I had quite a lot of expectations for this summer and they were all met, thanks by the wonderful people who run UC LEADS. So let me tell you some of them.

One of the reasons why I chose to do semiconductor research was to get a bit more lab-technique experience. I believe that it may due to me being close to graduating soon, but that can vary by major. I learned how to use atomic force microscopy (AFM), Nomarski imaging microscopy, how to measure semiconductor properties, honed some of my soldering skills, and many other lab techniques. The other thing that was incredibly satisfying was the project results!

My project focused on growing indium arsenide (InAs) on a gallium arsenide (GaAs) substrate for use of devices such as hard drives. The issue of InAs on GaAs is that the surface of InAs is usually rough and that causes a drop in semiconductor performance. Roughness comes from dangling GaAs atoms (referred as dislocations) that remain unconnected to the InAs atoms due to straining that the InAs layer experiences. This is due to mismatch of the lattice constants of InAs and GaAs of 6.1 Å and 5.6 Å, respectively. A simple way to say it, the gallium and arsenide atoms have more compact spacing than the atoms in the indium arsenide layer. This mismatch forces the InAs to strain, which eventually leads to defects and roughness. Since this type of straining depends on thickness of the InAs layer, we grew InAs with different thicknesses (this was done on a 500 nm GaAs buffer plus substrate). After we grew these samples, I took them to take AFM images and to measure their transport properties.

From left to right, the growth thickness of indium arsenide: 50 nm, 100 nm, 250 nm, and 500 nm

Those holes you see are the dislocations, which decrease in number as thickness increases. This will mean that semiconductor performance (measured by the electron mobility, which tells you how fast the conducting electrons are moving through the semiconductor) will increase. The question is, though, is the relationship between growth thickness and electron mobility. Below was surprising (to me, at least!):

Indium arsenide on gallium arsenide performance as a function of indium arsenide thickness

I (again, only speaking for myself!) did not expect it to be non-linear. Basically – in this case – the electron mobility doesn’t get much higher as thickness increases. I find it fantastic to find results such as these, especially since it can be used by members in the lab. This result made it great summer experience.

One of excellent suggestions that my mentor (Borzoyeh Shojaei) one that I will also recommend. If this is your second (or third, or fourth, etc.) lab, speak to your previous professors. It shows them your new skills, exciting results, and it will show them how much you developed ever since. This will indeed make it easier for them to recommend for other positions in their labs, in other people’s labs, in work, and in graduate school. You’ll surprise yourself how much you’ve changed too!

Hi everybody,

This lab has me doing a lot of things. Every day, there is a changeup. One day, I am taking measurements, the next day I learning about lab experiments, the following day I am work that requires some elbow greasin’, and then back to reading . . .

I’m a physics undergrad from UC Berkeley and I will be entering my final semester this coming fall. I work for Borzoyeh Shojaei in Chris Palmstrøm’s lab in the Materials Department and this is my fourth lab. I work in the same group as Vishaal Varahamurthy (see his post) and Ashton Meginnis, though our projects are different.

My project focuses on semiconductor growth. We study the growth of group III-V heterostructures, which are layered semiconductors made of more than one material from the boron group and the nitrogen group (such as indium arsenide, InAs). Each material has its own electronic band structure and varying the material of each layer provides different electronic properties. Simply put, this is a lot like making a sandwich. Each ingredient is layered and this serves as a way to bring out the best flavors. Interestingly enough in heterostructures, there are ways to see how well our materials are grown, just like the way we can tell how delicious a sandwich is.

We grow our semiconductors by molecular beam epitaxy, which shoots heated atoms to the heterostructure layer-by-layer. We control the amount of atoms by using shutters, just like the shutters of a window can block light. After we grow the heterostructure, we measure lots of electrical properties – such as electron mobility – through the Hall effect (watch this video to see what the Hall effect is). As the name electron mobility may suggest, we want this value to be high just like the way we want our sandwiches to taste good!

Molecular Beam Epitaxy: Semiconductors Are Made Here!

We connect it to the Hall effect apparatus, which applies different voltages, current, and magnetic field values to get the semiconductor’s stats, such as mobility (measured in cm²/(V·s)), the carrier density (1/cm²), sheet resistivity (Ω/square or ohms per aspect ratio), and the Hall coefficient (cm²/C). In order to make our numbers better, we need to accept the fact that our materials contain defects. It would be extremely difficult to make a perfect, defect-free semiconductor, so the science here is to understand these defects and which what my project is.

Hall Effect Apparatus

Although most of these defects lay in between materials, we can still see them under a microscope since the layer thickness ranges around a few hundred microns. I do this by Nomarski microscopy, an optical microscopy technique that reveals any slight deviation of time (called a phase difference) that it takes for a photon to travel from the material to the lenses. After adjusting Nomarski imaging settings, I note what I see on the surface and where it is. Although as much as I want for these defects not to exist, some of these are quite pretty.

GaAs Sample “Bo309”

This material is a gallium arsenide sample that Borzoyeh grew. The defects did hinder the measured mobility of 6900 cm²/(V·s), though not drastically. Compare this value to the typical value of 7900 cm²/(V·s) for GaAs at room temperature. Later in the summer, I’ll be working with atomic force microscopy (AFM), which will allow us to look at our samples at an even greater detail. Future measurements will also involve lowering the temperature of the material to see how well does the semiconductor really perform. And future results will give us answers to making the better semiconductor.