phenol) present in AMX are 2.6, 7.4 and 9.6, respectively (Putra et al., 2009). The maximum AMX removal was obtained at an initial wastewater pH of 6 (Fig. 1b). In a pH range around 3– 7 (i.e. pKa1< pH < pKa2, pKa3), amine functional groups will be present as –NH3+ and carboxyl functional groups will be present as –COO- (Putra et al., 2009). Therefore, the electrostatic force of attraction will be operational between the activated carbon (as pH less than pHpzc) and –COO- while the repulsive force will exist between positively charged activated carbon and –NH3+ present in AMX molecule. Moussavi et al. (2013) have also reported the optimum pH value of 6.0 for AMX adsorption. The adsorption of AMX on CAC, ACZn2.5 and ACZn2 can also be explained by various other mechanisms which are illustrated in Fig. S3. Due to the small molecular size of AMX (~13 Å), it could be penetrated into the pores of activated carbon. The reaction between amine and carbonyl groups present on AMX molecule and activated carbon, respectively, can

result in a compound with C=N functional group along with the release of water molecules 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

leading to chemisorption (Putra et al., 2009). The AMX adsorption can also be attributed to the hemiacetal and ‘donor-acceptor’ complex formations as discussed above for lignin adsorption. The dispersive interactions between π electrons from aromatic ring in an AMX molecule and graphene planes present on the surface of activated carbon may also be involved in AMX adsorption (Rivera-Utrilla et al., 2013). Hydrogen bonds between the carbonyl (or ether) surface oxygen atoms and hydrogen atoms in amine functional groups as well as hydroxyl groups present in phenol may also be responsible for its adsorption. Ding et al. (2012) suggested that polar interactions of antibiotics with the mineral phase available in the pores of the adsorbents may be another important mechanism. As another possibility, the hydrophobic interactions may occur between surface sites on the activated carbon with AMX molecules which should be in the neutral form in a pH range of 3 – 6 (Westerhoff et al., 2005). The hydrophobic interactions are determined by octanol/water coefficient (Kow) of an adsorbate. The value of log(Kow) for AMX (~0.87) implies the involvement of hydrophobic interactions to some extent. Insert Fig. 1 here 3.2.4. Adsorption isotherms studies Various two-parameter adsorption isotherm models (Langmuir, Freundlich, Temkin and Dubinin-Radushkevich models) along with a three-parameter Redlich-Peterson model were employed to fit the equilibrium adsorption data obtained for the two adsorbates. The isotherms can be represented by the following equations (Foo and Hameed, 2010): Langmuir isotherm: qe = Q.KL.Ce/ (1+KL.Ce)

(1)

Freundlich isotherm:

(2)

Temkin isotherm: qe = B1.ln KT + B1.lnCe

(3)

Dubinin-Radushkevich isotherm: qe = qm exp (-KD.ε2)

(4)

Redlich-Peterson isotherm: qe = KR.Ce / (1+ αR.Ceβ) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

(5)

where Q and KL = Langmuir constants, KF and n = Freundlich constants, B1 and KT = Temkin constants, qm and KD = Dubinin-Radushkevich constants and KR, αR and β = RedlichPeterson constants. In mathematical form, ε can be expressed as follows: ε = RT ln (1+1/Ce)

(6)

The determination of ‘separation factor (RL)’ assists in understanding the status of adsorption process (i.e., favourable or non-favourable). It is calculated using Langmuir constant and initial adsorbate concentration by the following expression: RL = 1/ (1+KL.C0)

(7)

The validity of the models was checked using different error functions: sum of square errors (SSE), average relative error (ARE), Hybrid fractional error function (HYBRID), Marquardt’s percent standard deviation (MPSD) and non-linear chi-square test (χ2) (Foo and Hameed, 2010). Out of these error functions, SSE and χ2 were based on the square residuals whereas ARE considers the difference between the experimental and calculated values of adsorption capacity obtained from various models. HYBRID and MPSD take the degrees of freedom (i.e. the number of data points minus the number of parameters within the isotherm equation) into account. The equations for various error functions are: SSE =

(8) (9)

χ2=

(10) (11) (12)

