[alancalverd]

As I said, a good photocathode will respond to a single photon, but a film won't (note "usually": x-ray film is easily fogged by natural background radiation because it is much more sensitive than ordinary camera film). Thus classic double-slit experiments like Taylor's Box are frankly questionable: they demonstrate the wavelike behavior of something over a long period, whilst replacing the film with a single-photon counter clearly demonstrates the instantaneously particulate behavior of the same thing.

[hamdani yusuf (OP)]

As I said, a good photocathode will respond to a single photon, but a film won't (note "usually": x-ray film is easily fogged by natural background radiation because it is much more sensitive than ordinary camera film).

The effectiveness of a photocathode is commonly expressed as quantum efficiency, that being the ratio of emitted electrons vs. impinging quanta (of light). The efficiency varies with construction as well, as it can be improved with a stronger electric field.

https://en.wikipedia.org/wiki/Photocathode#Construction
This fact reduces the credibility of claim that each detection in photocathode represents a single photon impact. The temperature dependence of detection rate as I mentioned earlier in photomutliplier only strengthen this suspicion.

[Bored chemist]

Astronomers used to "pre-fog" their photographic plates to lift the sensitivity above the RF level,

So, one presumes, do physicists monitoring low photon fluxes.

[Bored chemist]

This fact reduces the credibility of claim that each detection in photocathode represents a single photon impact.

How?

The temperature dependence of detection rate as I mentioned earlier in photomutliplier only strengthen this suspicion.

In what way?

[hamdani yusuf (OP)]

This fact reduces the credibility of claim that each detection in photocathode represents a single photon impact.

How?

The temperature dependence of detection rate as I mentioned earlier in photomutliplier only strengthen this suspicion.

In what way?

Ideally, every photon impingement is followed by a detection on the screen. Other factors such as electric field and temperature shouldn't significantly affect the result.

When the detections are significantly less than the photon, then there must be some undetected photons. On the other hand, when there are more detections than the photons, then there must be some other causes that trigger detections other than photons.

Alternatively, the detections are mainly caused by something else. The energy from light source only adds the chance that the detection is triggered.

The further away from the ideal condition, the result reduces the credibility of photon model and increases the credibility of the alternatives.

[Bored chemist]

You seem to be assuming that the surface of the photocathode is perfectly even and that the design and manufacturing process is perfect.

Not every photon initiates a cascade.
But most do.
The size of the pulse is pretty much identical for each photon.
It's very hard to explain that except in terms of a single electron being set free in the first place.

then there must be some other causes that trigger detections other than photons.

That's never been in dispute. It's called the "dark current".
Thermal energy of electrons is occasionally high enough to get them to escape (especially for detectors sensitive to longer wavelengths in the IR.
So, you would expect that higher temperatures increase the likelihood of an electron escaping the photocathode without being hit by a (light) photon.

Also, of course things like cosmic rays will occasionally set an electron free and cause a pulse.

The fact that not every event registered by a PMT is due to a photon, does not mean that photons don't cause events.

[alancalverd]

The effectiveness of a photocathode is commonly expressed as quantum efficiency, that being the ratio of emitted electrons vs. impinging quanta (of light). The efficiency varies with construction as well, as it can be improved with a stronger electric field.

https://en.wikipedia.org/wiki/Photocathode#Construction
This fact reduces the credibility of claim that each detection in photocathode represents a single photon impact. The temperature dependence of detection rate as I mentioned earlier in photomutliplier only strengthen this suspicion.

No. A single spark can ignite an entire petrol refinery or ammunition dump, so the explosion tells you that there has been one spark. Same with most radiation detectors: you rely on one photon event initiating a cascade so you get a big output pulse for each incoming  photon. The geiger counter is a particular example where the cascade causes the entire gas cloud to conduct so the output pulse amplitude is independent of the energy of the incident photon. A photomultiplier uses a sequence of photocathodes (cascodes) to produce a fixed energy-independent gain but the output pulse amplitude depends on the energy of  the incident photon, which determines how many electrons are emitted from the first interaction.

