In this paper, we present a theoretical study of a band engineered detector design, which significantly improves the performance of mercury cadmium telluride (HgCdTe)-based unipolar n-type/barrier/n-type (nBn) infrared (IR) detectors for the midwave IR and longwave IR spectral bands. This band engineered nBn detector is based on the assumption that the valence band offset, which is normally present between the barrier and n-type absorber layer, can be eliminated using a bandgap engineering approach. The valence offset is the key issue that currently limits the performance of HgCdTe-based nBn detectors. Eliminating the valence band offset allows the nBn detectors to operate at |V Bias | < 50 mV, thus rendering insignificant all tunneling-related dark current components and allowing the detector to achieve the maximum possible diffusion current limited performance. The developed model allows the device performance to be optimized by an appropriate design of the conduction band barrier to block the flow of majority carrier electrons, while allowing minority carrier holes photogenerated in the absorber layer to reach the contact layer unimpeded. Furthermore, because of the absence of tunneling-related dark currents, it is shown that band engineered nBn detector architecture exhibits a better performance at maximum allowed absorber layer doping density compared with conventional nBn detector.Index Terms-Device simulation, mercury cadmium telluride (HgCdTe), n-type/barrier/n-type (nBn) infrared (IR) detector, unipolar barrier detector.