Revealing the structure of complex biological macromolecules, such as proteins, is an essential step for understanding the chemical mechanisms that determine the diversity of their functions. Synchrotron based X-ray crystallography and cryo-electron microscopy have made major contributions in determining thousands of protein structures even from micro-sized crystals. They suffer from some limitations that have not been overcome, such as radiation damage, the natural inability to crystallize a number of proteins, and experimental conditions for structure determination that are incompatible with the physiological environment. Today, the ultra-short and ultra-bright pulses of X-ray free-electron lasers have made attainable the dream to determine protein structures before radiation damage starts to destroy the samples. However, the signal-to-noise ratio remains a great challenge to obtain usable diffraction patterns from a single protein molecule. With the perspective to overcome these challenges, we describe here a new methodology that has the potential to overcome the signal-to-noise-ratio and protein crystallization limits. Using a multidisciplinary approach, we propose to create ordered, two dimensional protein arrays with defined orientation attached on a self-assembled-monolayer. We develop a literature-based flexible toolbox capable of assembling different kinds of proteins on a functionalized surface and consider using a graphene cover layer that will allow performing experiments with proteins in physiological conditions.