There are a lot of situations where a research group may turn to an electron microscope to get information about whatever system they might be studying. Assessing the structure of a virus or protein, analyzing the morphology of a new nanoparticle, or examining the layout of a semiconductor all might require the use of one of these devices. But if your research involves the electron microscope itself, you might be a little more reluctant to tear down these expensive devices to take a look behind the curtain as the costs to do this for more than a few could quickly get out of hand. That’s why this research group has created their own electron detector.
Specifically, the electron detector is designed for use in a scanning electron microscope, which is typically used for inspecting the surface of a sample and retrieving a high-resolution, 3D image of it compared to transmission microscopes which can probe internal structures. The detector is built on a four-layer PCB which includes the photodiode sensing array, a series of amplifiers, and a power supply. All of the circuit diagrams and schematics are available for inspection as well thanks to the design being licensed under the open Creative Commons license. For any research team looking to build this, a bill of materials is also included, as is a set of build instructions.
While this is only one piece of the puzzle surrounding the setup and operation of an electron microscope, its arguably the most important, and also greatly lowers the barrier of entry for anyone looking to analyze electron microscope design themselves. With an open standard, anyone is free to modify or augment this design as they see fit which is a marked improvement over the closed and expensive proprietary microscopes out there. And, if low-cost microscopes are your thing be sure to check out this fluorescence microscope we featured that uses readily-available parts to dramatically lower the cost compared to commercial offerings.
10 thoughts on “Building An Electron Microscope For Research”
This is actually pretty interesting. Thought their spiel about “The closed attitude of commercial entities about how exactly the different parts of electron microscopes work, makes it even harder for newcomers in this field.” is nonsense. SEMs in general have been around for a long, long time and other than miniaturization and reducing the control cabinet to a PC not much has changed.
A lot of SEMs just come with secondary electron detectors and BSE (Back Scatter Electron) is an option. I may have to give this a try and build one.
My observations though, commercial BSE detectors allow you to turn on/off quadrants of the diodes which would be a nice add. Also you generally want to keep as much stuff out of the chamber that will outgas as possible. It would probably be better to have the diode ring separate from the amps because who knows what will outgas from the ICs, connectors, and whatever residue is left behind from soldering.
Now someone needs to come up with an open source room temp EDX. That would be very nice. I’s love not to need LN2 for my EDX.
Pump-down time and out-gassing of the PCB and SMT resistors and caps are also what I wanted to know. The circuitry is right from ap-notes and pedestrian but executed well. It would be nice to have gold wire-bonding available. I like the diodes removed from plastic tips.
Do you get en effect like side-lighting (shadowed features) from the quadrant view?
Yes, quadrant view allows for a more topographical view by yielding more “shadows”.
EDX systems with Peltier coolers are the way to go.
In fact doing an EDX build is the next thing we are looking into and that seems entirely feasible except maybe for hitting the ultimate resolution specs as compared to commercial solutions.
If you want a nice cheap high resolution stage, perhaps this might be a way to go about it,
I like how they repurposed a cheap photodiode, turning it into an electron detector by dissolving away its electron-blocking case and re-bonding the naked silicon to their PCB. Great hack.
Regarding the suggestion made to make an entire microscope, there are I think a few things that are much harder than this humble detector. The most challenging is vacuum or making good enough vacuum at an affordable cost for home hacking. (For me, finding a second hand turbo pump is not a solution as it doesn’t scale to another diy kid hat wants to the same). I would love to hear out of the box ideas about this. Electron gun is also tricky, but there are at least several examples on hackaday where someone makes a hot filament source from easily sourceable materials. Compared to other SEM builds I have seen on hackaday it would also be good to focus on intrinsic safety (high voltage, Xrays…) .
Not sure how high the vacuum requirement really is, the paper shows some pictures taken with a low vacuum setup (700 Pa), this can easily be reached with a low-cost single-stage pump, no need for an advanced setup with a turbo molecular pump to reach that pressure. Admittedly the image quality is significantly reduced with this low vacuum setup but it can still show much smaller detail than what would be possible with an optical microscope.
Still wondering how this is even possible at such a low vacuum level, according to the calculator on https://www.omnicalculator.com/physics/mean-free-path the mean free path is just 20 µm at a pressure of 700 Pa (assuming Nitrogen and a temperature of 300K). The whole setup can’t possibly be so small that the electron beam won’t hit lots of air molecules on the way to the sample.
Another challenge for a DIY setup is the whole electron beam generation/focussing, that also needs some engineering to get it right.
This calculator gives the mean free path for a molecule in the gas to collide with another molecule. Here we need to know the scattering of a fast electron with this ‘cloud’ of moving gas molecules. Typically during such an interaction the electron will change direction and lose some energy, but not all of its energy. I didn’t immediately found an online calculator but a way to look at it would be to compare the scattering in a thin slab of material with say a few cm of travel length through a low density gas and take into account only the projected density.
Please be kind and respectful to help make the comments section excellent. (Comment Policy)