crystals of potasium phoaphate incorporating the chemical amaranth dye

Amaranth Potassium Phosphate Crystals

Potassium phosphate crystal chemistry

This weekend I grew some potassium phosphate crystals with amaranth dye. I did this back in 2001 in Bart Kahr’s O-Chem class and remembered it recently. It’s a fun demonstration of the chemistry of crystal growth, the different chemistry of the crystal faces, and it’s pretty. I found Prof. Kahr’s paper[1] that gives a “foolproof recipe” and it did not disappoint. Even this fool could make it work.

As the crystals grow, each face of the crystal has a unique topology. The corners are growing with a different spacing of atoms than the faces, and the faces can be different from each other. Sometimes, the faces have the right spacing to allow a dye molecule to stick. In this case, there is a big difference between how well amaranth dye sticks to each face. So as the crystal rows, it only gets dyed in two quadrants.

We can learn about chemistry from crystals

Crystals are super useful to chemists. A good crystal of a chemical can be used to get x-ray diffraction data on the structure of the chemical. The most detailed structures are derived from x-ray diffraction data.

Knowing how molecules assemble into crystals is also really important to materials scientists. If you want to design a material from its atoms, you need to know how they are going to come together. I’ve been working on making an iron battery and reading up on battery chemistry. One of the interesting papers I read talked about designing a cathode material to hold sodium atoms. The chemists designed the “holes” in the structure to hold sodium atoms – and they needed to know how the other atoms would come together to make that shape.

Why chemically dyed crystals are cool

Of course, dyed crystals just look cool. Maybe that’s silly, but if you’re trying to teach organic chemistry, it’s good to have something visual and striking to hold on to. A lot of O-chem is solvents and white powder, so anything that sticks in the memory is a help.

The other reason I think that dyed crystals are so cool is that they dyes can be held still very precisely. One of prof. Kahr’s later papers used a crystal to hold a fluorescent dye in place at a specific orientation. Then they used a fluorescence microscope to look at single dye molecules[2]. I think that’s just really cool. I gather that they are more stable in the crystal than they are in solution.

I also made a time-lapse movie of the crystallization

Instructions (following [1])

  • Dissolve 17 g potassium dihydrogen phosphate (KDP) in 50 ml water with heating. Using a teflon stir bar helps.
  • Dissolve 4 mg amaranth dye in ~1ml of water and add to the mix.
  • Pour into a wide dish and allow to cool and evaporate slowly over ~4-24 hours.


1.Kahr, B. & Vasquez, L. Painting crystals. CrystEngComm 4, 514–516 (2002).

2.Wustholz, K. L., Kahr, B. & Reid, P. J. Single-Molecule Orientations in Dyed Salt Crystals. J. Phys. Chem. B 109, 16357–16362 (2005).

NORM 2017 Corvallis

I’m attending The Northwest Regional Meeting of the American Chemical Society in Corvallis. I just wandered around downtown. That was nice. My hope is to see some computational chemistry, commercialization, and nanoparticles tomorrow.

What tools are getting used for simulations? I’m especially interested in coarse-grained simulations of macromolecules. I see several Density Functional Theory talks and that should be interesting. Maybe folks from that world can point me in the right direction. Is anyone using tensorflow for such things?

There’s a panel on market-driven innovations. I would love to hear if people are funding academic labs through collaborations with industry. I feel like that would be a win-win, but I don’t know where to start there, either.

There’s also a bunch of analytical chemists giving MS talks and a “smart” nanoparticle talk. That’s just the morning session. I’ll have a hard time choosing.

If you’re in Corvallis and are reading this, do please shoot me a gmail (pballen). I’ll buy the first round at Tommy’s.

Negative and positive results in the quest for an all iron battery

Iron is cheap, and iron chemistry can be used to make a battery. If you want to buy a lithium-ion backup battery pack for a home solar system, it will cost as much as the solar panels. Effectively, a 24/7 solar system is about double the cost of a grid-tied system. The same is true for the grid itself. If the utilities want to move to cheap solar power, they will need to buy huge batteries. If utility companies tried this with lithium batteries, it would be such a big endeavor that it would mess with the lithium market. Iron is produced at such a huge scale that a move to grid-scale iron batteries wouldn’t completely alter the iron market.

I tried a dumb idea and it didn’t work. I tried to make an iron-oxide electrode for an iron battery. The idea was that iron oxide can be reduced to iron magnetite. That would be a cheap cathode for an all-iron battery. Plus, since iron oxide is a solid, it would stay where it was put and not diffuse over to the other electrode. So that would be nice, too.

Obviously (even to me at the time) iron oxide is an insulator, not a conductor. So if it is going to act as an oxidizing agent, it will need a path for electrons. Electrons can’t move through the iron oxide. They need to move through some other conductive material. So I embedded iron oxide particles in graphite.

2017-05-19 12_53_22-Krita

The result was nothing at all. The cell was dead on assembly. I could not detect the iron oxide reduction/oxidation with any instruments at my disposal. Other groups have reported the oxidation potential of the iron oxide nanoparticles. They put them in a suspension swirling near the electrode and that seemed to work. So maybe it’s possible, but I can’t get it to go. Iron oxide is out.

