(c) Dr Paul Kinsler. [Acknowledgements & Feedback]

Logo (c) Paul Kinsler Logo (c) Paul Kinsler

Queensland 1989-1994

Funding: UQ scholarship

Quantum Optics

I left Auckland and arrived in Brisbane in early 1989, to work on my PhD in the UQ Physics logoDepartment of Physics at the UQ logoUniversity of Queensland with my MSc supervisor, Peter Drummond, who had moved there in 1988.

It seemed natural to continue on from my MSc project, so during my time at UQ I ended up doing a pretty thorough analysis of the properties of the quantum parametric oscillator, some of which has now found its way into a couple of Quantum Optics textbooks. A quantum parametric amplifier is a photon converter -- it relies on using special non-linear materials (typically crystals) to convert high-energy photons into pairs of low-energy photons. Naturally, it can do the same trick in reverse, (re)combining pairs of photons into one higher energy one. If you trap these between some suitable mirrors, the photons bounce back through each other and form a quantum parametric oscillator. Even considered as a classical system this has some pretty complex behaviours, so I focussed on just a few of the quantum ones.

Firstly, this parametric oscillator (PO) undergoes a phase transition -- shine a weak laser (of the right frequency) on it and the photon pairs almost never exist. But make the laser beam stronger and you get a lot of photon pairs -- in either of two different states. Add in the random element of photons jumping in and out of the device through the mirrors, and this randomness can cause the device to hop (or tunnel) between the two states. Now this even happens in a classical PO, but the difference between classical noise and quantum noise means that the tunneling times scale very differently with the number of photons -- giving an (in principle) clear distinction between classical approximations and the quantum reality.

Next, these QPO's can make squeezed light -- light in which the intrinsic quantum uncertainty has been manipulated, manipulated in such a way that for certain kinds of measurements, the quantum uncertainty completely vanishes! This isn't free, of course, because for other measurements the uncertainty becomes correspomdingly massive. It turns out that the best sqeezing we can get from a QPO is at its thresold -- exactly at the point at which the device switches from having one stable state to two, as I descried in the previous paragraph. This was a tricky calculation, the first ever solution of a non-equilibrium quantum system at threshold.

It was while at UQ that I started using the internet, although I had known people to use email and so on when I was at Auckland. Mostly I just used email, chatted using IRC, and downloaded programs and stuff from a range of sites. I also remember Bruce Wielinga constantly downloading the latest linux kernels in order to make his laptop work better. Unfortunately the document formatting/ typesetting program latex didn't seem good enough to convince me to abandon MSWord when writing scientific papers and my PhD thesis. Consquently I now have unreadable legacy files instead of useful documents. I now write everything as either plain text, latex, or HTML, which will never be as useless as an archive of MS Word files, and is greatly preferable to the reams of paper based notes I have in my filing cabinet.

I recall one of the early seminars I gave at Queensland. I'd just finished my talk on the parametric oscillator when an applied mathematician in the audience makes a lethal criticism: "Your Hamiltonian is not bounded below". And indeed this measure of the energy in the system isn't, it is cubic function -- and so there are regions of its phase space where the system will run away to infinity. Needless to say, this isn't very physical and clearly points to a failure in the model. However, at the time I didn't even understand the question so I floundered hopelessly for a minute or two until Peter Drummond pointed out why the Hamiltonian was OK -- the "unbounded" behavour occurs in regions where the model is not designed to work anyway. Phew, I escape. But it's all very educational and character building.

Brisbane is a nice place to work -- the campus is nice, the winter weather is fine and sunny (if a bit cold), but I found the summer weather, while sunny, a bit warm for my tastes. There was a pretty dynamic group of staff there, not only in the quantum optics group but also the laser physics experimentalists. Still, all things come to an end and I handed in my thesis, and was awarded my PhD. I stayed on for a short while working as a postdoc, but I was applying for other jobs and looking for new challenges. I found them in the UK, in Sheffield, a rather colder place than Brisbane.

Research homepage; Publications; Conferences; History homepage;
Date=0320 20020106 1003 20000223 0610 0204 19990121 Author=P.Kinsler Created=1999