Erickson_ Christopher

							                     Erickson, Christopher
      High Temperature Vapor Cell for Calcium Absorption
                        Spectroscopy
            Faculty Mentor: Dallin Durfee, Physics & Astronomy Department

We have designed, built, and troubleshot a unique calcium vapor cell for use in the
development of a dual-species optical frequency standard with stability on the order of
one part in 10-16. This vapor cell’s unique design and high temperature allow for the
creation of a high-density calcium vapor for spectroscopy on the 1So – 3P1
intercombination line in calcium. The relative uncertainty of a laser wavelength locked to
the 1So – 3P1 transition in calcium can reach one part in 10-13 and can then be used in
tuning a second laser of a narrower line width to create the atomic clock.1

Using a simple vapor cell for this transition, even with a saturated absorption technique,
is not plausible due to low vapor pressures at temperatures below 400 C. At higher
temperatures other problems intervene including window coating, calcium loss, and
Doppler broadening. The use of a buffer gas to prevent window coating subsequently
introduces pressure broadening. Two groups have attempted to use a thermal calcium
beam in an effort to side step these problems, but have seen low signal to noise ratios.2,3
A third group developed a high temperature vapor cell at 400 C by allowing calcium to
coat mirrors inside the cell until they became calcium mirrors.1 They saw similar signal
to noise ratios as the thermal beam experiments, and their cell runs for only 3 days before
the calcium needs to be refilled.

Our vapor cell consists of three vacuum chambers designed to isolate hot calcium from
room temperature windows. The calcium chamber is separated from the outer view ports
by means of sapphire windows in clamps we designed. Bands heaters fit around the
windows to prevent them from getting coated at high temperatures and contain the
calcium. Originally, bellows were supposed to provide a valve-like opening to the
calcium chamber for pumping out. They allowed for the escape of calcium from the
chamber due to a weak metal to metal seal. At high temperatures calcium also managed
to escape past the sapphire windows. We modified the cell to eliminate internal leaks and
allow for an overall higher running temperature. The fix included replacing the bellows
with a pump out tube later sealed with silver solder, and adding copper gaskets to the
sapphire windows.

The modifications improved performance of the cell; however, we could only take data
for a week before calcium loss through the windows emptied the cell. Initial data sets
show high enough calcium density for 55% absorption of the laser light on resonance
with a full width at half max of about 2GHz and band heaters running at 600 C. These
initial sets also boasted large signal to noise ratios. After a week of taking data we found
a maximum absorption of about 10-15% with the band heaters running at 800 C and a
lower signal to noise ratio. The data also showed significant etalon effects in our setup.
We minimized these by angling the polarizer and filters in the laser beam, however, the
larger effects appeared to be inside the vapor cell and they are corrected for by a
parabolic fit in our data sets. Figure 1 shows data taken at the point calcium loss started to
become significant. A final problem became apparent when we opened the cell to refill
the calcium: calcium had reacted with the sapphire windows and had permanently etched
them, limiting our signal to noise ratio.

By designing this vapor cell with careful attention to the issues of window coating,
calcium loss, and the maximum operational temperature it can provide a much larger
signal to noise ratio than has been achieved. However, the issue of rapid calcium loss
and calcium reaction with the chamber windows has led us to further redesigns that have
not yet been tested. Currently we are in the process of replacing the sapphire windows
with calcium fluoride windows and adding beveled copper gaskets in order to create a
better seal. Calcium fluoride windows should react extremely slowly with neutral calcium
atoms even at high temperatures and beveled gaskets allow for a better seal against minor
surface defects in the window clamps. We expect that cell performance in terms of
calcium loss will be 100 times better with this new seal and allow the cell to operate well
for a year or more depending on usage.

Once running again, this cell will be used in heterodyne detection to calibrate a second
laser locked to a high finesse cavity. The extremely narrow line width of this second laser
will allow for the low partial frequency instability of a dual-species atomic clock. Future
plans for the cell also include laser-cooling applications.

Figure 1 – Absorption scan through the vapor cell. Band heaters were at 700 C. The difference in width
between the data and the simulation is evidence of calcium loss occurring in the cell, which has the same
effect as a lower cell temperature. Data before this set fit exceptionally well with the simulated points and
increasingly worse in sets taken later. By adjusting the effective simulation temperature we were able to fit
later sets equally well however.




1
  Mao-Sheng Huang, Mao-Hong Lu, and Jow-Tsong Shy, Rev. Sci. Instrum. 73, 3747 (2002).
2
  P. Kersten, F. Mensing, U. Sterr, and F. Riehle, Appl. Phys. B: Lasers Opt. 68, 27 (1999).
3
  A. S. Zibrov, R. W. Fox, R. Ellingsen, C. S. Weimer, V. L. Velichansky, G. M. Tino, and L. Hollberg,
Appl. Phys. B: Lasers Opt. 59, 327 (1994).