low voltage lithium recovery

[Ykick]

I cannot understand your calculations.

Test battery cell IR by measuring resting voltage, apply known load for about 10 seconds, measure voltage at that point. Subtract starting V from loaded V, divide by known Amps, multiply that result by 1000.

I have no idea what ohmn is - it’s milli-Ohm or mOhm as in less than 1.

For example - resting cell measures 4V, apply 5A load for 10s - measures 3.93V. Subtract that from 4V = .07V drop. Divide by 5A = 0.014 Ohm IR. Multiply by 1000 for a mOhm value of 14. Done

 

[whatever]

thanks for that, understand it, in my case I was reading the voltage from icharger screen, using amp and volts that icharger was giving me,
but my leads on the icharger have very high resistance ( around 1ohmn), and so the icharger is not giving me accurate values due to my leads....better leads needed.

 

[Ykick]

The only way I ever saw somewhat reliable IR measurement from an iCharger is with both the balance and MAIN power leads connected.

Never got comfortable with iCharger IR testing and easier for me to just snap a before & after pic from CellLog or similar while running a known Amp load for few seconds. As long as the Amp load is known, all the data a person needs to calculate individual cell IR in a multi-cell pack is fairly easy to grab.

 

[whatever]

sounds like a good method.
A note on icharger IR:
I changed my high resistance crappy leads for some short good ones: a major difference in IR reading.
I tried 3 different sets of leads to compare IR on icharger for one cell:
1. short leads : 76mohmn
2. longer leads( 30cm longer than 1): 116mohmn
3. my crappy long thin leads: 1443mohmn
I'm starting to think that the icharger actually gives quite an accurate IR reading but its greatly influenced by the resistance of
the leads used.
For example:
if I simply squeeze the aligator clips tightly onto the cell tabs whilst taking the IR reading, it reduces the IR reading by approx 10mohmn, simply by giving a lower resistance on those connections, so its pretty sensitive.

 

[whatever]

some more results from 3 cells in parallel:
discharging gave a total of approx 9.4ahr.
(once again the logview proved sensitive to running other programs whilst logging, the diplayed discharge graphs kept resetting,
though total ahr was kept on each new graph....bit of a pain.....solution = dont use computer for other apps while logview is running, on this laptop anyhow)
last portion of discharge graph of 3 cells in parallel below:


Once discharged I separated the 3p pack into individual cells.

cell 1: IR 80mohmn , rest volts 3.7v (after discharge)
cell 2: IR 88mohmn, rest volts 3.7v (after discharge)
cell 3: IR 79mohmn, rest volts 3.68v ( after discharge)
Charge graphs for cell 1 and cell 2 below ( charge capacity : cell 1 = 3.2ahr, cell 2= 3.25ahr)




On graph for cell 1, the purple lines for working out what percentage of charge occurs after constant voltage regime kicks in,
works out about 25% of the charge occurs in that last phase ( on cell 1).
On cell 2 the cv phase accounts for only 11% of the charge, I wonder if that difference can be used to say anything about the internals of the cell?
Since the total ahr when the 3 cells are in parallel is 9.4ahr I would expect the last cell to be charged should have around 2.95ahr capacity ( 9.4-3.2-3.25ahr)
I'll charge the last cell at half the rate ( 0.5amp instead of 1amp) and see if it improves the capacity the cell can hold.

post charge ir
cell 1 = 91ohmn
cell2 = 98mohmn

 

[okashira]

Stop using a 1s measument for IR. you need to use a 4-wire measurement, period.

icharger supports this. You need to make a balance connector with a separate wire going to the cell terminals.

