<p>Metastasis, a secondary tumor, behaves distinctly different from its state-of-origin. Aggressive metastasis comprises a diverse heterogeneous population; demonstrates varying molecular mechanisms. Hence, critical to investigate the heterogeneity of a metastatic cancer cell at a single-cell level. So far, available detection methods depend on bulk tumor samples lose the heterogeneity inherited in metastatic cancers, making them incapable of detection. Consequently, there exists no technique to detect metastatic cancers. Surface-enhanced Raman scattering (SERS), a label-free bioanalytical technique, can obtain the spectral fingerprints of biomolecules from cells. However, conventional SERS focuses on plasmonic metal structures depending on hot spots requires an expensive methodology and difficult procedures that result in low reproducibility and stability. As a new alternative, semiconductors recently evolved with good stability, reproducibility, and biocompatibility for SERS applications. However, limited due to low detection and enhancement efficiency. This thesis introduced a new phenomenon of defect-induced functionalization in semiconductor materials to transform non-SERS materials to SERS active quantum/nano-sized probes. The unique functionalization activates the probes by tuning the concentration of surface and subsurface defects for ultrasensitive molecule detection. Different types of defects such as oxygen vacancy, interstitial defects, and dopants impart plasmonic resonance in semiconductor probes in addition to the charge-transfer mechanism. Thus, we first explored a collective behavior of intrinsic defects in probes as a sensing platform for molecular-level detection. Incorporating defects gives semiconductors other distinct properties such as multiple wavelength activity, non-degradable SERS activity, and detection of low-cross section cancer biomolecules. In addition, surface defects impart anionic property, increase biocompatible nature for safe adherence and penetration, permitting intracellular readout of biomarker signals. The diagnostic transformation signals accurately pointed to metastatic cancer and differentiated them from other cancer and normal cells. This study then focused on the primary cause of the development of metastatic cancer called cancer stem cells. Very scarce in number, we then focused on developing the sensing platform to capture these populations and magnify the trace cues present. This study presents underlying key targets that lead to the development of most cancer-related deaths and enhances the chance to increase the cancer prognosis improving clinical outcomes.</p>