The final method we attempted for measuring the Zeeman effect was via spectrometry. We had initially been using this spectrometer, throughout each of the methods, to do things like verify the quality of our filters and verify the strength of specific spectral lines. From those uses, came the idea that perhaps the spectrometer could be another way to visualize the Zeeman effect.
We used the spectrometer on intensity gathering mode, meaning pointing the fiber optic probe at the atom discharge source could illuminate which wavelengths were being emitted, and at which intensities.
We observed the 546nm line of Mercury once again, and utilized a narrow wavelength green filter to cut out all intensities other than the line at 546nm. Note that we did not use the wide green as we did when imaging mercury. This is because the narrow green filter made the light intensity too dim for the camera, but did not pose the same issue for the more sensitive spectrometer. We can see how attenuating the filter is, as in the graph below, the 546 line intensity is completely oversaturated with no filter, while with the filter, it has a smaller intensity than even the smallest no-filter peaks.
We then compared two spectrometry measurements of the narrowly filtered mercury lamp with and without the external magnetic field. The measurement taken with the magnetic field was taken at maximum magnetic field strength. This correlated with a field strength of ~1.6A, or ~10,000 gauss.
We initially took this as a promising sign. Under the assumption that perfectly resolved lines would be too fine to view (hardware limitations, doppler broadening), we thought at least that we’d maybe be able to observe peak widening due to the generation of new peaks next to the original one. This graph seemed to indicate widening, but confusingly it also grew in intensity. Energy line splitting theoretically should not contribute to more intense peaks, as energy is moreso being “redistributed”. We took this as a sign to investigate the actual shape of the peak by rescaling with intensity. If the peak was actually wider in shape, when the intensity was rescaled we would observe the same shape in both situations.
Unfortunately, that is exactly what we observed. From this graph, we concluded that the spectrometer was indeed not fine enough to be able to observe even peak widening in the presence of a magnetic field.
This encouraged us to attempt to quantify the resolution limit of the spectrometer. If we want to be able to observe the Zeeman effect, we want the smallest resolvable wavelength to be:
Δλmin =< (hc/En) – (hc/En+ΔE)
If we apply this to the mercury line of interest, for hc/En = 546 nm, gΔmj = 0.5 to 2 (differs for different transitions), and an applied magnetic field of B ~ 10,000 G, the required resolution is anywhere from 0.005 to 0.02 nm.
Our OceanOptics USB4000 spectrometer has a resolution of about 0.3 nm, hence it is about 15x too coarsely grained to resolve even the largest splits of the anomalous Zeeman effect.
If this method were to be adapted, we believe that improved resolution would be possible by upgrading the hardware to a smaller diffraction grating slit.
Leave a Reply