This work demonstrates the implementation of thioglycolic acid (TGA) functionalized water-soluble NH2-molybdenum disulfide (MoS2) nanosheets as sensitive layers for toxic heavy metal ion sensing in drinking water. The sensor device was tested for several ions commonly found in drinking water, e.g., As (III), Pd (II), Hg (II), and Cd (II). When the sensor was exposed to individual ions, it showed negligible response toward elements such as As (III), Pd (II), and other commonly present metal ions. However, at 25 ppb, some cross sensitivity was observed between Cd (II) and Hg (II) ions. The mathematical analysis was performed to investigate the cross-sensitive behavior, and a model was put forward. It was evident from the higher attachment factor (N) for Cd concerning Hg (II), derived from the model which supported Cd's higher response when exposed individually (N-Cd = 4 x 10(16) cm(3)/ppb and N-Hg = 1.6 x 10(14) cm(3)/ppb). Furthermore, the band structure and electron concentration were plotted for individual Cd (II) and Hg (II) ions. Compared to Hg (II), Cd (II) efficiently depleted the amine-MoS2 and TGA junction. To better understand selectivity and response, the device was exposed to Cd (II) and Hg (II) ions simultaneously at different concentrations, which looked into the effect of the two interacting ions. A mathematical model was hence established. It was found that the absorption coefficient (K-Hg) was much higher (63 ppb(-1)) than Cd (K-Cd ) (0.38 ppb(-1)). The interaction factor (KHg-Cd), i.e., the absorption coefficient of Cd in the presence of Hg, was calculated as -0.086 ppb(-2), which is lower than KCd-Hg(-0.013 ppb(-2)), which indicated that Hg is more active toward adsorption in competition to the Cd ions. This competitive dominance for Hg (II) adsorption may be attributed to the affinity of thiol groups present in the TGA molecules for Hg (II) over Cd (II) ions. The device was repeatable after testing with five different devices with a standard deviation of Of 0.5322 and stable up to 50 days with +/- 3% response variation. Thus, the proposed sensor may be a suitable candidate for low-cost water quality monitors.