Removal of pharmaceuticals from aqueous samples by adsorption using pristine polypyrrole as adsorbent: kinetic, isothermal and thermodynamic studies

ABSTRACT Conductive polymers have been used as adsorbents to remove organic pollutants from aqueous samples. In this study, pristine polypyrrole (PPy) was used for the first time in the removal of potassium diclofenac (PD) and moxidectin (MOX) from aqueous solutions.PPy was properly characterised by scanning electron microscopy, thermogravimetric analysis, infrared spectroscopy, textural properties, wettability, and point zero charge. Through the characterisation results, it was observed that the adsorbent showed be favourable to adsorption, high temperature of thermal degradation, bands characteristic of the PPy,mesopore structures, and hydrophilicity. The results pointed that pH 6 favoured the removal, reaching percentages of 95.26% for PD and 99.75% for MOX.The kinetic data followed the model of pseudo second-order (PD: R2 = 0.872 and MOX: R2 = 0.887) with equilibrium reached in 20 min for PD and MOX. OThe best fit isotherm model was the dual-site Langmuir-Freundlich (R2 > 0.99), with maximum capacity adsorption of 221.23 mg g−1 for PD and 87.46 mg g−1 for MOX. Finally, the thermodynamic study showed that the adsorption process was spontaneous and endothermic. Through the reuse study, it was possible to show a good cost-benefit ratio when using the developed adsorbent material, since it can be used several times without losing its adsorption capacity. Finally, PPy showed great potential to be used in the adsorption of compounds with different physicochemical properties.


