As of early $\sim$2010's, the Japanese SPace Infrared telescope for Cosmology and Astrophysics (SPICA) space observatory will be launched. This actively cooled, cryogenic (4.5K), 3.5m telescope will operate in the mid and far infrared spectral regions. With its very high sensitivity, one of SPICA's aims will be the direct detection and characterization of extra-solar outer planets of nearby stars. The goal contrast ranges from $10^5$ to $10^6$ up to an angular separation of ${\sim}5$ arcsec. The relatively low angular resolution at MIR (5 to 20 $\mu$m) requires an efficient and robust coronagraphic mode working at cryogenic temperatures. In this presentation we describe several envisaged preliminary designs and assess their performance against the science goals and host telescope specifications. These are compared against numerical simulations and instrumental environment considerations, such as the need for an actively corrected wavefront.
