The Phoenix Mars Lander includes a Microscopy, Electrochemistry and Conductivity Analyser (MECA) instrument for the study of dust and regolith at the Martian arctic. The microscopy payload comprises an AFM and Optical Microscope (OM) to which samples are delivered by a robot arm. The setup allows imaging of individual dust and soil particles at a higher spatial resolution than any other in-situ instrument. A fully functioning test-bed of the flight microscopy setup within an environmental chamber to simulate Mars conditions was assembled at Imperial College, enabling characterization of the microscopes. Samples are collected on small disks rotated to the vertical position for imaging, with each substrate surface promoting different adhesion mechanisms. The vertical mounting necessitates good adhesion of particles to substrates. Moreover, to achieve safe operation and good AFM scans, a sparse field of particles is required. This work investigates models and experimental setups which consider the adhesion mechanisms of particles, including under Mars conditions. These models incorporate the forces from the AFM cantilever during scanning, particle-substrate adhesion and particle-tip adhesion. The solution offered to the problem of unstable particles is substrates with engineered features, micromachined in silicon, to trap and stabilise particles for AFM and reduce the loading of the sample to a suitable level. Various designs were investigated in a series of tests, and a final design was created for a substrate for AFM during the mission. The substrates were fabricated and incorporated on the sample wheel on Phoenix, now on Mars. The MECA results are discussed, focusing in particular on the characterization, calibration and cataloguing of samples using the Imperial College testbed. The best ways of obtaining data from the setup were investigated. These strategies were used during the Phoenix mission. Finally, the extant microscopy data acquired during surface operations are presented and the overall operations procedures discussed.