Introduction
The growing urbanisation and industrialisation has called attention to the risks to the environment, mainly regarding the emission of dangerous organic and inorganic pollutants in the water bodies [1].Contamination of water systems by organic contaminants represents a serious threat to the environment.The seriousness of this threat is due to its toxicity to animals and humans [2].In recent years, environmental problems such as contamination of the water supply with sewage have been strongly considered [3,4].Among the organic pollutants that have been causing concern are medications.
Pharmaceuticals were developed to improve human health and cure diseases.However, due to the high consumption of medicines, they are often found in streams that receive industrial, agricultural and domestic effluents, which can cause adverse effects to humans [5].
Among the most consumed pharmaceuticals are non-steroidal anti-inflammatory drugs (NSAIDs), such as potassium diclofenac (PD).Although these drugs are found in low concentrations in water, continued exposure to them can cause harmful effects to humans.Another class widely used are antiparasitic drugs, such as moxidectin (MOX), which can be used in veterinary medicine and also as a pesticide in agriculture [6].MOX can be transported to surface waters mainly by means of animals.
Water is very important for the whole world.It is essential in the production of clothing, food and maintenance of the environment and our health.Therefore, the elimination of emerging pollutants from water is of great importance [7].Processes such as photodegradation [8][9][10][11][12][13][14][15] and adsorption [16][17][18][19] have been studied extensively to remove wide range of organic pollutants from the aqueous medium.Adsorption has been considered a very effective method for treating water.The adsorption process is a simple process, easy to operate, high efficiency and the ability to be adapted to be selective or nonselective in nature [20,21].In adsorption, the adsorbate present in a liquid solution binds to the solid adsorbent through different interactions that can be physical or chemical and based on a mass transfer process [22].The efficiency of an adsorption process depends of adsorbent type.Then, the search for the development of efficient, economical, reusable and easily synthesised materials, capable of removing organic pollutants from aqueous solutions, is crucial.
Different types of adsorbents have been explored to eliminate organic pollutant from the aqueous medium such as carbon xerogels [23];iron III oxide/graphene oxide/citrus peel-derived magnetic bio-nanocomposite composite [24];functionalised biochar [25] and mesoporous molecularly imprinted polymer [26].Among these numerous materials, polymeric adsorbents based on conductive polymers have attracted attention due to their mechanical rigidity, pore size distribution, high surface area and viable regeneration [27].Some studies of adsorption with materials based on conductive polymers have been used in organic pollutants removal [28][29][30][31][32][33], proving that these materials are efficient in removing organic pollutants from the aqueous medium.For example, a composite based on polyaniline, polypyrrole (PPy) and cigarette filters was prepared and applied in removal of phenylbutazone from aqueous solution archived maximum adsorption capacity of 100.22 mg g -1 [28].Other composite of polyaniline, polypyrrole (PPy) and multi-walled carbon nanotubes showed maximum adsorption capacity of 159.9 mg g -1 for phenylbutazone removal [29].A cellulose fibre decorated with PPy have been synthetised for removal of potassium diclofenac from aqueous medium.This adsorbent material showed be an efficient adsorbent reaching adsorptive capacity of 210.07 mg g -1 [30].Some composites based on polyaniline with cellulose fibre [31] and multi-walled carbon nanotubes [32] were studied for adsorption of meloxicam from aqueous media showing maximum adsorption capacity of 169.5 mg g -1 and 221.2 mg g −1 , respectively.A mesoporous magnetic PPy was also synthetised and firstly used as an adsorbent for theremoval pollutants from aqueous medium.Ibuprofen (acid), caffeine (neutral), and bupropion (basic) were adsorbed with maximum adsorption capacities of 53.67 mg g −1 , 16.74 mg g −1 , and 24.72 mg g −1 , respectively [33].
Among these materials, pristine PPy is one of the most widely used conducting polymers because it is easily synthesised, with low cost, excellent environmental stability,biocompatibility and due to the presence of active nitrogen groups in its polymeric chain [34].The interactions between PPy and organic pollutants can occur in several ways, including through the group -NH + , promoting electrostatic interaction and also through the ion exchange process [35].
In this work, the pristine PPy have beeneasily prepared to be used for the first time in the removal of PD and MOX from the aqueous samples.The use of composites material can turn synthesis processes more expensive, time-consuming and unviable to be implemented in real situations of pollutant removal.The use of pristine PPy makes the process simpler, faster, cheaper and easier to implement in real situations.In addition, pristine PPy can be used for compounds with different physicochemical properties because it can to promote different types of interaction with organic compounds.Therefore, the main objective of this study is to compare the adsorption capacity of the pristine PPy for two analytes that have different structures and physicochemical properties (PD: pKa 4 and MOX: pKa 2.81 and 12.55).Thus, from the results found, it will be possible to verify whether the proposed material will also prove to be promising for use in the removal of other organic contaminants.First, PPy was characterised by scanning electron microscopy (SEM), thermogravimetric analysis (TGA) and differential thermal gravimetry (DTG), infrared spectroscopy (FTIR), textural properties, wettability and point zero charge (pH PZC ).After characterisations, adsorption studies were carried out evaluating parameters such as pH, contact time/ kinetics, concentration/isotherm and temperature.In the evaluation of kinetics, the models of pseudo-first order (PFO), pseudo-second order (PSO), Elovich and intraparticle diffusion were studied.For the study of isotherms, the models applied were Langmuir, Freundlich, Sips, single-site Langmuir-Freundlich and dual-site Langmuir-Freundlich.Finally, the material was subjected to adsorption/desorption cycles to assess its ability to reuse.

Instrumentations and conditions
A Hitachi Analytical Table Top TM3000 microscope (Tarrytown, NY, USA) with voltage acceleration to 20 kV was used to obtain the SEM images.A thermobalance (2950 Thermal Analysis Instrument, TA Instrument, New Castle, DE, USA) with temperature range of 25-1000°C, heating rate of 10°C min −1 and under nitrogen flow (50 mL min −1 ) was used to perform the thermogavimetric analysis (TGA).A Fourier transform spectrometer (Shimadzu, IRAffinity-1, Kyoto, Japan), operating from 4000 to 400 cm −1 , using the conventional KBr chip method was used to obtain the infrared spectrum of the PPy.The textural properties were determined by the methods of Barret-Joyner-Halenda (BJH) and Brunauer-Emmett-Teller (BET), using the instrument Quantachrome Nova 1200e and pre-treatment of 100°C for 24 h.A Nikon D90 camera with a 50 mm lens was used to assess the wettability of the adsorbent material.For this, a small amount of PPy was placed under a Petri dish and then a small drop of water was applied to the surface of the material.The wettability was determined by means of the contact water (θ) formed between the surface of the PPy and the drop of water.The pH PZC of the adsorbent was determined as follows: aqueous solutions at pHs 2, 4, 6, 8 and 10 were prepared, adjusting with HCl or NaOH (both at 0.1 mol L −1 ).Thus, the volume of 5.0 mL of each solution was placed in contact with 25 mg of PPy.This mixture was stirred for 1 min at 3000 rpm and left to stand for 24 h.At the end of this period, the pHs were measured again.The determination of pH PZC is done by means of a graph of pH initial versus pH final .The point at which these pHs are close is called pH PZC [36].

