Machine learning for the adsorptive removal of ciprofloxacin using sugarcane bagasse as a low-cost biosorbent: comparison of analytic, mechanistic, and neural network modeling.

Contamination with traces of pharmaceutical compounds, such as ciprofloxacin, has prompted interest in their removal via low-cost, efficient biomass-based adsorption. In this study, classical models, a mechanistic model, and a neural network model were evaluated for predicting ciprofloxacin breakthrough curves in both laboratory- and pilot scales. For the laboratory-scale (d = 2.2 cm, Co = 5 mg/L, Q = 7 mL/min, T = 18 °C) and pilot-scale (D = 4.4 cm, Co = 5 mg/L, Q = 28 mL/min, T = 18 °C) setups, the experimental adsorption capacities were 2.19 and 2.53 mg/g, respectively. The mechanistic model reproduced the breakthrough data with high accuracy on both scales (R2 > 0.4 and X2 < 0.15), and its fit was higher than conventional analytical models, namely the Clark, Modified Dose-Response, and Bohart-Adams models. The neural network model showed the highest level of agreement between predicted and experimental data with values of R2 = 0.993, X2 = 0.0032 (pilot-scale) and R2 = 0.986, X2 = 0.0022 (laboratory-scale). This study demonstrates that machine learning algorithms exhibit great potential for predicting the liquid adsorption of emerging pollutants in fixed bed.

Mathematical modeling and numerical simulation of sulfamethoxazole adsorption onto sugarcane bagasse in a fixed-bed column.

Having rigorous mathematical models is essential for the design and scaling of adsorption columns. In this study, the dynamic behavior of the sulfamethoxazole adsorption on sugarcane bagasse was studied and compared using analytical models and a theoretical mechanistic model. Initially, fixed-bed column tests were carried out at different flow rates and bed heights. Then, the experimental data were fitted with the most widely used analytical kinetic models, and their fit and fixed-bed parameters were compared with the mechanistic model. Of all analytical models analyzed, the Log-Gompertz model was the one that had the best agreed with experimental data. Although some analytical models fitted the experimental data accurately, their usefulness was questionable. Their parameters did not show a clear relationship with the change in operating conditions, and in certain cases had different behavior from that observed in experimentation. Conversely, the mechanistic model not only predicted the breakthrough curves with great accuracy in the initial and transition stage (R2 > 0.92; SSE < 0.06), but it also estimated relevant parameters. Additionally, the effects of the global mass transfer coefficient (Ki) and the axial dispersion coefficient (Dz) on breakthrough curves were studied using the mechanistic model. Increasing Ki increased the slope of the breakthrough curves with a faster adsorption rate. Similarly, high values of Dz produced lower adsorption capacities of the adsorbent; and it was established that the axial dispersion is relevant in SMX adsorption on SB. The theoretical model presented can be used for the design, scaling, and optimization of adsorption columns.

Comments on "Dynamic Adsorption of Sulfamethoxazole from Aqueous Solution by Lignite Activated Coke".

This paper presents a brief discussion with regard to the fixed-bed modeling results of a recent paper by Li et al. published in this journal.

Comments on "Treatment of malachite green dye containing solution using bio-degradable sodium alginate/NaOH treated activated sugarcane baggsse charcoal beads: Batch, optimization using response surface methodology and continuous fixed bed column study" and "Adsorption and oxidation of ciprofloxacin in a fixed bed column using activated sludge derived activated carbon".

This comment analyses the results of the fixed-bed column modeling from the studies published by Gupta and Garg (2019), and Das et al. (2020), which are the last two papers on this topic published in this journal.