Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
How to Build a Computer 29: Electron Microscopy
For the next couple of posts, we’ll be sauntering through the science of measurement. To put it simply, computer bits are really, really small. So as you wander through the world of building them how do you know you’ve made the thing right? Well, let’s start simple. You can just look at ’em. I could go on a great big tear about optical microscopy which is still an important subject, and relevant. The problem with it is that I just don’t find the subject very interesting. Still, you get some neat images.
This is my fingerprint, photographed on the background of one of them hard drive platters I ripped out of that drive in the video. FBI please ignore.
To understand why you need the electron microscope it helps to spend some time with an optical microscope. The majority of the time I spent looking at parts I spent looking through an optical microscope, not on the SEM. Largely because Chem Lab owned the SEM, and they get all fidgety when someone else touches their stuff. Briefly though, I think I can demonstrate the usefulness of an electron microscope with two images.
Don’t try and make sense of those circuits, this is a test part, not part of a functioning circuit, or even of this complete breakfast.
The image at the left is magnified 25 times. 10x by the eyepiece and 2.5x by the objective lens. You’re looking at copper traces on a dielectric (plastic) background which in turn sits on stainless steel. The small copper lines there are something like eight microns wide. The small red inset section is showing some topography. Because the copper isn’t flat in those spots the light from the scope gets reflected some other direction, and it shows up black in the image. (The black fleck at the bottom is dust of some sort, probably a skin flake. It’s unimportant to our story.)
The two inset images at the right are magnified 500x, same eyepiece with a 50x objective lens. Both inset images are showing the same black spots from the zoomed out picture. Because we’re at such a magnification you simply can’t get everything in focus at the same time. The top inset image has the surface of the copper, you can see the grains in the plating, but the corners on the right get all blurry; the surface is lower there. Which is why I took the next picture. I altered the focus a little, you can see clearly the black lines along the edge of the parts (where again the light is reflecting in the wrong direction to show up in the scope) and a bit of the surface of the dimples, although again they go to points on the far right.
Now let’s take a look at an electron microscope image of those two parts:
I left the lime green captions in so you can eye-strain trying to read them.
Huh, that’s quite a bit better focus, isn’t it? Bigger too. The best I could do optically was 500x up above. Here we’re at 1370x, and not even straining. I took photos at 10,000x to get my electron microscopy certification (in school. You’d think those Chem lab folks would be fine with me using their fancy equipment, but noooo.)
You’ll note too that we’re looking at it at an angle; the part was tilted at 60 degrees to get this sort of a look at it. You can kinda-sorta do that with a nice enough optical microscope, but you run into that depth-of-focus thing real quick. I couldn’t see the top and bottom of a maybe-five micron depression in the optical photo, here you can make out details easily ten times that distance from the point of interest.
Okay, but how does an electron microscope actually work? Well, much like an optical microscope uses light to look at things, an electron microscope throws electrons at things, and sees what sticks. Okay, you say that, but what does that actually mean? I tell you what, I’ll go over the hows briefly now, and we’ll go into it more deeply next time around.
Start with an electron gun. Sounds fun, even if low caliber. Okay, not quite like that. An electron gun shoots out a lot of high-velocity electrons. You take your stream of electrons, and you run it through some calibration magnets. The calibration magnets let you focus and aim the electron beam. The electron beam strikes the surface of the sample (being the thing that you’re trying to look at). And from there a bunch of things happen.
- Some of your electrons are going to hit the nuclei of the atoms they’re pointed at, and bounce nearly straight back. Those are called backscatter electrons.
- Some of the electrons are going to knock electrons out of their orbitals in the sample. These are called secondary electrons.
- Electrons are going to fall into the spots vacated by those secondary electrons. This will shoot x-rays all over the place.
- Some of those x-rays are going to knock other electrons out of their spots. These are called Auger (augh-jay) electrons.
