Colin Humphreys: The electron microscope

The easiest way to think of an electron microscope is to start by thinking of a light microscope. So a light microscope, you have a source of light, like a light bulb, and then there are lenses which focus that light onto the specimen, and you get wonderful images, but a light microscope has two disadvantages. One is the magnification is limited to about 10,000 times maximum, and that's because the resolution is limited. If you try to magnify more than that, you don't see anything more. And the other is you can't see inside material in optical microscopes, in light microscopes, unless they're transparent. So a piece of metal, for example, you just see the surface of it. With an electron microscope, you can magnify 10 million times, so you can actually see the atoms in a material, and you can see right through the material. These electrons penetrate right through the material, so you can image the interior of materials. And to give you an example, this is an atomic model of a high temperature superconductor, and in electron microscope, you can see all these atoms, and you can also analyse what all the different atoms in this material are, and that you can see in the interior of the material, and you can see defects inside the material, which are also important. For example, this turbine blade, which is in an aerospace engine, it's the hottest point of the engine. It contains 18 different elements. We know where these elements are because of the electron microscope, and without the electron microscope, we just wouldn't know where all these elements are and where the defects in these materials are as well. I came to use electron microscopes by accident. So my thesis project, and it was here, I was in this department as a PhD, my thesis project was to construct some equipment called X-ray topography equipment, and I arrived late to do my PhD, about, you know, just a couple of months late, and the workshops were fully occupied. So by the time a workshop technician became free, it was six months. So six months later, he started working to build this equipment, and we had drawings, but a lot was in his head, and after about nine months, we were halfway there, he died. And so I had to wait again, and another technician took over, and the upshot was, after two years, I had no equipment, I had no results, and I was feeling slightly depressed, and that's why a PhD is called Doctor of Philosophy, because you have to be philosophic. And one day, a professor walked into the lab called Professor Howey, he was in Dr. Howey, and he held up this electron micrographic picture, and he said, we've been puzzling how to interpret this picture, would any one of you like to think about this? And I had no results, and I was desperate, so I put my hands up and I said, I'll think about this. So I wrote a really complex computer program, and I explained the image in this picture, and I then didn't look back. So I then, you know, I spent a lot of time working on this program, my thesis was mainly about interpreting electron microscope images, I should say the x-ray equipment did finally work, that's also in my thesis, but that's how I got involved in electron microscopy by accident, and I was particularly involved in interpreting the images you get in the electron microscope. Let's say you put a leaf into an optical microscope, and you gradually turn up the magnification, it still looks like a leaf, you know, you start to see the veins, and you see them at higher magnification. If you put a piece of metal, say, in an electron microscope, and you turn up the magnification, it doesn't look like a piece of metal at all. And you see defects in it, which you didn't know existed, and if you really go to high magnification, you see the atoms in it. So the image looks completely different, and the key thing is to interpret that image, and say what's really in the object you're looking at. This electron microscope has a vacuum in the column, otherwise the electrons will get scattered by the air molecules, but to improve the vacuum, we're going to, pouring this liquid nitrogen, and around the specimen, there's something called a cold finger, because it's very cold, and the main problem with the vacuum, although it's a very good vacuum, it contains a lot of water vapour, and the water vapour condenses on this cold finger, and not on the specimen, because the specimen isn't cold, the specimen's room temperature, and this improves the vacuum by a factor of 10 times. So this then improves the images you get. In this electron microscope, we start by switching on the high tension, it's called, that's the accelerating voltage. So this microscope uses electrons, which have got an energy of 400,000 electron volts, and so we have a voltage generator, which generates 400,000 volts, it has to be extremely stable, so it's 400,000 volts to within one volt, and it keeps that constant over minutes, and then the electrons are accelerated through a potential difference of 400,000 electron volts, so when they come down this column, the specimen is inserted on this long specimen rod, so this long rod goes into the microscope, and it sits the specimen right in the centre of this microscope column, and the electrons hitting this