Temperature dependence of low-energy (photocathode) detectors is simply due to the thermal excitation of conduction band electrons in the cathode. If they have more thermal energy then it requires less energy to expel them from the surface so the gain of the target and each cascode stage increases with temperature. But you still only get one pulse per photon!   

[Bored chemist]

A photomultiplier uses a sequence of photocathodes (cascodes) to produce a fixed energy-independent gain but the output pulse amplitude depends on the energy of  the incident photon, which determines how many electrons are emitted from the first interaction.

No.
One  photo-electron per photon (or near offer).
Your head still seems to be stuck in the x-rays or something.

[hamdani yusuf (OP)]

It's very hard to explain that except in terms of a single electron being set free in the first place.

I don't dispute that. In popcorn analogy, I doubt that the activation of detector means that a  photon just hit it. It's explained in the end of the video I cited previously.

[hamdani yusuf (OP)]

Temperature dependence of low-energy (photocathode) detectors is simply due to the thermal excitation of conduction band electrons in the cathode. If they have more thermal energy then it requires less energy to expel them from the surface so the gain of the target and each cascode stage increases with temperature. But you still only get one pulse per photon!   

No. The temperature increases the rate of detection, whether or not the light source is on. The frequency required to increase the detection rate depends on the voltage between electrodes.

[alancalverd]

One  photo-electron per photon (or near offer).

If only! The best optical photocathodes have a quantum efficiency of around 30% with a definite wavelength dependence.

[Bored chemist]

One  photo-electron per photon (or near offer).

If only! The best optical photocathodes have a quantum efficiency of around 30% with a definite wavelength dependence.

OK, so 30% is the "near offer". Photodiodes do better (but the s/n ratio is worse)
However, this

but the output pulse amplitude depends on the energy of  the incident photon, which determines how many electrons are emitted from the first interaction.

is still wrong- at least for visible light.
If you want to say you get about 0.3 electrons per photon, that's fine. It just means that about 2/3 get absorbed as heat.
But it's pretty much independent of wavelength (and thus, photon energy) in the region where they actually get used.
eg
https://www.sciencedirect.com/science/article/abs/pii/S0168900206009247

You are muddling it with scintillation counters etc. where the output pulse from the PMT is proportional to the gamma photon energy.

[Bored chemist]

I doubt that the activation of detector means that a  photon just hit it.

Essentially all the evidence disagrees with you.

[alancalverd]

According to Hamamatsu (who know a thing or two about their products) photocathode emission is strongly wavelength-dependent

[Bored chemist]

According to Hamamatsu (who know a thing or two about their products) photocathode emission is strongly wavelength-dependent

And according to you the quantum efficiency is simultaneously less than one

The best optical photocathodes have a quantum efficiency of around 30%

And also more than 1

determines how many electrons are emitted from the first interaction.

It's never "electrons" i,.e. plural, is it?

[hamdani yusuf (OP)]

According to Hamamatsu (who know a thing or two about their products) photocathode emission is strongly wavelength-dependent

The glass window material contributes significantly to that dependency.

[hamdani yusuf (OP)]

I doubt that the activation of detector means that a  photon just hit it.

Essentially all the evidence disagrees with you.

Name one.

[Bored chemist]

I doubt that the activation of detector means that a  photon just hit it.

Essentially all the evidence disagrees with you.

Name one.

Einstein's Nobel prize,

[hamdani yusuf (OP)]

Einstein's Nobel prize,

That's a theoretical work. What's the experimental evidence? Why can't it be interpreted in another way?

[Bored chemist]

Einstein's Nobel prize,

That's a theoretical work. What's the experimental evidence? Why can't it be interpreted in another way?

For a start, that's silly.
The experiment on which the work was based is real enough; I did it as a high school experiment.
The evidence for the quantization of light isn't just one experiment.
It's a whole mass of interlocking ideas.
It solves the ultraviolet catastrophe.
It explains the photoelectric effect and, whether you like it ot not, it's the only credible explanation of photomultipliers (and their solid state equivalents- the avalanche photodiodes).
It's also a vital part of photochemistry.

So,

Why can't it be interpreted in another way?

Well, go ahead and try.
But make sure that your idea  matches all the observations in spectroscopy, electronics, cosmology and so on.