Sodium Ion Battery evaluation

[5U4GB]

Kind of Bizarre that these are being sold through normal channels with the true manufacturer hidden (Hakadi obviously isn't). Sodium Ion chemistries even from Chinese and Indian manufacturers seem quite diverse, would be interesting to know where these are really from.

And whether they really are Na-ion batteries or just rebadged Li-ion laptop pulls.  If it's Aliexpress I'd expect the latter, I guess you can use them to power your $14.95 128TB SSDs.

[Psi]

Be interested to see someone taking one apart and doing some chemical tests to confirm lithium or sodium

[IanB]

Be interested to see someone taking one apart and doing some chemical tests to confirm lithium or sodium

A simple test would be the weight of the cell. Sodium is over 3x heavier than lithium, so when comparing two cells of similar capacity, I would expect the sodium ion cell to weigh more.

[robint]

there's only a microscopic amount of alkali metal in these batteries, obtw I see Libs use an Aluminium can vs Lipofe use a steel can

[IanB]

there's only a microscopic amount of alkali metal in these batteries, obtw I see Libs use an Aluminium can vs Lipofe use a steel can

The size of microscope you need depends on how good your eyesight is 

A 2600 mAh lithium ion cell would contain about 0.7 g of lithium. The same capacity sodium ion cell would contain about 2.3 g of sodium.

Granted, a difference of 1.6 g in a cell weighing about 45 g is not going to be very easy to spot against statistical variation between different lithium cells, but we can at least say that a cell that is lighter than average is unlikely to be a sodium ion cell.

[IanB]

What about variations in cell wall thickness, end cap thickness, electrolyte density/fill ratio, cathode loading? Why would you assume those are all the same between two random cells?

I wouldn't. Which is why I said I wouldn't in the post above.

[robint]

Ladies behave, we seek to add to our fund of knowledge not watch cat fights

[Psi]

Someone send one to Nilered to check   

[IanB]

Someone send one to Nilered to check   

If you are prepared to destroy a cell or two, it is possibly not that hard to check. Lithium shows a characteristic crimson-red color in a flame test, while sodium has the unmistakable orange color of low pressure sodium lamps.

So if you hold a small piece of battery electrode material in a gas flame and you see a strong orange color, then it is probably sodium. But if instead you see a bright red color, then it is lithium. I don't know if lithium and sodium may be mixed in the same cell, but the sodium color is so strong it hides other colors. So if you see orange, there is sodium.

(It goes without saying, if you do dismantle a cell, make sure it is thoroughly discharged first, otherwise you may get a very different kind of flame.)

[SiliconWizard]

Yes, and obviously the base thing to look at is the BMS (if there's any inside the cells, which I rather hope there would). Should give you a pointer or two as well...
Speaking of which, I'm not completely discarding the idea that they may have put a BMS (if again there is any) in there that is modified to make the cell, from the user's POV, look like something else than Li-ion, while in fact it would just be Li-ion.

[5U4GB]

A 2600 mAh lithium ion cell would contain about 0.7 g of lithium. The same capacity sodium ion cell would contain about 2.3 g of sodium.

Granted, a difference of 1.6 g in a cell weighing about 45 g is not going to be very easy to spot against statistical variation between different lithium cells, but we can at least say that a cell that is lighter than average is unlikely to be a sodium ion cell.

In theory the way to do it would be to weigh (say) ten cells of each type, which would both magnify the difference in weight and smooth out any slight irregularities.  However I'm still not sure whether this would be valid, you'd then also need to weigh batches of ten Li-Ion cells from different manufacturers to make sure the intra-cell-type variations don't outweigh the inter-cell-type variations.

[Craig]

I've obtained a couple of the 18650-1500mAh cells from Hakadi.

This is a discharge test. It appears to correspond to the datasheet.

[robint]

many thanx, can you provide a charge test at say 1C
IMHO that looks like a very distinctive curve and easy to determine the DOD (4V to 2.5V) unlike the Lifepo which is substantially flat and needs a very accurate coulomb counter - big disadvantage.

Impressive, pls keep us informed of progress

Robin

[ROFLhoff]

I've bought some SIB as well back in November. Took about 2 weeks to get to the states.

I used this vendor on Alibaba and bought several of their "1300mAh" 18650 cells, and "10Ah" 33140 cells.
https://www.alibaba.com/product-detail/18650-sodium-ion-battery-3-1V_1600389785267.html?spm=a2700.shop_plgr.41413.6.7c0244c37lsHxY
https://www.alibaba.com/product-detail/Sib-32140-sodium-ion-battery-3V_1600860472916.html?spm=a2700.shop_plgr.41413.13.7c0244c3wfDeR3

I used an Atorch DL24 tester to characterize the 18650 cells. I made a custom charging board in the past (project linked on my hackaday.io account) and used it for the 18650 cells. I will seek out a better charger for the larger cells since I do want to perform 1C charging and characterizing.

As noted in my images, I am charging these cells at 1C (1300mA), using 0.1C (but more like 146mA in practice) as my cutoff, and discharging at 1C (1300mA) until 600mV.

