vanesch said:
EDIT: ah, I think I see your argument now: you seem to think that if a BH crosses a nucleus, it "must" absorb it, no matter how small the hole. But on the BH's scale, the nucleus is not a "massive filled thing", in the same way as for a fast electron, the atom is not a massive filled thing".
What you estimated was roughly the probability of a NEUTRON to interact in matter.
You have a point. Just like the atom is made up of a nucleus and some electrons within a lot of empty space, so the nucleus is made up of protons and neutrons, which are themselves made up of up and down quarks. My expectation is that the hole would have to actually hit one of the most fundamental particles to achieve anything. Its gravity, at this stage, is much to weak to draw a particle in against the electromagnetic or the nuclear force.
However, what is the size of the most fundamental particle? If an electron is fundamental, one might try to calculate a minimum size for an electron, based on the idea that its total mass must be larger (or, at any rate, not smaller) than the mass of its electric field. This gives me a diameter of roughly 1/100.000 of the atom diameter for the electron. The quarks, being more massive, and carrying less charge, might be smaller, say 1/10.000.000 of the atom diameter. Or they might not be that small; they carry other fields, which presumably have their own contributions to their mass.
The black hole would be much smaller still; essentially point sized relative to the quarks and electrons.
Would swallowing an electron cause the hole to attract and devour the nucleus (and the other electrons after that)? Perhaps not. If it formed by cosmic radiation in the atmosphere, it will have relativistic speed, and may not dwell long enough in the vicinity of the other particles to draw them in. But if it formed in the LHC, at a moderate speed, if might well have enough time to do so. And otherwise it would be attracted to any nucleus it passed, increasing its effective cross section until it had neutralized its charge.
At any rate, I wonder whether a black hole could get away after swallowing just one quark. Not only would it have a partial charge, it would also have a colour charge, and be strongly attracted to the remaining quarks. I would at least expect it to take the other two quarks of the same nucleon. If this nucleon is a neutron, this might have little effect. If a proton, the hole would again have charge, and attract electrons. Eventually it would catch one, whether from the same atom or from somewhere else. It would be the electrons that moved towards it, rather than vice versa, considering that the hole is initially 10.000 times as massive at as a proton, and twenty million times as massive as an electron.
But, yes, the cross section of 1/100,000 squared might be wrong. Suppose that it is essential for the hole to hit a quark, with a cross section of 1/100,000,000 squared, and suppose that even if it does so, it only takes one nucleon mass, not an entire atom. Instead of having to pass through ten billion layers of atoms, it would have to pass through ten million billion layers of atoms, or one hundred kilometers of solid mass. On one passage through the Earth the hole would acquire roughly 100 proton masses; it would need 100 orbits to double its mass, which it would complete in a week. The growth would not be linear, because the diameter of the hole itself would slowly increase, but it might be centuries before this became noticable.
However, there would be more than one hole. The holes would be in decaying orbits, ending up in a small region, where they would eventually meet. The diameter of a black hole is proportional to its mass, not to the cube root of its mass. So the ability of a combination of holes to increase its mass would become proportional to the square of the total mass. And when the mass reaches 10
30 proton masses, or a few kilograms, their gravity becomes strong enough to overcome interatomic (van der Waals) forces and draw neighbouring atoms in. From there, the growth is exponential.
I expected the "dormant" stage to last a few decades, but it may actually be a hundred thousand years.