# How to Build a Computer 33: Atomic Force Microscopy!

Atomic Force Microscopy is a refinement of that long and hallowed scientific tradition: poke it with a stick and see what happens. Picture, if you will, a blind man walking across the street. He taps the ground with his cane, profiling the height of the surface. That tells him where the curbs are; he doesn’t trip because he knows when to step up and step down. Now picture that blind man in a skate park, full of ramps and contours. He could, by painstaking effort, tap his cane up and down the entire area of the skate park and build up a picture in his mind where all the half-pipes lay, even though he can’t see ’em himself. Now picture him in that same skate park, doing kick-flips and grinding like a pro. Because that sounds awesome.

Three square microns of (highly ordered pyrolitic) graphite. A friend of mine measured this as part of a school project we worked on. This is after a metaphorical baseball bat to the head of mathematical smoothing.

Atomic Force Microscopy builds up a portrait of the surface of a thing by rubbing a tiny, tiny needle across it, and reading it like you’d read the grooves on a record. Heck, you could probably play it like a record too, only it’d sound all staticky because nobody bothered to lay down music on that spot to begin with. (Although…)

What exactly are we working with? You start with a cantilever. Here demonstrated by the use of a popsicle stick.

The probe tips are made with a drop of honey. I tried maple syrup first but it wasn’t viscous enough.

If you were actually building one of these you’d want it to be small; microns-wide.Rather than use a sugar-based needle you’d etch it out of whichever material. (Which material? There’s a diversity of options. Not going to cover ’em.) The tip itself gets down to a couple nanometers wide for a standard AFM, It can be less if you’re using one of the specialized variants. Oh, and build a lot while you’re at it; these things break easy. A quick internet search shows they cost about \$9 per, if you buy in bulk. The cheap ones. Replace ’em using a pair of tweezers and a steady hand.

How does it work? There are a number of variants on the general principle, but they share certain points in common. Take your cantilever and move it across the surface. The cantilever will bend upwards when your needle is crossing a bump, and downwards when it’s going over a dip (at that scale there are forces which cause these things to tend to stick to one another.) Bounce a laser beam off the back of your cantilever; the laser will pick up and amplify the deflection. Measure where the laser ends up with a photodetector and you’re golden!

What good is it? Over certain domains, you get better images than a scanning electron microscope. You also don’t need to destroy the sample in order to image it (whether you do anyway is your business. You vandal.) Me, I’ve never used these things much; never since I left the tech school. But let me tell you about that project I alluded to up above.

We were attempting to make graphene. Well, we never actually made graphene; that scotch-tape method isn’t as easy as they make it sound. But thinking ahead, if we made graphene how would we know? The usual method is to look for it under an optical microscope. We wanted to use an AFM to verify we were getting a one-layer-thick bit of graphite. And to make sure we could measure something as astonishingly small as the distance between one layer of graphite, well, we tried to do just that using a solid mass of graphite. Here’s the actual scan result:

Lighter areas are taller and darker areas are shorter. The garbage-bag-falling-into-the-can shape that results is an artifact of measurement. Really we’re looking at that y-thing. All that cranium-smashing math done to the 3D image is meant to remove that garbage curvature.

That one black line on the last scan? This profile is an image of that cross-section. We’re measuring the height of individual jumps to see how small of a distinctive jump we can get. Come to think of it this nanometer-level precision would come in handy looking at Beto O’Rourke’s poll numbers.

How did we do? The literature value for the distance between layers of graphite is 0.341 nm. The smallest height we measured was ~20% higher than that; which indicates we were probably measuring a step of exactly one layer of graphite. Why is it some fraction of a layer thick? I still wonder about that; I don’t have a good answer. But let’s take a quick detour back to the main subject so we can wrap this one up.

One variant on the Atomic Force Microscope, (it was invented first but it’s harder to grok), it’s called the Scanning Tunneling Microscope. In that case, you merely wave your probe over the surface. You keep an electric potential difference between your sample and your probe. Electrons will occasionally quantum mechanically tunnel from one to the other (yeah, this is why we didn’t start with this one). The number of electrons that tunnel will depend on the potential barrier, or in real words the distance between your sample and your probe. Measure the current running through your loop and you’ll have a measure of the profile of the surface. This makes an extremely sensitive method, and it can be used to image — and manipulate — single atoms.

Here we have a short video of IBM showing off. (They have a perfect right to brag, they invented the machine and the technique) What you’re looking at is an animation where the still photos are ‘drawn’ by carefully placing individual atoms on a cryogenic-temperature substrate. May I present to you a boy and his atom:

Join us fortnight next when we take a final stab at measuring micro- and nanoscale thicknesses in “Layers on the Wafer of my Mind” or “your Shaolin Shadowboxing is weak to My Thin Film Interference!”

This is part 33 of my ongoing series on building a computer, the cheap-as-we-can-get way. You may find previous parts under the tag How to Build a Computer. This week’s post has been brought to you by Wikipedia! Yes, once in a great while Wikipedia has has a human-readable article on a scientific topic, and this week we struck gold! Wikipedia!

Published in Science & Technology
Like this post? Want to comment? Join Ricochet’s community of conservatives and be part of the conversation. Join Ricochet for Free.

1. Member
Matt Balzer, Imperialist Claw
@MattBalzer

I was going to do an “I saw Atomic Force Microscopy open for X” gag but then I realized there was no need.

2. Reagan
GLDIII Temporarily Essential
@GLDIII

Someday we will learn how to build a computer…..once we learn the name of every tree in the forest. -NYTimes Review

3. Member
Judge Mental
@JudgeMental

Good, old IBM style soundtrack on the movie.  Right up there with their ideas on video.

