Capture and indirect detection of inelastic dark matter

We compute the capture rate for dark matter in the Sun for models where the dominant interaction with nuclei is inelastic--the dark matter up-scatters to a nearby dark ''partner'' state with a small splitting of the order of 100 keV. Such models have been shown to be compatible with DAMA/LIBRA data, as well as data from all other direct detection experiments. The kinematics of inelastic dark matter ensures that the dominant contribution to capture occurs from scattering off of iron. We give a prediction for neutrino rates for current and future neutrino telescopes based on the results from current direct detection experiments. Current bounds from Super-Kamiokande and IceCube-22 significantly constrain these models, assuming annihilations are into two-body standard model final states, such as W+W-, tt, bb, or τ+τ-. Annihilations into first and second generation quarks and leptons are generally allowed, as are annihilations into new force carriers which decay dominantly into e+e-, μ+μ-, and π+π-.