Out of boredom, I went and calculated the energy and volume of gas produced from a RMFFMJ (reactive metal filled full metal jacket) .45 round. (assumed as being 15 grams total; 7.5 grams sodium)
7.5g sodium = .32 moles
So, that's 3.6 liters of hydrogen gas (at room temp, mind you - it'll probably acctualy be 4.5 L at 100*C.)
The reaction with water liberates 44kj of energy (that's 59,000 foot-pounds)
That's enough energy to boil roughly 1 mole of water, giving us another ~30 liters of gas.
So... 34 liters of hot gas produced inside the body. That's going to make a mess.
(take all numbers as aproximate... I didn't really pay much attention to precision - but it's better than a factor of two, for sure.)
So someone calculated that a .45 acp bullet with half it's mass made of sodium metal would release about 69,000 ft lbs of enery, + about 400 fpe from the normal ballistic action of the bullet. So what's that, 3 or 4 times more powerful than a .50 bmg? Stepping into the power levels of a cannon without producing any extra recoil, and in a neat handgun size package.
/Doing the math for the fun of it. Previous energy calculations in this thread are way off as they ignored OAL and density.
If half of a .45 ACP bullet's mass was sodium (and the rest of it lead), you'd end up with a ~37 gr .45 ACP bullet. Dimensionally, the bullet would be the same as your run of the mill 230 gr bullet. The low mass is due to sodium being ~ 11 times less dense than lead.
~18.5 grains of sodium is a little over a gram (closer to 1.2 grams). Total chemical energy for that amount of sodium is 6200 J, which comes out to about 4600 ft lbs. Energy would be... well, who knows what it'd be for a 37 grain .45 ACP, but let's just say 600 ft lbs kinetic energy (Glaser like). Total energy would be ~5200 ft-lbs or so. .50 BMG puts out ~13000 ft-lbs of kinetic energy. So, total energy is not quite .50 BMG energy, but comparable to .470 NE.
Now it may just be me, but I wouldn't trust a bullet like this.
First, it would be an ineffective penetrator, you know, the abysmally poor sectional density and all. (Sectional density would be 0.026. For reference, 35 gr .25 ACP has a sectional density of 0.079, and 230 gr .45 ACP has a sectional density of 0.162.) At least there's no need to worry about overpenetration... I'd be surprised if such a bullet got more than 2-3" of penetration in gelatin. Not so surprised if it fragmented upon hitting gelatin.
The explosive effect likely wouldn't happen either. The body may be mostly water, but the heat of the sodium reacting with water would cauterize the area, stopping the reaction before the sodium would have a chance to ignite and explode the hydrogen gas generated from the reaction.
So, most likely, you'd end up with a very shallow, self cauterizing wound.
Cesium would be better, at least it'd get up to ~ 80 gr bullet weight. Still, you'd end up with the same results as for sodium, just with a little better penetration, a little radiation poisoning, and a higher long term cancer rate. Explosion odds are also worse than sodium's, not that sodium has much of a chance of exploding.
I'd take a nice 230 gr .45 ACP JHP over either of these any day of the week.