Published: 2024-10-31 | Categories: [»] Engineering, [»] Opticsand[»] Chemistry.

In my post on [»] absorbance spectroscopy I explained that a molecule could absorb an incoming photon of specific energy to promote its electrons to higher energy orbitals. This resulted in an absorbance spectroscopy spectrum which would tell us about the symmetry properties of molecular orbitals of the molecule under analysis. I, however, did not tell what happened to the electrons after they were promoted to higher energy levels. This is precisely the starting point of this post!

Just like with any other physical system, the promoted electrons will tend to lose their excess energy to return to their ground state. It’s not actually that they “want” to lose it, but they are just exchanging energy through other processes which have higher entropies – making them irreversible.

In Figure 1, I present the energy levels of a fictional molecule along with their transitions in a modified [∞] Jablonski diagram. The diagram presented here is highly simplified, and a fictional spectrum is also appended below the diagram to illustrate the photon-matter interactions.

Figure 1 – Modified Jablonski diagram of a fictional molecule

Among the different transitions shown in Figure 1 we can spot:

- The different electronic states Si themselves subdivided into vibrational states Vi. The index 0 refers to the ground state of either the molecule (S0), or the specific electronic state (V0).

- The absorption of a photon of high energy in the low-wavelengths UV region, promoting an electron from the ground state S0 to the excited state S2. In a Jablonski diagram, this is shown by a straight arrow. Here, I’m using dashed arrows to represent excitation/absorption processes.

- The vibrational relaxation of this absorbed photon from vibrational state V1 to the ground state. Vibrational relaxation is a very efficient relaxation mechanism in which the molecule loose energy by collision with surrounding molecules. Vibrational relaxation is so efficient that it usually happens before any other relaxation process.

- The internal conversion of electronic state S2 to S1. Internal conversion can occur between any electronic states but is favored when there is an overlap between the higher electronic state and the lower electronic state vibrational levels, as shown here with S2 and S1.

- The absorption of a photon of mid energy in the high UV region, promoting an electron from the ground state S0 to the excited state S1. Absorption can promote an electron to any vibrational state but may have different transition probabilities, represented here by various thicknesses of the line. The different absorption energies are translated into an absorption (excitation) spectrum at the bottom of Figure 1.

- Various relaxation processes inside electronic states S1 and S0.

- The emission of photons which makes the molecule go from the ground vibrational state of S1 to various vibrational states S0. It is worth noting that because vibrational relaxation is so efficient in releasing energy, it is much quicker than emission processes and all emissions occur from the vibrational ground state of S1. Similarly, absorption/excitation occurs at the vibrational ground state of S0.

In this post, we will focus our attention on emitted photons. It’s worth noting that this process is far less likely to occur than other processes of relaxation because it is slower and the probability of transition is given by the ratio of the emission process rate with the sum of all relaxation process rates. As mentioned above, the emission of photons is in competition with internal conversion processes and won’t occur between electronic states that share vibrational energy levels because internal conversion will usually be favored.

Something interesting in Figure 1 is that the energy of the emitted photon can be lower than the energy of the absorbed photon due to the various relaxation processes involved, and because any vibrational states of the lower electronic energy state can be reached. In our fictional example of Figure 1, this translates to photons absorbed in the UV-blue region of the spectrum, and emission in the green-red region. The emission process shown in Figure 1 is also independent on the initial excitation wavelength because S2 will decay into S1 through internal conversion. The fictional example of Figure 1 is actually representative of the behavior of quinine which has two absorption peaks at 250 nm and 350 nm, but emits at 450 nm independently of the incoming photon energy in the absorption process.

When a molecule absorbs photons of a specific energy and emits photons of a lower energy (higher wavelength), we talk about fluorescence of the molecule. There is a different type of fluorescence, known as phosphorescence, which involves longer reaction times, but the overall process is about the same. Finaly, because the same vibrational states differences are involved in the absorption and emission processes, it is common that the emission spectrum is a mirror image of the absorption spectrum (which is illustrated here in Figure 1 as well).

Fluorescence is an important spectroscopy technique, and it therefore deserves to have its own experimental setup on the website which is shown in Figure 2. The setup CAD files can be downloaded for reproduction under a CERN OHL V2 license [∞] here. The CAD also contains a cover, not shown in Figure 2, and which must be set in place before using the recommended 365 nm UV light-source. It is theoretically possible to use different light-sources, but I found the 365 nm LED from Thorlabs to be a good compromise between price and excitation properties. Note that fluorescence rarely occurs with excitation below 200 nm because these sources are powerful enough to break chemical bonds, leading to a release of energy which prevents relaxation through fluorescence.

