Lithium ion batteries are in all kinds of high performance devices. I made a little video talking about how they work. They store a good amount of energy, release it fast enough for a cell phone, and don’t blow up all that often. Typically, lithium or lithium-ion batteries need to be assembled without the presence of water or oxygen which makes them less than ideal for DIY. The old potato battery with lithium is a bad idea.
What about sodium batteries? Sodium is another alkali metal like lithium. It should work similarly, but it’s way cheaper. You can buy sodium hydroxide for $8 a pound at the hardware store. That’s about half elemental sodium by mass. That’s an order of magnitude less expensive than lithium. The price per watt-hour stored could get considerably lower. The lowest that prices can go is the price of materials, and sodium is cheap. For now, though, the cost of materials is not the biggest part of the final battery price. The cost of assembly, housing, and associated electronics is a bigger share of the pie, so it makes sense to work on those first.
What is not so good about sodium batteries? Sodium metal stores less energy per atom, so you get a lower voltage. It’s also a lot heavier, so you get less energy per unit mass. Sodium is more explosive in contact with water than lithium. It’s harder to pack into electrodes. Where graphite holds lithium and lets it migrate, the equivalent for sodium doesn’t work so well. As you charge a LiFePO4 battery, you move lithium into a graphite cathode for storage. Lithium slips between the layers of graphite, but sodium doesn’t fit.
If those problems could be overcome, sodium batteries would be great for stationary storage. Sodium batteries would be heavy (no good for mobile) but cheap (good for large scale). Aquion, Faradion, and GE are all working on it. Several articles from academic labs have come out very recently showing off sodium battery technology. So what are the major hold-ups? Two reviews talk about the issues, and they are all challenging.
I’m interested in iron batteries. Iron is cheap, ubiquitous, and I think the sheer volume of iron available may make it a good candidate for grid storage. So I’m playing with it a bit. Check out the vlog:
I had an idea to build an index of refraction-based detector for an electrophoresis set-up. When light passes through a prism, it bends. That’s how prisms cast rainbows. The angle at which the rainbow shows up depends on the material of the prism. If the prism is made of glass, you get a rainbow at one angle. If the prism is made of water, you get a different angle. We can use this phenomenon to detect changes in the material. We need a prism that we can fill with different materials. Then we can see how the angle changes. I built such a prism out of acrylic with my laser cutter. It looks like this:
I haven’t started flowing things through the prism yet. I’m still figuring out how to detect the angle change. But there definitely is an angle change. I can see the difference in angle between when I have air in the prism and when I have water in the prism. Have a look – that’s air on the left and water on the right:
So now the question is how tiny a difference can I detect? Adding small amounts of sugar to the water will change the index of refraction a tiny bit. That should change the angle by a small amount as well. If I can measure the angle very precisely, I should be able to detect very small changes in the sugar content of the water.
I’ll try to get my students to build generation 2 of the device. Design, build, test learn. The work will teach them.
There is a neat paper in this issue of the journal Lab on a Chip. The Tuteja lab out of Michigan developed a clever way of making droplets using a laser cut jig and an open platform. I mean literally open (not like open source). They make water in oil droplets that float across the surface of a hydrophobic chip. It reminds me a lot of the beer sphere.
I caught this image of a beer sphere suspended on its surface tension on the surface of a glass of beer. It was so persistent that I was able to get my camera and come back to the table to take a picture.
The Slo-Mo guys got really good footage of surface tension droplets. It’s an interesting phenomenon. Surface tension prevents a water droplet from merging with the water surface. If you have an oil surface, the water will not merge at all. Surface tension isn’t needed. But it looks similar.
I meant to get in the lab this morning and fabricate some devices, but I am not feeling motivated. I ramble a bit about that in the video today. I’m very worried about science funding. That’s dumb. Can’t do anything about it. It’s worrying for nothing. But I am worrying anyway. That’s four hours of my life I will not get back.
Focus on science. Rise above.
“Monodispersed microfluidic droplet generation by shear focusing microfluidic device,” is from 2006. It’s a study on the design of the flow focusing droplet generator. It explores the role of flow rate and pinched geometry on the droplets. At the time, it wasn’t completely clear what these droplets could be used for. I was looking at them as little storage containers for cells.
In 2016, digital PCR was a clear application of this technology. A PCR reaction was segregated into lots of little droplets. Each droplet either has a DNA molecule or does not. As a consequence, the PCR reaction makes it go green or not. Instead of trying to interpret different levels of green fluorescence (which is relatively hard to quantify), the scientist can just count the bright droplets (much easier to quantify). “Centrifugal micro-channel array droplet generation for highly parallel digital PCR” presents an unconventional droplet generator to make lots of little droplets for that application.
The application I’m working toward is a little different. I want to make particles based on these droplets. The little particles will have a sensor on them so that we can detect what is happening near to the particle. The particle might then respond by glowing green or by releasing a drug. Similar particles have applications in cosmetics and lubricants. I think that we can make them smarter. We can apply them to research (reporting cell environments) diagnostics and maybe therapeutics (some day).
One of the reasons I am trying to develop these techniques for acrylic microfluidics is to shorten the design-build-test feedback loop. With the PDMS techniques I worked on and grad school the loop is about a week long. It takes a day to design, a few days to get the photomask, and a day or two to fabricate. If nothing goes wrong, I use to be able to get a design tested on day five.
With acrylic based microfluidics, the loop is much shorter. I can modify a design, cut it with the laser, fabricate the chip, and have it tested all within three hours. Three hours is a significant improvement over five days.
This morning I came in and cut a new chip with some new parameters. I was hoping to see a narrower width channel which would produce smaller droplets. I didn’t get that, but I can try again (tomorrow not next week). Plus, I learned something. I need to add a viewing area to the chip to see the droplets better.
I made a little video of some of the process. I skipped the bonding step in the video because it’s just 10 minutes of waiting while the hot press does its work.
Here are parts 1 2 3 4, 5 and 6.