Success stories
Roger Reeves: Spectroscopy
The new Laser Lab at Canterbury University
Bringing a nanoscopic world to light
Let us embark on a nanoscopic journey into the depths of a semiconductor crystal. The vast empty space is punctuated with positive ion cores, their regular formation stretching to a seeming infinity in all directions. Around these ion cores electrons swarm in clouds or “orbitals” of various shapes and sizes, each ion drawing the electrons in like a city does people. The electrons arrange themselves in pairs, each with a different energy and occupying a different orbital and the orbital energies are arranged in bands separated by band gaps. Low energy electrons are held tight to the core but as you move out the electrostatic attraction weakens and you start to find orbitals with ‘holes”, evidence of escaped electrons. The energy gap between the last full electron band (the “valence band”) and the “conduction band” beyond acts like a city wall, which electrons must climb to get free.
The advancement of semiconductor technology depends on the characterisation of band gaps and understanding the behaviour of electrons in newly fabricated materials. Unfortunately in real life it is not so easy to journey through this nanoscopic world but techniques have been developed to go where human eyes cannot. Roger Reeves uses spectroscopy to shed some light on the subject.
Photoluminescence
“Photoluminescence is the light beam emitted by semiconductors mainly when electrons and holes recombine. The photons that come out are very characteristic of the material”
Fire a high power laser at a piece of semiconducting material and valence electrons will ricochet out of their orbits. These high-energy electrons will collide with further valence electrons. More and more will be released in this way until the average energy drops to the bottom of the conduction band. Usually at this stage attraction to the hole overwhelms the electrons and they recombine emitting photons of the band gap energy. The recombination shows up as a peak on the emission spectrum usually at the band gap energy.
“I use Photoluminescence to characterise semiconductors and look for their fundamental light emitting properties.”
At first glance analysing the PL emission spectra appears trivial - Roger assures me it is not. Occasionally electrons get lost in mid-gap states before recombining with holes. They can get stuck in electron traps caused by added impurity atoms or defects in the crystal structure or in very good quality crystalline materials excitons can form. (This is when electrons orbit a hole just as they orbit a proton in a hydrogen atom creating series of excitonic energy levels just below the conduction band.) The result is that photons are emitted at energies lower than the band gap energy and ‘mystery’ peaks appear in the emission spectrum sometimes partially obscured or covered by the dominant band gap peak or by each other. As Roger pointed out “peaks don’t come with labels attached” and it is his job to solve thismystery and work out the implications of his findings in terms of the crystal structure.
Photoluminescence. Eg is the band gap energy. On the left a high energy laser photon dislodges an electron from its orbit. The electron loses energy until it reaches the bottom of the conduction band. The right hand diagram shows two possible transitions. On the left the electron combines immediately with a hole in the valence band emitting a photon of energy Eg. On the right it gets stuck in a 'mid-gap' state emitting a lower energy photon.
Photoconductivity
Photconductivity is when you apply an electric field to the sample as you expose it to light. Both electrons and holes move to create a current. The trick to this is evaluating whether defects are involved as electrons and holes can jump from defect to defect. Roger has had a lot of success lately in seeing photoconductivity in Indium Nitride samples.
Equipment
“To do photoluminescence measurements you need a laser to create electron hole pairs and a spectrometer to measure the intensity and wavelength of photons coming out.” The equipment Roger has developed allows for measurement of photoluminescence from ultraviolet wavelengths (suitable for large band gap materials like Gallium Nitride) all the way into the infrared (for small band gap materials like Indium Nitride).
In the past Roger has had to rely on people giving him samples to make measurements on but with the recent completion of the MacDiarmid funded Pulsed Laser facility at Canterbury University he will be able to make his own.