This is some of the data I've gathered so far:

Note: Horizontal Axis is in Ah, not mAh

This graph is useful for people that want to roughly gauge how a cell would perform if charged to a lower voltage.
I'm curious about the upper plateau on the 4.2V and 4.3V charges. I imagine an unwanted secondary chemical reaction is happening inside the cell. It's neat to see this large increase in capacity but I do not have the equipment to run hundreds of tests quickly to measure cycle life loss at these higher voltages.

The change in slope below 800mV is due to the design of the DL24. The discharge path contains a set of series-connected power diodes and they limit the current flow at lower voltages. If I short them out, then I imagine the discharger could bring the cells to effectively ground but I would lose the reverse insertion polarity protection.


Note: Horizontal Axis is in Ah, not mAh

My setup for testing cold temps is a spare minifridge. It gets down to -1°C in the main compartment, and down to -10°C in the icebox. My battery holder rests on the floors of this fridge, so there isn't a means to keep the cells from warming up in my setup.


Note: Horizontal Axis is in Ah, not mAh

One thing I've noticed in these tests is that the cells do warm up when discharged at higher currents. Like, they rise 30°C from start to finish. This may explain the voltage rise in the 3C and 5C tests. If I could measure the energy going into the cells while charging, I could estimate coulombic efficiency. While charging at room temp, the cells barely rise in temp at 1C charge rate.



Also for the people doubting if they suspect receiving worn down lithium cells, do note that a proper SIB will not use a copper current collector for the anode. If you open a cell up and only see aluminium foil current collectors, you can be assured that you didn't receive a lithium cell. There is no need to determine if the electrolyte contains lithium salts.


I hope this information is helpful.

[robint]

WOW @ROFLhoff  That is amazing semi-pro test results.  Its quite rare to get such quality and attention to technical integrity.
Indeed that is very curious charging to 4.2, 4.3 V produces nearly 20% increase in capacity - but at what penalty (life span?)

Be interesting to see what your equivalent Charging curves reveal.  Much emphasis is placed on fast charging (for EV needs) - not my area of interest as a Solarist.

Lifepo's are already blowing away Lipos on safety and life span grounds as they become competitive.  As production lines get switched over to SiBs then their prices should tumble.  Could well happen within a year?

I wonder if SiBs will have  more reasonable cell voltage characteristics when connected in series chains (given all the complexity with cell monitoring and balancing needed for Lipos).

Current Wh/kg is only ca 50% of Lipos but BYD claims to be putting them in one of their models.  Imagine an EV where the batteries last 20 years.  EV's are already shown as needing far less regular maintenance (except tyres/brakes) - whats left to go wrong (assuming plastic components are properly made) - New driver seat after 500,000 miles

Please keep up the excellent work and keep us informed - IMHO this is groundbreaking
Robin

[robint]

this reply is cos the board doesnt yet recognise your new handle and you may not get notified yet

[ROFLhoff]

Thank you for the kind words @robint !

I am developing a BMS for future batteries based on the ISL94202. The arbitrary control on balancing voltages and FET controlling are nice. A shame the cells are so cheap and the controlling ICs are not!

[robint]

Hmm I see the development kit is £540  - ouch  but the ic is £5 - but I cant handle those flow solder pieces, not geared for it and looks so easy to screw up with my eyesight

ps

Ive followed your lead and ordered 8 x 18650s SiBs

be glad to exchange test results by PM if you wish

[tszaboo]

I've bought some SIB as well back in November. Took about 2 weeks to get to the states.

Very nice info indeed. They don't look like Li-Ion cells at all.
Did you measure the ERS and the short circuit current of the cells?

[T3sl4co1l]

What's the charging curve?

The difference between the two, is a hysteresis loop, the area of which is lost energy -- thus we can calculate total efficiency.  How it varies with dis/charge rate (internal resistance vs. electrode overpotential) will be interesting to see.

Tim

[Psi]

...@120Wh/kg (could be 200) its 1/4 weight of a LAB...
...Current price of 100Ah 24v same as LAB but half size 1/4 weight 20kg vs 80kg...
...up to 3000 cycles (claimed) vs 500 cycles LAB...

What is the LAB type cell talked about in the first post?
Google failed me trying to figured it out and even chatGPT has no idea.

Lithium Air Battery maybe?

[IanB]

What is the LAB type cell talked about in the first post?
Google failed me trying to figured it out and even chatGPT has no idea.

Lithium Air Battery maybe?

Lead acid battery?

[eneuro]

This is a discharge test. It appears to correspond to the datasheet.

I bought these batteries and I'm looking for a datasheet 

I am particularly interested in their resistance to higher temperatures and the risk of spontaneous combustion in the event of overcharging, because I would like to use them as an energy buffer to power directly the AVR controller, charged with very small a few mA currents directly from the 230VAC mains socket without a transformer using 3.9Vz Zener 1N4007 1000V diodes and a few resistors 

[Psi]

What is the LAB type cell talked about in the first post?
Google failed me trying to figured it out and even chatGPT has no idea.

Lithium Air Battery maybe?

Lead acid battery?

weird, i've only ever heard them called SLA AGM or GEL. never LAB.

[Craig]

I bought these batteries and I'm looking for a datasheet 

The datasheet should be attached to reply 14.