Conductive and Piezoresistive Thermoplastic
A group primarily from the University of Warwick have published a libre-licensed, open-access article describing the capabilities and method for creating a conductive thermoplastic that exhibits piezoresistive behavior. Excellent work. Way to go!
I wanted to publish my thoughts on some things before joining the workforce fulltime. Like many others, I’ll be under a contract where everything I think and create must be reported to and given ownership by the corporate overlords who will be unable to bring instances of the ideas to market (see The Innovator’s Dilemma). While the place I’m going to work for doesn’t fabricate their own chips, but they do do plenty of prototyping, so I’d like to dump some of my ideas, despite them not yet being reduced to practice, before my thoughts belong to someone else. Honestly, I resent the practice and can’t wait to build up enough money to survive on if I found a startup that doesn’t make money for the first five years.
Probably the most useful next step would be to hone Rhys Jones’s solder deposition methods. It might be useful to explore additional solder types than were covered in his original study, and while these would be pricey, the material properties of indium (low melting point) and silver (better flow) may be worth investigating as potential solutions to the deposition issues. While I’ve yet to come across a strong argument for why depositing solder would be better than etching copper, solder does have safety requirement and cost advantages over milling copper. I suppose working with acid may have more burdensome safety requirements than soldering.
So if we can deposit electrically conductive metallic material, why not try depositing magnetic metallic material? Zinc oxide with the proper doping may serve as a useful ferromagnetic material. With it, one could fabricate both the ferromagnetic part of a motor. With conductive material laid out in a way that creates an inductor, one could also create the electromagnetic part of a motor, allowing a RepRap to replicate its own motors!
So the semester is over and I was not able to accomplish all I set out to. If I could do it over, I would concentrate a lot more heavily on one specific issue, such as fabricating conductive traces. While certainly interesting, and hopefully helpful to otheres through this blog, the survey of possible fabrication methods for conductors, electrical components (resistors, capacitors, inductors), and electronic components (diodes, transistors), was quite broad and ended up being rather theoretical. I hope, however, that any future students continuing this line of research will be able to work on physical proofs of concept more immediately, thanks to an already constructed RepRap and a useful collection of previous work.
Some of my recent work has been to compile information on printing electronics on the RepRap wiki. I’m hoping to provide a collection of historical and ongoing attempts at printing electronics, be it wiring and traces, electrical components, or semiconductor devices. I also made some updates to the Desktop Semiconductor Foundry article on the Future Wiki on Wikia. The original article was very optimistic about a laser sintering technique similar to burning a CD or DVD, but its predictions of (expensive) devices existing by 2010 are to my knowledge no quite accurate. Hence I toned down some of the speculation when adding inkjetting and thermoplastic extrusion techniques to the list of possibilities.
Under the direction of my advisor, I made cadmium sulfide in the lab from cadmium chloride and sodium sulfide. There are several more steps necessary, however, to actually put it to use. In its current condition it is precipitate in solution, which can be centrifuged and extracted to yield a powder. The powder will need to be resuspended in a solution suitable for ink-jetting or glued together somehow. A popular conductive polymer for that gluing together is Alq3, which we happen to have on hand. Its melting point is too high (above 300 °C) to extrude it, though. Thus, we’ll probably be going the ink-jetting route. I’ve been wrestling with which approach might be better–ink jetting or extruding–but now the question is answered, at least until we find a suitable conductive thermoplastic.
So the order from MakerGear finally arrived and I’ve begun to put together the Prusa Mendel. It’s quite a lot of fun, and reminds me of building with Legos.
So the graphite-resin-ethanol paint we requested a sample of eventually did come a while back. I’ve been playing with it, and it seems great for what it does. The alcohol doesn’t evaporate as nicely when trying to apply more than a thin film, however. Also, the resistance of a thin film isn’t quite as low as one might want for a conductor. I think a graphite and resin material could do quite nicely for resistors, though.
So, I was looking up what it might take to make just a resin and graphite compound myself. The third hit for “melting point of resin” on Google is the Wikipedia page for rosin. Not far down in the article, it’s mentioned that rosin is used as flux in solder. It now seems painfully obvious, but why not stick a strand of solder down the RepRap extruder head? It turns out that Rhys Jones did exactly that two years ago.