I made a better cell with Iron (III) EDTA as the oxidizing agent. It’s soluble so that makes things work better. I used a graphite felt as a current collector and it worked just great. The energy density is low (as expected) but it works.

2017-06-01 08_33_03-Krita.png

The next step is to optimize and stack up a bunch of cells. I think it’s getting close to being an “open source battery.”

I’ve been vlogging about this, if you want to watch progress in almost real time, have a look.

All Iron Battery: Iron metal/Iron(III)EDTA cell

I have been working on building an all-iron battery. A cheap storage solution needs to happen soon: the pacific northwest has such an abundance of electricity that Bonneville Power needed to shut down wind turbines to deal with the surplus. For stationary batteries, weight and performance are less important than cost.

An iron battery is likely to be cheap. Iron is a very cheap material and iron chemistry doesn’t require special handling for air-sensitivity or toxicity. I ran the numbers. If the world used all of its lead, lithium or iron to build a battery, lead and lithium would be small as a percentage of the grid. Iron looks pretty good. We could make a grid scale iron battery without disrupting the iron market.

So how would an iron battery work? First, we need an anode. That’s easy: we can use the Edison Cell anode made of iron and iron hydroxide. It’s an old technology and is very robust.  We also need a cathode. We need an iron chemical that can be easily reduced. That means some form of Iron(III). Rust, basically. A solid would be great, and a conductive solid would be even better. I tried iron oxide. That didn’t work. It can absorb an electron, and it’s solid, but it’s not conducive and I couldn’t get significant current.

I gave up on having an all-solid chemistry. I decided to try Iron(III)EDTA which is nicely soluble in water. I made the first version of this cell that seemed to work. I tested the design with an iron/copper cell. Half of the Edison Cell and half of the Daniell Cell. I made a salt bridge with 1M KCl and agarose. The result gave 0.9 V.

For all-iron chemistry, I kept steel wool on the anode side. Then I replaced the copper with iron/EDTA. I made a ~0.1 M solution of Iron (III) EDTA in pH 8 buffer (a little Tris/acetate that I had lying around; probably not ideal). I connected up the cathode with a bit of nichrome wire (I would prefer to soak the iron solution into conductive graphite, but I started simple). That gave 0.5 V. So it works, at least a little. Now I need to make it better.

I also just wanted to mention that I “vlog” about this and other things every Monday to Friday. I’m sometimes talking about other projects, but for whatever reason, I’m feeling a bit obsessive with this one. I hope this becomes a somewhat practical battery that folks could make according to a DIY video. Maybe that would be useful to somebody (even if only as a science fair project).

Organ-on-a-chip for safer drugs sooner

We need a new way to test drugs and supplements for safety. The current method is slow and uses animals. Nobody likes animal cruelty. Plus, animal biology is not exactly the same as human biology. There are unpredictable differences. Organ-on-a-chip is a way to culture human cells in a device that mimics the structure of an organ. The device is made of a clear plastic so that scientists can watch the cells under different conditions. If this approach works, it will allow for faster and more accurate safety tests without using animals.

By Timothy.ruban - Own work, CC BY-SA 3.0 - 2017-04-13 06_44_38-Conceptual Schematic of a Human-on-a-Chip - Organ-on-a-chip

Safety is critical. Before a drug enters human safety trials, it is tested on two species of animals. Even so, strange things can happen when moving to a new species. If a compound is not dangerous to rats or dogs, it can still be dangerous to people. BIA 10-2474 is such a compound: it killed a safety trial participant in France.  Conversely, theobromine is safe for humans but dangerous for dogs. If theobromine had been a drug candidate, and it had failed safety tests in dogs, it would have been regarded as too risky to try in people. This despite the fact that it is actually safe. Undoubtedly, there are safe drugs that have been rejected for reasons that are not applicable to people.

With an organ-on-a-chip approach, it may be possible to test drug candidates on human cells in a way that reports whether the compound is actually safe for humans. A 2010 paper in science talks about building a mini-lung that can be used to investigate whole-organ responses like inflammation. That’s not something that shows up in a simple tissue culture model; it requires multiple cell types and structures.

The FDA is now Testing ‘Organs-on-Chips’ Technology according to an FDA blog post:

On April 11, 2017, FDA announced a multi-year research and development agreement with a company called Emulate Inc. to evaluate the company’s “Organs-on-Chips” technology in laboratories at the agency’s Center for Food Safety and Applied Nutrition, one of a number of FDA efforts to help evaluate this chip technology. The flexible polymer organ-chips contain tiny channels lined with living human cells and are capable of reproducing blood and air flow just as in the human body. The chips are translucent, giving researchers a window into the inner workings of the organ being studied.

That’s encouraging. If the FDA determines that organ chips give results that are comparable or better than animal results, we might see lower regulatory hurdles for new drugs. Faster, better approvals are good for patients and investors. Plus, nobody likes animal testing.