Batteries and fuel cells are probably out because of the long duration of an Earth-Mars mission.
Here are some mass and power output numbers for radioisotope thermoelectric generators (RTGs); the most mass-efficient of these run about 5W/kg, so you're looking at about a ton of RTG to run your thruster full up. RTGs are attractive for outer planetary missions because solar energy falls off with the square of distance, but we're only going to Mars, so it's likely that solar is a better option.
According to Wikipedia, solar arrays can produce as much as 300W/kg at Earth distance. It's not clear from that article how practical cells are at that power to mass ratio, or how much of the mass is structural/deployment mechanics versus actual cells, so your mileage may vary considerably if you're going beyond "back of the envelope" design of the spacecraft, but let's go with that number.
At Mars orbit, 1.666 AU, you're only going to get about 35% of that, about 100W/kg -- still a lot better than RTGs! So your solar array is going to mass on the order of 50kg. You might want to take extra solar panels as insurance against degradation, or just run the ion thruster at lower power later in the mission if necessary.
If you're very patient, there may be a sweet spot where a smaller solar array plus a large battery can run the thruster at full power for limited periods of time for less mass investment than a big solar array. This will mean that, e.g. orbital adjustments will have to be done in more, shorter burns than if you can sustain the thruster constantly.