Uncertainty in light energy

[Raskolnikov]

The shortest laser pulse ever produced is roughly between 3~10 fentoseconds. Due to the time-energy uncertainty principle, since we know so precisely the time period in which this phenemenon has occured, we need to have a big uncertainty in the energy of this photons. My question is: can't we produce such a short laser pulse and use the photoelectric effect to determine it's energy? If not, why not? If so, doesn't that violate the uncertainty principle?

Edited April 10, 2011 by Raskolnikov

[swansont]

The pulse isn't made up of just one photon, it's made up of many, and the HUP is telling you the spread of photon energies in the pulse.

[Raskolnikov]

But since it's a LASER, all the photons must have the same energy, because they are all in the same exact quantum state. So it doesn't matter if the pulse is made of more than one photon, with the photoelectric effect we should still be able to calculate the energy of such photons since we know how many photons are produce and that they all have the same exact energy.

[swansont]

But since it's a LASER, all the photons must have the same energy, because they are all in the same exact quantum state. So it doesn't matter if the pulse is made of more than one photon, with the photoelectric effect we should still be able to calculate the energy of such photons since we know how many photons are produce and that they all have the same exact energy.

 

It's a common misconception that lasers emit photons of exactly the same energy; all lasers have a linewidth, which itself is a result of the HUP. If they did not, there would be no way to narrow a laser pulse — the methods depend on dispersion effects.

 

Regardless, the photoelectric effect would not be a good way to measure photon energy.

[Raskolnikov]

Why not?

[swansont]

You ionize the electron. That doesn't discriminate between photon energy. You could measure the electron energy, but this approach requires precisely knowing the work function. Consider a mediocre laser with a linewidth of 1 MHz, but that's out of a frequency above 10^14 Hz. Do you know the work function to a part in 10^8?

[Raskolnikov]

Good point. Thank you. So how can the energy of a photon be determined more precisely?

[swansont]

There are a number of spectroscopy techniques. Grating spectrometers, resonant cavities, comparison with a known spectral line, and others.