 

[whatever]

not sure I understand okashira, you mean put the 3 cells in series and then take IR using 3s balance connector?
I notice in monitor mode, if I squeeze aligator clips tightly there is no difference in voltage value, if I do same measuring IR ( squeeze aligator clips tightly), the IR value changes, indicating IR reading is very sensitive to the resistance of connectors/leads.
Discharge graph of cell3 at 1/2 current of the other two cells, predicted capacity was 2.95ahr, capacity measured from graph is 3.1ahr, possibly a slight increase in capacity, though the capacity measurement probably doesn't take into account the efficiency of storage of the current into the cell. So cant say if low current charge was beneficial for capacity.
Total capacity discharged of the 3 cells in parallel was 9.4ahr, total ahr back into the cells when charged individually is 3.2+3.25+3.1=9.55ahr.
its possible that additional 0.15ahr put back into the cells accounts for losses in the cell during charging. That would mean approx 2% of the charge is not stored ( lost as heat or other process), or an efficiency of about 98% during charging is stored. That seems quite a high value.
The percentage of charge time in constant voltage phase was only 3% for this cell.

glitch in middle of graph is where I wiggled the aligator clips to make sure connection was good

 

[whatever]

I've been researching dimethyl carbonate ( dmc), one of the previous research papers I posted used it to remove lithium products from the sei layer in order to rejuvenate cells.
The good news is dimethyl carbonate is a fairly benign/safe chemical to use. Basic safety prodecures required ( gloves/eye protection/dont breath fumes too much, respirator not a bad idea or well ventilated area).
There is bad news though:
it likes water, so its necessary to keep moisture out, water is bad news for lithium batteries even at very low levels as many byproducts are produced within the cell.
The other issue is where to buy it from for a reasonable price, I haven't solved that one as yet, but it is readily available but the prices i've found are too high.
The lithium byproducts that build up on the sei layer ( cause swelling and loss of capacity) are soluble in dimethyl carbonate and apparently can be removed by flushing with dmc.
If I can get a hold of some dmc I'll try soaking cells completely submersed in dmc, that would be simplest method to keep water away ( no need for argon gas environment). Overtime the lithium products that dissolve into the dmc should migrate out of the cell ( by diffusion).
After that new electrolyte would be needed.

 

[dnmun]

you do not seem to understand what the SEI layer is.

the electrolytes and the flourine in the lithium flourophosphate react with the carbon of the electrode and grow a surface layer of these flourinated carbon/electolyte compounds on the surface of the carbon.

since the reaction of the flourine with the carbon is oxidizing and it releases exothermic heat then the SEI layer is created during the forming charge process to stabilize the surface of the carbon electrode so that there is space for the lithium ions to flow around the obstructions of the surface SEI layer to the carbon where it is intercalated into the carbon matrix.

there is no such thing as an unreacted carbon/electolyte interface that one would wish to create. the SEI layer is essential to the stability of the electrode when the cell is charged up. have no idea what all this stuff you have written means either.

 

[whatever]

funny........neither do I!
There are various byproduct reactions that occur ( i'm just reading research papers and trying to determine if there is anyway to rejuvenate low capacity cells), they interfere with the sei layer, dmc can dissolve those byproducts, lithium and fortuitously some of the byproducts dissolve in dmc.

 

[whatever]

this link
http://www.lithiumbatteryresearch.com/pdf/SEI1.pdf
Goes into great detail of what byproducts of lithium are found on the sei interface by experiment. And shows their solubility in dmc.
Heres the intro:


Note the sentence:
" sei films on the surface of the negative electrode taken from a commercial battery after soaking in dmc for 1 hour suggested that the films can dissolve"

I would think its valuable information to know that dmc may be used to rejuvenate batteries, there are links I've given in this thread previously where others have shown that, theres a couple of patents also that use this process. I've not seen it posted in the forum previously. ( new electrolyte of course needs to be added, or alternatively lithium metal dissolved in dmc can be added, as used in one of the previous links posted). Ideally lipf6 should be added after doing a dmc flush out.

 

[whatever]

I'll make a prediction:
a cell that is swollen due to sei layer build up ( dendrites etc), once soaked in dmc the swelling will reduce.

A cell that is swollen due to gas build up is a different matter.

( this paper has a good picture of dendrite growth, and a good explanation of how it occurs)
http://authors.library.caltech.edu/45658/7/jz500207a_si_001.pdf

 

[whatever]

heres a good picture of what sei layer might look like ( its complex)



from this page
https://www.liv.ac.uk/chemistry/research/hardwick-group/research/

 

[Punx0r]

How do you propose to wash the electrodes with this solvent and then remove it without opening the cell?

