What exactly does "IMR" mean?

Battery Guy

Enlightened
Joined
Apr 28, 2010
Messages
807
Location
Portland, Oregon
Greetings Everyone,

I have seen a lot of threads recently on "IMR" lithium-ion cells. DFiorentino and myself have posted some test results recently (seen here and here, respectively). I also have seen a lot of comments recently that indicate a general misconception of what IMR cells actually are, especially with respect to rate capability and safety. The purpose of this thread is to clarify the issues surrounding the "IMR" designation for lithium-ion cells. I will try to keep the techno-jargin to a minimum, but please feel free to ask clarification questions if I lose you.

Unlike most battery chemistries whose name defines the chemistry of the anode and cathode, the term "lithium-ion" refers to an ever growing class of cell chemistries that utilize different materials to host lithium-ions in the charged and discharged state of the cell. Most lithium-ion cells use graphite or hard carbon as the negative electrode host material (typically referred to as the anode), but the selection of the positive electrode material (typically referred to as the cathode) is much more varied. The following three-letter designations for lithium-ion cell denote what cathode material is used in the cell:

ICR = LiCoO2 (also called LCO)
IMR = LiMn2O4 (also called LMO)
IFR = LiFePO4 (also called LFP)

If you take away one thing from this thread, remember this: The term "IMR" simply indicates that LiMn2O4 is a major component of the cathode. It does not necessarily indicate that a cell has exceptional rate capability or improved safety characteristics, properties often incorrectly associated with cells with the IMR designation.

You can stop reading now, but if you want that statement explained, please read on.

The selection of a particular cathode material generally has implications for the performance (capacity, energy density, rate capability, etc...), reliability (cycle life, calendar life) and safety characteristics. The impact of LiMn2O4 on these properties is discussed below.

Performance

The crystal structure of LiMn2O4 has three-dimensional tunnels that enable rapid diffusion of Li+ through LiMn2O4 particles, and therefore lithium-ion cells made with this material can have better theoretical drain rate capability. Conventional lithium-ion cells use LiCoO2 or variations of that material. Cells with LiCoO2 cathodes are typically designated with the term "ICR". The LiCoO2 type materials have a layered crystal structure that only enables 2D diffusion within the layers, so the overall Li+ diffusion rate is slower. Interestingly, LiFePO4, which is commonly used in high rate cells (think A123), has 1D diffusion of Li+ and an overall low diffusion rate. The only way to make LiFePO4 work is to make the particles very small (i.e. "nano") to reduce the distance that Li+ needs to diffuse. "Why not just make LiCoO2 or LiMn2O4 nano size to get even better rate capability?" I hear you ask. The only reason nano LiFePO4 works is that this material has a lower oxidizing potential (which is why the cell voltage is lower) and therefore it does not chemically react with the electrolyte. LiCoO2 and LiMn2O4 actually oxidize the electrolyte slowly with use, so using a nano-particulate cathode with these materials would greatly accelerate aging due to the higher surface area of the cathode, and the cells would have terrible cycle life.

Now back to LiMn2O4. This material has a lower crystalline density and therefore a lower inherent capacity (mAh/ml) than LiCoO2. If you do a direct replacement of LiCoO2 with LiMn2O4, the capacity of the cell will be reduced by ~15%. But here is the interesting thing: simply doing a direct replacement with LiMn2O4 will not significantly increase the rate capability of a cell. In order to get a real increase in rate capability, you need to design a cell with higher electrode surface area and thinner electrodes because Li+ diffusion in the cathode particles is not the only bottleneck to getting current out of the cell. Using thinner electrodes further reduces the capacity of the cell because you will have a higher fraction of "inert" materials like the separator and current collector. This is why IMR cells typically have ~40% lower capacity than their ICR cousins.