Figure 2 – Experimental setup

The setup works using a back-scattering implementation where the UV light-source is collimated using a 400 µm solarization resistant fiber and a doublet achromat lens (right hand-side of Figure 2). The light is focused again inside the experimental cuvette using another doublet achromat which also serve the purpose of re-collimating the emitted fluorescence signal (left hand-side of Figure 2). This signal is then split using the cube beamsplitter and re-imaged on the analysis fiber using a third doublet achromat (top of Figure 2). Since the fluorescence can be fairly low for some molecules, it is necessary to clean up the Rayleigh back-scattered UV signal using a high-pass filter placed between the beamsplitter cube and the third doublet achromat lens. A XY translation stage is used to allow centering of the two fibers which can be done using a camera as detailed in the assembly procedure provided with the CAD files.

I however ran into experimental issues with the setup of Figure 2 as I noticed some fluorescence of the setup itself which can be seen in Figure 3. The quick falloff below 400 nm is due to the high-pass filter FEL0400 that I used to clean up the 365 nm source. I tried to identify the origin of this background fluorescence without much success but it’s clearly not due to the fiber themselves and does not seem to be due to the beamsplitter cube or the FEL0400 filter. After contacting Thorlabs about this issue, they mentioned suspecting the AC127-019-A lenses which are made of N-BAF10 and N-SF6HT and recommended replacing them with N-BK7 lenses. I did not test this solution due to a lack of budget, but you can try replacing the AC127-019-A lenses with LA1074-A plano-convexes N-BK7 lenses. Here, I used a complex background subtraction method to remove this background fluorescence from the measurements, but it was clearly not optimal.

Figure 3 – Background fluorescence of the setup

I tested the setup of Figure 2 using three common fluorescent molecules in water solution: quinine (Figure 4), fluoresceine (Figure 5), and rhodamine (Figure 6) which fluoresces in the blue, green and red region respectively when illuminated with the 365 nm UV source. Sourcing the three molecules is not very difficult but you should be aware that they are extremely potent dyes and only a few milligrams is already enough for our experiments. Quinine is the most straightforward to get because it is used in some beverages like Schweppes [∞] tonic water. I found rhodamine the trickiest to handle as it formed a mist as I poured it from its original container bag into a PE bottle and contaminated my lab (and my clothes, and my nose) despite being very careful. I would therefore recommend purchasing liquid solutions directly if possible.

Measuring the samples is relatively easy with the cover in place. The exposure time used for the results presented here was 5 seconds although it depends on the concentration of your samples. Since the concentration of quinine in tonic water was fixed, I adapted the dilution of the others to match the fluorescence levels of quinine. As explained above, the spectra were polluted with background fluorescence that I removed from a pure water measurement at the same exposure time and by subtracting the water fluorescence spectrum to the others. To avoid any scaling issues, I normalized all spectra to the background fluorescence peak height near 400 nm before subtraction.

Figure 4 – Quinine fluorescence spectrum
Figure 5 – Fluoresceine fluorescence spectrum
Figure 6 – Rhodamine fluorescence spectrum

Because the spectra of Figure 4, Figure 5, and Figure 6 all represent visible colors, I decided to plot them on the xyY chromaticity diagram of Figure 7 as discussed in my former post on [»] color theory. In addition to the experimental chromaticity, shown as circles in Figure 7, I also included the dominant wavelength which is found as the intersection of the line joining the neutral point E (1/3, 1/3, 1) to the experimental point with the monochromatic locus. The chromatic purity, ρ, represents the saturation of the color and is found as the distance of the experimental point to the monochromatic locus from the E point.

In absorption experiments, the neutral point is often chosen as the color of the illumination source but with emitted light, it is common to use the E neutral point. It is interesting to note that the neutral point of the sRGB standard is not E but (0.3127, 0.3290, 1), reason for which the chosen neutral point in Figure 7 does not appear white.

Figure 7 – Quinine, fluoresceine and rhodamine shown on a chromaticity diagram

If you have read my post on [»] chromaticity diagrams, you may have spot that the three colors of quinine, fluoresceine and rhodamine fluorescences form a gamut about as large as the sRGB standard. A gamut is an addressable color space formed by a set of primary colors where each color comprised in its convex hull can be generated by a proper mixture of the primaries.