Synthesis of PPy
For the synthesis of the PPy, 13.5 gof FeCl 3 .6H 2 O were solubilised in 150 mL of ultrapure water.Subsequently, 1.75 mL of pyrrole was added dropwise to the solution prepared above.This mixture was left under constant magnetic stirring for 3 h.After this time, the resulting solution was filtered under vacuum to give a black precipitate.The resulting material was washed with portions of ultrapure water and methanol (1: 1, v/v) and oven dried at 60°C for 24 h [37].The yield obtained from this synthesis was approximately 98%.
Figure S1 shows a PPy synthesis scheme that begins with the oxidation of the pyrrole monomer, leading to the formation of a radical cation (Step 1).Later, there is the coupling of two radical cations and through a charge transfer, two protons are eliminated forming a neutral dimer (Step 2).The dimer oxidises more easily than the pyrrole monomer and therefore is converted to a new radical-cation and thus, successively, chains of oligomers that form the PPy are generated (Step 3).

Adsorption experiments
The adsorption experiments were carried in 50 mL Falcon tubes, containing 25 mg of PPy and 5.0 mL of solution, with agitation at 250 rpm in a shaker.The studies were carried out separately for PD and MOX.In the evaluation of parameters such as pHand contact time/kinetics, solutions of PD and MOX were prepared in the concentration of 0.1 mg mL −1 .For the thermodynamic study, the concentration was 0.5 mg mL −1 .The evaluated pHs were 2, 4, 6, 8 and 10, with the solutions stirring for 1 h.The evaluation of the contact time was performed in 5-90 min, using solutions adjusted to the best pH for each analyte.Concentrations in the range of 0.025-1.5 mg mL −1 were used to determine the adsorption isotherms.The study of the effect of temperature was carried out at temperatures of 298, 308 and 338 K.After stirring, all solutions were filtered using Millipore Millex-GV® hydrophilic 0.45 μm PVDF filter and analysed by HPLC-DAD.All analyses were performed in triplicate (n = 3) and the averages were used to plot the graphs for each parameter evaluated.The amount adsorbed from each analyte can be calculated by Equation 1 [38].
C i and C f (mg L −1 ) correspond to the initial and final concentrations of the adsorvate (PD or MOX) in the aqueous solution, respectively; V (L) is the volume of solution and m (g) is the mass of PPy used.
In the kinetic study, the models evaluated were PFO, PSO, Elovich and intraparticle diffusion.The isotherm models studied were Langmuir, Freundlich, Sips, single-site Langmuir-Freundlich and dual-site Langmuir-Freundlich.The kinetic and isothermal data were adjusted using non-linear equations described by the literature.The best models are found through the evaluation of the best coefficients of determination (R 2 ) and the comparison between the q e values determined experimentally and through the equations.In addition, kinetic data provide the residual root mean square error (RMSE) which also assist in determining the best kinetic model [33].
In the thermodynamic study, the distribution coefficient (K d ) can be calculated by Equation 2. As the adsorption was carried out in an aqueous solution with very low concentration of the analytes, the K d value to be converted into dimensionless values can be multiplied by 1000 mg L −1 [21].

Reuse of the PPy
After the adsorption studies, the ability to reuse the material was evaluated.For this, PPy was subjected to seven cycles of adsorption-desorption and at each cycle the amount adsorbed from the analyte was calculated.It is important to mention that at the end of each adsorption, the material was washed with acetonitrile, followed by ultrapure water for its regeneration.At the end of the cycles, it was evaluated if there was a loss in the adsorption efficiency of the material and if it could be used several times, reducing the cost of analysis.

