Spectroscopy on trapped ions to probe a possible variation of alpha
The remarkable precision of frequency comb lasers can be used to see if the constants of nature are really contant. Within our group we have several experiments on this topic. E.g. experiments concentrating on the finestructure contant and the ratio between proton and electron mass. In another experiment a molecular fountain is constructed for ammonia to test if the proton/electron mass ratio is changing in the present epoch.
The experiment described here is searching for a variation in the fine structure contstant alpha on a cosmological timescale (so billions of years). We do this by calibrating transitions of ions in the UV spectral region. The same transitions (but then strongly redshifted) can be observed with telescopes as absorption lines in the light from 'quasars' (very bright objects). With a telescope you effectively look back in time. It can be billions of years, depending on the distance of the absorbing cloud. If alpha would change over this time, then some transitions would shift up in frequeny, while others would go down or stay the same. By comparing transitions from many different ions, one can try to see if alpha has changed. Some experiments claim to have seen a change [1,2] in alpha in this way on a cosmological timescale. A different way to see if alpha is changing now, is to perform ultra-high precision spectroscopy and repeat the experiment on a timescale of a year. In this way one can see if alpha is changing in the present epoch. The latest experiment on this sees no change on a level of 10-17 . Clearly more and accurate measurements are needed over different timescales to sort out this issue.
In our case we want to compare the present positions of the lines with those a few billion years ago. In the picture above made at ESO, you see an example of a part of the sky where absorption measurements were performed. On the right is a typical absorption measurement that was taken with the same telescope Most of the absorption lines you see are actually from atomic and molecular hydrogen. The reason why there are so many lines is that there are many objects with different speeds in the line of sight of the quasar, so that the same spectral lines appears many times in the spectrum at different places due to Doppler shifts. However, at somewhat longer wavelengths careful analysis of the data shows that there are also other absorption lines by many atoms and molecules. The laboratry ('earth value') of these lines are not always accurately known yet to be included in the analysis if alpha has changed.
We have set up an experiment with an ion-trap to calibrate some of the transitions so that they can be included in the analysis. The first element we have done is calcium. Calcium+ ions can be stored in a so-called Paul trap. Such a trap is very convienent as it localizes the ions and allows to perform laser cooling on them. Below is a schematic 3D picture of a linear Paul trap. On the right you see a photo of the real thing taken through one of the vacuum viewports.
The present trap consists of four rods and two rings. Two rods are on ground, while the other two (on opposite sides) have an AC voltage of a few hundred volts at a few MHz. The rings have a static voltage of a few volts positive. Ions (positive) in this construction are trapped along the axis by the rings due to the positive voltage on them. Radially the ions will see a 'pseudo potential' from the alternating voltage. This only works for a limited range of frequencies and voltages, but it can confine the ions if properly set. Laser cooling is then employed at 397 nm to make the ions cold. Without cooling the ions heat up in the trap due to 'micromotion' from the AC field in combination with the Coulomb interaction between the ions. Below is a figure of the excitation and cooling scheme we employ. The laser at 866 nm is used for re-pumping the atoms back to the ground state in case they get stuck in the 3d2D3/2 state. On the right you can see what happens if you monitor the fluorescence while scanning the 397 nm cooling laser over the resonance. If the laser cooling is not sufficient then you will see a broad curve (red). With sufficient cooling to compensate micromotion heating, a phase transition can take place so that a 'ion crystal' is formed. In this case the temperature drops dramatically (to 1 K or less, blue curve), which is very useful for precision spectroscopy.
We have calibrated the same transition as the cooling transition by sending in a very weak (mircowatts) beam as a probe. A schematic of the setup is given below. By calibrating the probe laser with a frequency comb, we were able to calibrate the transition with an accuracy of 1.7 MHz. This is more than an order of magnitude better than achieved before, and certainly sufficient to compare with telecope observations.
The actual resonance line looks like this:
An article about this experiment has been accepted recently:
Frequency metrology on the 4s 2S1/2 - 4p 2P1/2 transition in 40Ca+ for a comparison with quasar data,
A.L. Wolf, S.A. van den Berg, C. Gohle, E.J. Salumbides, W. Ubachs, and K.S.E. Eikema,
To appear in Phys. Rev. A.
Below is a picture of the Calcium-ion experiment and the PhD student who is working on this: Anne Lisa Wolf.
 J.K. Webb et al., Phys. Rev. Lett. 82, 884-887 (1999)
 M.T. Murphy, J.K. Webb, V.V. Flambaum, Mon. Not. R. Astron. Soc. 345, 609-638 (2003)
 T. Rosenband et al., Science 319, 1808-1812 (2008)
This project is a collaboration with the Dutch standards institute (NMI), and part of a FOM metrology program.
Questions? Contact: K.S.E.Eikema@vu.nl