“The new laser has a power of 100 megawatts. This is about 50 times the power that it takes to run the whole university! Fortunately, while the university requires this power on a continuous basis, we use such a power for only a tiny fraction of time – actually 0.000000001 seconds in each pulse to be exact. That way we obtain the huge laser power without taking all the electricity. The next surprising number comes from the intensity when we focus that power down to a small spot on the surface of our target. The intensity can be as high as 10 17 watts/m 2, which is more than 10,000,000,000 (10^10) times larger than the light intensity at the surface of the sun. Given that we know the sun’s temperature is several million degrees, it’s not surprising that our laser can melt almost anything! Of course reality is not quite this dramatic, as our focussed laser spot is only a fraction of a millimetre and thus the amount of melted material is quite small. Nevertheless, this is exactly what we want for the material growth. A well controlled region of molten material from which the vapour can be used to grow a near-perfect thin film!”
This super powerful laser will open up a world of new research possibilities.
Research Collaborations
“Photoluminescence is a characterisation tool like microscopy. It’s expensive to have this equipment. You don’t need to have one in every single university.” This means that Roger’s expertise is in hot demand. He is involved in several collaborations working on Zinc Oxide (ZnO). In the past he has done a lot of work on Gallium Nitride (GaN) and GaAs.
For the past two years Roger has been investigating thin films of Indium Nitride (InN) with Steve Durbin from the Electrical and Computer Engineering Department at Canterbury University. Steve uses molecular beam epitaxy (MBE) to grow the films atom by atom onto a substrate. The aim is to get a really good quality crystal but in practice this is not so easy. Defects caused by contamination in InN films have been so overwhelming that photons are no longer emitted at the band gap energy. The scientific community has been in confusion over what the actual band gap of InN is. Roger and Steve’s research seems to suggest a small band gap (around 0.7eV) but as yet they are still perfecting the fabrication process. The lack of evidence of excitons in emission spectra indicates that the samples are still full of defects.
Rare Earth Ions, Insulating Superlattices and spintronics
In some materials the photoluminescence peaks are so broad they smother any band edge details or slight shifts in energy. Other materials don’t display photoluminescence at all. In these cases a trick can be used. Small quantities of other elements implanted into the intrinsic material can cause very sharp energy levels within the band gap, which act like nanoscopic probes into the crystal structure. You can tune a laser so that only the electrons in these implanted atoms are excited. What you get out is a photon characteristic of the implanted atom. The beauty of rare earth ions such as Europium (Eu 2+) is their sensitivity to their environment. If there is any difference in the local crystal structure the sharp spectroscopic lines shift in energy acting like a magnifying glass. “Even the slightest shift can be detected”.
Roger and his collaborators in Russia have come up with a unique application for rare earth ions - using them to probe the structure of insulating superlattices. A superlattice is like a club sandwich of thin films of different materials. Although the technology for semiconducting superlattices is very advanced their insulating counterparts have received very little attention. Recently their potential use in memories, spintronics and lasers has begun to spur interest.
Insulating superlattices are made in Russia and sent to Roger where the interrogation begins. The way atoms form crystals near the surface of a material is quite different to in its interior. Superlattice layers might be only a few atoms wide so the crystal structure is very hard to predict. By implanting Europium ions into the lattice (“they go in quite nicely”) you can detect these differences. The technique is very precise, you can control which atomic layer the Europium goes into so you can tell exactly where the surface type structure gives way to the bulk structure.
A thin film containing five atomic layers. Implanted europium ions fit into the lattice without disturbing the surrounding structure. Their emission spectra vary depending on which atomic layer they are in.
One way to fit more data into smaller areas as is the demand of modern technology would be to control and utilise the spin as well as the charge of electrons. As yet no one has achieved great success in the area but Roger and his Russian collaborators have an idea. One of the insulating materials Roger has been looking at is Manganese Fluoride (MnF 2), which in bulk form is magnetic below 67 ° K. In thin films however it looks to be magnetic up to 100 ° K but only up to the tenth or twelfth monolayer. In a superlattice you could have the majority of material in the new ‘thin film’ structure. “We may be able to get a magnetic (non- metallic) material at a temperature higher than liquid nitrogen.” That’s pretty significant but it doesn’t stop there. By alternating MnF 2 layers with layers of semiconductor in a superlattice you could create a magnetic semiconductor, which is pretty close to the holy grail of spintronics. It is early days still but the prospects are exciting.