Saturated absorption spectroscopy is a really neat spectroscopy trick used on a vapor cell with hot atoms that creates a transmission plot with only spectral features from the atoms that are moving perpendicular to the laser (v‖ = 0). Reread that

Fig. 1 An illustrative example showing the transmission plots for a two-level atom. On the left is the Doppler profile ... On the right is the transmission plot for a saturated absorption setup.
e use a single laser beam, we would expect to see a Doppler broadened spectrum from these atoms that is Gaussian in shape. Saturated absorption spectroscopy uses two laser beams, resulting in a small Lorentzian feature on top of the Gaussian shape, as shown in Fig. 5.1. The small Lorentzian feature comes only from those atoms that have zero speed in the direction of the laser.
This is how we do it: we send two laser beams into a vapor cell from opposite directions, see Fig.2 The laser beam that starts on the left and moves to the right has a small amount of power. We call this laser the probe beam. In saturated absorption spectroscopy, we measure the transmission of the probe beam. The other laser starts on the right and is moving to the left and has a large amount of power. We will call this laser the pump beam. In most experimental setups, the probe beam and the pump beam originate from the same laser. The laser can be split into two paths using, for example, a λ/2 plate and a polarizing beam splitter, see Sect. 1.5. One would adjust the orientation of the λ/2 plate so that the probe beam has less power than the pump beam.

Fig. 2 Saturated absorption spectroscopy needs two laser beams: a probe beam and a pump beam. Not pictured is the photodiode with which we monitor the transmission of the probe beam.

Fig. 3 An example of a Doppler profile simulated using parameters for a transition in europium-156 atoms with a temperature of 400 Kelvin
To explore how this technique works, we will use our simple two-level atom. If the pump beam were not present, we know the transmission of the probe beam looks like Fig. 3. This plot is calculated using the mass of a europium-156 atom and a vapor cell at 400 Kelvin. The natural linewidth of the transition is about 25 MHz, which is much smaller than the Doppler width of 320 MHz.
Now, let’s add in the pump beam. Our ultimate goal is to determine how the transmission plot of the probe beam changes with the addition of the pump beam. Let’s start by thinking about the transmission of a laser through the vapor cell when the laser frequency is at
, see Fig1. Since both the pump and the probe beam come from the same laser, they have the same frequency. The only difference is that they are moving in opposite directions. We want to ask the question: Which atoms interact with each laser beam?
As a reminder, the Doppler shift for an atom moving with velocity component v‖ is given by the formula:
(1)
where v‖ is negative if the atom is moving towards the laser source and positive if it is moving away.
I find it useful to use numbers, so let’s say that the frequency of the laser is set to
, and
is 200 MHz below
= 652.0000 THz. Since we have two laser beams moving in different directions, we are going to use the magnitude of v‖ and the descriptors “to the left” and “to the right” for the following discussion.
Let’s start with the probe beam, which is moving to the right. Since the frequency of the laser is below resonance, we know that atoms which interact with the probe beam have to be moving to the left with a specific v‖ so that, according to those atoms, the Doppler effect shifts the laser frequency into resonance. That means the atoms would have to be moving at:
(2)
to absorb light from the probe beam.
Now let’s think about the pump beam, which is moving to the left. That means the atoms that absorb light from the pump beam must be moving to the right. The math is the same, but the direction of movement is opposite:
(3)
Spend a few minutes on the above argument to make sure it all makes sense.
Here is the important take home message: When the frequency of the laser is at fA, different atoms interact with the probe beam and the pump beam. Both lasers are losing photons, but they are losing photons to different atoms. Since we are monitoring the probe beam transmission, the probe beam transmission is the same whether the pump beam is on or off. Again, this is a very important concept so make sure it makes sense before moving on.
Your turn! The laser frequency is now at frequency
, which we will assume is 200 MHz higher than
= 652.0000 THz. What velocity does an atom need to have to absorb light from the probe beam? From the pump beam? The answers are in the footnotes.
The conclusion for when
is the same as when
: When the frequency of the laser is at
, the probe beam transmission is the same whether the pump beam is on or off. The trick happens when the laser frequency is at fr. The Doppler shift is 0, so both the probe and the pump beams interact with the same atoms. When the laser frequency is at
(or any frequency except
), the pump and the probe lasers interact with different atoms. When the laser frequency is at
the two laser beams compete for the same atoms.
To explore this more, let’s do a thought experiment. First, we either block or turn off the pump beam so that there is only a probe beam. The probe beam frequency is set to the resonance frequency, and we’ll assume it hits a single atom at rest. Let’s say the probe beam has 10 photons that pass by the atom for every lifetime of the excited state. From those 10 photons, the atom absorbs 1 photon reducing the probe beam transmission to 9 photons; this is a 10% reduction in probe beam transmission. Now we turn the pump beam back on. The pump beam has more power than the probe beam. Let’s say the pump beam provides an additional 990 photons. The atom will randomly pick 1 photon from a possible 1000 photons (10 from the probe and 990 from the pump). Most likely the atom is going pick a photon from the pump beam. Since all 10 photons make it through, the transmission of the probe beam is larger when the pump beam is on. Every once in a while, the atom will randomly absorb from the probe beam, decreasing its transmission percentage. However, the transmission of the probe beam is, on average, larger when the pump beam is present. If we increase the number of atoms in the vapor cell, each atom with v‖ = 0 will randomly absorb from either the pump beam or the probe beam.
The conclusion is: When the laser frequency matches the resonance frequency, the transmission of the probe beam is larger when the pump beam is present.
What is really important is that this only happens for the atoms that have v‖ = 0. For any other velocity, the probe beam transmission is exactly the same whether the pump beam is on or off. Let’s recap all of this in a table. To make things easier, we are going to define “moving to the left” (towards the probe beam) as negative and “moving to the right” as positive.

A simulation of the transmission of the probe beam with the pump beam off (left plot) and with the pump beam on (middle) is shown in Fig. 5.4. If we subtract the two plots (right), we are left with a spectral feature with a full width half maximum equal to the natural linewidth of the transition. This is the same plot as the absorption plot from the thought experiment that we did in Sect. 3.1, which was an absorption plot

Fig. 4 The transmission plots of the probe beam for a 2 level atom with just a probe beam (left), the probe beam and a pump beam (middle), and the difference between the two transmission plots (right)
for atoms at 0 Kelvin. We call this plot a saturated absorption plot. Neat trick, huh!