Hydrogen dissolved within metals leads to embrittlement of the metal, and can lead to brittle and premature failure. While this embrittlement might occur due to hydrogen introduced during the production process, hydrogen also gets absorbed due to reactions of the metal with its environment. Specifically, due to reactions between electrolytes (such as water) and the metal, hydrogen ions attach to the metal surface and the metal corrodes, with these processes in turn altering the state of the electrolyte. Capturing this coupling is paramount to accurately predict hydrogen uptake and the resulting failure \cite{Hageman}. One common manner in which fracture processes can be included is the phase-field method. Instead of directly representing fractures and cracks within the geometry, the presence of fracture surfaces is indicated through an indicator function. While this allows for complex crack patterns to be represented, the coupling with other physical phenomena becomes non-trivial. The transport of ions within cracks is commonly included through an empirically or experimentally determined diffusivity \cite{Wu}, with this diffusivity having little to no relation to the actual crack geometry. Similarly, assumptions on the manner in which (electro-) chemical reactions are included limit the accuracy of the predicted hydrogen uptake and embrittlement. Here, a phase-field model will be presented which includes the diffusion and reactions within an electrolyte-filled crack in a consistent manner. By formulating the governing equations for a discrete fracture before transferring them to the phase-field framework, the dependence of the geometry and environment is retained. By comparison to alternative models and discrete fracture simulations, it is shown that this greatly enhances the accuracy of the predicted hydrogen uptake. Finally, application to hydrogen embrittlement is presented, highlighting the ability of capturing electro-chemical phenomena coupled to crack propagation.