In this Optics Express article, Zhao et al report on the creation of the fastest pulse of light ever recorded, a mere 67 as in duration. Light pulses from lasers are good candidates for producing very fast strobing events, as coherent light can be easily manipulated in both the time and frequency domains. The latter is helpful because of a result from Fourier analysis: an event which is very short in the time domain must contain a very broad range of frequencies. The short event occurs at a point in time where all these frequencies are in phase, so that their electric field vectors add together constructively. As this time passes, more and more components begin adding together destructively, and the electric field dwindles to zero. The more frequencies that are available, the faster the pulse can be made to go to zero after the peak, and the shorter its total duration. This mathematical result can be realized in practice using a Ti:sapphire laser oscillator, which can simultaneously amplify a very broad range of frequencies. By careful tuning of the oscillator, nearly all of these frequencies can be brought into phase at a single time point. Pulses of around 10 femtoseconds can be achieved with this method.
However, a Ti:sapphire crystal only amplifies red and near-infrared wavelengths, with a gain center around 800 nm. An 800 nm wave has a cycle length of 2.7 fs. Thus, even a half-optical-cycle pulse – the shortest pulse that could possibly be created with this center frequency – would have a duration in excess of 1 fs. To create attosecond pulses, shorter wavelengths are required. The authors achieve this by focusing a high-intensity femtosecond pulse from a Ti:sapphire system into neon gas. A nonlinear process in the gas called high-harmonic generation produces replicas of the pulse at higher frequencies – all the way up into the Extreme UV (XUV) region. Due to their higher center frequency and broader bandwidth, these up-shifted replicas can be much shorter than the pulse which created them. The previous shortest pulse of 80 as was created with this method.
To obtain an even shorter pulse, Zhao’s group employed several innovations. For one, they used a technique known as “double optical gating” to reduce the fraction of the infrared driving pulse which could actually produce harmonics, effectively shortening it. They also pioneered methods for removing the “attochirp”, a type of temporal aberration that causes some XUV frequency components to be slightly out of phase at the pulse peak. Physical arguments show that the biggest offenders are the lowest and highest frequencies in the harmonic spectrum. The authors used a zirconium filter to remove the lowest frequency components; transmission through the metal also improved the phase-matching of mid-range frequencies. By adjusting the pressure of the neon gas, they found they were able to eliminate the highest frequency components as well, producing a shorter pulse in spite of the reduced total bandwidth.
Measuring such short pulses is as much a challenge as making them, as no shorter event is available to use as a yardstick. However, nonlinear optical effects combined with the mathematics of “phase retrieval” can provide temporal information using only replicas of the pulse itself. Two separate phase retrieval techniques were applied, and found to agree within 2 as, verifying that this is indeed the shortest pulse ever created.
You must log in to add comments.