The lignin and AMX adsorption systems followed ‘L curve’ pattern (Figs. 2 and 3, 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

respectively) as per the classification suggested by Giles et al. (1974). This means that more the occupied active sites on activated carbon, difficult will be the further adsorption. Freundlich, Langmuir and Redlich-Peterson models were found reasonable fit to the experimental data in accordance with the higher R2 values and lower values of error functions in most of the cases. However, Temkin isotherm exhibited fair fit only in case of lignin adsorption on both waste derived adsorbents (i.e., ACZn2 and ACZn2.5 adsorbents. DubininRadushkevich model failed to provide an adequate fit to the experimental data. The values of various isotherm parameters are presented in Table 2 while the error functions are given in Table 3 of supporting information. To determine whether the adsorption was favourable or not, n and RL values were calculated using Freundlich and Langmuir isotherm equations, respectively. Redlich-Peterson isotherm combines the elements from both Freundlich and Langmuir isotherms and suggests that the adsorption mechanism is hybrid. The relative adsorption capacity of CAC was higher than ACZn2 and ACZn2.5 as indicated by the KR values. The corresponding β values were ranged between 0.57 – 0.61 for lignin and 0.89 – 0.91 for AMX. The αR values confirmed the higher affinity of the two adsorbates towards CAC. Insert Fig. 2 here Insert Fig. 3 here Insert Table 2 here Insert Table 3 here 3.2.5. Adsorption kinetics studies In order to describe the adsorption kinetics, the time based adsorption capacity (i.e. qt) data was plotted in accordance to the pseudo first and second order reaction, intra-particle diffusion and Elovich models (used mainly for describing chemisorption). Apart from this,

the effect of pore diffusion on the adsorption process was observed using Weber-Morris 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

intra-particle diffusion model. The equations for various kinetic models are (Singha and Das, 2013): Pseudo first order: log (qe - qt) = log qe – [(k1/2.303).t]

(13)

Pseudo second order: qt = t.k2.qe2/ (1+ t.k2.qe)

(14)

Intra-particle diffusion model: qt = kid.t0.5 + C

(15)

Elovich model: qt = (1/β) ln (α.β) + (1/β) ln t

(16)

where k1 and k2 = pseudo first and second order rate constants, respectively, kid = intraparticle diffusion rate constant, α = initial adsorption rate and β = desorption constant. The temperature dependence of rate constant and activation energy (Ea) were predicted by performing additional adsorption runs at higher temperatures (i.e., at 35 and 45°C temperatures). The Arrhenius equation was used to determine activation energy of the adsorption process. The parameter provides information about the nature of adsorption process (i.e. physical or chemical). For physical adsorption, the Ea values are usually smaller (generally less than 4.2 kJ/mol) due to the adherence of adsorbate molecules on the porous solid surface by weak van der Walls forces (Hu and Hu, 2013). In contrast, chemical adsorption is caused by stronger forces and hence needs much higher activating energy. Arrhenius equation can be presented as follows: ln k = ln A – (Ea/R.T)

(17)

where k = rate constant and A = frequency factor. The coefficient of determination (i.e. R2 and SSE values) shows the validity of pseudo first and second order reactions for the experimental data for lignin and AMX adsorption (Figs. S4 and S5). The activation energy for lignin and AMX adsorption was 3.5 – 4.5 and 12 – 22 kJ/mol, respectively. The attachment of lignin is primarily attributed to the physical adsorption whereas AMX adsorption on the synthesized and commercial adsorbents can be

ascribed to a combination of physical and chemical adsorption. The determination of 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

thermodynamic parameters (discussed in the next subsection) provides better insight to the nature of adsorption process. The intra-particle diffusion model was also applied to both the adsorption systems to determine the rate controlling step (Figs. S4 and S5). The plot between qt against t0.5 showed three linear segments indicating the occurrence of adsorption process in the same number of steps. The external surface adsorption (i.e. film diffusion stage) could be identified by the initial rapid step while the rate controlled step (second stage) showed the continuation of adsorption at slower rate. The equilibrium stage (i.e. horizontal line) suggested the completion of the adsorption process. The intra-particle diffusion rate constant (kid) was found to be higher, in the range of 16.8 – 17.5 mg/g-h0.5, for lignin adsorption system compared to AMX adsorption (1 – 1.3 mg/g-h0.5). Elovich plot was found to fit well with the kinetic data for both the adsorbates as evidenced by the high values of correlation coefficients (Table 4). From α and β values for various adsorbents, it can be suggested that the rate of adsorption and desorption on/from waste derived adsorbents was comparable to CAC. Insert Table 4 here 3.2.6. Adsorption thermodynamics The thermodynamics of the adsorption systems was studied by performing the adsorption runs at three temperatures (i.e. 25, 35 and 45°C). The equilibrium constant (K), Gibbs’ free energy change (ΔG°), entropy change (ΔS°) and heat of adsorption (ΔH°) were calculated using the following formulae: ΔG° = -R.T ln K

