Contributor Post Created with Sketch. Recommended by Ricochet Members Created with Sketch. How to Build a Computer 30: SEMsational


This is a continuation of last time’s discussion on Electron Microscopy. In that one, we covered the question of why you’d want one of these and gave a summary of how you’d work one. Take some electrons, throw it at your sample, and watch what bounces off for information. Sounds so simple when we put it that way, right? This week we’re talking about what happens when you actually buckle down to do it in practice.

Taken from Chem lab, when there weren’t any chem techs around to stop me.

Okay, just looking at the thing isn’t doing me much good. What’s going on there, and why? Start from the top. That bottle on the left? That’s for liquid nitrogen, used in the x-ray detector. (Neat! Why do we want to detect x-rays? That’s a subject for a future column.) The cylinder on top marked “GEMINI” is your column; the electron gun is in the top, and the rest of it contains the magnets for focusing and directing the electron beam. The cube-ish box it’s sitting on is your sample chamber; the front pulls out to reveal the stage where you’d put your puck holding the samples. The dark grey table surface is granite, used to lend stability to the whole apparatus. The cabinet it’s sitting on contains electronics and the vacuum pumps. Now let’s get to how all that works together.

Take Some Electrons

This is for all of you who want to know how to build your own electron gun, with which to crush those who oppose you. To bounce our electrons off of the sample we’re going to have to start with some real high-energy electrons, and to start with some real-high energy electrons, we’re going to have to energize them ourselves. Thus, we build an electron gun.


Three essential parts of your electron gun. You’ve got your hot cathode (in dark blue), you’ve got your Wehnelt cylinder (in light blue), and you’ve got your anode (in red). Your what, your what, and your what? Okay, start with the cathode. You heat up a filament, it gives off a whole buncha electrons. That’s straightforward enough. The electrons want to head towards the anode. Also reasonable. You control the strength of your electron gun by the electrical potential bias between your cathode and your anode. (Uh… the bigger difference between positive and negative the more energy your electrons get.) Generally, I’ve done SEM work between 5,000 and 15,000 volts.

But that’s not all; you’ve also got to direct your electrons. That’s what the Wehnelt cylinder is for. Okay, that’s the fancy name for it. You know what it actually is? A bowl. A metal bowl with a hole drilled through the center of it. You put a negative charge on it, that’ll repulse electrons. It’s a smaller charge than the positive charge on your anode. The anode accelerates the electrons, the Wehnelt cylinder directs them. The electrons end up dropping out the hole in the bottom, at a high energy.

Throw Some Electrons at your Sample

Here’s a question for you; what’s the difference between the atoms in your sample and the atoms in the air around it? The answer, of course, is that you care about what’s going on in your sample. Any electrons that bounce off the air get lost at best and contaminate your data at worst. Oh, and your electron beam is going to impart a charge to your air molecules, which in turn is going to distort the electron beam. Clearly we can’t do this with air mucking things up.

Okay, encase your electron beam in a stainless steel chamber (betcha feel like a real mad scientist now!) and let’s pump it down to hi-vac. Vacuum systems will get their own post eventually, but a quick word here. You start with a roughing pump to get most of the air out. It takes you from ~750 Torr (atmospheric pressure) down to ~10^-2 Torr, or roughly one part in ten thousand. From there you turn on the high-vac pumps. These things drop the pressure to ~10^-5 or 10^-6 Torr, which is where you want to operate. One part in 10 million. Which still leaves roughly a quadrillion air molecules in the chamber, but I guess that’s few enough to manage.

Now, running your samples under a vacuum can cause issues too. Most organic things tend to have water in them. Water will boil off in vacuum. So, for example, if you want a nice shiny picture of red blood cells you have to prepare them by hardening the membrane so they don’t all pop like balloons in the vacuum. (As mine did, when it got time to do that lab in school. Still don’t know what went wrong.) This, incidentally, is why chem lab nixed scoping my sausage.

And Watch what Bounces Off

What happens when the electrons hit your sample? Well, they go bouncing all over the place, providing you with your images, as we covered last time. And your sample charges. You dump a whole bunch of electrons onto a thing, it’s going to charge up. And if it charges up that’s going to interfere with your electron beam and… well, you get the point. To make sure it doesn’t charge overly much, well, here’s a picture of a SEM sample mount:

You can always find a spare SEM puck in the ‘clean these so we can use them again’ tray. Cleaning these is always someone else’s job.

