At the beginning of scientific investigations more than 200 years ago spectroscopy was carried out using sunlight, then radiation from thermal filaments and only in the last 40-50 years using lasers. Coherent radiation is only really available from a laser source and since the majority of lasers are of a fixed wavelength they are not really suitable for spectroscopy. Except of course tunable diode lasers, dye lasers, OPO's etc.

FTIR spectroscopy is near universally carried out with thermal sources (e.g. glowbar) emitting out to wavelengths of many microns.

Coupling a thermal source such as a filament or arc lamp into a waveguide is possible and is done widely. Similarly, coupling a LED into a waveguide is straight forward. There are no real difficulties there except that it is possibly quite lossy and you wont get a high coupling efficiency because of the usually extended size of your source and the fairly high divergence of a LED emission.

Remember some basic rules of optics,

The product of divergence x spot size is constant,

meaning that if you have a source with inherently high divergence such as a LED, you cant focus it down to a small spot size unless you use a very short focal length lens, however the majority of waveguides have a limited NA and ligt launched into it under a larger angle than the NA is lost very quickly.

You mention coupling to a waveguide and 1.5 to 3 micron?

What material are you thinking of to use in your waveguide?

There aren't many common materials which guide at 3 microns although you can use capillaries with internal reflective coatings.

There are some very bespoke optical fibres (hollow core PCF) which guide in this wavelength range but their band width is never larger than a few hundred nm and you would need several types of HC-PCF to cover your 1.5-3 micron range.

The spectral area you are trying to address is of great interest to many spectroscopy applications, even further out to 5 micron, since many species have their main vibrational bands in this region, however that also means that there has been a lot of research activity in this area and there already exists a large body of work.

Good luck with your idea !

Is it equipment- or theory - driven question?

Sorry for the vagueness. I meant to use (1500nm to 3000 nm - Broad band LEDs + thermal sources) This is what I have in mind. A source->coupled to a waveguide & a tunable filter-> detector. (Assuming that Spatial coherence is present & I can couple it to a waveguide)

IT is for a proposal. Right now a theory driven question. Are there similar systems already existing?

In general, Information content of spectroscopic measurements doesn't depend on the degree of coherency of the source unless you perform some tricky and very specific experiments that rely apriori on coherence presence. I agree with Jacopo.

I'm sure something similar does exist. I hope that you've thought trough the way of coupling a 3-micron light to a waveguide (multimode, optics etc.). Other than that, the idea may fly for the proposal. But remember, that the devil is in the details:)

At the beginning of scientific investigations more than 200 years ago spectroscopy was carried out using sunlight, then radiation from thermal filaments and only in the last 40-50 years using lasers. Coherent radiation is only really available from a laser source and since the majority of lasers are of a fixed wavelength they are not really suitable for spectroscopy. Except of course tunable diode lasers, dye lasers, OPO's etc.

FTIR spectroscopy is near universally carried out with thermal sources (e.g. glowbar) emitting out to wavelengths of many microns.

Coupling a thermal source such as a filament or arc lamp into a waveguide is possible and is done widely. Similarly, coupling a LED into a waveguide is straight forward. There are no real difficulties there except that it is possibly quite lossy and you wont get a high coupling efficiency because of the usually extended size of your source and the fairly high divergence of a LED emission.

Remember some basic rules of optics,

The product of divergence x spot size is constant,

meaning that if you have a source with inherently high divergence such as a LED, you cant focus it down to a small spot size unless you use a very short focal length lens, however the majority of waveguides have a limited NA and ligt launched into it under a larger angle than the NA is lost very quickly.

You mention coupling to a waveguide and 1.5 to 3 micron?

What material are you thinking of to use in your waveguide?

There aren't many common materials which guide at 3 microns although you can use capillaries with internal reflective coatings.

There are some very bespoke optical fibres (hollow core PCF) which guide in this wavelength range but their band width is never larger than a few hundred nm and you would need several types of HC-PCF to cover your 1.5-3 micron range.

The spectral area you are trying to address is of great interest to many spectroscopy applications, even further out to 5 micron, since many species have their main vibrational bands in this region, however that also means that there has been a lot of research activity in this area and there already exists a large body of work.

Good luck with your idea !

I believe you might find our latest work interesting.

http://dx.doi.org/10.1364/OE.21.024985

To Robert

I am not aware of the possibility of coupling light from a thermal source into a waveguide unless it is highly over-moded (light pipe). If the waveguide is single mode, the efficiency will be to my bet understanding, almost null.

If you have a reference on coupling light from a thermal source into a waveguide I would appreciate it.-Thanks!

Shlomo

You are perfectly correct, the launch will be highly ineffective since you will only be able to launch into the N.A. of the fibre.

However it is possible, and we regularly do this with single mode fibers by using a microscope objective with the appropriate NA (NA for a SMF-28 or equivalent at 1550nm is ~0.12) . We use typically use a x5 or x10.

Sorry no reference, just general lab practice.

source tungsten filament eg. 12V 20W

distance to microscope objective typically 160 to 180 mm depending on tubus length of the objective. Aperture in back of objective ~ 6.5 mm diameter

fibre in focus of microscope lens (focal length typ. 32mm for X5 )

When using x10 we overfill the core of a SMF-28 but the excess radiation outside of the acceptance angle is quickly lost and when we use fibers in excess of 2-3m we don't need to worry about the light in the cladding .

If the fibre is shorter we use a cladding mode stripper to clean up the transported light. (strip jacket and dip fibre in Glycerol)

In some cases when we want to do simple transmission measurements we just use very high NA fibers (e.g. imaging fibres with 50 micron core, 5 micron clad, NA ~0.4) and a large area detector after the sample.

Thanks a lot Robert. Seemingly this Researchgate business is working...

In principle, coherent laser sources are not mandatory for spectroscopy experiments. But it depends on the experiment. In your case, teh source you mention might work, but it depends on the application.