(18)

ΔG° = ΔH° – T.ΔS°

(19)

ln K = (ΔS°/R) – (ΔH°/R.T)

(20)

where R is the universal gas constant and T is the temperature in Kelvin.

ln K at a particular temperature was obtained by drawing a plot between ln (qe/Ce) and qe 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

(Khan and Singh, 1987). ΔS° and ΔH° were calculated from the plot between ln K vs (1/T). The positive values of ΔH° indicated the endothermic nature of the lignin and AMX adsorption systems (5.6 – 13.9 kJ/mol and 8.5 – 16 kJ/mol, respectively). This suggests that the adsorption process will be better at higher temperature (Zhou et al., 2014). ΔS° was found in the range of 30.9 – 62.4 and 35.2 – 55.9 J/mol-K for lignin and AMX adsorption, respectively, which indicates an increase of entropy (or in other words randomness or disorder), i.e., the occurrence of favourable adsorption. Higher disorder results in better energy dispersion. Besides, the positive ΔS° increases the randomness on solid-liquid interface with the loading of adsorbate molecules on the surface of commercial and waste derived activated carbon (Angin, 2014). The negative free energies (ΔG°) values also confirm the favourable and spontaneous adsorption for both the adsorbate molecules on commercial as well as synthesized activated carbon materials. The adsorption process was found to be more spontaneous (i.e., higher negative ΔG° values) with increase in temperature from 25 to 45°C. It also confirms the endothermic nature of the adsorption process. Since ΔG° values for the two systems were found between -20 and 0 kJ/mol, the removal of adsorbates from the wastewater may be considered primarily due to physical adsorption (Feng et al., 2011). The values of thermodynamic parameters are presented in Table 5. Using the isosteric method, the differential enthalpy of adsorption (i.e., the change in the internal energy of an adsorption system) upon the addition of an excess amount of infinitesimal surface can be determined. The apparent isosteric heat of adsorption (ΔHst,a) at constant surface coverage was calculated using Clausius-Clapeyron equation (Srivastava et al., 2006; Kyzas et al., 2014). The equilibrium concentration (Ce) at constant amount of the adsorbed compound is measured from isotherm data at different temperatures and the slope of a plot between lnCe versus (1/T) at constant qe values gives ΔHst,a.

ΔHst,a = R [d lnCe]/d(1/T) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

(21)

The adsorption isosteres for lignin and AMX removal on the various adsorbents are shown in Figs. 4 and 5, respectively. ΔHst,a values for lignin adsorption system at constant qe values (i.e. 20, 25, 30, 35, 40 mg/g) were found in the range of 2.7 – 15.8 kJ/mol whereas it varied between 3.0 – 11.4 kJ/mol for AMX adsorption system. The ΔHst,a varied with surface loading thereby indicating that the adsorbents have an energetically heterogeneous surface (Srivastava et al., 2006). Insert Table 5 here Insert Fig. 4 here Insert Fig. 5 here 3.3. Stability of sewage sludge derived adsorbents The evaluation of the stability of sewage sludge based adsorbents is required for their practical application. Therefore, a few continuous long-term adsorption experiments were carried out in duplicate with both the synthesized adsorbents (ACZn2 and ACZn2.5) for lignin and AMX solutions (initial concentration = 100 mg/l). The test was conducted for 120 h duration at the corresponding optimum pH and adsorbent dosage for both the synthetic wastewater solutions. No appreciable reduction in adsorption capacity shows that the adsorbents were stable for the extended test duration (Fig. S6). 4. Conclusions Sewage sludge based adsorbents showed reasonable performance for the removal of lignin and AMX from synthetic wastewater. The adsorption process was governed by a number of mechanisms which may include electrostatic forces, donor-acceptor complex formation and hydrogen bonding, depending upon the operating conditions. Redlich-Peterson isotherm was found the best among various isotherms. The lignin and amoxicillin adsorption data showed

better fit to the pseudo first and second order models, respectively. Adsorption process was 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

endothermic and the spontaneity increased at higher temperatures. The recycling of sewage sludge as adsorbent can reduce its transfer and disposal costs significantly. Acknowledgements The authors are thankful to Sophisticated Analytical Instrument Facility (SAIF), Department of Chemical Engineering and Material Sciences and Metallurgical Engineering (MEMS) of Indian Institute of Technology (IIT) Bombay, Mumbai, India for their kind support for analysing the adsorbents. Annexure A. Supplementary data Supplementary figures (Figs. S1 to S6) are added as Electronic annex. References 1.

