Comprehensive experimental study on microbial desalination cells for biological wastewater recovery, seawater brine treatment, and simultaneous power generation: applications of machine learning and thermal image processing

Microbial Desalination Cells (MDCs) have emerged as a groundbreaking solution for addressing water scarcity, wastewater treatment, and sustainable energy production simultaneously. In this study, we present an advanced experimental and analytical approach to optimize MDC performance by integrating machine learning predictions, thermal imaging analysis, and electrochemical monitoring under diverse environmental conditions. We conducted five distinct tests using wastewater from urban treatment plants and saline water from the Caspian Sea and the Persian Gulf to examine the interplay of key operational parameters, including salinity, COD, pH, TDS, internal resistance, and polarization behavior. The results reveal several significant insights. Higher initial salinity substantially enhances desalination efficiency, achieving a record salt removal rate of 96 %. We identified an inverse relationship between COD removal and desalination efficiency, indicating a trade-off between organic matter degradation and salt removal. Oxygen availability proved to be a critical determinant of MDC performance; its presence increased voltage significantly, peaking at 1099 mV. Power generation reached its maximum at an optimal current density, producing a peak power density of 0.143 mW/cm².Thermal imaging analysis uncovered a direct correlation between heat distribution, ion migration, and microbial activity, offering valuable insights into system efficiency and energy losses. The integration of machine learning models yielded highly accurate predictions, closely matching experimental data and providing a scalable pathway for MDC performance optimization. Collectively, these findings establish MDCs as a transformative technology for renewable water and energy solutions with strong potential for real-world applications.