In principle Silver behaves as a free electron medium (see Drude-Lorentz model) in the far infrared and its reflectivity stays at very high values (98.5% or more) all the way to the microwaves and beyond. The only explanation I have for the low reflectivity above 10 microns is that the surface is contaminated by other substances

Some commercial IR reflective slides have a thin SnO2 coating, and the SnO2 vibrational modes, which are in that region, could absorb some of the light.

Thanks for your replies!

These aren't commercial reflective slides (they were deposited by evaporation in-house), but I've been told that they are definitely only a single layer of Ag, perhaps a few hundred nanometres thick. Possibly they have become contaminated, or the evaporant material or coating chamber was contaminated.

Needs more investigation, I'll let you know how it goes.

Thanks again.

Check with the supplier of the mirror, typically there is a protection layer on top.

There's definitely no protective layer, these were made in my department using only Ag, so there must be some other problem.

If the Ag films were made in-house by evaporation, I am sure that you will have a superficial layer of oxide on top of your film. Ag forms an absorbing oxide. A few nanometers is sufficient to absorb some light and lead to a decrease of the reflectance.

customer-deposited silver has short life time. What you can try is to make a thicker Ag layer in order to have high reflectance beyond 10um..

Sounds like contamination of your film is the issue. I suggest checking your evaporator for leaks and verify your pressure gauges are working correctly.

Pure silver reacts very fast with ppb amount of H2S in the air, forming a thin Ag2S layer.

it will be probably due to contamination and forming an oxide layer

As silver gets oxidized very easily

From past experiences I have with depositing silver, I can tell you that the reflectivity does not change above 10 microns so it is possible that there is some contamination. However, usually the source of the problem is a calibration error in the detector or instrument used in the measurement. My reasoning is that a contamination will affect ALL wavelengths not just those above 10 microns.

Here is a source that may be helpful for you.

http://refractiveindex.info/?group=METALS&material=Silver

The important take away from these graphs is that the values of n and k are monotonically increasing. Therefore reflectance is increasing as well.

Alas, I found may authoritative reference on the matter:

M.A. Ordal, "Optical properties of metals in the infrared and far infrared". Univ. of Missouri-Rolla. Sept. 1985.

Ross,

This is not normal. Pure siliver should be a perface reflector for long wavelength than 10 um. You can referre to early Wolfe and Zissis' book of Handbook of Infrared.

Without careful protection of your reference mirror, optical properties could be detoriated resulting in degraded reflectivity.

On the otherhand if the Ag layer is thin comapred to the IR wavelength, interference could occur. As a result, the refectivity depends on the film thickness and protiety of the host substrate.

See A. D. Rakic, A. B. Djuriˇic, J. M. Elazar, and M. L. Majewski,

Appl. Opt. 37, 5271 (1998). It's more up to date, with a very interesting fit

of the permittivity. I just made a webpage for that. I'm planning to release quite a few codes in a near future, this is a beginning.

http://elma.univ-bpclermont.fr/moreau/progs.html

I've measured several mirrors over the 7-13µm band on a different instrument. Noisy data (attached) aside, this is much closer to what I'd usually expect from these reference mirrors. Samples 3 and 4 are old mirror samples, a bit scratched up and duller in appearance.

It appears the problem lies with the original spectrophotometer I was using, possibly a detector or internal mirror contamination.

Thanks to everyone who contributed.