Angin, D., 2014. Utilization of activated carbon produced from fruit juice industry solid

waste for the adsorption of Yellow 18 from aqueous solutions, Bioresource Technol. 168, 259–266. 2.

Andersson, K.I., Eriksson, M., Norgren, M., 2011a. Removal of lignin from wastewater

generated by mechanical pulping using activated carbon and fly ash: Adsorption isotherms and thermodynamics. Ind. Eng. Chem. Res. 50, 7722-7732. 3.

Andersson, K.I., Eriksson, M., Norgren, M., 2011b. Removal of lignin from wastewater

generated by mechanical pulping using activated carbon and fly ash: Adsorption kinetics. Ind. Eng. Chem. Res. 50, 7733-7739. 4.

Baccar, R., Sarra, M., Bouzid, J., Feki, M., Blanquez, P., 2012. Removal of

pharmaceutical compounds by activated carbon prepared from agricultural by-product. Chem. Eng. J. 211-212, 310-317.

5. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Boualem, T., Debab, A., Yuso, A.M., Izquierdo, M.T., 2014. Activated carbons

obtained from sewage sludge by chemical activation: Gas-phase environmental applications. J. Environ. Manage. 140, 145-151. 6.

Carabineiro, S.A.C., Thavorn-amornsri, T., Pereira, M.F.R., Serp, P., Figueiredo, J.L.,

2012. Comparison between activated carbon, carbon xerogel and carbon nanotubes for the adsorption of the antibiotic ciprofloxacin. Catal. Today 186, 29-34. 7.

Ding, R., Zhang, P., Seredych, M., Bandosz, T.J., 2012. Removal of antibiotics from

water using sewage sludge and waste oil sludge derived adsorbents. Water Res. 46, 40814090. 8.

Feng, Y., Yang, F., Wang, Y., Ma, L., Wu, Y., Kerr, P.G., Yang, L., 2011. Basic dye

adsorption onto an agro-based waste material-Sesame hull (Sesamum indicum L.). Bioresource Technol. 102, 10280–10285. 9.

Foo, K.Y., Hameed, B.H., 2010. Insights into the modelling of adsorption isotherm

systems, Chem. Eng. J. 156, 2-10. 10.

Giles, C.H., Smith, D., Huiston, A., 1974. A general treatment and classification of the

solute adsorption isotherms. J. Colloid Interf. Sci. 47, 755-765. 11.

Giri, A.S., Golder, A.K., 2014. Ciprofloxacin degradation from aqueous solution by

Fenton oxidation: reaction kinetics and degradation mechanisms. RSC Adv. 4, 6738-6745. 12.

Gupta, A., Garg, A., 2015. Primary sewage sludge derived activated carbon:

Characterisation and application in wastewater treatment. Clean Technol. Envir., DOI 10.1007/s10098-014-0895-4 (In press). 13.

Hu, S.H., Hu, S.C., 2013. Pyrolysis of paper sludge and utilization for ionic dye

adsorption. BioResources 8, 1028-1042. 14.

Khan, A.A., Singh, R.P., 1987. Adsorption thermodynamics of carbofuran on Sn(IV)

arsenosilicate in H+, Na+ and Ca2+ forms. Colloid. Surface. 24, 33-42.

15. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Kyzas, G.Z., Deliyanni, E.A., Matis, K.A., 2014. Graphene oxide and its application as

an adsorbent in wastewater treatment. J. Chem. Technol. Biotechnol. 89, 196-205. 16.

Lal, K., Garg, A., 2015. Catalytic wet oxidation of phenol under mild operating

conditions: development of sludge pathway and sludge characterization. Clean Technol. Envir., 17, 199-210. 17.

Moussavi, G., Alahabadi, A., Yagmaeian, K., Eskandari, M., 2013. Preparation,

characterization and adsorption potential of the NH4Cl-induced activated carbon for the removal of amoxicillin antibiotic from water. Chem. Eng. J. 217, 119-128. 18.

Putra, E.K., Pranowa, R., Sunarso, J., Indraswati, N., Ismadji, S., 2009. Performance of

activated carbon and bentonite for removal of amoxicillin from wastewater: Mechanisms, isotherms and kinetics. Water Res. 43, 2419-2430. 19.

Ragnar, M., Lindgren, C.T., Nilvebrant, N.O., 2000. pKa values of guaiacyl and

syringyl phenols related to lignin. J. Wood Chem. Technol. 20, 277-305. 20.

