There’s a simple reason why high-tech gadgets like PCs, TVs and smartphones are so cheap: they’re mass-produced. By spreading out huge engineering costs over equally huge production volumes, the cost per item can remain quite low. The flipside to this is that devices with only a small niche market can be extremely expensive even when they seem quite simple.
[Baird Bankovic], an undergrad student at Penn State University, discovered this when he was working with a scanning tunneling microscope (STM). He noticed that the machines used to make STM probes, a pretty straightforward process, cost north of $7500. This inspired him to make a cheap STM probe etching machine using simple homebrew parts.
If you’re not familiar with scanning tunneling microscopy, here’s how it works in a nutshell: a very sharp tungsten needle is positioned a few nanometers above the sample to be analyzed, and a small voltage is applied between the two. Through an effect known as quantum tunneling, a small current then flows between the probe and the sample. By observing this current and scanning the probe across the sample, a three-dimensional picture of the surface is obtained with sub-nanometer-level resolution.
One of the many factors that determine the quality of the image is the sharpness of the probe. Because a very sharp probe is extremely fragile and prone to oxidation, they are typically made on-site by dipping a piece of tungsten wire into an etchant in one of those costly machines.
That’s exactly what [Baird]’s device does: a Petri dish on a 3D printed frame holds a volume of sodium hydroxide solution, while a jackscrew Z-stage moves a probe holder up and down. A piece of tungsten wire is dipped into the solution and a voltage is applied to start the etching process. Because most of the etching happens at the liquid’s surface, the wire gets progressively thinner at that point until it snaps and the bottom half drops off. When this happens, the current through the wire changes rapidly, which triggers the machine to pull the wire out of the solution and stop the etching process.
The resulting probes have a well-defined sharp tip with an estimated width of about 50 nanometers — pretty impressive for such a simple setup. The entire hardware design is open source and available on [Baird]’s GitHub page for anyone to replicate. Nanometer-sized needles might only seem useful for those with a professional STM setup, but they also come in handy for all kinds of homebrew atomic-scale imaging experiments.
I need something like that to sharpen TIG electrodes.
XD
I love it when someone realizes they can make a niche product for far less.
Really cool how so simply these probes are made. When I first heard about them, I thought they were machined -_-
They are! This is a chemical machining process! ;)
It’s always nice if people find simple ways to do things that were previously difficult or expensive.
I seem to recall that years ago, it was done by breaking off some brittle material, in the hope the point breaks in a way that leaves it sharp.
I am a bit surprised by the way the etching works. Apparently it works faster near the surface, but I don’t know why. Another thing I don’t really understand is why the etching is stopped as soon as the tip breaks off. My intuition suggests that the very tip will get sharper if the etching is prolonged a bit more, but keeping it too long will make the tip stubby and blunt.
I guess you could also experiment with varying the height during the etching to create a better defined taper on the point. If you have a taper first, then etching further will always result in a sharp point.
Yeah, I used a crude STM setup in an undergrad lab assignment in the year 1999 AD, and I had to make fresh probes by drawing a wire. IIRC I just used pliers, having seen a demo of what it looked like when the wire had snapped in the way that forms a (very) sharp tip. I was a chemistry student, so I was used to doing similar stuff with hot glass tubes, but it certainly wasn’t a technique that takes years to master or anything.
The yield of usable probes was quite low, and you had to just try keep trying them with a known sample until you got a decent image. If this was your day job, a more reliable method would easily be worth $7,500. But for hobby purposes, the pliers totally work.
I remember when an STM was being described at a talk ages ago at the Univ. of N. Carolina (CH). When someone asked how the probes were made, they said they just used wire cutters, cutting at a steep angle. I assumed there had to be a certain amount of technique involved, but I was surprised that it could be done with simple tools.
I have current monitoring on the etcher, and when the tip breaks off below the surface, the etching current also drops. When the current has a large change, the machine stops the etching, ensuring the tip does not get etched blunt. :)
Have you tested this?
I don’t believe the “tip gets etched blunt”. In fact, in the old days old files used to get sharpened by etching a layer off. If you go on etching too long, then of course you will etch the whole tip away and it gets blunt. But as long as the majority of the tip is a taper, (and assuming even etching) then the tip will stay sharp, or it will become sharper if the first break of left the very tip a bit blunt.
Just imagine the shown picture of the tip, but with a slightly blunt tip, and then draw a line on paper where the edge would be if etch of around 10% more of the diameter. This will result in a sharp tip. It certainly is worth to try to verify on your real tips. At the very least, the time at which you turn off the current and stop / slow the etching is not as critical as you may think.
Can anyone explain why etching happens faster at the surface?
Hi PWalsh,
A product of the etching is tungsten trioxide. As the tungsten wire is etched, tungsten trioxide falls out of solution and coats anything below. The bottom of the submerged wire is coated with tungsten trioxide from etching happening above. This results in the submerged wire being coated in electrically insulating tungsten trioxide; only the very top is not. Thus etching happens preferentially at the surface of the etchant.
Cheers – Baird
Just what I needed for my record player ;)
😂
A bit too fragile I suppose. I was going to throw one on, but then realized how much difficulty it would be to balance the tone arm to reduce the weight on the tip to keep from bending.
What no knife sharpening nerds in the comment section? What about Electro chemical Metal Blade sharpening? Or Reverse 3d etching in a Metal Plate maybe Form molding?
I might be thinking of another instrument, but this seems like the thing that somebody told about once where the probes can be made by simply cutting a piece of the wire with a pair of pliers – often enough, so the story went, the tip was just that one lonely atom. Maybe it was an atomic force microscope now that I think of it…
Hi Andrew,
This is the same instrument you are thinking about. People also cut platinum iridium wire with pliers, and hope the tip is sharp. This works well for a lot of people, and I used to do that; however for our research we need to position STM tips on a sample within a micron or so, and having the long, well-defined geometry of etched tips allows us to more accurately position the tips.
Atomic force microscopy (AFM) typically uses a piece of diamond on the end of a semi-rigid cantilever. Then the cantilever is vibrated at its natural resonant frequency, while a laser measures the vibration amplitude. As the diamond on the end of the cantilever comes close to a surface, the van der Waals forces bring the cantilever out of resonance, which can be realized by a drop in amplitude on the laser’s deflection.