Investigating Optical-Field-Induced Currents in GaN using Ultrafast Lasers

Abstract (Swedish)

This study saw the development of an experimental setup capable of generating and measuring optical-field-induced currents in a variety of nanodevices fabricated specifically for this project. Each device design features two metallic contacts, closely separated by about 5 micrometers, deposited onto a semiconductor or insulator substrate. The region between these two contacts is a junction, being where the laser is focused to generate, and consequently measure, the current. In total, 15 separate substrate samples had nanodevices fabricated onto them. Most devices were made using gold etching, placing Au on mostly GaN substrates, but some SiC and SiO2 substrates were also used. Devices were also fabricated using a lift-off procedure, allowing for a Ti/Au-GaN device to be made.
Two separate laser systems were used to investigate the generation of optical-field-induced currents. One, a titanium-sapphire laser generating a field asymmetry by ultrashort pulses, the other a ytterbium laser generating it by a fundamental and second harmonic superposition. However, while currents were measured with both laser systems, several characteristics of the signal raised further questions. The phase modulation of the Ti:Sa laser did not result in a current oscillation, nor did the device only produce a current when the pulse was polarised in the direction of the junction, instead generating a current even with a perpendicularly polarised pulse. On the other hand, the Yb laser measurements seemed to confirm that the current direction has a polarisation dependence, since no current was generated with perpendicularly polarised pulses, and when reversing the polarisation there was a sign change in the current. However, modulating the relative delay of the two pulses from the Yb laser generated no current oscillation, unlike what was expected from the model. Furthermore, the high intensities required to induce a current would ablate both the gold and substrate materials, destroying the devices. Likewise, the phase modulation of the Ti:Sa also saw no change in current.
To substantiate the experimental efforts, a model derived from Bloch equations developed by Khurgin[1] to estimate the photoinduced charge produced by a single, or two cross-polarised, laser pulses on a device was recreated. It was also expanded upon to cover two-colour experiments, like the wave superposition used in the Yb laser experiments. Finally, a fully featured GUI was written to more easily control the parameters of the model, having the potential to quickly create and compare experimental results to the expected modelled outcome.

Popular Abstract

Laser Activated Electronics
Transistors are often considered the most revolutionary electrical devices of this century, attributed as the building block of all modern technology. It is these tiny devices that control the flow of electrons in all your electronics. In your computer alone, there are hundreds of millions of these transistors, all working together in a complex circuit. Previously, improving the operational speed of these devices needed only that they were made smaller, but a limit is now being reached in how small they can be made. However, what if we didn’t need to make them smaller to improve their speed? What if we could activate them using the fastest thing in the universe, light!

Currently, what limits the operating speed of conventional transistors is not their size, but the time it takes for electrical signals to reach the device in order to switch it from a conducting (on) to a non-conducting (off) state. We communicate with these devices using physical wires, where real electrons need time to build up enough charge before the signal switches the state of the transistor, the fastest of such a charging time is approximately 100 picoseconds (10,000,000,000 times shorter than a second). With current methods, this means that operation of computer devices cannot exceed the GHz regime, which is currently the highest speed that modern CPUs can reach. Recent developments, however, have demonstrated that extremely short laser pulses can be used to activate devices with attosecond response times, a thousand times faster than with physical wires. This research could therefore see the development of small electronic devices that can operate in the THz regime, a thousand times faster than the ones we can make now.

How did we test this phenomenon?
This project is a fundamental study into limits of the switching speed of electronics, albeit a perhaps unconventional way to do so. We made over 40 nanodevices that fit the description of previous successful experiments, most of which used Gold on Gallium Nitride, like the device in the figure. Then we shone different lasers in the region between two gold contacts, something we call a junction, after which we tried to measure if a current was produced. Ultrashort laser pulses are asymmetric, meaning it is stronger in one direction of the wave. Theory states that the electrons should move in the same direction as the stronger part of the laser pulse because of a complex mechanism called Quantum Interference. Consequently, no current should be measured when the direction of asymmetry is not across the junction, something we confirmed in our experiments. We also wrote a computer model capable of estimating the kind of currents we should be measuring in our experiments.

How can this be used in the future?
With greater response times and transparent substrates, potential uses include faster photodetectors, transparent electrodes, biosensors, and light controlled nanomachines. While THz operation is the short-term goal, it should, in principle, be possible to reach potentially upwards of PHz operation speeds. Since this technology currently requires a high-power laser, it is unlikely that consumer products, like CPUs, will see its use, but it paves the way for further research into alternate methods of overcoming the limitations that prevent faster electronics from being made.

• author: Flodgren, Vidar
• supervisor: Anders Mikkelsen; Fabian Langer
• organization: Synchrotron Radiation Research; Department of Physics
• course: FYSM60 20191
• year: 2019
• type: H1 - Master's Degree (One Year)
• subject: Physics and Astronomy
• keywords: GaN, gallium nitride, nanoscience, material science, ultrafast lasers, Yb Laser, Pharos Laser, titanium sapphire laser, device fabrication, lift-off, gold etching, optical-field-induced currents, modelling, matlab
• language: English
• id: 8992477
• date added to LUP: 2019-08-20 09:26:53
• date last changed: 2019-08-20 09:26:53