Electrochemical CO2 reduction (eCO2R) is expected to play a central role in achieving net-zero and negative-carbon technologies. Designing and optimizing electrochemical systems, such as CO2 electrolyzers, requires a multiscale approach that links the elementary reactions at the catalyst surface with the local chemical environment and the coupled transport processes, both in solid and liquid electrolytes, and within the often complex porous architecture of the catalyst and its supporting layers.

The first part of this thesis considers a multiscale model for eCO2R on a planar Ag electrode, coupling atomic-scale information with continuum transport via a microkinetic model that includes electrolyte effects across scales. The H-cell transport model is based on gmPNP equations, modified for Donnan potential, effective diffusion, and water self-diffusion in ionomers with TMA+ fixed charge. Reactivity is described by a microkinetic model incorporating cation transport from the outer Helmoltz plane to the reaction plane, surface microenvironment concentrations, and DFT-derived kinetic constants with partially hydrated cations. Applied to Ag electrodes in various bicarbonate electrolytes, the model reproduces current-voltage trends and FEs, showing that optimal performance requires balancing CO2 diffusion and cation accumulation. In ionomers, fixed organic cation microenvironments mitigate some transport limits, but water availability and transport remain critical. In MEA-type setups with saturated KHCO3, K+ is still found at the interface, contributing to microenvironment formation.

The following part of this work uses focused ion beam-scanning electron microscopy (FIB-SEM) tomography to reconstruct the 3D porous structure of materials with resolutions down to 4 nm. Chapter 2 details the workflow from sample preparation and embedding in epoxy, through image acquisition and segmentation, to morphological analysis, illustrated using a Bi2O3 pre-catalyst. An updated method, including Cu-based metal complexes in the epoxy, enables simultaneous imaging of an Ag-based catalyst layer (CL) and underlying microporous layer (MPL). Chapter 3 presents the 3D reconstruction and characterization of an Ag CL, along with 2D segmentation of both CL and MPL, extraction of morphological parameters, and pore-scale simulations to determine MPL transport properties. A coupled multi-physics pore-scale and homogenized gmPNP model evaluates the effect of wetting regimes on GDE performances. Chapter 4 applies direct pore-scale simulations (single physics) to three CLs, two sputtered Cu-based and one Ag nanorod-based, quantifying anisotropic transport properties and revealing discrepancies with empirical models. The results provide a robust morphological dataset for homogenized device modeling.

This work addresses key aspects of eCO2R through a multiscale approach. A multiscale model is developed to explain electrolyte effects beyond alkali-based electrolytes, informing the design of microenvironments. FIB-SEM tomography, combined with simulations, enables detailed pore-scale resolution of operando conditions and 3D morphological characterization of CLs. Directly extracting tensors of effective physical properties from real CL structures improves the accuracy of homogenized device-scale GDE models. Integrating atomic-scale kinetics, pore-scale morphology, and device-scale transport offers a robust framework for optimizing eCO2R devices, as discussed in the outlook.