Characterisations
SEM was used to evaluate the surface morphology of the adsorbent material.Figure 1(a) and 2(b) show the SEM images of PPy at magnifications of 500× and 2000×, respectively.The images show particles connected together, forming an agglomerated structure with a granular surface.
It is important to mention that when the contact angle is θ > 90°, the material is considered hydrophobic; if the angle is θ < 90°, it is hydrophilic and θ ≤ 10°, the material is  considered super-hydrophilic [39,40].It was possible to notice in Figure 2(c) that the drop of water spread completely under the surface of the material, forming a thin liquid film, indicating that PPy has super-hydrophilic characteristics.This may be a favourable characteristic for the adsorption of the analytes, since the greater interaction of the aqueous matrix with the surface of the material may also provide a greater interaction of the analytes that are solubilised in the sample, providing better adsorption.
The thermal analysis of PPy was evaluated by TGA and it is illustrated in Figure 1(d).It was possible to observe two mass loss events: the first occurs up to 100°C, with about 10% loss of mass.It can be attributed to the loss of moisture or other volatile compounds that were not consumed during the synthesis of the material; the second event is attributed to the thermal degradation of the polymer chains.It occurs at 100-670°C, with total carbonisation of the material above this temperature.From the TGA, it was concluded that PPy has a high temperature of thermal degradation and favours the adsorption studies that are carried out at temperatures of 298-338 K.
The characterisation of PPy by FTIR is shown in Figure 1(e).It was possible to notice bands referring to C = C and C-N around 1525 and 1445 cm −1 , respectively, attributed to the asymmetric and symmetric stretching vibration of the pyrrole ring [41].The bands around 1301 and 1155 cm −1 can be attributed to the C-H vibration in the ring plane and the C-H vibration outside the plane shows its characteristic band at 883 cm −1 [42].
As given in Figure 1(f), PPy adsorption isotherm was of type -IV based on Brunauer's classification, which showed a mesopore structure of adsorbent, i.e. 2.0 nm < pore diameter < 50 nm, in which the pore size distribution, calculated from desorption data using the BJH model, suggested an average value of 5.80 nm.In addition, the pore volume 0.26 cm3 g −1 and BET surface area 27.22 m 2 g −1

Determination of pH PZC and effect of pH in adsorption
In determining the pH PZC of the adsorbent, it was possible to analyse the behaviour of the charges on the material surface, indicating the pH value at which the balance between positive and negative charges is zero.This value can be determined using a graph of pH initial vs pH final and pH PZC is the point at which initial pH and final pH are equal or close.Figure 2(a) shows that the pH PZC of PPy was 2.15.Above this pH PZC , the polymer surface will be negatively charged and below this pH PZC , the PPy surface will be positively charged [43].From these conclusions and the study of pH, it is possible to analyse how the interaction between the adsorbent surface and the analyte will occur.
Therefore, some studies on the effect of pH on the removal of organic pollutants can be found in the literature [44,45].The pH is a parameter that, in most cases, interferes with the adsorption efficiency.In these studies, aqueous solutions adjusted to pHs 2, 4, 6, 8 and 10 were analysed and the results are shown in Figure 2(b).PD (Figure S2(a)) is a NSAID with pKa 4 and at pH 2 it is in molecular form [46].This pH is close to pH PZC , that is, the surface of the PPy has no charge.In this case, PD and PPy may be interacting through weaker interactions, such as Van der Waals forces.At pH> 3, PD and PPy are negatively charged and considering the mechanism involving Coulomb forces (attraction between opposite charges), at pHs above pH PZC less PD adsorption would occur, however this does not happen.The adsorption remains constant throughout the studied range, with no significant difference between them.It can be concluded that the adsorption of PD in PPy is not influenced by pH.The results obtained can be attributed to different interactions between adsorbent and adsorbate, involving intermolecular dipole-dipole forces, iondipole, hydrogen bonds and π-π interactions [47].Figure S3(a) shows the interaction between the functional groups of the PD and PPy.Based on this principle, the pH chosen for the next analyzes involving the PD was 6.0, as it is close to the neutral pH of the aqueous solution.
MOX (Figure S2(b)) is a milbemycin with two pKa values (2.81 and 12.55) [46].It can be seen in Figure 2(b) that the best adsorptions occur in the pH range of 4-10, approximately 99% removal.In this range, PPy is negatively charged and MOX is in molecular form [46]. Concluding that the material and the analytes may be interacting through Van der Waals forces, hydrogen bonding or π-π interactions.Figure S3(b) shows the interaction between some functional groups involved in the adsorption process between MOX and PPy.As with PD adsorption, the aqueous solution of pH 6.0 was chosen for the next steps because presented adsorption around 100% and it is more close to the pH of the water.