Rivera-Utrilla, J., Sanchez-Polo, M., Ferro-Garcia, M.A., Pardos-Joya, G., Ocampo-

Perez, R., 2013. Pharmaceuticals as emerging contaminants and their removal from water: A review. Chemosphere 93, 1268-1287. 21.

Singha, B., Das, S.K., 2013. Adsorptive removal of Cu(II) from aqueous solution and

industrial effluent using natural/agricultural wastes. Colloid. Surface. B 107, 97-106. 22.

Smith, K.M., Fowler, G.D., Pulket, S., Graham, N.J.D., 2009. Sewage sludge-based

adsorbents: A review of their production, properties and use in water treatment applications. Water Res. 43, 2569-2594. 23.

Srivastava, V.C., Swamy, M.M., Mall, I.D., Prasad, B., Mishra, I.M., 2006. Adsorptive

removal of phenol by bagasse fly ash and activated carbon: Equilibrium, kinetics and thermodynamics. Colloid. Surface. A 272, 89-104.

24. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Westerhoff, P., Yoon, Y., Snyder, S. and Wert, E., 2005. Fate of endocrine-disruptor,

pharmaceutical, and personal care product chemicals during simulated drinking water treatment processes. Environ. Sci. Technol. 39, 6649–6663. 25.

Xu, G., Yang, X., Spinosa, L., 2015. Development of sludge-based adsorbents:

Preparation, characterization, utilization and its feasibility assessment. J. Environ. Manage. 151, 221-232. 26.

Yadav, A., Mukherji, S., Garg, A., 2013. Removal of chemical oxygen demand and

color from simulated textile wastewater using a combination of chemical/physiochemical process. Ind. Eng. Chem. Res. 52, 10063-10071. 27.

Yadav, B.R., Garg, A., 2014. Catalytic wet oxidation of ferulic acid (a lignin model

compound) in the presence of non-noble metal-based catalysts at mild conditions. Chem. Eng. J. 252, 185-193. 28.

Zhu, X., Liu, Y., Qian, F., Zhou, C., Zhang, S., Chen, J., 2014. Preparation of magnetic

porous carbon from waste hydrochar by simultaneous activation and magnetization for tetracycline removal. Bioresource Technol.154, 209-214. 29.

Zhuang, H., Han, H., Hou, B., Jia, S., Zhao, Q., 2014. Heterogenous catalytic ozonation

of biologically treated Lurgi coal gasification wastewater using sewage sludge based activated carbon supported manganese and ferric oxides as catalysts. Bioresource Technol. 166, 178-186. 30.

Zou, J., Dai, Y., Wang, X., Ren, Z., Tian, C., Pan, K., Li, S., Abuobeidah, M., Fu, H.,

2013. Structure and adsorption properties of sewage sludge-derived carbon with removal of inorganic impurities and high porosity. Bioresource Technol. 142, 209-217.

List of Figures 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Fig. 1. Effect of initial pH on the percent removal of the model compounds: (a) lignin (adsorbent concentration = 2.5 g/l) and (b) AMX adsorption (adsorbent concentration = 3 g/l) with initial adsorbate concentration = 100 mg/l Fig. 2. Equilibrium data fit to various adsorption isotherm models for lignin adsorption on various adsorbents: (a) ACZn2, (b) ACZn2.5 and (c) CAC [Experimental: E, Freundlich: F, Langmuir: L, Temkin: T, Dubinin-Radushkevich: DR, Redlich-Peterson: RP] Fig. 3. Equilibrium data fit to various adsorption isotherm models for AMX adsorption on various adsorbents: (a) ACZn2, (b) ACZn2.5 and (c) CAC [Experimental: E, Freundlich: F, Langmuir: L, Temkin: T, Dubinin-Radushkevich: DR, Redlich-Peterson: RP] Fig. 4. Adsorption isosteres for lignin adsorption on various adsorbents: (a) ACZn2, (b) ACZn2.5 and (c) CAC Fig. 5. Adsorption isosteres for AMX adsorption on various adsorbents: (a) ACZn2, (b) ACZn2.5 and (c) CAC

List of Tables 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65

Table 1. Major characteristics of the sewage sludge derived adsorbents and CAC Table 2. Isotherm parameters for lignin and AMX adsorption systems Table 3. Values of error functions obtained for various isotherm models Table 4. Kinetic constants and activation energy for lignin and AMX adsorption systems Table 5. Determination of thermodynamic parameters for lignin and AMX adsorption systems

Table 1. Major characteristics of the sewage sludge derived adsorbents and CAC Adsorbent