That’s a small disk of Stainless Steel. On top of the stainless steel there’s a layer of double-sided carbon tape. Carbon tape is conductive. $32 for a roll on eBay right now. You stick your sample on top of the carbon tape. Done? Not yet. Is your sample conductive? All of it? To ensure you’ve got a conductive sample you sputter coat on a thin (couple nanometers thick) layer of metal. The metal is conductive, it will ground out excess charge. Mostly. If you keep your electron beam focused on your sample too long it’ll still charge and screw up your pictures.

The top image is brighter because the beam has been allowed to rest on it longer, charging the sample. Probably. Adjusting brightness, contrast and focus is a continual battle on these scopes; it could have a different shade for other reasons too.

These are only some of the joys that brighten the lives of the noble SEM technician. An oddly shaped sample or an injudicious tilt of your sample stage can break your backscatter electron detector. (Which, incidentally, is another reason the Chem Techs are so protective of their scope.) Or some clueless engineering type can request several thousand images for his pointless project. (Helped with that one I did.) Or some wacko could come down asking you to scope his sausage… Right, moving along.

We’ll take a detour out of materials characterization and the SEM in particular next time. I casually mentioned sputter coating a minute ago, and I’d like to spend a little more time on what that means and how it’s done. Then briefly back to the SEM as we cover X-ray vision. Join us fortnight next for “Volleys of Argon” or “Iridium Dreams”.

This is part thirty of my ongoing series on building a computer, the Betsy Ross way. You may find previous parts under the tag How to Build a Computer. This week’s post has been brought to you by Old Glory! Yes sir, she’s a grand old flag indeed. Old Glory!

[First – Silicon] [Previous – Electron Microscopy] [Next – Sputtering]

There are 8 comments.

Become a member to join the conversation. Or sign in if you're already a member.
  1. Matt Balzer, Imperialist Claw Member

    Me, I’m regretting that you didn’t take the suggestion as to how to do it with food.

    • #1
    • April 20, 2019, at 11:51 PM PDT
  2. Judge Mental Member

    Hank Rhody, Meddling Cowpoke: You’ve got your hot cathode (in dark blue),

    My cathode is moderately attractive at best.

    • #2
    • April 21, 2019, at 12:28 AM PDT
  3. Judge Mental Member

    And the obligatory PIT comment:

    I saw Scoping My Sausage open for the Sex Pistols at CBGB’s in ’79.

    • #3
    • April 21, 2019, at 12:29 AM PDT
  4. Randy Weivoda Moderator
    Randy Weivoda Joined in the first year of Ricochet Ricochet Charter Member

    Just curious, Hank. Do you decide on your topics one episode at a time, or do you have a grand outline all the way to the end?

    • #4
    • April 21, 2019, at 3:00 PM PDT
  5. Hank Rhody, Badgeless Bandito Contributor

    Randy Weivoda (View Comment):
    or do you have a grand outline all the way to the end?

    You give me entirely too much credit for planning.

    • #5
    • April 21, 2019, at 6:19 PM PDT
  6. Randy Weivoda Moderator
    Randy Weivoda Joined in the first year of Ricochet Ricochet Charter Member

    Hank Rhody, Meddling Cowpoke (View Comment):

    Randy Weivoda (View Comment):
    or do you have a grand outline all the way to the end?

    You give me entirely too much credit for planning.

    When you started, did you think you would have 30 articles with many more to come, or did you think a couple dozen would do it?

    • #6
    • April 21, 2019, at 7:37 PM PDT
    • 1 like
  7. Hank Rhody, Badgeless Bandito Contributor

    Randy Weivoda (View Comment):

    Hank Rhody, Meddling Cowpoke (View Comment):

    Randy Weivoda (View Comment):
    or do you have a grand outline all the way to the end?

    You give me entirely too much credit for planning.

    When you started, did you think you would have 30 articles with many more to come, or did you think a couple dozen would do it?

    The former. When I started I was planning on going into more depth than I already am. Turns out computers are complicated though.

    • #7
    • April 21, 2019, at 7:41 PM PDT
  8. Gary McVey Contributor
    Gary McVey Joined in the first year of Ricochet Ricochet Charter Member

    This is really good. Scanning Electron Microscopes are a source of constant hilarity, to be sure. I’m not piling on the requests for more food references, not at all, but I have to let the wine wear off before I comment wisely, and frankly it would have helped if I’d ordered the home delivery cheese and sausage platter from RightFoods. 

    • #8
    • April 21, 2019, at 8:54 PM PDT

Comments are closed because this post is more than six months old. Please write a new post if you would like to continue this conversation.