Effect of contact time and adsorption kinetics
The contact time was evaluated to find the time necessary for the material and analyte to balance and the adsorption to remain constant.The evaluated times were 5-90 min and the results are illustrated in Figure 3.It can be noted that the amount adsorbed increased rapidly at first until reaching equilibrium with 20 min of contact for both analytes.In the initial stage, there are a large number of sites available for adsorption and after a while, the empty sites are more difficult to be occupied due to the repulsive forces between the solute molecules on the solid and bulk phases.Therefore, a higher removal rate is achieved at the beginning and after equilibration, adsorption becomes slower [17].The amounts adsorbed at equilibrium were q e = 18.76 mg g −1 and q e = 19.96mg g −1 , for PD and MOX, respectively.The experimental data obtained were submitted to four kinetic models: PFO, PSO, Elovich and intraparticle diffusion.These models are represented by non-linear equations described below [48,49]: where q e (mg g −1 ) and q t (mg g −1 ) are the amounts of PD and MOX adsorbed per unit mass of adsorbate at equilibrium and at time t, respectively; k 1 (min −1 ) is the PFO rate constant; k 2 (g mg −1 min −1 ) is the PSO rate constant;α (mg g −1 min −1 ) andβ (g mg −1 ) represent adsorption and desorption rates, respectively; k id (mg g −1 min −1/2 ) is the intraparticle diffusion constant and C represents the measurement of the surface thickness of the adsorbent.
The PFO and PSO models were used to adjust the experimental data and the Elovich and intraparticle diffusion models were used to better understand the mechanism in the adsorption process.The values obtained for each parameter are summarised in Table 1 and the graphs for each model are illustrated in Figure 4.The results in Table 1 confirmed the better fit of the kinetic results with the PSO model, showing higher values of R 2 and lower values of RMSE and indicating that the adsorption process is controlled by a 2 nd order equation.In addition, the q e values calculated according to the PSO model were closer to the experimental q e values, while a greater difference was found between q eexp and q ecal using the PFO model [50].Figure4(a) and 4(b) presents the adsorption kinetics using PFO and PSO, respectively.
In the Elovich and intraparticle diffusion models, Figure4(c) and 4(d), respectively,there are two stages in the adsorption process: the first stage consists of rapid adsorption at the beginning up to 20 min of contact, that is, the adsorption increases with increasing time and the second stage corresponds to equilibrium, in which the adsorption of the analytes remains constant with increasing time.The α values, in the first step, obtained from the Elovich plot were higher than the β values in the adsorption of PD and MOX, indicating that the adsorbed PD and MOX molecules onto PPy were sufficiently held and may not be easily desorbed [51].There was no adjustment of data in the second stage when Elovich Equation 5 was applied.In the intraparticle diffusion model, the higher k id values for PD in the first stage of the process, suggest that this molecule diffused more quickly than the MOX molecule in the initial process.In the second stage, k id was too small for PD and the equation did not fit for MOX, as there was an equilibrium stage [52].Figure 4(d) shows that the graph did not pass through the origin and showed high C values (Table 1), indicating that the diffusion process of the boundary layer was also part of the adsorption process [53].

