diy solar

diy solar

My adventures building a Zinc-Iodine battery

I've been trying to increase the cycling stability of the batteries, as most of the batteries I made were either dying due to shorts after 5-10 cycles or their capacity was fading quickly after the first 10-20 cycles.

I tested a battery using a CCP cathode (150um) with a Whatman 42 filter (200um) paper as separator and a Zinc anode. The electrolyte is made from 15m ZnCl2 + 5m KI + 5m NaCl in white vinegar (4% v/v Acetic acid). The idea of the white vinegar is to increase ion mobility, since the dipole moment of acetic acid is larger than that of water. Note that the electrolyte itself is already very acidic (pH < 2), so the acetic acid is all protonated under these conditions and merely provides a higher dipole moment media.

I wanted to test stability. This means reaching a higher cycle number quickly, for this reason, the battery was charged to 1.3V at 10mA, discharging to 0.5V. Previous attempts had shown poor results under these conditions. The results up until now are below (first 100 cycles):

1634589392629.png1634589408901.png
1634589422609.png1634589437452.png
The CE is stable and mostly above 98% and the EE is above 80%. So far capacity has been quite stable, although it is clear that some instability still remains. Since the battery is being charged to a low potential at a relative high current density (7.75mA/cm2), the energy density is quite low, currently at around 9.66Wh/L. I want to be able to reach a cell that is stable for >500 cycles before I aim for higher capacity, as higher capacity and densities are meaningless if a cell is simply unstable.

The achievement of a CE > 98% is already a significant improvement from the original paper. Both the use of vinegar and the use of a Zn metal anode (instead of the graphite anode of the Swagelok cell) have been critical in the improvement of the battery.
 
I stopped the battery after 138 cycles, when it started showing accelerated capacity loss. I also fixed a small error in my EE calculation, so the average EE is actually around 78%, not 80%.

1634601690746.png
 
Last edited:
I just assembled a battery using a CCP cathode, a Whatman 42 filter separator, a zinc anode and an electrolyte using 19m ZnCl2 + 5m KI in white vinegar. I want to see the effect of a higher ZnCl2 concentration, as I am probably using a lower purity ZnCl2 compared to that used in the article. I will be charging to 1.3V at 5mA, discharging to 0.5V.
 
Last edited:
I just assembled a battery using a CCP cathode, a Whatman 42 filter separator, a zinc anode and an electrolyte using 19m ZnCl2 + 5m KI in white vinegar. I want to see the effect of a higher ZnCl2 concentration, as I am probably using a lower purity ZnCl2 compared to that used in the article. I will be charging to 1.3V at 5mA, discharging to 0.5V.
This cell charged to much higher capacity (1.2mAh) but died due to a short - probably caused by dendrites - after 3 cycles. It seems devices with higher capacity die due to dendrites when using a 200um Whatman 42 separator quiet quickly, the formation of dendrites is the most likely culprit. Makes me think if this is one of the main reasons why the paper never charged at currents this high. No mention of dendrites is ever made in the paper.

I am now testing a cell on a 22m ZnCl2 + 5m KI + 5m NaCl electrolyte in distilled water (CCP cathode, W42 separator, zinc anode), charging to 1.4V at 5mA. Charging to 1.4V will make dendrites appear faster if they are indeed present. This electrolyte has dramatically lower water activity which should also lead to significantly lower dendrite formation. We'll see if that's actually the case.
 
Last edited:
Increasing the ZnCl2 increases the electrical resistance a lot, without preventing the shorting that happens. The battery made with 22m ZnCl2 died after a few cycles. Increasing ZnCl2 to 30m gave even worse results as virtually all cell capacity is lost due to the huge increase in resistance (as reported in the paper).

I have never ran a test of the ZnCl2 15m + KI 5m in distilled water with a Zn anode, so I will be running that (which will be the exact same setup of the paper) to see if I still get the shorting problem. I will charge to 1.3V and discharge to 0.5V at 5mA.
 
These are the results for a cell with ZnCl2 15m + KI 5m in distilled water, CCP cathode, Zn anode, W42 separator. Charged to 1.3V, discharged to 0.5V at 5mA. I ran it for 6 cycles without any shorting appearing. The lower ZnCl2 concentration allows for much better ion mobility, which means the capacity increases significantly. This cell had a max energy density of 31.7Wh/L at this current density.

1634732527795.png
I then constructed the same cell, but using a vinegar as the solvent, which is running now.
 
