Why is the resistance of light-dependent resistors wavelength-dependent?

Question

The resistivity of the semiconducting CdS in LDRs is dependent on the intensity of the incident light. This makes sense, given that the more photons (with sufficient photon energy) hit the material, the more electrons are excited into the conductivity band, and resistivity decreases.

However, it also depends on the wavelength (photon energy) of the light. I don't see why this is the case, given that in the photoelectric effect, only a certain energy is required to excite an electron, and if the photon energy is higher, the electron will simply have higher kinetic energy). Thus it would seem that the number of electrons released is not wavelength-dependene.

I found this question, which seems to cover a similar topic, but it doesn't deal with LDRs or semiconductors specifically.

Answer 1

Note that intensity (measured in Watts/square meter) is the product of the number of photons and their individual energy. I can get the same intensity with 100 photons of 800 nm or 50 photons of 400 nm.

Secondly - the photoresistive effect is a bulk effect, and the probability of the successful interaction of the photon with the material will depend on its penetration, and on it having the "right kind" of interaction with the material. When there are all kinds of absorbing materials (starting with the window of the LDR device) before you even get to the active region, and when further there are different absorption mechanisms in the material (not all of which will result in photo-induced conduction) you see there is really no reason to believe that the change in resistivity would NOT be wavelength-dependent.

Finally, according to Lambe et al, "Recombination Processes in Cadmium Sulfide", Phys.Rev 98, 985:

It is found that the observed spectral response can be explained in terms of the dependence of lifetime of conduction electrons on the wavelength of exciting light.

In other words - depending on the energy of the incoming radiation, the electrons generated may have a shorter or longer life in the unbound state: this will determine how much "conducting" they can do.