As an experiment, I therefore decided to try different mixtures of quinine, fluoresceine and rhodamine and reported them in a chromaticity diagram. The results are shown in Figure 8 along with the color space boundaries formed by the pure compounds. Obtaining a desired color was trickier than initially expected but I managed to create a mix that had a nearly white fluorescence signal. You may notice that some points in Figure 8 falls outside of the gamut which is impossible is standard color theory. I doubt this is a chemical interaction process between the various compounds, although it’s still a possibility, and I think it’s an artefact of the background subtraction process. Nonetheless, it was a “fun” experiment to try out which I haven’t heard of previously and which illustrates an interesting link between spectroscopy and color science.

Note that the phenomenon of having a white fluorescence from a mixture of multiple fluorescence signals is not uncommon at all because it’s the exact process used to produce white light in fluorescent tubes! These tubes actually generate UV which is turned into white light by the fluorescence of many different metal compounds placed on the inner surface of the tube. The difference with the experiment of Figure 8 is that we are using liquid solutions here whereas fluorescent tubes work using solid compounds. It is also easier and safer to reproduce the experiment of Figure 8 with students than messing up with fluorescence tubes (don’t do it).

Figure 8 – Various mixtures of quinine, fluoresceine and rhodamine shown on a chromaticity diagram

If you don’t have access to the chemicals used for the experiment of Figure 7 and Figure 8, you can still measure fluorescence very easily with fluorescent pens. In a second experiment, I bought a set of cheap fluorescent pens, shown in Figure 9, from my local grocery store. Fluorescent pens work through the deposit of fluorescent compounds on your sheet of paper that will fluoresce with ambient light. Because the compound will emit light, it won’t mask the text behind it – reason for which it made the quick popularity of these pens in office work.

Figure 9 – Cheap fluorescent pens bought from a local grocery store

To extract the fluorescent compounds of the pens, just dip the tip of the pens in water for a few minutes. This will drain the compounds out of the tip and replace it with water. After this operation, the pen will still hold a lot of fluorescent compounds in its reservoir, but the tip will be saturated with water and the pen won’t work anymore. To drain the water out and refill the tip with fluorescent compound, just press the tip for a minute on a soft tissue paper and let it sit for a few more minutes after. You can reproduce the experiment many times with only one set of pens, which makes it very suitable for usage with students, too.

Out of the set of Figure 9, I was able to measure all of the fluorescent compounds but the blue one. This can be explained by either the fluorescence background of Figure 3, a much weaker signal for the blue pen, or by the fact that the blue pen might not be fluorescent at all.

The results are plotted in Figure 10 on a chromaticity diagram. Although I have no information on the actual chemicals used for fluorescence, we can spot a few things on the graph. First, despite the colors of Figure 9 takes a full rainbow range, the results are clustered in two different regions: green and red. The green pen is actually blue-green, the yellow one peaks at green dominant wavelength and the pink, orange and red ones are all in the orange-red region of the chart. The spectral purity of the green and yellow pens are relatively low and might form a line leading to pure fluoresceine (check on Figure 7) which make me think that they might be a mix of fluoresceine with another chemical in the blue region of the chromaticity chart but I don’t have any other evidence to support that claim.

Figure 10 – Fluorescent pens of Figure 9 shown on a chromaticity diagram

A lot more can be said about fluorescence and its brother phosphorescence that I have not discussed here yet. But I wanted to keep this post short and focus on the experimental aspects as well as the setup that can be used to measure fluorescence. Don’t hesitate to share your comments on the [∞] community board to let me know if you enjoyed the post :)

I would also like to give a big thanks to Young, Sebastian, Alex, Stephen, Lilith, James, Jesse, Jon, Cory, Karel, Sivaraman, Samy, David, Themulticaster, Michael, Shaun, Tayyab, Kausban, Kirk, Marcel, Onur, Dennis, Benjamin, M, Sunanda and Natan who have supported this post through [∞] Patreon. I also take the occasion to invite you to donate through Patreon, even as little as $1. I cannot stress it more, you can really help me to post more content and make more experiments!

[⇈] Top of Page

You may also like:

[»] In-line Absorption and Fluorescence Sensor

[»] DIY Raman Spectroscopy

[»] The Perception of Colors

[»] Absorbance Spectroscopy

[»] Dynamic Range Analysis of a 350-700nm Spectrometer