It is also worth noting that the original WiS article did not use carbon cloth, but a carbon fiber paper. I have bought some additional carbon materials to test:

Spectracarb 2050A-0850 - 10cm x 10cm
Size: 10cm x 10cm
AvCarb© MGL280 - 10cm x 10cm
Size: 10cm x 10cm
Toray Carbon Paper TGP-H-090
Wet Proofing: 1% Wet Proofing
Size: 10cm x 10cm
Spectracarb 2050A-1050 - 10cm x 10cm
Size: 10cm x 10cm

These were all bought from fuelcellearth.com. I will therefore have 4 cloths, 1 felt and 4 carbon fiber papers to test this chemistry on.
 
These are the results for a cell with ZnCl2 15m + KI 5m in vinegar, CCP cathode, Zn anode, W42 separator. Charged to 1.3V, discharged to 0.5V at 5mA. I ran the battery for 39 cycles, when decay became too pronounced.

1634761881477.png

Also, got the carbon papers today! So I will be doing a test with an MGL280 paper next.
 
Last edited:
These are the results for a cell with ZnCl2 15m + KI 5m in distilled water, MGL280 cathode, Zn anode, W42 separator. Charged to 1.3V, discharged to 0.5V at 5mA. Curves are cleaner than for a carbon cloth, EE is higher and CE is similar. There is an evident decay in capacity after 27 cycles.

1634812817712.png
 
I noticed that the original paper barely studies the effect of KI concentration, as they say that 5m is the highest concentration they are able to reach. However, I am able to easily dissolve 7.5m KI in the ZnCl2 solution, so I'm going to try that and see the effect it has on the battery properties. After that I'll study the effect of lower KI.

I have now assembled a new cell with ZnCl2 15m + KI 7.5m in distilled water, MGL280 cathode, Zn anode, W42 separator. I will charge to 1.3V, discharged to 0.5V at 5mA.
 
These are the results for the cell with ZnCl2 15m + KI 7.5m in distilled water, MGL280 cathode, Zn anode, W42 separator. Charged to 1.3V, discharged to 0.5V at 5mA. The cell died after 9 cycles due to what seems to be a dendrite related short.

1634833816791.png
 
I am now doing a test with new cell with ZnCl2 15m + KI 2.5m in distilled water, MGL280 cathode, Zn anode, W42 separator. I will charge to 1.3V, discharged to 0.5V at 5mA. This will allow us to see the effect of lower KI.
 
I am now doing a test with new cell with ZnCl2 15m + KI 2.5m in distilled water, MGL280 cathode, Zn anode, W42 separator. I will charge to 1.3V, discharged to 0.5V at 5mA. This will allow us to see the effect of lower KI.

This cell failed on the first cycle due to dendrites :unsure: Lower KI does lead to a ton of polarization, my guess is due to Iodine depletion at the battery's cathode.
 
Last edited:
I have put the graphite electrodes of the Swagelok cell in ethyl alcohol to clean them. The alcohol immediately turned very yellow, a sign that the graphite has absorbed a lot of iodine through all the experiments. I have therefore decided to move to my Titanium electrodes, as the graphite seems to interact a lot with the iodine which has probably contributed to cell instability over time (as the electrode mass is much larger that the battery's cathode/anode and it is probably "soaking up" iodine from each device).

I will now do a test with the most successful electrolyte up until now (15m ZnCl2 + 5m KI + 5m NaCl in white vinegar (4% v/v Acetic acid)) a zinc anode, W42 separator and a Spectracarb 2050A-1050 cathode, using the Ti electrodes in the Swagelok cell. I will post the results as they are obtained ?
 
  • I don't think white vinegar is a good idea. It has a complex composition and contains protic acid which may offer activity H+.
    I have put the graphite electrodes of the Swagelok cell in ethyl alcohol to clean them. The alcohol immediately turned very yellow, a sign that the graphite has absorbed a lot of iodine through all the experiments. I have therefore decided to move to my Titanium electrodes, as the graphite seems to interact a lot with the iodine which has probably contributed to cell instability over time (as the electrode mass is much larger that the battery's cathode/anode and it is probably "soaking up" iodine from each device).

    I will now do a test with the most successful electrolyte up until now (15m ZnCl2 + 5m KI + 5m NaCl in white vinegar (4% v/v Acetic acid)) a zinc anode, W42 separator and a Spectracarb 2050A-1050 cathode, using the Ti electrodes in the Swagelok cell. I will post the results as they are obtained ?
 
Back
Top