One of the most promising applications of X-ray Free Electron Lasers (XFELs) is the imaging of isolated particles, such as proteins, using single-particle X-ray diffractive imaging (SPI). This technique can provide high-resolution structural information on individual particles and facilitate the study of dynamic processes at the nanoscale. SPI, employing gas phase injection through an aerodynamic lens stack (ALS), has attracted significant attention due to its low background scattering and suitability for high-rate data collection. Despite these advantages, these SPI experiments encounter several challenges, especially with smaller and lighter biomolecule particles. These include low signal strength, limited collected datasets, high background scattering, and issues with sample compatibility in delivery system. In this doctoral thesis, I address the latter three challenges by developing and optimizing traditional electrospray-based gas phase sample delivery systems for SPI at XFELs. My research aims to enhance particle transmission efficiency, reduce background scattering, and expand the conductivity range of these systems to enable high-resolution imaging of smaller biological particles. I have developed three modified electrospray systems based on the traditional system to improve SPI at XFELs: enhanced electrospray, helium electrospray (He-ES), and coaxial helium electrospray (CHeES). The enhanced electrospray, upgraded from the traditional system by exploring different neutralizers and geometries, achieves an eightfold increase in particle transmission efficiency by employing a VUV neutralizer and optimizing the counter electrode's orifice size. This enhanced system achieves over 40% particle transmission from solution to the X-ray interaction region. The He-ES system uses a 3D-printed nozzle to reduce N2 and CO2 usage compared to traditional electrospray while ensuring stable sample delivery. It enhances particle delivery efficiency tenfold for 26 nm-sized biological particles and decreases gas load in the interaction chamber by 80%. Lastly, the CHeES system uses a coaxial 3D-printed nozzle to accommodate a broader conductivity range up to 40 000 µS/cm-eight times higher than traditional systems, and to lower background noise using He-ES technique. In tests at the European XFEL, the CHeES system notably lowered background noise by more than threefold in helium mode. My findings indicate improvements in transmission efficiency, background noise reduction, and sample versatility in SPI experiments, potentially enhancing both data quality and quantity. These advancements could yield higher-resolution structures and expand the scope for studying diverse biological and material science samples. My research has broader implications for structural biology, as obtaining higher-resolution structures is crucial for understanding the atomic structure of proteins and other biomolecules.