“Why would anyone need more than eight colors?”

4. Contributor
Gary McVey
@GaryMcVey

Good, old IBM style soundtrack on the movie. Right up there with their ideas on video.

“Why would anyone need more than eight colors?”

Agreed about the soundtrack. I enjoy the music in old IBM ads, often sounding like a New England-based quintet that plays semi-classical with a semi-whimsical flair.

5. Contributor
Gary McVey
@GaryMcVey

One question: if we’re “reading” atoms, how come the atoms making up the background don’t show? In other words, my dumb question is how could you distinguish between the foreground and background atoms?

6. Member
Judge Mental
@JudgeMental

One question: if we’re “reading” atoms, how come the atoms making up the background don’t show? In other words, my dumb question is how could you distinguish between the foreground and background atoms?

Pay no attention to that man behind the curtain.

7. Contributor
Gary McVey
@GaryMcVey

Hank does a great job of explaining what it does and how it does it, but I have to admit, it’s one of those technologies that seems like it shouldn’t work in practice: the needle could never be fine enough, the signal would be submerged in random Brownian noise, the laser wouldn’t be able to reflect fluctuations that are vastly tinier even than the wavelength of the beam.

8. Contributor
Gary McVey
@GaryMcVey

I knew that someday a wise, gifted teacher could finally explain all this stuff.

I just didn’t expect him to be 34 years younger than me.

9. Member
Judge Mental
@JudgeMental

Good, old IBM style soundtrack on the movie. Right up there with their ideas on video.

“Why would anyone need more than eight colors?”

Agreed about the soundtrack. I enjoy the music in old IBM ads, often sounding like a New England-based quintet that plays semi-classical with a semi-whimsical flair.

I was referring to their hardware.  Decent graphic and sound cards were after market stuff.

10. Contributor
Gary McVey
@GaryMcVey

I bought a third party add-on for my Atari that bragged it could give me 256 colors. Probably at 240 x 320.

11. Contributor
@HankRhody

One question: if we’re “reading” atoms, how come the atoms making up the background don’t show? In other words, my dumb question is how could you distinguish between the foreground and background atoms?

Think ‘braille’. You need to work with a flat background area so that the individual atoms you’re using as pixels in your film stick out. After that, think of it in terms of ‘focus’. A Scanning-Tunneling Microscope doesn’t drag the tip along the surface like the AFM I spent most of the article describing; it waves it over the material at a pre-set height and measures the strength of the signal received. You scan over the pixel atoms (which are sticking out on top of the surface) and you get a stronger signal than off of the background. Up the contrast on your computer image of the signal you’re getting and you’ve effectively left the background out of focus.

IBM also produced a ‘making of’ video, which doesn’t answer your question. Or discuss the soundtrack. I’ll link it here for completeness’ sake.

12. Contributor
@HankRhody

Hank does a great job of explaining what it does and how it does it, but I have to admit, it’s one of those technologies that seems like it shouldn’t work in practice: the needle could never be fine enough, the signal would be submerged in random Brownian noise, the laser wouldn’t be able to reflect fluctuations that are vastly tinier even than the wavelength of the beam.

Among other things yeah. From that video I learned that they’re actually shuffling carbon dioxide molecules around, (I’ll forgive them the extra two atoms, but note that this leaves the door open for someone to make a slightly smaller movie). That, however, gives rise to a different question. What did they do with the water molecules?

Start with your surface in a chamber that’s (let’s say) two liters large. At atmospheric pressure there are about six times ten to the twenty-second power atoms of atmosphere in that chamber. Some proportion of that is going to be water vapor, which tends to stick to things. Especially things at cryogenic temperatures, the temperatures necessary to keep your atoms stuck to your surface so you can make your movie.

So pump it down, from atmosphere to low vac to high vac to ultra-high vacuum. You’re left with one atmospheric molecule out of every trillion you started with. But run the numbers; you’ve still got ten billion molecules of atmosphere in there. None of those are water? None are randomly sticking to the surface and photobombing your animation?

They probably did a couple purge cycles where they pump it all the way down and fill it with carbon dioxide and pump it down again to drive out most of the other gasses. Then you run an initial scan on your surface, see how much carbon stuck, and that’s the number of pixels you get to work with. But all that’s speculation. Once again, there’s a whole world of stuff hidden here.

13. Contributor
@HankRhody

Gary McVey (View Comment):
the laser wouldn’t be able to reflect fluctuations that are vastly tinier even than the wavelength of the beam.

This is also something I’ve never gotten a good answer on. If you make clumps of gold small enough strange things happen. One thing is that gold changes color. You get a bright red color out of it. Actually, the medieval method for making stained-glass windows uses gold nanoparticles for that lovely red color.

Why does it change color? I don’t know. How does color even make sense when you’re talking about a particle that’s smaller than the wavelength of the color you’re describing? I don’t know either. Something they said they’d show me in the nanoscience progrma, but never seemed to get around to it.

14. Contributor
Gary McVey
@GaryMcVey

It’s reassuring when even young Obi-Wan Kenobe says, “Midichlorians? That crazy stuff never made any sense to me either.”

The big difference is Hank’s brand of alchemy produces tangible, useful, reproduce-able products.

15. Inactive
SkipSul
@skipsul

IBM’s animation reminds me of Douglas Adams’s account of how the Infinite Improbability Drive was invented, and what happened to the poor scientist who had the breakthrough – lynched by other scientists for being a smart-ass.  That kind of animation is extremely clever.