I can understand "nuclei excitation states" in two different ways.
- If you mean excited states of the neutrons and protons in the nucleus, I really have no idea, mainly because these are generally unachievable in an engineered system. These excitations are in the gamma-ray spectrum and just way too high-energy and destructive to be useful, so no one has tried to harness them for quantum information processing.
- If you mean the "orientation" of the spin of the nucleus, which can be easily manipulated with reasonably weak low-frequency magnetic fields (like we already do classically for medical imagining purposes), then this is actually a pretty good storage medium. Frequently people think of the electron spin as the "networking card" because it can easily interface with photonic qubits, and think of the nuclear spin as the storage medium because nuclear spins can retain quantum information for many minutes (because they are so well isolated). One problem is that transferring the data from the electron spin to the nuclear spin is relatively slow and noisy. But "progress is being made" albeit less quickly than we would like.
My claim to authority here is that I have designed various control protocols for such hardware during my postdoctoral time. But I had it easy, most of my work includes statements roughly equivalent to "assuming the progress of the last 20 years continues, we will have good qubits in 10 years if we use the protocol suggested here". To be clear, there are technologies that will be workable sooner (or already are), they just have different tradeoffs in cost and other requirements.
- If you mean excited states of the neutrons and protons in the nucleus, I really have no idea, mainly because these are generally unachievable in an engineered system. These excitations are in the gamma-ray spectrum and just way too high-energy and destructive to be useful, so no one has tried to harness them for quantum information processing.
- If you mean the "orientation" of the spin of the nucleus, which can be easily manipulated with reasonably weak low-frequency magnetic fields (like we already do classically for medical imagining purposes), then this is actually a pretty good storage medium. Frequently people think of the electron spin as the "networking card" because it can easily interface with photonic qubits, and think of the nuclear spin as the storage medium because nuclear spins can retain quantum information for many minutes (because they are so well isolated). One problem is that transferring the data from the electron spin to the nuclear spin is relatively slow and noisy. But "progress is being made" albeit less quickly than we would like.
My claim to authority here is that I have designed various control protocols for such hardware during my postdoctoral time. But I had it easy, most of my work includes statements roughly equivalent to "assuming the progress of the last 20 years continues, we will have good qubits in 10 years if we use the protocol suggested here". To be clear, there are technologies that will be workable sooner (or already are), they just have different tradeoffs in cost and other requirements.