ACZn2

ACZn2.5

CAC

Carbon (% dry

62.25

69.85

92.06

502.65

510.80

514.12

0.289

0.297

0.296

4.58

4.67

4.67

Nature

Non-amorphous

Non-amorphous

Amorphous

Minerals

Quartz (Q) and

Quartz (Q) and

Quartz (Q)

Feldspars (F)

Feldspars (F)

O-H, C=C, C=O and

O-H, C=C, C=O and

O-H, C=C, C=O and

C-O

C-O

C-O

basis) BET surface area (m2/g) Total pore volume (cm3/g) Average pore diameter (nm)

Functional groups

1

Table 2. Isotherm parameters for lignin and AMX adsorption systems Adsorbent

Langmuir isotherm

Freundlich isotherm

Temkin isotherm

D-R isotherm

Lignin Q = 71.43, KL = 0.04, RL = 0.09 – 0.35, MPSD = 5.56

AMX Q = 90.91, KL = 0.01, RL = 0.27 – 0.65, MPSD = 3.83

Lignin Kf = 6.84, 1/n = 0.51, MPSD = 11.64

AMX Kf = 1.87, 1/n = 0.71, MPSD = 4.55

Lignin KT = 0.39, B1 = 17.33, MPSD = 2.8

AMX KT = 0.12, B1 = 18.33, MPSD = 11.92

Lignin qm = 51.21, KD = 10.39, MPSD = 20.48

AMX qm = 39.29, KD = 44.78, MPSD = 24.67

Lignin KR = 18.24, αR = 1.15, β = 0.58, MPSD = 2.27

AMX KR = 0.99, αR = 0.017, β = 0.91, MPSD = 2.27

ACZn2.5

Q = 76.92, KL = 0.06, RL = 0.06 – 0.24, MPSD = 8.92

Q = 83.33, KL = 0.01, RL = 0.21 – 0.57, MPSD = 3.43

Kf = 8.83, 1/n = 0.47, MPSD = 5.35

Kf = 2.49, 1/n = 0.67, MPSD = 4.8

KT = 0.50, B1 = 17.55, MPSD = 7.75

KT = 0.15, B1 = 18.27, MPSD = 12.12

qm = 52.77, KD = 5.46, MPSD = 25.58

qm = 40.89, KD = 31.44, MPSD = 24.81

KR = 18.39, αR = 1.26, β = 0.59, MPSD = 3.78

KR = 1.27, αR = 0.023, β = 0.92, MPSD = 1.67

CAC

Q = 83.33, KL = 0.09, RL = 0.04 – 0.19, MPSD = 12.69

Q = 90.91, KL = 0.024, RL = 0.14 – 0.45, MPSD = 4.09

Kf = 9.66, 1/n = 0.47, MPSD = 1.93

Kf = 4.05, 1/n = 0.64, MPSD = 4.19

KT = 0.58, B1 = 18.1, MPSD = 13.03

KT = 0.24, B1 = 19.81, MPSD = 12.84

qm = 53.95, KD = 3.63, MPSD = 26.69

qm = 46.15, KD = 12.88, MPSD = 26.69

KR = 18.52, αR = 1.33, β = 0.61, MPSD = 0.97

KR = 2.23, αR = 0.036, β = 0.9, MPSD = 4.36

ACZn2

R-P isotherm

Units: Q - mg/g; KL - l/mg; Kf - (mg/g)/(mg/l)1/n; KT - l/mg; B1 - l/g; KR - l/g; αR - l/g; qm - mg/g; KD - mol2/kJ2

2

Table 3. Values of error functions obtained for various isotherm models Adsorbent

