Skip to content

Week of August 11th, 2025

General plan

What is the question we are trying to answer? Fundamentally, we want to demonstrate that this synthesis method produces narrow, stable molecules with low inhomogeneous broadening.

  • Scan from 779.5-781.5 nm at 50 nW. Is the inhomogeneous broadening ~100 GHz? Are there a good fraction of molecules around 60 MHz wide?
  • If so, pump the crystal with ~ 10 mW for 10 minutes to shift things around and hopefully stabilize them a little bit.
  • Select a few of the brightest molecules that are close to 60 MHz and are not in extremely dense regions. Go to each one and park the laser on it. Watch the counts on the APD. Try to get an idea of how stable each one is.
  • Find the one that seems the most stable. Now you have a bright, stable, narrow molecule, and the idea is to characterize it.

Characterization

  1. Saturation curve

You want to go from about 1 nW to 1 uW in 11 steps at least (1, 2, 4, 8, 16, 32, ...). Of course there is no point in measuring a power if you see no signal at all so you can start at the lowest power that gives you a lineshape you can measure. For each power, scan over the molecule. 10 GHz should be a fine scan width. I would recommend scanning 200 or 500 MHz/sec. If it is very stable you can go even slower. Use the lowest dark count APDs we have and get confident that your background is limited by the dark counts. So you should have 50 dcps when you block the beam. You can use ND filters for power control if you like, although they are cumbersome. Another option is to put a HWP before the fiber coupling on the laser side of the table. Because there is a PBS in the fiber launch, rotation of the HWP converts to power. This should give you 30-40 dB of control. It's much easier to get a lot of points. You can use an ND filter for the lower end of the power spectrum of course. It is nice to have a beam pickoff and powermeter in the setup and attached to the computer. With the spectroscopy setup, you can put the powermeter on one side of the 90:10 beamsplitter. That way you have plenty of signal to be confident in the relative power going into the objective. (1 nW in the objective is 10 nW on the powermeter, which is above the noise of the powermeter if the doors are closed). I would also recommend installing a PBS after the triplet collimator on the spectroscopy setup if there is not one already. You want to be confident that your polarization is stable. Also keep an eye on your power fluctuation. If it is more than a few percent we will have to stabilize that as well some how.

2. Zero-power linewidth

At this point I would recommend doing some brief data analysis. Fit a Lorentzian to each lineshape from (1). Plot the height of the Lorentzians as a function of power and the width. Examine the data and see how noisy it is. If it is very clean and fits to the theoretical equations, then we can proceed. Otherwise we will have to troubleshoot. If the fits are clean, then the linewidth vs. power plot will give you the zero-power linewidth. It would be best if this were <=50 MHz. If it is >60 MHz, we may want to consider finding another molecule or trouble-shooting the sample more.

3. Stability

I know you have a lot of stability measurements, but if it is not too much trouble and you already have the code, I would set up a scan (1-5 GHz at 1 GHz/sec) and scan repeatedly over the molecule for 1-2 hours. We will average this data and fit a Voigt profile to get the Lorentzian component of the broadening and the gaussian component. This will give us a very precise measurement of the spectral stability.