Tardigrades, carrot carotenoids and the DIY Raman spectrometer (VII)

Newcomers to our DIY Raman spectroscopy experiment are recommended to look back and begin reading at our February- and March issues.

For all the others we are going to give a summary of our experiences with laser long-pass filters. Please keep in mind that the long-pass filter must be able to suppress the (simply reflected) main laser signal as it would overwhelm the spectrum. But it should at the same time let pass the much weaker (information bearing) Raman signals at longer wavelengths. As our green laser is emitting at exactly 532, it might sound a good idea to choose a cutoff wavelength very close to the laser, e.g. 540 or even 535 nm.

But at this point things are becoming tricky: the laser itself has a typical bandwidth of about 1 nm. The filtering process will always be a compromise because cheap 535 nm filters might not be able to cleanly cut out the 532 nm laser signal. This is a worst case scenario as the residual laser radiation will spoil the spectrum completely. On the other hand filters which cut at a greater (safety) distance from the laser will hide some Raman signals close to the laser. For example, a long-pass filter cutting at 550 nm will provide no Raman signals below a Raman shift of ca. 615 cm^-1. As a consequence we would be able to perceive the main Raman signal of Calcite (at a Raman shift of ca. 1082 cm^-1) but not the main signal of Quartz at ca. 458 cm^-1 (cf. fig. 8 and fig. 9, below).

Raman filters which are cutting strictly can be very expensive (in the range of ca. 1,000 US$ and more). But for first experiments a much cheaper laser like the one shown below might be sufficient as well:

Fig. 1: Raman long-pass filter, cutting everything below 550 nm. This is a so-called interference filter, not a much cheaper "optical" filter. It is made by the THORLABS company, model FEL0550.
Price ca. 100 US$.

At the moment when it becomes clear that your experimental setup is actually working you might switch to a more expensive filter. We chose one which cost us roughly 250 US$. The long-pass filter should be positioned between the projection eye-piece and the spectrometer. We simply place it upon the upper lens of the eye-piece. All the following spectra were recorded with the 250 US$ filter in place (please note that there are some more technical hints at the very end of this page):

Fig. 2: Raman spectrum of a carotene capsule from a grocery store. This is an ideal specimen for experimenting and testing your first setups: the carotene Raman bands are exceptionally strong due to a so-called resonance effect. The carotene capsule was cut in half and the carotene applied to a slide, then covered by a regular cover glass. It will stick. But please be careful when working with the grocery store carotene as it might cause a terrible mess on any other surface due to its intense color! When everything is all right you will get the carotene Raman signals, namely ca. 1,006, 1,148 and 1,519 cm^-1. Our setup is not very elaborate, the accuracy range being about +/- 5 Raman shift numbers. Nevertheless we will be able to recognize typical patterns which can be compared with reference spectra from the internet.

Fig. 3: The - classical - Raman spectrum of diamond. A single, very strong band at a Raman shift of ca. 1,330 cm^-1. Commercial spectrometers will deliver this signal within fractions of seconds. The diamonds used here are actually very small, ca. 1 mm. So you will be able to get them cheaply but they are really nice to look at under a dissection microscope. Easy-going practice specimen.

Fig. 4: Raman spectrum of a facetted Zircon. Please note the two strong bands in the Raman shift range between 2,000 and 3,000.

Fig. 5: Easy-going as well - Vanadinite (a lead vanadate), a classical collector's mineral with two Raman bands at about 353 and 825.

Fig. 6: A further mineral - the Crocoite (PbCrO₄). Main bands at Raman shift values of about 350 and 833. This specimen type is recommended on J.M. Derochette's fine web site as well (siehe sources, below).

Fig. 7: Raman spectrum of sulphur, with a strong band at ca. 463 cm^-1.

Fig. 8: Raman spectrum of a Quartz splinter, main band at ca. 458 cm^-1. Some noise in the spectrum. This spectrum might be used to identify Quartz grains from your marine tardigrades' micro aquarium ;-).

Fig. 9: Raman spectrum of a Calcite crystal. Main band at ca. 1,080 cm^-1 (symmetrical stretch of the carbonate ion with its tetrahedron geometry). A further, weaker band can be seen around 750 cm^-1. Some noise in the spectrum as well.

Fig. 10: Raman spectrum of our "Rocharium" (a minimalistic sea-water micro aquarium). This micro aquarium is made from polystyrene. You will be able to compare it with other polystyrene spectra in the internet. The "aromatic" styrene rings are causing a sharp band at ca. 1,000 whereas the aromatic C-H-stretching is the cause for the Raman bands above 3,000.

Technical Supplement

Obviously, it is possibly to spend much more time and money, to work with other lasers or better interference filters. On this pathway you will definitely reach more splendid results (cf. links to more sophisticated projects below).
The spectra shown here were recorded by help of the following means:

-- 540 nm long-pass interference filter (Ebay, producer unknown)
-- Laser cleaning filter (ca. 100 US $) placed in front of the laser pointer
-- Spectrometer from Science Surplus, 1800 lines diffraction grid, 50µm slit
-- Basic software "Spectrum Studio" V. 1.2, came with the spectrometer
-- Measurement times up to about one minute (sum of several runs)
-- Microscope objective: Nikon M Plan 10x/0.25 and sometimes 40x/0.65

Absolute wavenumber fit and wavenumber resolution are relatively low and some spectra are showing strong noise. Nevertheless the spectra can serve as a basis for comparison with internet sources, e.g. with the impressive "RRuff" Raman database (online). When comparing the general appeal of your Raman spectrum with the internet Raman spectra you should be able to find positives (clear fits) and exclusions (clear non-fits). The main advantages of our DIY system are its very low price and the possibility to use it in combination with an incident light microscope.
When using a 100x objective and the field diaphragm of the microscope in order to confine the area under investigation you might reach an analysis spot size as small as 30 µm (as in the case of the Echiniscus tun).
Further potential applications are the identification of gems, in particular in order to check for real diamonds, but also a check for poisonous lead white on your window frames and the identification of synthetic resins.
But here is one serious drawback which must be mentioned: in many cases the green laser might trigger some disturbing fluorescence which as a rule will overwhelm and spoil your spectrum. So there will be many samples out there which cannot be analyzed.

Fig. 11: Some like it big, others like it small. A Raman spectrometer might be cheap or tremendously expensive. The appropriate choice of a spectrometer is depending on many factors. It is up to your brain to decide!

Internet resources
-- The "RRuff" minerals' Raman spectra database
-- An excellent DIY Raman spectroscopy site run by J.M. Derochette
-- and a further high-level amateur site: Uranglasuren.de

© Text, images and video clips by  Martin Mach  (webmaster@baertierchen.de).
The Water Bear web base is a licensed and revised version of the German language monthly magazine  Bärtierchen-Journal . Style and grammar amendments by native speakers are warmly welcomed.

Main Page