Today, one of the major goals of modern astronomy is the search for other habitable worlds and the presence of life on them. Crucial in this search is the atmospheric characterisation of small, rocky planets orbiting in the habitable zone around solar type stars. The LIFE initiative will be able to perform atmospheric characterisation of a sizeable subset of these planets in the mid-infrared (mid-IR) wavelength regime (5-20 m). The mid-IR is an important bandwidth as it contains some important atmospheric biosignatures. Extremely sensitive and highly efficient detectors are required to detect the faint signal from these small exoplanets. Current state-of-the-art detectors based on semiconductor technology are unable to meet these requirements. Microwave Kinetic Inductance Detectors (MKIDs) are superconducting pair-breaking detectors able of single-photon detection with no readout noise or
dark current. This makes MKIDs a promising candidate for mid-IR detectors for the LIFE initiative. In this thesis we investigate what development is necessary to meet the detector requirements set by the LIFE initiative. We also investigate how the performance of MKIDs can be reliably measured in the mid-IR.

Currently, there are no single-photon counting MKIDs designed for the mid-IR. Measurements are done with two MKID devices that originally have been designed for the near- and far-IR bandwidths. Prior to this work the near-IR detector has shown single-photon counting 1545 nm and the far-IR detector at 38 m. In this work we show the single-photon counting ability of MKIDs 3.8 and 8.5 m. This is the first time that single-photon counting has been shown at 8.5 m. The resolving power (8/?8) at 8.5 m is found to be about 4. Experiments are planned at 18.5 m for which a setup has been designed with a cryogenic black-body radiator as the source. This is the longest wavelength required for the LIFE spectrometer. We also perform an optimisation of the near-IR detector geometry to see if a realistic device can be made that is sufficiently sensitive to 18.5 m radiation. The results show that a realistic design could in theory be made but this strongly depends on how the detector is limited by the noise.

Next steps are to design a dedicated MKID for the mid-IR to determine its efficiency and dark current. This will also require us to improve the current measurement setup as measurements show that we suffer from thermal background radiation which limits the detector performance.