Safety and Reliability

An additional property of LiMn2O4 is that it cannot be overcharged. When a "true" (and I will explain why I put that in quotes later) IMR cell is charged beyond ~4.25 V, no additional capacity will go into the cell. The voltage will simply spike. This is exactly the same case for LiFePO4 cells. Cells that use LiCoO2 and its layered-metal oxide cousins can be overcharged, leading to significant safety issues when these cells are charged above their specified voltages. Cells with LiMn2O4 and LiFePO4 cathodes can be damaged by overcharging (oxidation of the electrolyte leads to increased internal resistance, loss of capacity and shorter cycle life), but overcharging to higher voltage does not make them less safe. This inability to overcharge, combined with the lower total energy density, is why lithium-ion cells with LiMn2O4 and LiFePO4 are generally considered to be "safer".

The last property of LiMn2O4 that I will mention is that in its pure form it tends to have very poor cycle life. There has been a tremendous amount of research that has gone into improving the stability of LiMn2O4 to improve the cycle life of IMR cells, and this has produced cells with reasonably good long term performance. However, the easiest way to get great cycle life out of LiMn2O4 is to blend it with a separate cathode material that contains nickel. This includes the layered cathode materials Li(Nix,Cox,Mnx)O2 (aka NCM or 333), Li(NixCoyAlz)O2 (aka NCA) and Li(NixCoy)O2 (aka NCO). I won't bore you with the details, but it turns out that the oxides containing nickel acts to change the local chemistry around the LiMn2O4 particles and helps to improve stability.

The end result is that most cells called "IMR" actually have a significant fraction of nickel-containing layered metal oxide as a co-cathode. This means that these cells can in fact be overcharged, thus minimizing the perceived safety advantage of the cell.

Summary

Not all IMR cells are created equal. One cannot and should not make generalized statements about their performance, reliability or safety characteristics.

I hope this helps dispel some of the mystery surrounding these cells.

Cheers,
BG
 

archimedes

Moderator,
Staff member
CPF Supporter
Joined
Nov 12, 2010
Messages
15,792
Location
CONUS, top left
Awesome post - thank you !

When I first saw the title, I was surprised that it was "in the form of a question" ... like Jeopardy!, I guess :)
 

Battery Guy

Enlightened
Joined
Apr 28, 2010
Messages
807
Location
Portland, Oregon
Awesome post - thank you !

When I first saw the title, I was surprised that it was "in the form of a question" ... like Jeopardy!, I guess :)

Hey archimedes,

Glad you liked the post. The title was designed to help folks find the post using the search function. I like the Jeopardy analogy!

Cheers,
BG
 

rlcrisp

Newly Enlightened
Joined
Sep 6, 2010
Messages
15
So now the natural follow up question is how the different IMR cells which are relatively common in the cpf world compare to each other - maybe real world results that conclusions can be drawn from rather than dictation (no "X is better than Y"). Say maybe AW, Shao, etc. I know DFiorentino has done some of this work but it would be interesting to see an analysis from an expert to speculate as to how those cells embody the different possible IMR implementations. =)
 

Battery Guy

Enlightened
Joined
Apr 28, 2010
Messages
807
Location
Portland, Oregon
So now the natural follow up question is how the different IMR cells which are relatively common in the cpf world compare to each other - maybe real world results that conclusions can be drawn from rather than dictation (no "X is better than Y"). Say maybe AW, Shao, etc. I know DFiorentino has done some of this work but it would be interesting to see an analysis from an expert to speculate as to how those cells embody the different possible IMR implementations. =)

DFiorentino's tests are great, and provided the "spark" that I needed to get going with my lithium-ion cell tests.

I just placed an order for a complete set of the Shao cells. Glad I did actually, because I was not aware that Shao had an 17650 IMR cell. I sincerely hope that it holds up well to ~3.8 A discharge because I would love to use three of these in a small modded light with an 1185 bulb.

Cheers,
BG
 
Last edited:

Mr Happy

Flashlight Enthusiast
Joined
Nov 21, 2007
Messages
5,390
Location
Southern California
On the subject of high performing cells and hotwires, I have some K2 Energy LFP26650EV cells sitting at home waiting for me to do something with them. On the datasheet these are capable of continuous discharge currents up to 12 A (but recommended to keep to 3 A for maximum life). I've been thinking of a bulb configuration to use with them, with perhaps three cells in series for 9.6 V. Any ideas on this? It might be a squeeze to fit 3 in a Mag 3C body, but I imagine a 3D body would work. Then I wonder if the voltage would drive say an 1185 at full brightness, or if a different bulb would be a better match?
 