 

[whatever]

attached is a picture of the positive end of the cells I intend to experiment with.
They have a stainless steel metal casing, so it would be extremely difficult to open the cell without damaging the internals
( it would be possible).



On the positive end there are two holes, one larger hole is made from some sort of thin metal, which is easy to puncture,
its probably a pressure relief device. There is also a tiny hole which has been sealed with resin ( or some sort of glue).
Theres a couple of ways I can tackle the problem:
Simplest method: simply open the pressure relief hole, place the whole cell into dmc. Let it sit there possibly a couple of days.
As the lithium products dissolve into the dmc they should migrate out of the cell ( there will be a concentration gradient between the inside of the cell and the dmc outside the cell, eventually the lithium products will migrate outwards until that gradient is equalised). Depending on the volume of dmc the cell is soaked in, will determine the final concentration of lithium products within the cell.
Better method: more work involved:
drill or grind a small opening at the negative end of the cell, using a syringe push dmc through the cell, probably best to fill the cell with dmc and let it sit for some time, then use syringe to push new dmc through cell. I'm guessing that will be a better flush than the simpler method.

The overall idea is to remove as much as possible of the sei layer, and any lithium or lithium products in the cell.
So you start with a relatively clean graphite and cobalt oxide ( or manganese oxide or iron phosphate) lithium layer. In my case its lithium cobalt oxide and graphite.
After the flush out then new electrolyte LiPF6 can be added ( alternatively a lithium metal strip in dmc can be discharged to replemish the lithium, as outlined in previous paper I linked to, where they successfully used this method to replace lithium
http://ma.ecsdl.org/content/MA2010-03/1/812.full.pdf).
In that link they use an a123 2.2ahr ( the white cylinder cells used to find them in dewalt packs), where the bottom of the cell can be removed ( no need to open the whole cell up).

 

[whatever]

Assuming the process works, if a new sei layer is required to be formed, then apparently that takes 3 full charge/discharge cycles.
info from this patent http://www.google.com/patents/US20080066297

Typically, this requires two to three cycles of full charge/full discharge. Each cycle lasts approximately 10 hours, so forming the SEI layer during charging can consume up to 30 hours in the battery manufacturing process. It can also irreversibly consume up to 20% of available lithium ions, which reduces the battery's energy.

 

[whatever]

A few other bits of info,
there are some exothermic and some endothermic reactions involved when lithium products dissolve into dmc, I found a paper which studied this, but its too complex for me to understand if the reactions would produce excess heat, so if anyone tries it,
some caution needed, just in case excess heat is produced.
There is also a patent that has some good links to patents aimed at lithium battery rejuvenation.
http://www.google.com/patents/US20050244704
One thing it mentions is that the batteries have to be fully discharged before trying to rejuvenate them ( the lithium can react with oxygen and water is still in charged state).

 

[whatever]

just an obsevation:
the dud cell that I did some charge/discharge tests on earlier in the thread:
the cell eventually become extremely expanded, since its in stainless steel shell with gas release valve I assume the pressure was from sei layer deposits or from lithium crystals causing expansion.
I was a bit worried it might internally short, so I discharged it, very very slowly, I used a bike rear light ( 5 red leds), the current was very low ( I cant remember exact value but was between 60 to 90ma). It took quite a few days to discharge, during this time the swelling of the cell decreased markedly

 

[whatever]

heres a very informative lecture series ( in video format) on lithium batts by an academic
https://nanohub.org/courses/IMSB/01...ialreactionsdendritesinrechargeablebatteriesi

 

[whatever]

found a thesis online, which uses a123 20ah cells, its mainly about temperature effects on a123 performance, but has lots of other charge/discharge data, alot of useful info for anyone interested in technical side of a123 20ah cells. It discusses electrolytes briefly, its quite likely that a123 20ah are using dimethyl carbonate and ethyl carbonate as electrolyte, probably there own blend with additives, its also likely that rejuvenating using dmc/ec flush, then replemish electrolyte with commercial Lipf6 would also work on a123 cells.
https://uwspace.uwaterloo.ca/bitstream/handle/10012/7350/Lo_Joshua.pdf?sequence=1

This safety data sheet from a123 gives the electroytes used ( typical of most lithium based batteries on the market)
http://www.simrad.com/www/01/NOKBG0...A123_System_Safety_Data_Sheet.pdf?OpenElement
So I'm quite confident that dmc could be used for rejuvenation( one paper I read found that ec is better to rejuvenate lithiums,
so either dmc or ec could be used, or a mixture). The data sheet also confirms commonly used lipf6 is used as electrolyte.

