The increasing energy demand in our society has stimulated intensive research in the development of sustainable and renewable energy sources to lessen our strong dependence on fossil fuels. Hydrogen is a clean, storable, and high-energy density energy carrier, and is a promising sustainable solution to achieve an environmentally friendly fuel economy. Electrochemical and solar-driven photoelectrochemical water splitting is regarded as one of the most promising approaches to utilize renewable energy to product hydrogen fuel, yet Pt remains the best electrocatalyst for hydrogen evolution reaction (HER), the high cost of which ultimately limit the scalability of such technologies. Layered transition metal dichalcogenides (TMDCs) is a family of compounds that has attracted widespread attention due to their broad range of applications in electronics, optoelectronics, sensing, energy storage, and catalysis. My research has primarily focused on understanding the chemistry of MoS2 and related compounds, and developing rational approaches to enable these materials for efficient electrocatalytic and photoelectrochemical (PEC) hydrogen evolution. We demonstrated highly efficient and robust photocathodes based on heterostructures of chemically exfoliated metallic 1T-MoS2 and planar p-type Si for PEC-HER. Photocurrents up to 17.6 mA/cm2 at 0 V vs reversible hydrogen electrode (RHE) were achieved under simulated 1 sun irradiation, and excellent stability was demonstrated over long-term operation. Building upon the 1T-MoS2 groundwork, amorphous ternary compounds MoQxCly (Q = S, Se) were then developed as excellent catalysts for HER. The preparation of MoQxCly requires much lower temperature and easier fabrication, yet the PEC performance of MoSxCly-based photocathode is even better than 1T-MoS2-based photocathode. Moreover, when MoSxCly is incorporated with n+pp+ Si micropyramids (MPs), we demonstrate the highest current density ever reported for Si-based photocathodes. Furthermore, to fully harness the potentials of MoS2 and utilize it for a broader range of applications, we demonstrate covalent functionalization on the basal plane of 2H-MoS2 via thiol conjugation, despite the general belief that the basal plane is too inert for functionalization. We correlate the degree of functionalization to the amount of sulfur vacancies on MoS2 basal plane, and successfully demonstrated the preparation of MoS2-PbSe quantum dot heterostructures using a bi-functional dithiol linker molecule.