Infrared Reflection Methods

14. Infrared Reflection Methods

Everyone who uses infrared spectrometry is familiar with transmission methods; films of liquids and gums between KBr flats, free standing films of polymers, Nujol mulls and KBr discs – most of us have used them all. In Edition 1 we considered theory and practice in mulling and KBr disc making, but there are alternative methods of sampling.

Many years ago pioneers demonstrated that excellent spectra could be recorded using reflection methods. As time went on, a bewildering range of these appeared, but in a sense they were of very limited value and only of real interest to the specialist. The problem with them all was that the reflection process was inefficient and the older dispersion instruments were not sensitive enough to give good quality results – or to be more correct, unless used with considerable skill and backed by experience, the results were unacceptable.

15. In the 70’s and 80’s, FTIRs penetrated first the research and then the routine laboratory. These instruments had vastly improved sensitivity and as a result the reflection methods got a shot in the arm. The FT instrument did not make these methods possible, although some of the accessory manufacturers imply that it did, but rather its development made them routine.

What are these reflection methods? Basically there are three – specular reflection, diffuse reflection, and attenuated total internal reflection, although all of them have been subdivided and developed into seemingly newer and more advanced techniques. To assist readers who are not familiar with them, I will now very briefly and simplistically describe each. In articles to follow, this brief introduction will be expanded.

16. Specular Reflection
Everyone is familiar with this – the mirror is an example. Light incident on a smooth surface is reflected at the angle with which it hits the surface.

Note that the angle of incidence and reflection is conventionally referred to the normal, not the surface itself. Don’t ask me why – it just is! So glancing incidence – light slithering along the surface itself – has a high angle of incidence, say 85°.

17. Light reflected off a surface is not identical in characteristics to that which is incident. Some attenuation is normal. The nature of the reflector itself causes this attenuation. Thus, my car is red – white light is severely attenuated in the blue/green by the paint, so it appears to me to be red (I conveniently forgot to mention the dirt and the rust). Also, the reflection process causes polarization. White light from all sources is non-polarized – the electric vector of the electromagnetic radiation is not controlled. Reflection is more efficient for light polarized normal to the reflector than it is for radiation polarized parallel to it. So we can redraw the first diagram:

18. Thus, to the observer, the reflection process has caused the radiation to be somewhat polarized. Many of us exploit this effect when we buy Polaroid sunglasses. The Polaroid sheet in the glasses is set to transmit only the horizontally polarized light. Hence the eye sees 50% of the light reaching the Polaroid, (in fact a bit less than 50% due to absorption losses), thus reducing the light level as is desired. However, the reflected light is more severely attenuated so that reflection off smooth surfaces seems to the wearer to be dimmed. That is why Polaroid sunglasses reduce the glare of the sun reflected off water or polished surfaces and makes them ideal for motorists or fishermen (you really can see the fish more easily when wearing Polaroid rather than any other sunglasses).

19. If you illuminate a surface with infrared radiation, the light will be attenuated and polarized exactly as it would be in the visible. The efficiency of reflection is related to the refractive index of the reflector. The theory will come up later but it is sufficient here to accept that as a sample absorbs, its index of refraction changes as well. As a result, infrared absorbers show an efficiency of reflection that wanders up and down with the absorbance. Hence, the reflected light contains information in it about the absorption spectrum. It would be nice if the reflection spectrum was the same as the absorption one, but it is not. It is related to it. Here is a comparison

20. There is an obvious connection between the two but the reflection spectrum is hard to recognise. Fortunately, computer processing of the reflected spectra can unscramble this lot and produce an output very close to the conventional absorption spectra. The software (or algorithm) is due to Kramers and Krönig and almost all FTIRs have it as part of the software suite. So we have –

Highly polished metal mirrors can reflect ~97% of the light falling on them but fairly smooth organic ones (say a lump of plastic) will only reflect 10-25%, but this is ample for a modern FTIR. Accessory manufacturers make Specular Reflection accessories and most enable users to vary the angles of incidence and reflection at will.

21. Sampling in specular reflection can be a problem. If the sample readily transmits, the reflection can be weak and of less diagnostic value. The method works best on ‘optically dense’ samples e.g. carbon loaded plastics where the carbon soaks up the transmitted radiation and the reflection bounces off the surface.

If a sample is oriented – say a mineral or polymer – the absorption spectrum is dichroic, i.e. it is direction dependant, hence the reflection efficiency is dichroic as well and as a result measurements can yield orientation data. If the reflection is carried out at high angles of incidence, grazing angle reflection information can be acquired from very thin films.

22. So, the message is clear – reflection is valuable and versatile. The problem is that unless a diagnostic protocol has been worked out in a particular case, it is difficult to use, requires computer processing, and is certainly not a universal method for the casual user.

A variant on specular reflection is transflectance, although it has been called other names in its time. If you have a thin sample – a piece of film or a liquid film – you can examine it like this


23. Two processes occur – reflection off the top surface and also transmission and reflection off the mirror, followed by a second transmission. If the reflection off the top surface could be eliminated, the light collected would be a true absorption spectrum typical of a sample of thickness 2t/Cos(i). The reflection off the mirror, as we have seen, is much more efficient than our organic material hence the combined radiation collected in the experiment illustrated above is predominantly the transmitted stuff and many users ignore the reflection off the sample surface and assume their spectrum is due to transmission.

24. Diffuse Reflection
This technique is often called DRIFT – Diffuse Reflection Infrared Fourier Transform – but the Fourier Transform part of the name is irrelevant. Diffuse reflection occurs off rough surfaces. The reflected radiation comes off in the specular direction and in all other directions as well.

The specular component arises very much like it does off a smooth surface but the diffuse bit occurs to at least some extent from penetration and multiple reflection/absorption inside the bulk of the specimen. The result is that if one can separate the two, the diffusely scattered radiation resembles an absorption spectrum, the specularly reflected part, the spectrum described in paras 16 to 23 above.

25. When it appeared, DRIFT seemed to be the answer to the analyst’s prayer – a simple, almost universal analytical tool – but it transpires that it is good in the hands of the experienced expert. Sampling is fiddly and requires skill. Alexander Shchegolikhin and Olga Lazareva have provided a piece to follow which tells you more about DRIFT and a variant they find more useful – DRAFT (and I’m not going to tell you what DRAFT stands for!). Thank goodness it didn’t come out as DRAUGHT!

REF: Int.J. Vib. Spect.,[] 1, 4, 14-25 (1997)