The use of evaporative cooling in the gas diffusion layer (GDL) of polymer electrolyte fuel cells has been proposed as a simple yet impactful solution to simultaneously provide the functionality of cell cooling, thanks to the dissipation of the latent heat of the fluid, and of humidification of gases via water vaporization. However, the underlying multiphase heat and mass transport processes in the GDL porous matrix are still elusive, which affects the cooling performance of the process and leads to a lack of reliable design tools for the structure and morphology of the GDL. In this thesis, I performed pore-level analyses of heat and mass transport in GDLs upon development of novel modeling tools, according to three sequential steps: (i) effective heat conduction through a real GDL geometry; (ii) heat and mass transport through an artificial lattice with a static liquid-gas interface; (iii) mass transport of water adopting an interface-resolving method. First, pore-level thermal conduction simulations of commercial GDLs were conducted, utilizing the exact GDL structures obtained by X-ray tomography. Correlations for effective thermal conductivity were extracted as a function of saturation. An energy sink term, accounting for the evaporative cooling, was added, providing a quantification of the conductive heat transfer in GDLs with evaporative cooling. The compression of the GDL and the water distribution in the GDLs was found to be a key factor for thermal conduction and evaporative cooling ability. Second, the heat conduction model was augmented by enabling convective transport within the gaseous phases, mass transfer, and local evaporation at the water-gas interface. The GDL structure was replaced by an artificial lattice. A three-dimensional model was developed, solving the Navier-Stokes equations, species transport and energy conservation equations in the gas domain, and heat conduction in the solid fibers and water phase. Evaporation at the liquid-gas interface was modeled using kinetic gas theory. It was demonstrated that porosity and fiber size had a small effect on evaporative cooling. On the other hand, operating conditions such as carrier gas velocity and type led to large changes in the evaporation rates as they increased the removal of vapor through convection and diffusion, hence increasing the evaporation rate and cooling effect. Larger operating temperatures led to larger saturation pressures, thus driving an increase of the evaporation rate. The artificial lattice structure was then modified to investigate a novel mixed hydrophilic/hydrophobic coated GDL. The mixed coating allows for separating liquid and gas transport paths in the anode. The results indicated that patterned GDLs provide increased evaporation capabilities compared to traditional, non-patterned GDLs, as the water-gas interface is nearer the gas channel, leading to removal of evaporated vapor and enhanced evaporative cooling. Third, X-ray tomography was used to reproduce an artificial lattice, and the water distribution and meniscus shape were predicted using a Volume Of Fluid approach, comparing the obtained interface from different initial water distributions, representing different water injection processes into the patterned wettability GDL. This thesis showed the impact of water distribution on conductive heat transport, and the importance of removal of water vapor at the water-gas interface to enhance evaporative cooling.