Last edited:

45/70

Flashlight Enthusiast
Joined
Oct 9, 2005
Messages
2,800
Location
Rural Ohio
I missed this thread earlier today, nice writeup, BG.

I have to say, to this day, every time I see "IMR", even here on the Forum, the first thing that comes to mind is "improved military rifle". This was one of the first smokeless powders developed by, I believe DuPont, in the late 1800's and still available and used by the military, as well as civilians in various forms today. After reloading handgun and rifle cartridges for over 35 years, I guess that's why it still comes to mind first.

Also, as pertains to Li-Ion cells, I thought the "I" meant lithium, the "M" manganese, and the "R" round. Or, is the "R" for rechargeable?

Dave
 

xul

Enlightened
Joined
Aug 28, 2011
Messages
336
Location
MD
properties often incorrectly associated with cells with the IMR designation.
If you say IMR or SLA or alkaline you can come with an 'average' performance but with a wide tolerance on this average value because it's not very specific.
As you become more specific the tolerance band should narrow.

That's why some performance is shown as 'typical.' The definition for this word is up for grabs but the word 'average' or 'mean' has a clearly defined meaning.
Also, giving max or min values says nothing about the distribution of values for the units that are sold.
 

Battery Guy

Enlightened
Joined
Apr 28, 2010
Messages
807
Location
Portland, Oregon
Also, as pertains to Li-Ion cells, I thought the "I" meant lithium, the "M" manganese, and the "R" round. Or, is the "R" for rechargeable?

I believe that you are correct. For example, Molicel refers to their LiMn2O4 prismatic cells as "IMP", where the "P" stands for prismatic.
 

Battery Guy

Enlightened
Joined
Apr 28, 2010
Messages
807
Location
Portland, Oregon
On the subject of high performing cells and hotwires, I have some K2 Energy LFP26650EV cells sitting at home waiting for me to do something with them. On the datasheet these are capable of continuous discharge currents up to 12 A (but recommended to keep to 3 A for maximum life). I've been thinking of a bulb configuration to use with them, with perhaps three cells in series for 9.6 V. Any ideas on this? It might be a squeeze to fit 3 in a Mag 3C body, but I imagine a 3D body would work. Then I wonder if the voltage would drive say an 1185 at full brightness, or if a different bulb would be a better match?

While I have never been a huge fan of the LFP chemistry, I have to admit that the really flat discharge curve even at high currents is advantageous for hotwire setups.

Tough call on selecting a bulb that would match a 3S string of the K2 Energy LFP26650EV cells. Looking at LuxLuthor's charts, it seems that an ROP 3853H bulb might work well, although you might need to worry about :poof: without a soft start. Looks like it could be borderline. If it does not instaflash, it would be a wonderful option for a 3D body.

Let me know what you try. I am currently using Eneloops in a 9AA->3D holder for my Mag85. It would be nice to compare that configuration to a 3S string of K2 LFP26650 cells with a 3853H bulb. It would certainly save money on the battery holder and on bi-pin adapter.

Cheers,
BG
 

Mr Happy

Flashlight Enthusiast
Joined
Nov 21, 2007
Messages
5,390
Location
Southern California
Let me know what you try. I am currently using Eneloops in a 9AA->3D holder for my Mag85. It would be nice to compare that configuration to a 3S string of K2 LFP26650 cells with a 3853H bulb. It would certainly save money on the battery holder and on bi-pin adapter.
Hmmm. Time to stop thinking and start ordering some bits...
 

Mr Happy

Flashlight Enthusiast
Joined
Nov 21, 2007
Messages
5,390
Location
Southern California
While I have never been a huge fan of the LFP chemistry, I have to admit that the really flat discharge curve even at high currents is advantageous for hotwire setups.