The main time when you’re looking to get a picture out of SEM you’ll be using the secondary electrons. Here’s a simulation of how the high-energy beam electrons pass through the more sedate electrons already present in the sample:
The thing about secondary electrons is that they’ve got much less energy than the primary (from the electron beam) electrons. The electron beam is going to penetrate a way into your sample before it slows down into nothing. But those secondary electrons from the deepest part of your sample? Not much is going to happen with them. They won’t have the energy to get out. The ones nearer to the surface will win free, which means they can get into your detector. All that adds up to the SEM beam producing a pretty good look at the top surface of your sample. Which makes it real good for looking at the surface topography, that is, coming up with pretty pretty pictures like these.
He’s reusing pictures again. What do we pay this guy for anyway?
Okay, but how do you get a picture? First, you gotta raster your beam across the sample. That is, you’ve got a couple of scanning coils up in the column, and they’ll bend the beam back and forth over the course of the area you’re scanning. It hits each spot in sequence and generates secondary electrons on each spot in sequence. These secondary electrons get picked up by your (what else?) secondary electron detector. In the olden days, these secondary electrons were used to in turn raster a pattern onto a cathode ray tube, producing an automatic television image of your sample. Rastering electron beams is how TVs did it for decades. Scanning Electron Microscopes still scan things that way, but these days the imaging is all digital.
But how does a secondary electron detector detect electrons? What do you do with backscatter electrons or the x-rays? How do I build an electron gun, with which to slay my enemies? Join us fortnight next for the answers to these questions in more when we continue on the topic of electron microscopy in “Novac, Lo vac, and Hi Vac” or “Doomsday device you say?”
This is part twenty-nine of my ongoing series on building a computer, the self-reliant way. You may find previous parts under the tag How to Build a Computer. This week’s post has been brought to you by Garage Wood. Those leftover bits of lumber hanging around your garage have a million and one uses. Garage Wood!
[First – Silicon] [Previous – Deconstruction] [Next – Electron Microscopy II]
Published in Science & Technology
Have you tried offering them sausage and cheese?
SEMs are also good for images of a bug’s eye. Computers and bugs are made for each other.
I have to say, I’m learning a helluva lot more from Rhody University than I’d expect to get from a college course in the subject.
(Santa Monica College Syllabus Spring 2019)
Re-Introduction to Cyber Basics For Belligerent Older Men Who Think They Invented This Stuff. OLD-1967, Prerequisites OL-000, Studying Memory Loss Among Older Males, and WUHU-001, Computer Jargon of the Mid-Twentieth Century.
When, oh when, is Ricochet command authority going to strike a food home delivery deal? Hank could support his web logging based on a cut of the take. Every time we mention food, we should be able to click on a link to have it delivered. Naturally, Ricochet gets a cut too. Like Whole Foods, Amazon, and the Washington Post; synergy.
Eat Right–Read Right
I’d say it’s more of a mosey.
Does Raster Blaster run Electron Town?
Here’s hoping No-bark Noonan shows up.
“You gotta raster your beam across the sample” could get eyebrow-waggling laughs as a bit of artfully phrased euphemism.
EDIT: Phrasing. Sometimes you just have to read it on screen to see it.
Three or more guys, equipped with beer, can make anything sound dirty.
Dr. George Brown’s “And Part of Which I Was: The Odyssey of a Research Engineer” is a pompous, egotistical, but info-laden and fascinating book. Brown was, in effect, quarterback of the successful RCA–NBC team that developed electronic color television between 1948 and 1953. This was a forced march of engineers on an investment scale that almost matched the onset of sound movies, a quarter century before. When it comes to his own accomplishments, the doc is windier than a typhoon. One legitimate brag was RCA’s work in product-izing electron microscopes. It’s also a thoughtful defense of capitalist risk and profit.
Some say Doc Brown should have been committed. But I say he should be commended.
I don’t know about that but I’d be willing to try.
Great Scott!
How does any of this get us to a building a working computer in our garage? This story reminds us of ancient texts -NYTimes Review
In the begins God said let there be light, and he order up one standard Big Bang.
It was a warm summers evening….