This study aimed to compare two adsorption modes for tetracycline (TC) removal on granular activated carbon impregnated with copper sulfate (AC-Cu), solid modification proposed by Costa et al. (2020), these being batch, through modeling the equilibrium isotherm with applying minimum square method and Bayesian statistics, and continuous, through the breakthrough curve analysis also using Bayesian statistics. For the Bayesian statistics, the methodology used in the present work follows that proposed by Oliveira et al. (2023a). The goal was to assess the potential scalability of the process. This work showed that using a more reliable mathematical approach can change the values of the isotherm parameters and indicated that the ideal operating conditions, for continuous adsorption of TC, were C_0TC = 80 mg·L^–1, Q = 6 mL·min^–1, and Z = 1 cm, resulting in a 95% TC removal with an adsorption capacity of the bed equal to 6.87 mg_TC·g_AC-Cu^–1, around three times higher than that of the unmodified AC (2.08 mg_TC·g_AC^–1). The saturation time of 107 min shows an efficient process, making it viable for implementation on a larger scale. For the elucidation of the mechanisms, the Clark model proves to best represent the process and the Log-Gompertz model is practically favorable for industrial implementation since its parameters provide valuable information about the effectiveness of the process depending on the configuration of the system. An underestimation of the maximum adsorption capacity of the bed was observed by leaving the parameter as a random variable, by approximately 46.72% and 58.95% for the Thomas and Yan models, respectively, leading to an erroneous application of the process on a larger scale. Additionally, the Langmuir model overestimated the value of q_max by approximately 21.63%. Finally, the study highlights the importance of operating conditions and precise parameter estimation in optimizing the adsorption process and that the adsorption mechanisms change depending on the type of operation, in this case for the batch mode, the Langmuir isotherm, and for the continuous mode, the Clark model. These findings provide valuable insights for the development and implementation of environmentally sustainable and effective water treatment strategies.

中文翻译：

---

通过对改性活性炭的批量和连续吸附优化四环素的去除：贝叶斯视角和高级建模分析


本研究旨在比较 Costa 等人 （2020） 提出的固体改性硫酸铜 （AC-Cu） 浸渍颗粒活性炭上去除四环素 （TC） 的两种吸附模式，它们是批量的，通过应用最小平方法和贝叶斯统计对平衡等温线进行建模，以及连续的，通过突破曲线分析也使用贝叶斯统计。对于贝叶斯统计，本研究中使用的方法遵循 Oliveira 等人 （2023a） 提出的方法。目标是评估流程的潜在可扩展性。本研究表明，使用更可靠的数学方法可以改变等温线参数的值，并表明连续吸附 TC 的理想操作条件为 C_0TC = 80 mg·L^–1，Q = 6 mL·min^–1，Z = 1 cm，从而去除了 95% 的 TC，柱床的吸附容量等于 6.87 mg_TC·g_AC-Cu^–1，大约是未改性 AC （2.08 mg_TC·g_AC^–1） 的三倍。107 min 的饱和时间显示了一个高效的过程，使其可以进行更大规模的实施。为了阐明机制，Clark 模型被证明最能代表过程，而 Log-Gompertz 模型实际上有利于工业实施，因为它的参数提供了有关过程有效性的宝贵信息，具体取决于系统的配置。通过将参数保留为随机变量，观察到对床的最大吸附容量的低估，约为 46.72% 和 58。Thomas 和 Yan 模型分别为 95%，导致该过程在更大规模上的错误应用。此外，Langmuir 模型高估了 q_max 的值约 21.63%。最后，该研究强调了操作条件和精确参数估计在优化吸附过程中的重要性，并且吸附机制根据操作类型而变化，在本例中为间歇模式的 Langmuir 等温线，对于连续模式为 Clark 模型。这些发现为制定和实施环境可持续和有效的水处理策略提供了宝贵的见解。