1
$\begingroup$

Let's consider a real source emitting a sequence of single-photons (like for example a N-V colour centre pumped with a pulsed laser with highly accurate frequency of the pulses). I want to characterize the source. In particular I want to directly verify experimentally that in each of the sequential time slots at most 1 single-photon is present, not more than one at the relevant wavelength range.

Ideally a detector that tells that the light is absorbed and how much energy was absorbed in that event is what I need. Is something like this existing today?

Improve this question
$\endgroup$
7
  • 1
    $\begingroup$ Since usually a large number of photons is used for the excitation, the established procedure verifies the single-photon character by inspecting the emission of the source instead of its effect on the excitation pulse. Is there any constraint in your specific experiment which makes you prefer investigating how much light is absorbed? $\endgroup$
    – A. P.
    Commented Jan 4 at 17:21
  • $\begingroup$ @A.P. Thanks for your question. I'm asking if a single-photon detector/apparatus exists that provides information about each single emission event from the source. The information should be "I got a single-photon" or "I got more than 1 photon". $\endgroup$
    – Ang
    Commented Jan 4 at 17:32
  • $\begingroup$ I recently answered a similar question: "What does $\tau$ mean in second-order correlation function?" The asker there was learning from a tutorial by qutools, which I can recommend. For characterizing single-photon sources, see Ch. 2.2 therein. $\endgroup$
    – A. P.
    Commented Jan 4 at 17:46
  • 1
    $\begingroup$ Also see here: "Can people create single photon in the laboratory?" $\endgroup$
    – A. P.
    Commented Jan 4 at 17:51
  • $\begingroup$ Consider also EMCCD. The QE is high and the detected electron generates many electrons above read/dark current noise. $\endgroup$
    – PhysicsDave
    Commented Jan 5 at 2:46

1 Answer 1

Reset to default
0
$\begingroup$

What you want is a photomultiplier tube.

This is an evacuated glass chamber whose active surface is coated with a thin layer of a metal with good electrical conductivity and a small "work function," which is the energy required for a photon to eject an electron. This "photocathode" is held at a large negative potential, so that ejected electric are repelled from it and travel towards a conductor at a less negative potential, usually a few hundred volts of difference. The accelerated electrons crash into the surface and liberate several more electrons, which travel towards the next electrode, where they multiply again. By the time the pulse reaches the anode, a single photoelectron has become perhaps $10^7$ electrons, a charge of about $10^{-12}$ coulombs, which is a microamp of current if the pulse duration is a microsecond. With appropriate pulse-shaping electronics you can see the photon pulses on an oscilloscope, and use them as logic triggers for some photon-counting system.

There are also "avalanche photodiodes," which do the same sort of thing, but all in silicon.

The measurement you describe would be challenging. Any single-photon detector will have a "dark current" of electrons from the photocathode when the detector is not exposed to any light source. The "detection efficiency" is also finite: not every photon that sees the cathode will send an electron into the vacuum.

Improve this answer
$\endgroup$
2
  • 4
    $\begingroup$ Photomultipliers have mediocre quantum efficiency, so if two photons reach the photocathode, it is unlikely to emit two electrons. APDs have better quantum efficiency, and you might be able with a carefully designed APD system to distinguish between one and two photons with some efficiency (but nowhere near perfectly). $\endgroup$
    – John Doty
    Commented Jan 4 at 18:30
  • $\begingroup$ Finite efficiency will be an issue in any such experiment; the analysis will involve some estimate of the statistical power of the dataset. $\endgroup$
    – rob ♦
    Commented Jan 4 at 22:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.