Adsorption isotherm
An adsorption isotherm assesses the distribution of the adsorbate molecules between the liquid and solid phases at various concentrations.The experiments to determine the isotherms were carried out in aqueous solutions at pH 6 fortified with PD and MOX in the concentration range from 0.025 to 1.5 mg mL −1 , 20 min of contact time, 25 mg of PPy and temperature of 25°C.The results obtained experimentally showed that an increase in the concentration of the analytes leads to more interaction, indicating an equilibrium reaction leading to a more effective removal [54].The maximum amount adsorbed of PD and MOX was q e = 224.73mg g −1 and q e = 86.44 mg g −1 , respectively.The adsorption equilibrium data were applied to various isotherm models in order to predict the adsorption behaviour of PD and MOX in relation to PPy.In this study, the models applied were Langmuir, Freundlich, Sips, single-site Langmuir-Freundlich and dual-site Langmuir-Freundlich, represented by Equations 7, 8, 9, 10 and 11, respectively [33,55].
where Q (mg g −1 ) corresponds to the maximum adsorption capacity of PD and MOX per gram of PPy;C eq (mg L −1 ) is the equilibrium concentration; n is the surface heterogeneity factor: n > 1 shows favourable adsorption, indicating that the PPy adsorption capacity increases with increasing initial concentration;K L (L mg −1 ), K F (L g −1 ), K S (L mg −1 ), K LF (L mg −1 ) and K 1 + K 2 (L g −1 ) represent the affinity constants of the Langmuir, Freundlich, Sips, single-site Langmuir-Freundlich and dual-site Langmuir-Freundlich equations, respectively.
Langmuir's model assumes that adsorption occurs at equal adsorption sites located on the material surface, that is, all adsorption sites are energetically homogeneous.In this model, adsorption occurs in monolayers and the adsorbed molecules do not migrate to the unfilled places of the adsorbent, becoming localised [56,57].The Freundlich model characterises the surface of the material as heterogeneous and adsorption occurs in multilayers in the adsorbent [58].This model assumes that the binding depends on whether the adjacent locations are already occupied or not [59].
When the equation combines elements of Langmuir and Freundlich, the Sips, single-site and dual-site Langmuir-Freundlich models are formed.Sips and single-site Langmuir-Freundlich models involve the same theory, but are demonstrated with different equations.In low concentrations of adsorbate the equilibrium data are adjusted to the Freundlich model and when the concentration of adsorbate is high, they are reduced to the Langmuir equation with adsorption in monolayers [60].Therefore, these models are based on the theory that in the adsorbent there are active homogeneous and heterogeneous sites and adsorption occurs in one of these sites.Unlike Sips and single-site Langmuir-Freundlich, the dual-site Langmuir-Freundlich model assumes that adsorption can occur in both homogeneous and heterogeneous sites of the adsorbent [28].
The graphics referring to the models are shown in Figure5(a) and 5(b) for adsorption of PD and MOX, respectively and the parameters calculated for each one, together with the coefficient of determination (R 2 ), are listed in Table 2. Based on the values of R 2 , the model that best fitted the data for the two analytes was the dual-site Langmuir-Freundlich (R 2 > 0.99).In addition, the two analytes showed values of n 1 + n 2 > 1 (PD: n = 5.21 and MOX: n = 28.40),showing favourable adsorption.The maximum adsorption capacities were Q max = 221.23 mg g −1 for PD and Q max = 87.46mg g −1 for MOX.This values are close to the values determined experimentally: q e = 224.73mg g −1 and q e = 86.44 mg g −1 for PD and MOX, respectively.This information confirms the adequacy of this model to the experimental data, suggesting that the adsorption process unites Langmuir and Freundlich elements with adsorption on homogeneous and heterogeneous surfaces of the PPy.It is noteworthy that Q max in PD adsorption was about 2.5 times higher than Q max in MOX adsorption.This may have occurred because the PD is a smaller molecule than the MOX molecule, adhering more easily to the sites available on the material surface.

Thermodynamic study
The adsorption of PD and MOX in PPy was analysed at three different temperatures: 298, 308 and 338 K.The thermodynamic process was evaluated using the parameters of Gibbs free energy (ΔG), enthalpy (ΔH) and entropy (ΔS) that help to describe the energy changes involved in the adsorption process.These parameters can be found using Equations 12 and 13 [61].
where R is the ideal gas constant (8.314J mol −1 K −1 );T is the absolute temperature (K) and K d is the adsorption equilibrium constant.The ΔH and ΔS parameters can be calculated from the slope and intercept values obtained from linear fitting of Vant Hoff plot between ln K d vs 1/T.
Through Figure 5(c), it can be seen that the amounts adsorbed at 338 K were q e = 97.57mg g −1 and q e = 93.08 mg g −1 for PD and MOX, respectively.Figure 5(d) shows the good linear fit of the graph of ln K d vs 1/T with R 2 > 0.95 for the two analytes.According to Table 3, the ΔG values for both PD and MOX became more negative as the temperature increased, showing the spontaneity of the adsorption process.The slight decrease in ΔG values with increasing temperature indicates that the adsorption process was more favourable at higher temperatures [62].It is believed that when ΔG values are within the range of 0 to −20 kJ mol −1 , as in this study, there may be a physical adsorption process with weaker interactions between material and analyte, facilitating the desorption process [58].Table 3 shows ΔH = 36.45kJ mol −1 and ΔH = 23.82kJ mol −1 for PD and MOX, respectively.The positive values of ΔH indicate that the adsorption of the analytes in the PPy occurred via an endothermic process, absorbing heat.The entropies of PD and MOX in PPy were 131.65 J K −1 mol −1 and 83.33 J K −1 mol −1 , respectively, concluding that the randomness of the solid-liquid interface increased during the adsorption process [63].The values of ΔG more negative and ΔS greater for PD indicate that the adsorption of this analyte in PPy is more favourable than the adsorption of MOX, as expected.

