Energy harvesting terminals are emerging as favorable alternatives to conventional terminals for wireless (sensor) networking due to environmental concerns and their potential to extend the network lifetime. Stochastic availability of energy for such networks calls for new network design insights on power control and scheduling, particularly in multi-user settings with interference. This paper considers an asymmetric interference channel with two transmitters and two receivers, and seeks optimal power policies to maximize sum capacity in an energy harvesting setting. It is shown that in the asymmetric interference case, the optimal sum capacity for the channel can be found by iteratively employing single-user optimizations, and the corresponding single-user problems are solved using modified water-filling algorithms such as directional water-filling and generalized water-filling. The performance of the proposed iterative algorithm is demonstrated through numerical results.