We have investigated i) the formation of gravitationally bounded pairs of gas-giant planets (which we call binary planets) from capturing each other through planet-planet dynamical tide during their close encounters and ii) the following long-term or
bital evolution due to planet-planet and planet-star {it quasi-static} tides. For the initial evolution in phase i), we carried out N-body simulations of the systems consisting of three jupiter-mass planets taking into account the dynamical tide. The formation rate of the binary planets is as much as 10% of the systems that undergo orbital crossing and this fraction is almost independent of the initial stellarcentric semi-major axes of the planets, while ejection and merging rates sensitively depend on the semi-major axes. As a result of circularization by the planet-planet dynamical tide, typical binary separations are a few times the sum of the physical radii of the planets. After the orbital circularization, the evolution of the binary system is governed by long-term quasi-static tide. We analytically calculated the quasi-static tidal evolution in later phase ii). The binary planets first enter the spin-orbit synchronous state by the planet-planet tide. The planet-star tide removes angular momentum of the binary motion, eventually resulting in a collision between the planets. However, we found that the binary planets survive the tidal decay for main-sequence life time of solar-type stars (~10Gyrs), if the binary planets are beyond ~0.3 AU from the central stars. These results suggest that the binary planets can be detected by transit observations at >0.3AU.