Reuse
The reuse of the material is an important factor from the environmental point of view and also in relation to the cost-benefit of the process, planning possible commercial applications of the adsorbent on a large scale for water treatment [41].Acetonitrile was used as an elution solvent.After seven cycles of adsorption-desorption, it was found that the material maintained its adsorbent capacity.In the first cycle, the percentage adsorbed from PD was 98.86 ± 0.99% and from MOX was 99.05 ± 4.61%, while the percentage adsorbed in the seventh cycle was 99.92 ± 1.67% for PD and 99.01 ± 4.34 for MOX, as  shown in Figure 6.Therefore, PPy can be effectively reused for seven consecutive adsorption-desorption cycles, without losing its original capacity.

Comparison with other studies
Table 4 shows some adsorption studies that use diclofenac as a pollutant present in the aqueous medium [29,42,[64][65][66].No studies have been found to remove MOX present in aqueous solutions.Analysing the data, it was possible to observe that the developed study came into balance with the analytes present in the aqueous solution more quickly when compared to other studies, requiring only 20 min.The best kinetic model for most studies is PSO and the isothermal model that best fits the data varies with each work.Finally, one of the most important parameters is the maximum adsorption capacity of the material.Analysing Table 4, it can be concluded that the PPy presented high Q max values, demonstrating to be an excellent material to be used in the removal of organic compounds present in the medium aqueous.In addition, it is worth mentioning that this is a new work for removing MOX from aqueous solutions.

Conclusions
Thiswork clearly demonstrated that PPy is an excellent adsorbent in the removal of PD and MOX from the aqueous solution.As PD and MOX are analytes with different physicochemical properties, it is possible infer that this adsorbent can also be efficiently used to remove other organic pollutants.The high percentage of adsorption over a wide pH range (> 95%) is also a great advantage of this study, showing a good interaction between adsorbent and adsorbates.In addition, it also showed rapid adsorption kinetics, with equilibrium being achieved in only 20 min.Kinetic studies revealed that the PSO model fitted better to the experimental data.As for the study of isotherms, the experimental results showed that the Langmuir-Freundlich dual-site model showed better fit, presenting Q max (PD) = 221.23 mg g −1 and Q max (MOX) = 87.46mg g −1 .The greater adsorption capacity of the PD is due to its smaller chemical structure, so it can be more easily adsorbed on the adsorbent material.In the thermodynamic study, the negative ΔG° and positive ΔH° values revealed that the adsorption process was spontaneous and with endothermic nature.In addition, PPy showed sufficient potential to be reused several times without impairing its high adsorption capacity.This makes the process simpler, cheaper, faster, more efficient and more applicable in real situations for removing organic contaminants.From this information, it is possible to conclude that the material developed has great potential to be used in the removal of different organic compounds for the purpose of water purification.

Figure 2 .
Figure 2. (a) pHPZC; (b) study of pH on adsorption of PD and MOX.

Figure 3 .
Figure 3. Evaluation of contact time on adsorption capacity of PPy.

Figure 5 .
Figure 5. Adjusted isothermal models for (a) PD and (b) MOX; (c) Effect of temperature on the adsorption of PD and MOX by PPy; and (d) Van't Hoff regressions for thermodynamic parameters.

Table 1 .
Kinetic parameters obtained from the adsorption of PD and MOX by PPy.

Table 2 .
Isothermal parameters obtained from the adsorption of PD and MOX by PPy.

Table 3 .
Thermodynamic parameters obtained from the adsorption of PD and MOX by PPy.