Success!

Although today (Friday) was one of the more eventful days with a cocktail party in the afternoon and a going-away party later in the day, we managed to finally see a change in THz signal when blocking/unblocking the excitation beam.

We switched again to the spectroscopy microscope and arranged a series of mirrors to end with our excitation beam incident on the Si wafer that we were testing. The better part of the time spent to set this up was on getting the optical path of the excitation beam and the THz beam to match (with the excitation pulse incident just before the THz pulse). This included setting up a delay for the excitation pulse (and the reason for trying to time the two pulses was in case the carrier lifetime was not long enough).

We next took a scan with the excitation beam blocked and the THz beam being chopped. We saved this scan and then autoscaled the lock-in to set it at zero. Then we turned off the THz chopper and chopped the excitation beam, and allowed it to be incident on the Si. On the lock-in we saw a change from zero to about 1-1.5 mV when we blocked and then unblocked the excitation beam. We then found a focusing lens and used it to focus the excitation beam on the Si.

We then then tried to optimize the signal and were able to read just under 2 mV on the lock-in. Also, after moving the delay line of the excitation beam, we concluded that the optical path probably does not matter since the lifetime of the charge carriers is long enough.

The last thing that we did is take two scans -- one with the THz beam chopped and the excitation beam blocked and the other with the THz beam NOT chopped and the excitation beam incident and chopped. We found that the amplitude of the second beam (the beam with the excitation beam incident) was slightly lower than that of the full THz signal through the Si. We also looked at the spectra of the two scans and again, the second signal showed a slight decrease in amplitude, though still not very prominent.

From this point, we need to optimize this effect, rearrange the path length of the excitation beam (since path length does not matter), put the excitation beam on a translational stand so that we may scan across the surface of the semiconductor, and try to take scans which better contrast the difference between the two signals (which optimizing the system should do).