have got almost the speed of light, they're travelling through incredibly fast. So if you pointed this column at the moon, the electrons would get there in a few seconds, right, going that fast. So they zip through the specimen, and there are lenses above which focus the electrons onto the specimen, then after the specimen, there are more lenses down here, and they magnify the image of the specimen, and so on the screen, you can get an image which is magnified 10 million times. And so we see, and we start by looking at lower magnification, and then we just work up in magnification to see really the atoms or almost atomic resolution on the screen. There's a lot of computer control, so if you just look here, there's illuminated LED panels here, lots of computer controls. Initially, it was all mechanical, everything was mechanical, so you had to operate the strength of the focusing lenses, you turn dials which would mechanically change the currents going through them, and now it's nearly all electronic, so a huge difference. Much of my work is on a material called gallium nitride. I've got, well, these LEDs you buy in the shops now, these are all made from gallium nitride, and this is a material which doesn't exist in nature, and we can design the material, and the electron microscope is critical at seeing what we're doing. With these LEDs, for example, I'll show this one again, let's show a slightly less bright one. This is an LED, which is a blue LED, which we grew in the department here. If I switch it on first, so you can see this blue LED, this LED operates because it has very thin, what are called quantum wells of material, and these are so thin, 10,000 would sit side by side in the width of a human hair. Now, you can't see these in an optical microscope, you have to have an electron microscope to see them, so to develop these new materials, the electron microscope is absolutely essential. Without the electron microscope, you wouldn't have these new materials. What we're seeing here is an image of an LED chip, and these are the quantum wells which are emitting the light, and the image, this is a magnification of 150,000 times we're looking at, and these very thin quantum wells, so these dark stripes, they're just 10 atoms thick, and it's from this very thin amount of material that the brilliant light comes from our quantum well, from our LEDs, and I can just move along these quantum wells, and just moving the specimen now, and we're interested in the quality of these quantum wells, and you can see some imperfections in the quantum wells as we move along, and we are reporting back to the growers some of these imperfections, so that we really get much more efficient LEDs, but without the electron microscope, we couldn't even see these quantum wells. So, no hope in an optical microscope, a light microscope, so this microscope is absolutely essential for seeing these quantum wells, and absolutely essential for developing this low energy lighting from LED materials in our homes and offices. What we're looking at here is a piece of LED which has been put into the electron microscope, and the LED is gallium nitride, so this is gallium nitride material here, and this particular LED is in fact grown on sapphire, so it's like any commercial LED you buy. So this is sapphire down here, and of course sapphire is quite expensive, and this is the gallium nitride we've grown on sapphire, and these lines are the dislocations I talked about earlier, so these are defects in the crystal structure, and if we just move across here, we'll see some other lines, some straight lines here, and these are what are called stacking faults, so they're different sorts of defects, and it's important to understand where these are, and their distribution in the material, and if we now just move, let's follow these dislocations going down, because they go through the material from top to bottom, and let's just follow these down, and then the quantum wells, which are where the light's emitted from, which are very thin and very narrow, they are just near this interface, so this is vacuum out here, this is the edge of the specimen, and the quantum wells are here, and unfortunately today, and this sometimes happens in science, the intensity of the electron beam has faded away, so there's something wrong with the filament, and so we don't have enough brightness to actually see the quantum wells, but if we, yesterday it was operating wonderfully, we could see the quantum wells, but we can see, instead of seeing those, we can see these defects called dislocations, and you can see how they run through the material, and they're quite complicated, and they affect the light-emitting properties of this material, and we really need to understand these dislocations, and also these stacking faults, which we saw just here, these straight lines here are stacking faults, and again, these affect the light-emitting properties of the material, so again, what we're doing, when we look at this electron microscope, we're seeing these defects, and we're understanding how they affect the properties of the material, and without electron microscopes, we couldn't image these dislocations at all, so this is a key instrument for looking at the defects in these materials.