High-performance removal of anti-inflammatory using activated carbon from water treatment plant sludge: fixed-bed and batch studies

Activated carbon from water treatment sludge (WASC) was employed as adsorbent material to remove the 20 anti-inflammatory Nimesulide (NM) from aqueous solutions. NM adsorption was performed in batch and 21 fixed-bed systems, evaluating pH, adsorbent dosage, adsorption kinetics, equilibrium isotherm, continuous 22 adsorption, and simulated effluents. The kinetic data were best fitted to the Elovich model and Intraparticle 23 diffusion reaching the equilibrium at 120 min. Langmuir model presented a better description of the 24 equilibrium data with the maximum adsorption capacity ( q max ) of 274.99 mg g -1 from NM adsorption by 25 WASC. The adsorbent was tested in two simulated hospital effluents and proved to be an excellent 26 adsorbent for removing NM from an aqueous solution with the presence of salts, sugars, and other 27 inorganics. Finally, WASC was applied in fixed-bed NM adsorption obtaining the adsorption capacity of 28 217.28 mg g -1 .

in medicine (Lima et al. 2013), which is a result of its great efficacy compared to similar drugs such as 48 ibuprofen, diclofenac, and piroxicam (Singh et al. 2001). The presence of NM in high concentrations was 49 reported by a study carried out in a sewage treatment plant in Greece (Papageorgiou et al. 2016

52
The adsorption process has been widely applied in the removal of emerging contaminants from 53 effluents due to its several advantages (Streit et al. 2020), such as efficiency, flexibility, low cost, simplicity

66
Thus, the objective of this work was to apply activated carbon from water treatment plant sludge 67 in the adsorption of anti-inflammatory NM. The batch analysis was done by evaluating initial pH, adsorbent 68 dosage, adsorption kinetics, and isotherm. The study also emphasized the treatment of simulated hospital 69 wastewater and the fixed-bed adsorption of NM by WASC, as a proposal for industrial application.

84
The reactive groups present on the surface of WASC before and after the NM adsorption were 85 obtained by the Fourier Transform Infrared Spectroscopy (Shimadzu IR Prestige-21, Japan) technique. The 86 spectrum was performed in the range of 500-4500 cm −1 with a resolution of 4 cm −1 .    2² Factorial Design (FD) was applied at a 95% confidence level to understand which factor 101 (adsorbent dosage (Ad (g L -1 )) and initial pH) would have a major impact on adsorption processes and how 102 these parameters would interact between them. This method was also used to develop a mathematical model 103 to describe the adsorption process.

108
The FD was carried out with independent variables at 3 levels and it was arbitrated 3 repetitions   Four kinetic models were tested to describe experimental data: the pseudo-first order, pseudo-113 second order, intraparticle diffusion, and Elovich kinetic models were adjusted to experimental data. These

115
(1 + 2 ) where t is the time of contact (min); qt is the amount of NM adsorbed at time t (mg g -1 ); k1 is the pseudo-116 first order rate constant (min -1 ); k2 is the pseudo-second order rate constant (g mg -1 min -1 ); kid is the rate 117 constant for intraparticle diffusion (g mg -1 min -1/2 ); C is related to diffusion resistance (mg g -1 ); α is the 118 initial rate of Elovich model (mg g -1 min -1 ); and β is the Elovich model constants (mg g -1 ).

119
Langmuir and Freundlich isotherm models were applied to analyze the equilibrium experimental 120 data. These models are represented in Equations 7 and 8, respectively.

124
All model parameters were defined by nonlinear regression, using the quasi-Newton method. The 125 adequacy of the models was analyzed by variance analysis (ANOVA) checked by Fvalue > Ftabled, were Fvalue 126 is the regression coefficient. Furthermore, the models were checked by using the Chi-square (X²) and  In this study, two simulated hospital effluents were produced to evaluate NM adsorption by WASC 134 in the middle of sugars, high salt, urea, and other inorganics commonly found in hospital waste effluents 135 (Saucier et al. 2015). Table 2 shows the composition of the simulated effluents.  The fixed-bed tests were performed on glass columns using 1 g of WASC. Two columns were 141 used, the internal diameter of column 1 is 5 mm and the bed height 50.4 mm, column 2 has an internal 142 5 diameter of 9 mm and 28 mm of bed height. The NM solution (200 mg L -1 ) was used to feed the bed with 143 a flow rate of 10 mL min -1 . Thomas and Yoon-Nelson models were fitted to the experimental fixed-bed 144 data for an estimate of the kinetic parameters, they are represented in Equations 11 and 12, respectively.

145
were kth is the Thomas rate constant (mL mg -1 min -1 ); qmax is the maximum adsorption capacity of Thomas 146 model (mg g -1 ); Q is the flow rate (mL min -1 ); kyn is the Yoon-Nelson rate constant (min -1 ); and τ ℎ is the 147 time required for 50% solute breakthrough (min).   Analyzing the results plotted in pareto charts it can be inferred that the pH is the most significant 168 effect on the adsorption capacity and efficiency of NM. Fig. 1(a) indicates that all parameters were 169 significant for adsorption capacity (p>0.05). The pH was the parameter with the most pronounced negative

173
From the adsorption efficiency presenting in Fig. 1(b), the more pronounced effect was the pH   Fig. 1(a-b) and 182 Table 3 indicates that the best pH condition for a higher was at initial pH 8. The higher the pH, the

186
The kinetic behavior of NM adsorption by WASC and the adjustments of the pseudo-first order, 187 pseudo-second order, Elovich, and intraparticle diffusion mathematical models are presented in Fig. 2.

202
The kinetic parameter values obtained from the fitting of the kinetic models to experimental data 203 are presented in Table 4.

223
The equilibrium experimental data of NM adsorption onto WASC and the adjustment of the 224 isotherm models are presented in Fig. 3 and their parameters in Table 5.  The isotherm curves (Fig. 3)

236
indicate that WASC is capable to adsorb NM anti-inflammatory.

237
The Langmuir isotherm is the best model for describing the experimental data, considering the

268
Such occurrence is in agreement with the kinetic curves presented in Fig. 2, where it was observed that the 269 adsorption of NM does not occur in an instantaneous way, such behavior in fixed-bed confirms that the 270 adsorption of NM is mainly controlled by chemisorption (Patel 2020). This means that greater bed heights 271 will allow for better performance because the unused fraction of WASC is reduced. WASC saturation time 272 was longer for column 1 than for column 2, this is also related to the greater use of the adsorbent bed and 273 consequently greater adsorption capacity.

274
To describe the dynamic behavior of the column, the Thomas and Yoon-Nelson models (Equations 275 11 and 12) were fitted to the experimental data.

9
Based on the results of this work, it is possible to infer that the NM-WASC interaction is mainly 299 controlled by intraparticle diffusion and chemisorption, but also presents attraction forces of H bonding.

318
Availability of data and materials

319
All data generated or analyzed during this study are included in this published article and its supplementary 320 information files.

322
The authors declare that they have no competing interests.