Remember the high school lab experiment where a pinch of sodium at room temperature was dropped into a beaker of cool tap water. The result was a violent chemical reaction, a very exothermic one, and the reaction product was sodium hydroxide, NaOH. The raw Na was so reactive it had to be stored in a kerosine bath to keep moisture away from it.

Now imagine thousands of pounds per minute of molten sodium coming out of a nuclear reactor vessel and going into a heat exchanger where only a thin steel pipe wall separates it from superheated steam, still in liquid form at high pressure, so the steam can then be sent on to a turbine.

Unlike a normal nuclear reactor, which may take hours to undergo a gradual China Syndrome meltdown, the breeder would fail catastrophically, in an explosion of radiaoctive debris and highly corrosive lye.

Sodium had to be used because breeding Plutonium requires "fast neutrons" and water slows them down too much. I actually visited the demonstration breeder in Playa Rincon, Puerto Rico in the early 70s. It was a mature technology, they had been doing it since the 60s, but no one ever though it was ready for full scale industrial use. It was strictly an experimental facility, around 400 MW capacity if I remember correctly.

If I'm not mistaken, the second nuclear sub (SeaWolf) reactor was a liquid sodium cooled jobber. As far as I know, it was the only one ever built.

Hopefully, they've gotten the bugs out of this technology by now.