Doubt on 'Specific Detectivity' for photovoltaic IR detector

[lei]

By the mathematical definition of 'Specific Detectivity',
D*= √AΔf / NEP

where NEP is the noise equivalent power, A is the detector's active area, and Δf is the bandwidth.
Its is said that D* is normalized with A and Δf, and D* is therefore independent of A and Δf, D* is the intrinsic property of the detector material.

But I found in the specification of some photovoltaic IR detectors made of (Hg,Cd,Zn)Te semiconductor. by looking at its spectral response, the Specific Detectivity decreases as the operational wavelength of the detector increases, please see the attached image, between curve 3 and 8.
Can anyone help me to explain why it is so? Thanks

 

[steer]

Re: Doubt on 'Specific Detectivity' for photovoltaic IR dete

To get different spectral response the proportion between Cd, Hg, Zn, Te in the material is changed. Thus material is different in these curves. So D* differs too.

 

[lei]

Re: Doubt on 'Specific Detectivity' for photovoltaic IR dete

Hi, Steer, Thanks,
I also guess probably due to the material changes,

But I am still wondering why D* decreases as the operational wavelength of the detector increases, can the detectors be made such that D* and operational wavelength both increase at the same time?

It seems there is some tradeoff between D* and the operational wavelength, I just want to know why it is so in the physics?
Can anyone help to explain? Thanks.

 

[steer]

Re: Doubt on 'Specific Detectivity' for photovoltaic IR dete

In order to improve D* at longer wavelength one has to cool down the detector.
The physics is simple: for longer wavelength one needs narrower bandgap material. Narrower bandgap means higher thermal generation. More thermal generation means more leakage, thus more noise. Cooling the detector solves the problem, at least partially.
Or use a different physical principle that limits thermal generation somehow.