Langmuir isotherm

Freundlich isotherm

Temkin isotherm

D-R isotherm

R-P isotherm

lignin R2 = 0.99, χ2 = 1.45, ARE = 3.86, SSE = 27.33, HYBRID = 5.56

AMX R2 = 0.99, χ2 = 0.1, ARE = 2.51, SSE = 2.85, HYBRID = 4.18

lignin R2 = 0.98, χ2 = 1.77, ARE = 7.76, SSE = 99.54, HYBRID = 11.64

AMX R2 = 0.99, χ2 = 0.21, ARE = 3.44, SSE = 8.3, HYBRID = 5.73

lignin R2 = 0.99, χ2 = 6.01, ARE = 1.95, SSE = 6.01, HYBRID = 3.26

AMX R2 = 0.98, χ2 = 0.78, ARE = 7.48, SSE = 15.58, HYBRID = 12.47

lignin R2 = 0.89, χ2 = 4.68, ARE = 12.96, SSE = 217.78, HYBRID = 21.6

AMX R2 = 0.87, χ2 = 5.33, ARE = 16.28, SSE = 185.79, HYBRID = 27.13

lignin R2 = 0.99, χ2 = 0.05, ARE = 1.28, SSE = 2.27, HYBRID = 3.2

AMX R2 = 0.99, χ2 = 0.03, ARE = 1.05, SSE = 1.03, HYBRID = 2.62

ACZn2.5

R2 = 0.99, χ2 = 1.03, ARE = 6.02, SSE = 53.61, HYBRID = 10.03

R2 = 0.99, χ2 = 0.06, ARE = 2.3, SSE = 2.61, HYBRID = 3.84

R2 = 0.99, χ2 = 0.3, ARE = 3.54, SSE = 10.65, HYBRID = 5.35

R2 = 0.99, χ2 = 0.23, ARE = 3.52, SSE = 9.47, HYBRID = 5.87

R2 = 0.99, χ2 = 0.61, ARE = 5.54, SSE = 25.62, HYBRID = 9.23

R2 = 0.98, χ2 = 0.67, ARE = 7.23, SSE = 14.44, HYBRID = 12.05

R2 = 0.84, χ2 = 8.7, ARE = 16.14, SSE = 434.38, HYBRID = 26.9

R2 = 0.86, χ2 = 5.74, ARE = 16.1, SSE = 210.87, HYBRID = 26.83

R2 = 0.99, χ2 = 0.09, ARE = 1.95, SSE = 3.44, HYBRID = 4.88

R2 = 0.99, χ2 = 0.02, ARE = 0.96, SSE = 0.6, HYBRID = 2.39

CAC

R2 = 0.98, χ2 = 2.76, ARE = 7.96, SSE = 153.59, HYBRID = 12.69

R2 = 0.99, χ2 = 0.13, ARE = 2.73, SSE = 5.85, HYBRID = 4.54

R2 = 0.99, χ2 = 0.05, ARE = 1.21, SSE = 2.14, HYBRID = 2.02

R2 = 0.99, χ2 = 0.2, ARE = 2.87, SSE = 9.81, HYBRID = 4.79

R2 = 0.97, χ2 = 1.63, ARE = 8.72, SSE = 61.5, HYBRID = 14.53

R2 = 0.98, χ2 = 1.07, ARE = 7.74, SSE = 26.42, HYBRID = 12.9

R2 = 0.84, χ2 = 1.63, ARE = 16.86, SSE = 338.42, HYBRID = 28.1

R2 = 0.84, χ2 = 8.02, ARE = 16.86, SSE = 338.42, HYBRID = 28.1

R2 = 0.99, χ2 = 0.03, ARE = 0.58, SSE = 0.52, HYBRID = 1.44

R2 = 0.99, χ2 = 0.11, ARE = 2.62, SSE = 3.8, HYBRID = 6.55

ACZn2

3

Table 4. Kinetic constants and activation energy for lignin and AMX adsorption systems Adsorbent

qe,exp (mg/g)

ACZn2

ACZn2.5

CAC

Lignin

AMX

Lignin

AMX

Lignin

AMX

32.8

22.67

34

24

34.4

27

Pseudo first order model qe,cal (mg/g)

43.35

16.22

46.24

14.69

43.75

17.06

K1 (h-1)

0.58

0.06

0.62

0.07

0.65

0.08

R2

0.94

0.96

0.91

0.94

0.97

0.96

SSE

18.55

5.2

25

10.83

14.57

12.35

Pseudo second order model qe,cal (mg/g)

43.48

25.64

43.48

27.03

43.48

29.41

K2 (g/mg-h)

0.01

0.01

0.01

0.02

0.01

0.02

R2

0.75

0.99

0.79

0.99

0.84

0.99

SSE

19

0.98

12.84

1.02

11.78

0.65

Weber-Morris (Intra-particle diffusion) model kid (mg/g-h0.5)

17.29

1.23

17.45

1.02

16.85

1.21

R2

0.99

0.99

0.99

0.99

0.96

0.97

α (mg/g-h)

24.51

30.78

26.67

61.71

29.79

75.16

β (g/mg)

0.07

0.24

0.07

0.26

0.07

0.24

R2

0.98

0.99

0.99

0.97

0.99

0.99

21.41

4.27

16.21

3.75

12.65

Elovich model

Arrhenius equation Ea (kJ/mol)

