I met with Antoine for a little bit today (Tuesday) and we talked about where I am and where I am headed. It looks like the next couple of weeks will be a lot of reading. He wants me to have a solid understanding of what I am working with before actually doing anything in the lab -- which I think is a good plan.
I spent the morning reading an article by Yeubin Wang, et al, about periodic optical delay based on a tilted parabolic generatrix helicoid reflective mirror. Some notes that I took on this article inclde what an optical delay line (ODL) is and a little bit about what other types of ODL there are (ranging from linearly scattering retroreflectors to multipass cavities). From what I understand, in a very general sense, an ODL is used to lengthen the path length of light, as this will alter the phase of the light and will in turn have an effect on diffraction and interference.
More importantly, I spent the better part of the day re-reading an article from a Japanese group for which the title of today's post is named after. This article describes a new (2001) technique for spatially profiling a THz beam with the Si wafer technique that I mentioned a little in Thursday's post (and which John commented on). After talking to Antoine early in the day about this technique and re-reading the article, I understand it much better now.
The setup is rather simple (and I am trying to get a sketch to upload and it is not working). We split the beam form our Mode-locked Ti:Sapphire laser and send part to the emitter and part to the detector (which both have similar setups to what I have described before with GaAs and metallic strips). The THz radiation is then focused using parabolic mirrors onto an n-type Si wafer with charge carrier density 5.6 x 10^14 cm^-3 and thickness of about .5mm, which seems to be located at the focal distance from the parabolic mirrors.
As John mentioned in his response post, this THz radiation should penetrate the Si sample, and indeed most of it will. The reason for this is that the THz radiation will have energy of a few meV and thus cannot move electrons up to the conduction band. Also keep in mind that there will be some reflection off of the surface of the Si wafer.
In order to attain a profile of this beam, this group has come up with a technique in which a CW laser diode of 820nm is placed on a translational stage and aimed directly at the spot where the THz beam is incident on the Si wafer. The angle of incidence is 35 degrees and the beam is first passed through an optical chopper which is running at 450Hz chopping frequency.
The photons of this 820nm laser have enough energy to excite the electrons of the Si wafer into the conduction band. What this does (as John alluded to in his last post) is creates a semi-localized point on the wafer which is a semi-metal. This in turn makes that point more reflective to the THz beam and so at that point there is a loss of total power of THz signal which passes through the wafer.
Since the laser diode is on a translational stage, we are able to scan the laser across the area of the Si wafer. By taking measurements of how the total power of the THz signal changes with the location of the incident 820nm laser, we can build up a profile of the THz beam.
The group that used this technique report using this for three different frequencies of THz (or near THz) radiation. For each case, they report Gaussian beams (with some variation in x- and y- width).
The few physical problems with this setup are that there may be some sort of Fabry-Perot effect with the THz signal and also diffusion of charge carriers near the localized metallic point. Other problems with this technique are that I have not found any sources which cite this one, suggesting that this technique maybe has not been used much.
My goal is to be able to recreate this setup and try to get a profile of our THz signal. There is also the possibility of using another method if an easier one comes about. As of now it looks like this is the way to go for profiling my THz beam.
Subscribe to:
Post Comments (Atom)
Alex,
ReplyDeleteOk, this has started to make a little more sense, but I have some more questions for you.
I am curious about the way in which the group in the paper distinguished the three different frequencies, since you say they were measuring total power transmitted through the Si and not the time domain waveforms using a gated receiver. (i.e., if they measured a time-domain pulse that corresponded to each position of the 820 nm laser spot, then they could see, by taking FFT's of the pulses, that the lower frequencies of the THz pulse extended out to larger diameters than the higher frequencies.)
However, if they were measuring total power, then they wouldn't know which part of the spectrum of the THz pulses was decreasing when they saw lower power transmission. Can you tell how they were measuring total power? One way would be with a bolometer, but you seem to indicate that they use a pulsed probe beam with a photoconductive receiver to detect the THz pulses. Plus, they chop the second, cw 820 nm pump beam, so they are presumably using a lock-in detection technique. Perhaps this could be employed with a bolometer, but I am more inclined to think that they are measuring the time-domain waveforms (which would be electric field), then squaring that electric field and integrating that signal over time to get power.
Anyway, I hope this is clear and gives you a few additional ideas of things to look for in that paper. Please feel free to try and describe what you find, as that sort of "teaching," or at least explaining, is a great way to learn.
- John
John,
ReplyDeleteSince I spent the better part of Tuesday looking over the article again and trying to find the answers to these questions, I will write about it in my entry for Tuesday, June 16 (which will have a posting date of Wednesday the 17th).
Alex