I have just arrived to start a postdoc at the Smithsonian Institution Naitional Museum of Natural History. One of the greatest places on Earth. No pressure…
I have arrived in Washington,DC to start a new postdoc, but more on that later. One of my first hurdles will be a spot inspection of my Raman spectroscopy knowledge. Somewhere, at sometime, I have been promised a quiz. “So, tell me what you know about Raman spectroscopy”. Very few people look forward to tests, and we all approach different tasks with different levels of anxiety. For some people, working in a new lab with an exam looming over you might seem like dining at Damocles’ table.
Well, please allow me to prepare my answer while I rearrange the cutlery.
There are three possible interactions of light with matter: light can be transmitted straight through, light can be absorbed, or light can be scattered. Raman spectroscopy is concerned with the latter, and is generally achieved with a mechanically simple experiment. 1) A sample is placed in the path of a monochromatic laser, 2) the sample and light interact, 3) scattered light is detected, and 4) a spectrum is collated. This description, of course, is like saying “magic happens here” repeatedly from steps one through four, so let us go into a bit more detail. I have picked up the following information from Smith and Dent (2005), which is the most user friendly text I have ever read on Raman spectroscopy.
1) Light source
The use of a monochromatic laser is very important. Monochromatic means that the laser is producing a single colour, which in turn means that the photons of light hitting the sample have a single wavelength (λ), and hence a single energy value (E = hcλ-1). The end goal is to monitor how the sample changes the laser energy, so it is important to have a single, precise wavelength to start with.
2) Light interaction
How do these changes occur? As brushed past above, an incident photon can have three types of interaction with matter. For Raman spectroscopy, we are not interested in photons that are transmitted or absorbed (although we need to be aware of them). Instead, we are interested in a third type of interaction called “scattering”. A good working definition for scattering is “…a deviation in the trajectory of a particle caused by an interaction with some heterogeneity in the medium it is travelling through…” Consider a photon produced within, and fired from, a laser. The photon travels along a straight path and hits our sample. This particular photon has struck mass, so it won’t be transmitted through the sample, and it does not have that very specific energy value that corresponds to the difference between ground and excited states of the molecule, so it won’t be absorbed. Instead, the photon and electron cloud meet and briefly coalesce into a complex of energy and matter, promoting the molecule into a “virtually excited state”.
There are three fates for a molecule in a virtually excited state. First, and most commonly, the energy will exit the complex having only deformed (i.e. polarised) the electron cloud. Electrons have very little mass, so the polarisation requires an insubstantial amount of energy from the photon. The photon will then depart the sample with pretty much the same amount energy it arrived with. This type of interaction is called Rayleigh scattering, and the insubstantial loss of energy means it is a type of elastic scattering.
Instead of simply deforming the energy cloud, the energy in the virtually excited complex might be used to move nuclei in the molecule. This particular fate is very rare, with national lottery odds (one in 106 to 108 incident photons). Once moving, a nucleus in a molecule will run on invisible rails – like the world’s simplest rollercoaster. The specific movements of nuclei (called vibrational modes) have different effects on the shape of a molecule. From the virtually excited geometry, the nuclear rollercoaster will deform (i.e. polarise) the shape of the molecule according to a specific vibration mode. The type of atomic nucleus (e.g. carbon, nitrogen, phosphorus), the bond strength between the nuclei, and the specific path that the nucleus takes each influence how much energy is subtracted from the virtually excited complex. The photon (temporarily interacting with the molecule as a virtually excited complex) will eventually scatter away, having lost the energy required to facilitate a vibrational mode. This is called Stokes scattering, and the change in energy means it is a type of inelastic scattering.
Alternatively, a highly excited molecule might give energy to the incident photon. This third fate of the virtually excited state is the rarest of all. The gain in energy means it is also a type of inelastic scattering, and it is called anti-Stokes scattering.
3) Light detection
Photons that have been inelastically scattered are carrying information about a sample. This information is in the form of lost or gained energy. In order for that information to be interpreted, it must first be detected. A notch filter, a diffraction grating and a detector are placed in the path of the scattered light. A notch filter is a carefully engineered crystal (i.e. photonic crystal) that absorbs the frequency of the incident laser light (and Rayleigh scattered light, which has the same frequency). From Smith and Dent (2005), “…a filter which collects most of the light within 200 cm-1 of the excitation frequency is regarded as sufficient…” Wavenumbers (cm-1) are a unit of energy measurement, and hence a notch filter in a Raman system is designed to “block” the incident laser light and light with very similar energies. The reason for this lies in the rarity of inelastic scattering: very few of the photons emitted from the laser end up being inelastically scattered, and they could easily be lost in the torrent of elastically scattered photons.
After passing through the notch filter, the photons must then travel through a diffraction grating (I am focussing on dispersive, and not Fourier Transform instruments). A diffraction grating separates light of different energies, turning a narrow beam of mixed energies into a broad beam with a range of energies. Like a prism splitting white light into a rainbow.
The diffracted beam is then channelled towards a detector. Most detectors used in dispersive Raman spectrometers are highly sensitive charge coupled devices (CCDs), like in digital cameras. Which makes sense: the CCD in a camera is used to detect light with different energies, which are in turn used to produce a map of different colours (i.e. a photograph). The CCD in a Raman spectrometer also detects light with different energies, but the energy differences were caused by scattered, rather than absorbed and reflected light (mostly).
4) Light measurement
The final step in Raman spectrometry is interpreting the detected photons. Because we used a monochromatic laser, each of the photons had the same energy when they were emitted. The energy of the photon changed while it was interacting with the sample – energy was subtracted to generate a vibrational mode in ground state molecules, and gained from vibrational modes in excited state molecules. The photons were inelastically scattered away from the sample with a different energy value than when they started with. The scattered photons eventually meet a detector, which determines the energy value of that photon. By knowing the energy that the photon started with, and the energy that the photon finished with, it is a simple matter to calculate the energy change that took place during the interaction with the molecule. Each of these energy values are collated during the Raman experiment, building up a spectrum of energy differences. Given that vibrational modes and atomic nuclei control the energy taken from or given to an inelastically scattered photon, we can use the Raman spectrum to interpret the chemistry of the sample.
It’s really that simple.
Smith, E. and Dent, G. 2005. Modern Raman Spectroscopy, a practical approach. John Wiley and Sons (Chichester).