3 ms·

Terahertz waves range in length from 1 micrometer to 1 millimeter. These are far too large to be able to image things at the scales of crystals. (That's part of

Terahertz waves range in length from 1 micrometer to 1 millimeter. These are far too large to be able to image things at the scales of crystals. (That's part of why X-rays are used for crystallography - they provide higher resolution than visible light, because the wavelength is shorter.)

amelius 10y ago
Petahertz (or higher) is not an option, I presume?
- kragen 10y ago
  Petahertz electromagnetic radiation is what's currently used for diffraction crystallography; it's conventionally known as "X-rays" (or "gamma rays" if it comes from radioactive decay instead of from Bremsstrahlung.)
  - amelius 10y ago
    Yes, can we use time-domain techniques on this frequency range?
    - kragen 10y ago
      You want to measure the instantaneous amplitudes of a wave packet in the petahertz range at above the Nyquist frequency? I don't want to say it's impossible, but I think you'd have to improve the capabilities of our instruments by about four orders of magnitude, which means you would have to invent several new major technologies. Maybe if you digitize the time-domain pulses using some kind of controlled nuclear reaction? Electronics are far too slow (although I don't really understand why, so maybe that can be fixed). X-ray-stimulated nuclear reactions are going to be tricky to get to happen at all, much less to chain together into computational circuits, although in that case maybe you can use α particles or neutrons to carry the information around between the computational elements, simplifying the task somewhat. It seems like, however you manage to measure it, the quantization of the radiation is going to pose really significant problems to measuring its amplitude accurately as a function of time, but maybe you can get around that by repeating the pulses a lot of times. But you'll need to make sure that you're emitting pulses of the same shape, rather than just pulses with a similar frequency spectrum and envelope but varying relative phases among the frequencies. On the other hand, if you just want to measure the time-of-flight for X-ray wavepackets to sub-picosecond precision, without worrying about the time-domain waveform within the packet, I don't think that gets any harder or easier depending on the "carrier frequency" of the wavepacket — X-rays should be just as easy as terahertz light. You do need a few terahertz of bandwidth, so you aren't going to be able to do this with bursts of gigahertz microwaves. X-rays' interaction with the material you're trying to image might be less or more convenient; their cross-section for backscattering is pretty small, but maybe you could scan the book from different angles and use total internal reflection to find interfaces between materials in the book with different refractive indices. This would also reduce your requirements for time-domain precision down into the hundreds of GHz range, which is a lot easier to deal with.
      - eternauta3k 10y ago
        > Maybe if you digitize the time-domain pulses using some kind of controlled nuclear reaction? Electronics are far too slow Terahertz TDS already samples at frequencies beyond the capabilities of bare electronics (look at my other comment on this story).

amelius 10y ago

Petahertz (or higher) is not an option, I presume?

kragen 10y ago
Petahertz electromagnetic radiation is what's currently used for diffraction crystallography; it's conventionally known as "X-rays" (or "gamma rays" if it comes from radioactive decay instead of from Bremsstrahlung.)
- amelius 10y ago
  Yes, can we use time-domain techniques on this frequency range?
  - kragen 10y ago
    You want to measure the instantaneous amplitudes of a wave packet in the petahertz range at above the Nyquist frequency? I don't want to say it's impossible, but I think you'd have to improve the capabilities of our instruments by about four orders of magnitude, which means you would have to invent several new major technologies. Maybe if you digitize the time-domain pulses using some kind of controlled nuclear reaction? Electronics are far too slow (although I don't really understand why, so maybe that can be fixed). X-ray-stimulated nuclear reactions are going to be tricky to get to happen at all, much less to chain together into computational circuits, although in that case maybe you can use α particles or neutrons to carry the information around between the computational elements, simplifying the task somewhat. It seems like, however you manage to measure it, the quantization of the radiation is going to pose really significant problems to measuring its amplitude accurately as a function of time, but maybe you can get around that by repeating the pulses a lot of times. But you'll need to make sure that you're emitting pulses of the same shape, rather than just pulses with a similar frequency spectrum and envelope but varying relative phases among the frequencies. On the other hand, if you just want to measure the time-of-flight for X-ray wavepackets to sub-picosecond precision, without worrying about the time-domain waveform within the packet, I don't think that gets any harder or easier depending on the "carrier frequency" of the wavepacket — X-rays should be just as easy as terahertz light. You do need a few terahertz of bandwidth, so you aren't going to be able to do this with bursts of gigahertz microwaves. X-rays' interaction with the material you're trying to image might be less or more convenient; their cross-section for backscattering is pretty small, but maybe you could scan the book from different angles and use total internal reflection to find interfaces between materials in the book with different refractive indices. This would also reduce your requirements for time-domain precision down into the hundreds of GHz range, which is a lot easier to deal with.
    - eternauta3k 10y ago
      > Maybe if you digitize the time-domain pulses using some kind of controlled nuclear reaction? Electronics are far too slow Terahertz TDS already samples at frequencies beyond the capabilities of bare electronics (look at my other comment on this story).