4.19

4

Table 5. Determination of thermodynamic parameters for lignin and AMX adsorption systems Adsorbent

ACZn2

ACZn2.5

CAC

ΔH° (kJ/mol)

ΔS° (J/mol-K)

ΔG° (kJ/mol)

Lignin AMX

Lignin

AMX

Lignin

AMX

5.66

30.9

35.23

-3.52

-0.04

35

-3.88

-0.32

45

-4.4

-0.70

-4.38

-0.72

35

-5.02

-1.20

45

-5.81

-1.84

-4.72

-2.27

35

-5.46

-2.58

45

-5.74

-2.99

T (°C)

25

25

25

11.84

13.83

10.49

15.99

8.56

54.57

62.39

5

55.89

36.23

% Lignin removal

100 90 80 70 60 50 40 30 20 10 0

(a)

ACZn2 ACZn2.5 CAC 2

3

4

5

6

7

8

9

10

11

12

pH

100

% AMX removal

90

(b)

80 70 60 50 40 30

ACZn2

20

ACZn2.5

10

CAC

0 2

3

4

5

6

7

pH Fig. 1.

8

9

10

11

12

80

80

(a)

(b)

70 60

50

50

qe (mg/g)

60

40 30

40 30

20

20

10

10

0

0 0

20

40

60

80

100

0

20

Ce (mg/l)

40

60

Ce (mg/l)

80

(c)

70 60

qe (mg/g)

qe(mg/g)

70

50 40

30 20 10 0 0

20

40

60

80

100

Ce (mg/l) E

F

L

T Fig. 2.

DR

RP

80

100

70

70

(a)

60 50

qe (mg/g)

50

40 30

40 30

20

20

10

10

0

0

20

40

60

80

0

100 120

0

20

Ce (mg/l)

40

60

(c)

60 50 40

30 20

10 0 0

20

40

60

80

100

120

Ce (mg/l)

E

80

Ce (mg/l)

70

qe (mg/g)

qe (mg/g)

(b)

60

F

L

T

Fig. 3.

DR

RP

100 120

3

3

2.5

2.5

2

2

ln Ce

3.5

ln Ce

3.5

1.5

1.5 1

1

0 0.0031

0.5

(a) 0.0032

0.0033

(b)

0 0.0031

0.0034

0.0032

1/T

1/T (K-1)

0.0033

0.0034

(K-1)

3.5 3 2.5

ln Ce

0.5

2 1.5 1 0.5 0 0.0031

(c) 0.0032

0.0033

0.0034

1/T (K-1) 20 mg/g

25 mg/g

Fig. 4.

30 mg/g

35 mg/g

40 mg/g

4

4

3.5

3.5

3

3

2.5

2.5

ln Ce

4.5

ln Ce

4.5

2

2

1.5

1.5

1

1

(a)

0 0.0031

0.5 0.0032

0.0033

1/T

0 0.0031

0.0034

(K-1)

(b) 0.0032

1/T

0.0033

(K-1)

4.5 4 3.5 3 2.5

ln Ce

0.5

2

1.5 1 0.5 0 0.0031

(c) 0.0032

1/T

20 mg/g

25 mg/g

0.0033

0.0034

(K-1)

30 mg/g mg/g

Fig. 5.

35 mg/g

40

0.0034

Highlights  Transformation of sewage sludge into adsorbents by thermo-chemical process  Removal of recalcitrant pollutants (lignin and amoxicillin) by adsorption process  Possible adsorption mechanisms for model pollutants  Detailed isotherm, kinetic and thermodynamic studies

GRAPHICAL ABSTRACT

SEM image of sewage sludge derived adsorbent Sewage sludge derived adsorbent

Mechanistic aspects
Surface reactivity
Donor-acceptor interactions
Electrostatic interactions
Hydrogen bonding
Polar interactions
Van der Walls interactions
Hydrophobic interactions

Isotherms, Kinetics and Thermodynamics Adsorbent ACZn2

ACZn2.5

Lignin or AMX molecules

CAC

Lignin KR = 18.24 Ea = 4.19 ∆G⁰ = -3.52 KR = 18.39 Ea = 4.27 ∆G⁰ = -4.38 KR = 18.52 Ea = 3.75 ∆G⁰ = -4.72

AMX KR = 0.99 Ea = 21.41 ∆G⁰ = -0.04 KR = 1.27 Ea = 16.21 ∆G⁰ = -0.72 KR = 2.23 Ea = 12.65 ∆G⁰ = -2.27

KR in (l/g), Ea and ∆G⁰ in (kJ/mol)