Even though it is hard work, it is quite satisfying to drill holes in concrete. Apparently, my degree covers everything from construction to theoretical quantum physics. Good to know that I’m getting a well-rounded education.
We have a new paper in the latest issue of Nature, about remote entanglement of two ions. Here’s a couple press releases: Michigan press release, and Maryland press release. We were also on Slashdot. You can read the paper here (pdf). Unfortunately, I’m not on this paper—this was the other project in the lab.
So, what’s this all about? Well, suppose we had two computers, and we wanted to transfer information from one to another. Typically, this would be done by sending bits across the internet (or wireless network, etc.). Now, what if we wanted to do this with a quantum computer? Well, now we have two types of information: classical and quantum. We could just read out the information of one computer, and send it to the other. The read-out process collapses the quantum information into regular classical bits. We could easily send this to the other computer easily and work with it (The classical bits could trigger the control logic of the other quantum computer).
Now what if what the other computer needed was the quantum information? We need a way to transmit this quantum information. It turns out one way of transferring this information is through quantum teleportation. This requires quantum entanglement.
Suppose I had three qubits, q0 q1 q2, where q1 and q2 are entangled. I give you qubit q2, and want to transfer the information in q0 to you. I perform a joint measurment on both qubits that I have (q0 and q1). When I tell you the result of this measurement, then you can perform a quantum operation on q2 to recover the information in q0. The two entangled qubits are like a quantum information bus. But we still need classical information to determine how to recover the information.
What we have acheived in the lab is a means of entangling these two qubits without needing them to be right next to each other. In my description above, I initially put q1 and q2 into an entangled state, and gave you one of them. Now, you could have your own qubit and live in Michigan, and I have my own in Maryland. We can then ‘open up’ the quantum channel by entangling the two qubits, allowing transfer of quantum information.
How is this done physically? First, we have a single ytterbium ion in an ion trap, and excite it to an excited state. The ion has two decay channels. If it decays to the qubit state |0>, it emits a photon of one color (say ‘blue’). If it decays to qubit state |1>, it emits a photon of a different color (’red’). In this method, we have generated a single photon entangled with a single ion. So, we have both ions do this, giving us a photon entangled with the first ion (q1), and a photon entangled with the other ion (q2).
We then interfere these photons on a beamsplitter. If a photon hits a beamsplitter, it can go one of two ways. We have detectors at each output port to detect the photon. Now, if the two photons interfere on the beamsplitter, we will only coincidence detection if the photons are of a different color. So we detect a ‘red’ photon and a ‘blue’ photon. But we don’t know which ion emitted which photon.
What we do know is that one ion is in |0> and the other is in |1>. That’s the key. These two ions are now entangled. And the quantum communication channel is open.
I have an article in latest issue of the IEEE Spectrum, explaining what we do. The IEEE Spectrum is the magazine for the Institute of Electrical and Electronics Engineers. This obviously justifies my work as an EE in a physics lab. Check it out, I hope it is readable for everyone. Thoughts? (Looking at Chris and Mike in particular)
Last Friday I was attempting to squeeze more power out of the Raman BBO, and proceeded to lose all the power. So I spent this past weekend fixing it. Fun. But I managed to accomplish what I had set out to do. All is well and good.
Or so I thought. Turns out, when I turned on the Microwaves for the EO (phase modulator), the power dropped significantly. Apparently, the free-spectral-range of the cavity wasn’t matched to the microwave sidebands, and hence weren’t building up.
Some explanation:
To double the frequency of our laser light (change color from infrared to blue, or blue to ultraviolet), we shine the light into a cavity with the nonlinear crystal (this crystal changes the color). A cavity basically is a set of mirrors aligned such that the light makes a complete circuit and ends up exactly where it started, heading in the same direction, and building up inside the cavity.
For an optical cavity, there are a discrete set of colors that will build up in the cavity. The spacing between these adjacent colors is called the free-spectral-range. So here’s the problem I had: without the phase modulator on, I only had one color in the cavity, and it built up just fine. Got a decent amount of power out of it. Turning on the phase modulator creates additional colors of the light. And these different colors were not in the set of colors that build up in the cavity.
So, basically, I had to re-align the cavity such that these additional colors build up. And that is why I’m posting this at 3am.
This past week the lab went down to the University of Maryland to see where we will be spending the next few years of our life. First impressions: The campus is quite pretty—the buildings have a red-brick southern-style architecture. I liked that. Unfortunately, the interior of the buildings leave much to be desired. Also, there wasn’t much in the way of food nearby or a collegetown. As Dan told me, College Park is a strip-mall with a university attatched. Perhaps it is because DC is nearby.
Unfortunately, I haven’t found a place to live yet. The cost of living is higher than in Ann Arbor, so I’m looking for a cheap place. I guess that means I’ll have to take another trip down there before moving.
Now back to the 13+ hour workdays.
I’ve been quite non-existent the past couple weeks. Undoubtedly because of two things. (1) I have no internet at home. The person whose wifi I’ve been using must have left for the summer. Hence, I have no signal at home. With three months left in Michigan(!), I don’t care. It’s obvious that I’ve become a product of the ‘information age.’ What did I do before the internet? Ah yes. Read. (2) Because I only have a couple months left, I’ve been working ridiculous hours, hoping to finish this experiment before leaving. 12+ hours for the past few weeks is not fun. (sometimes, it’s 15 or 16 hours). Additionally, Dan—the senior grad student on my project—just defended his thesis, and will be leaving the lab shortly. That means I’ll be in charge of this project, and I don’t know what I’m doing.
And things are not peachy in the lab. We have a new student on my project, and I have to explain everything all over again. We have very little power in our Raman laser, and Yisa just chipped the BBO crystal today (which is extremely expensive). He’s too cavalier with optics. But I figured out why I was only getting ~90% qubit detection fidelity—apparently some idiot turned off the RF generator for the doppler cooling beam. Thus, we weren’t cooling the ions. Alas, the initialization beam got mis-aligned. The fun never ends.
And there’s a needle there:
Even though I was an Electrical Engineer, I never was one for electronics. But since I’ve been in a lab, and needing it to run experiments, I’ve become a recent convert. I blame Cornell’s ECE 210/215. Those classes (when I took them) were more about “what’s the voltage here” and “the current there,” rather than developing electronics to do things.
For example, in this lab, I needed a PID controller for Laser frequency stabilization. Basically, the frequency of the laser drifts, and we need to control it, using a feedback loop. Basically, the PID controller adds a correction to the current in the laser diode (which determines the frequency of the laser) . This correction is a sum of a proportional signal (i.e. what’s going on right now—are we too high? or too low?), an integral signal (in the long-term, what’s its tendency to do?), and a derivative signal (right now, are we drifting away from where want to be? or towards?).
This provides an actual purpose to the circuit. Can I build a proportional signal?—yes, that’s just an amplifier with adjustable gain. I can build an integrating circuit, and a differentiating circuit as well. But before, I never had any reason to build them. They were just simple circuits that did what they did, and that was that.
That’s what was missing in my intro circuits class. I never felt that the circuits we made were to do things. We made circuits, and tested to see how they behaved. Which, quite frankly, didn’t interest me at all.
