Bouncing Light to See the Invisible

Kate Snyder
Department of Chemistry
Stanford University
March 2002

If you hold a glass of apple juice up to a light, the light appears dimmer because some of the molecules in the juice are absorbing the light. Some wavelengths of light are absorbed more than others so the solution appears colored. If you dilute the apple juice by putting a small amount of it in water, you might not be able to see its color anymore. This does not mean it's no longer absorbing light, it is just absorbing so little that it is hard for your eye to see it. In my research, I try to measure the number of molecules that absorb light (absorbers) in a solution so dilute it appears invisible. Accurately measuring the amount of absorbers in a dilute solution is vital in many fields of science. For instance, environmental chemists may measure the amount of pollutants in the water runoff from a factory. Forensic scientists may measure small amounts of toxins in the blood.

The most commonly used technique in science for determining the amount of an absorber in solution is called "absorption spectroscopy." Traditionally, this technique is performed by shining light through a reference container, which holds just the solvent, and also through a sample container, which holds both the solvent and absorber. The sample container will absorb more light than the reference container. The difference in the amount of light absorbed is related to the amount of absorber present. The problem with this technique is that light sources, even lasers, flicker slightly so you end up with an error in your measurement. This error prevents detection of very small amounts. Generally speaking, if your eye can't tell a color difference between the reference and sample, then neither can this technique.

To solve this problem and measure the absorption of a very dilute solution, I use a variation on a technique called Cavity Ring-Down Spectroscopy (CRDS). CRDS involves the use of two highly reflective mirrors facing each other (optical cavity) and a light detector after the mirrors. A laser pulse hits the back of the first mirror and a small amount of light makes it through the mirror coating. This light continues along until it hits the other mirror where most of it is reflected back, but some of it goes through and hits the detector. The light continues bouncing back and forth between the mirrors, letting a tiny amount of light through each mirror each time it hits. The detector sees a smaller amount of light after each bounce. The length of time (decay time) it takes for all the light to exit the optical cavity is determined by how well the mirrors can reflect and whether or not molecules in the cavity are absorbing some of the light. Thus, the decay time of an empty cavity will be longer than that of a cavity filled with absorbers. The difference in these decay times tells us the amount of absorbers present. In CRDS it doesn't matter if the light source flickers, and since the light goes through the sample many times, we can detect smaller amounts of absorbers.

Although CRDS sounds perfect for measuring absorbers in a dilute solution, it has only been applied to measuring gases. This is because if you place a glass container full of liquid between the mirrors the light is lost too quickly to determine the decay time. The light goes away so quickly because each surface of a glass container reflects 4% of the light and the beam goes through four surfaces at each pass. So after a few bounces all the light has left the cavity. Ideally, we determine the decay time based on hundreds or thousands of bounces. I'm working to develop a modified container that doesn't reflect light so we can measure liquid samples. From physics you can calculate a specific angle for each surface that minimizes the reflection of light. By adjusting the type of light I let through, I can make the light pass through the container without any reflection at all. Essentially the container is made invisible, but the absorbers in the liquid are still detected. With this improvement, smaller amounts of material can be detected than with the traditional method, thus improving the results of any user of absorption spectroscopy.