I’ll test out how well lead-free (mostly tin) solder adheres to ABS, but my hopes are pretty high for finally finding a good enough conductor to move along to the semiconduction. In other news, I’m still waiting on the Mendel we ordered in February, but I may be able to work with the DREAMS Laboratory before it comes. With their mechanical engineering experience with MakerBot 3D printers, they may be able to easily shoot down some of my likely impractical ideas, and suggest alternatives. Also, when the machine does come, I can probably use some help putting the thing together, so I’ll post on several mailing lists asking if anyone wants to join in the fun.
While we never received a response received a delayed response to our request to sample the one carbon company’s graphite in ethanol solution, I was able to find powdered graphite at the local art store and we mixed it with water and ethanol in different bottles.
Depositing the substance out of a pipette onto an ABS substrate, the ethanol spread rather easily. Depositing strips of mostly uniform width was possible, but additional material added while the first layer was still wet made the strip wider. In order to achieve thickness with the ethanol, several passes or a channel in which to deposit the solution will probably be necessary. In contrast to the ethanol, the water droplets balled up and it was easy to deposit a thick blob. However, it was very difficult to give the blob much length as once a droplet touched the main blob, it was pulled in.
I ran across some notes about surface tension and printing, but in the end in may just be easiest to first construct a channel in which to deposit the solution with graphite powder. It is important to note that the ethanol needs much less time (around 15 minutes) to dry and so is probably the preferred solvent.
Sticking multimeter probes on top of the graphite deposited with ethanol, I measured resistivity on the order of 1000 Ohms. Pressing the probes on top of the dry powder rubbed the powder away in the case of the thin ethanol-deposited strip. In the case of the thick, water-deposited blob, the pressure cracked off pieces, and I wasn’t able to get a reading in time. I’ll conduct a next round of testing, which will hopefully yield more trustworthy numbers from a better designed test, and will report back next week.
In order to fabricate semiconductor devices like diodes or transistors it is of course imperative to have a substrate on which to do the fabrication. Additionally, to do anything with the devices, electrical connections are necessary. The conductor forming the connections would preferably be made out of some cheap and easily manipulable substance.
Not wanting to wait until after our RepRap order was shipped, assembled, and printing, we ordered sheets of ABS to test out as the substrate. Biodegradable integrated circuits made on a PLA substrate could be useful for systems like remote wireless sensors that you don’t want recover. However, I think the more commonly useful case would be recycling technical materials into new systems. Wouldn’t it be great if hardware flaws could be fixed by melting the device down and printing out a fixed revision, or that outdated piece of electronics could be made into the newest version with little or no extra material? That goal for recycling and upcycling, as well as worries that a biodegradable substrate might degrade from one test to the next, drove the choice of ABS over PLA.
To investigate the thought that carbon may fit in well as a good, cheap conductor, we’ve also requested a sample of one carbon company’s graphite in ethanol slurry. Since the graphite may potentially slough off under rough conditions, sealing it in place will likely be necessary. Some sort of electrical insulation would have been necessary anyway, so that may not prove much of an inconvenience. Putting graphite in the polymer may also be possible, and if the concentrations can be easily varied, this would be especially useful for higher-value resistors.
Given just a substrate and conductive material, it is possible to fabricate resistors, capacitors and inductors. This post will briefly review some of the patterns currently used to create these components on printed circuit boards and integrated circuits.
While every trace on a printed circuit board has some resistivity, for a trace to exhibit a ratio much greater than one with the other traces, it has to be significantly longer than the others. Planar resistors can look just like the zigzagging symbol for a resistor. The second option for making resistors is to mix the conductive material in different ratios with the polymer. This would probably look like slightly discolored but otherwise identical fused filament.
American resistor symbol by Wikipedian K. Bolino (PD)
Fused filament fabrication with conductive filaments
Like the case was with a resistor, printing out the symbol for a capacitor with a conductive material can make a crude planar capacitor. More common in integrated circuits, however, is the metal-oxide semiconductor capacitor.
MOS Capacitor
Planar inductors are also possible to fabricate. Conductive material laid down in a spiral can function as an inductor.
Planar Inductor by Wikipedian Cyril Buttay (CC-BY-SA)
Forays Into Desktop Fabrication 