 

[4LivesPerGallon]

Don't know much about that sub-type of Lithium.

However, my group has revived "non-rechargeable" CR123A "primary" batteries hundreds of times, and here's what we know:

- Measuring internal resistance in a battery requires an expensive AC battery analyzer. Simple multimeter doesn't work.

- A primary lithium battery that is discharged below 2.5V cannot be revived.

- Any primary lithium battery's remaining capacity can be doubled with a variable power supply, if you know what you are doing; and, are extremely careful; and, aware that any battery can explode, or catch fire. (Justin L. had some violent explosions with Nicads or NiMh cells, I forget which.)

- If using a variable power supply, protect it from "going nuts" with a hospital-grade power isolation filter ($250 and up)

- Your variable power supply must be "lab grade", for your safety.

- We don't charge the primary CR123A's past 3.55 volts. (Maybe this is what the Tesla car is doing.)

- When a battery doesn't accept a charge, it obviously has high internal resistance, and checking with an AC internal resistance meter (yup, got one) isn't necessary.

 

[whatever]

can you give any more info on what your doing with variable power supplies to increase capacity?

 

[whatever]

Just came across another article:
http://www.researchgate.net/publica..._Methods_for_Spiral-wound_Lithium_Ion_Battery
The whole article can be viewed on that link, its a study of capacity recovery in a123 cell type batteries. It mentions the previous study I post where an a123 cylinderical cell ( 2.2ahr) had the lost lithium replaced by discharging the cell to a piece of lithium metal ( in electrolyte).
The paper above replicates that study. Both those papers are trying to replace the lost lithium. It does confirm that standard ec:dcm:lipf6 can be used with a123 cylinder type cells. So it appears to me that the best method will be to remove lithium products in the cell via dmc:ec mixture ( ec should remove more of the lithium products than dmc according to one paper), then to put new electrolyte ( lipf6 1 molar in ec:dmc or similar).
Flushing with dmc:ec should remove alot of the sei layer products and any lithium dendrites, puting in new electrolyte afterwards should be like working with a new cell, the effectiveness of the process I'm not sure, I"ve only found two papers that try this method. Depending on that state of the graphite layer ( eg amount of cracking etc) when the process is done will influence amount of capacity recovered ( lots of other factors involved also).
Best price I've found for dmc and ec is $75usd per 5kg ( approx 5litres), though I'm yet to find a reasonably priced method to get it sent from china. Its a low hazard chemical so should be a cheap method to get it sent from china.
It looks like lipf6 can be bought as a powder, so should be relatively easy to get sent from china. I haven't tried to get price of lipf6 as yet, it can be added to dmc:ec to make approx 1 molar concentration.

 

[whatever]

Here is an a123 patent filed july last year ( published may this year). Its very detailed about electrolyte formulation.
http://www.freepatentsonline.com/y2015/0064549.html
Since it was only filed last year, it would likely represent the formulation a123 might use in the future. It does confirm that ec:dmc are used, with other additives. The patent is about new additives to reduce gasing, and some for extreme temperature use. Particularly interesting is at the bottom of the page where they give two examples of formulations and do some comparative tests.
The patent is worded in such a way that a wide variety of electrolytes are covered in the patent, so there is a lot of information not particularly relevent.

 

[whatever]

One thing I have learned from all this research, is that each type of lithium battery on the market appear to use the same basic chemistry. Whether its cobalt oxide lithiums, manganese oxide lithiums, or iron phosphate lithiums, they all work on the same principle and common electrolytes work with each type.
So the basic formulation for electrolyte is lipf6 dissolved in ec:dmc. There are lots of different additional chemicals used for various purposes, but you could say that each type of lithium battery is using a common basic principle.