There's a few problems with ion propulsion from LEO to the Moon that I hadn't considered. The first problem is in providing the electrical power required.
Last week I suggested using solar panels. The problem with that idea is that the spacecraft will be travelling through the Van Allen belts, where there is a high level of ionising radiation. Ionising radiation destroys solar panels, so it's best to minimise the amount of time spent in the Van Allen belts.
Here's a rough approximation. The ion thruster I described last week (Isp of 3800 s) had a thrust of 0.2 N . Suppose the spacecraft has an initial mass of 20,000 kg (20 tonnes). That means the acceleration from this 5 kW thruster will be of the order of 1x10-5 ms-2. It'd take a long, long time (~ 20 years) to make a velocity change of 6.33x103 ms-1 at that rate of acceleration. 20 years, a large number of which would be spent in the Van Allen belts.
So what about increasing the number of thrusters? Let's say 2 years for a transfer is reasonable. Then 10 thrusters are needed, corresponding to 50 kW of power needed. If a solar array is degraded by 10-50% on a single transit through the Van Allen belt - and my spacecraft is spending quite a while there - I would need an initial capability of 100 kW. Unfortunately, the technology doesn't currently exist to make a deployable array that big. Additionally, the spacecraft would only be able to make one trip from Earth to the Moon.
Assuming that somehow there's been such a deployable array constructed and used, and since these (very heavy) arrays have been taken all the way to the Moon, it wouldn't be unreasonable to want to use them again. It would make sense to land them and use them to power a base. Only problem is, they'd need to be undeployed first. A good demonstration of the reason for that is as follows: take a strip of card (a bookmark, perhaps). Make a fist, and grip the middle of the strip between your thumb and the side of your index finger. With the strip horizontal, hold your fist a couple of inches above a table and allow it to drop. Watch what happens to the strip as your fist hits the table. Now imagine what would happen to the deployed solar array under an even larger shock. No-one's ever built a multiply-deployable array either: there's another challenge.
There is one possible solution to the Van Allen problem: run the solar panels at a high temperature (~ 400 K), when, in theory, they may be able to self-heal.
Next issue is mass. Once again, assuming 20 tonne initial mass and mf / m0 = 0.843, there is a requirement 3.14 tonnes of fuel. I'll round that to 3.5 tonnes to allow for the tankage factor. Typical specific masses for solar electric propulsion (SEP) systems are of the order of 20 kg/kW (I really think this is seriously underestimated), so a 100 kW system would have a hardware mass of 2 tonnes. That leaves 14.5 tonnes for payload and subsystems for lunar landing.
Now, what happens when the system scales up? A Saturn V was capable of sending 47 tonnes to the moon, and a hundred tonnes to LEO. What happens then?
 JPL Advanced Propulsion Technology Group, 'Advanced Propulsion Concepts', Island One Society