Tough call on selecting a bulb that would match a 3S string of the K2 Energy LFP26650EV cells. Looking at LuxLuthor's charts, it seems that an ROP 3853H bulb might work well, although you might need to worry about :poof: without a soft start. Looks like it could be borderline. If it does not instaflash, it would be a wonderful option for a 3D body.

Let me know what you try. I am currently using Eneloops in a 9AA->3D holder for my Mag85. It would be nice to compare that configuration to a 3S string of K2 LFP26650 cells with a 3853H bulb. It would certainly save money on the battery holder and on bi-pin adapter.

Cheers,
BG

OK, a follow up to this question. I didn't try the configuration above, but I have put a 3854-H bulb with two of the LPF26650P cells in a Mag 2D with lots of resistance fixes. It works pretty well. The bulb gives a very white light with plenty of power, and while it won't win any prizes it seems like a very practical upgrade to an otherwise boring light.

I have a couple of LFP26650EV cells on charge to compare with the LFP26650P's. The EV's have a useful amount of extra capacity that should be useful where practicality is concerned.

As far as the details go I had to replace the tail spring with a smaller recessed one due to the length of the cells and I bypassed all the steel springs with soldered copper wire. I made a plastic sleeve that exactly matched the diameter of the cells to the bore of the light. Then I did the obvious replacement of the reflector and glass with heat resistant items.

If there's a complaint, the beam pattern is not round. It's a shame the bulb has a horizontal instead of a vertical filament. The result is that the reflector tends to project the filament shape and you get a kind of wide flat hot spot. On the other hand you just rotate the light in your hand until the hot spot is horizontal and you get a kind of "car headlight" beam pattern.

It's the wrong forum for this, but the light from incans is so much "nicer" than from LEDs...
 

ZNickey

Newly Enlightened
Joined
Dec 19, 2009
Messages
37
I know this is bumping an old thread, but the information is great here and applies to a question which belongs here...

So now it is 2014...since this thread, Panasonic IMRs have seen a major reformulation (as have AWs and other users of those cells). How has cycle life improved?

Likewise, the Panasonic ICR18650 has resulted in 18650s with ridiculous amounts of capacity...my AW 3400 mAh 18650 pushes my Fenix PD32UE far beyond what disposable lithium batteries can, which not too long ago was unheard of!

Additionally, I've historically used AW IMRs to power high-drain devices such as personal vaporizers which need a battery capable of taking the heavy draw/high amperage. When I started, ICRs sometimes could not handle the load, and even in cases that they could, the harsher conditions would result in a dramatically reduced service life. Since I started, the ability for ICRs to handle heavy draws has improved significantly (almost to the point of being able to use all ICRs which I would prefer due to the capacity boost). From my usage of the latest AW IMR cells, I've gotten the better part of 600 cycles on them before the service life was below 80%...and they are used in an application that is exceptionally harsh on batteries (drawing major amps and quickly draining the battery as in recharge it at least once a day and sometimes more.)




So how do IMRs, ICRs, and IFRs stacking up today with all of the changes? Should new information be included here?

This is a very helpful thread!!!
 

StorminMatt

Flashlight Enthusiast
Joined
Oct 30, 2012
Messages
2,263
Location
Norcal
While I have never been a huge fan of the LFP chemistry, I have to admit that the really flat discharge curve even at high currents is advantageous for hotwire setups.

Speaking of that awesome discharge curve you get from LiFePO4, I have always wondered a couple of things. First of all, given that LiFePO4 operates in the same manner as other Li-Ion batteries, what is it about LiFePO4 that gives it that nice, flat curve? Also, might it be possible to make changes to 3.7V chemistries to give a similar flat discharge curve (albeit at a higher voltage)?
 

magellan

Honorary Aussie
Joined
Feb 3, 2014
Messages
5,001
Location
USA
I only just saw this. Thanks BG, that was great. I've copied it into my permanent notes on batteries and flashlight info. