The way I understand it (going by an old “How it works” encyclopaedia dated 1977
) the key is to seperate the Uranium-235 from the U-238. U-235 can allow fission, 238 doesn’t. (Generally speaking- it’s an old book. '38 can, however, capture neutrons and convert to Plutonium.)
'35 is only 0.73% of naturally occurring Uranium (the rest being U-238) and most reactors need it to be 1.5 to 2.5% to be able to use it as a fuel- “weapons grade” is purer still, roughly 85 to 95%. The Hiroshima weapon used roughly 110 pounds of it, so you can see that making one bomb’s worth of Uranium takes a great deal of raw ore.
Now the problem is, the two isotopes are essentially identical, save for what, three electrons? Neutrons? (Someone will correct me, I’m sure.) So it’s not just finding a needle in a haystack, it’s more like finding a #2 needle in a stack of #3 needles. You can’t seperate them chemically, they have to be seperated mechanically.
This seperation process, to concentrate the '35, is complex and costly, involving substances like hydrogen fluoride, tributyl phosphate, nitric acid, zirconium, magnesium, and a great deal of complex machinery like gas centrifuges. Considering the size of the machinery, the chemicals used, the systems needed to store, transport and use the chemicals, and the support systems, it’s all a setup about the size of a small oil refinery.
It’s large enough that it’s very difficult to hide such a setup underground, but without on-site inspections, the setup looks the same from a satellite’s view whether the place is simply making fuel rods for a reactor or making enriched “weapons grade” material.
And once you have the fissile material, I understand that making a single-stage nuke is absurdly easy. The Hiroshima weapon was not a whole lot more than a stout pipe, two chunks of enriched Uranium, a handful of an explosive, and an altimeter-detonator.