Optical lattice clocks offer the possibility to combine accuracy in the 10-18 range together with exquisite short stability, 10-16 for a measurement time of 1 second or even better . Clocks with such level of accuracy and stability largely outperform the existing primary frequency standard based on the laser-cooled atomic fountain geometry and on an atomic transition in the microwave domain. Optical clocks allow fundamental physics tests with unprecedented accuracy  and open the way to new applications such as Earth gravitation potential mapping . The ultimate limitation to the performance of optical lattice clock is still under investigation. Nonetheless, one limiting systematic shifts is clearly identified already: the blackbody radiation shift, the shift of the clock frequency due to the interaction of atoms with the ambient thermal electromagnetic background. At the temperature of 300 K and in fractional terms, this frequency shift is of the order of -5.5×10-15 for strontium (Sr) and -2.6×10-15 for ytterbium (Yb). Consequently, this effect must be controlled to much better than the percent level for an accuracy of 10-17, a highly challenging task. One motivation for considering mercury (Hg) is its low susceptibility to blackbody radiation . At 300 K, the corresponding fractional frequency shift is only -1.6×10-16, ∼16 times smaller than for Yb and ∼34 times smaller than for Sr. Hg is also interesting for its high sensitivity to a putative variation of the fine structure constant. Hg has 7 natural isotopes, 6 of them with abundance above 6%, 2 fermions and 5 bosons, which are all candidates for an optical lattice clock. Using Hg for an optical lattice clock remains however a significant technical challenge given that most of the laser wavelengths necessary to manipulate atoms and to probe the clock transition are in the deep ultraviolet range of the electromagnetic spectrum.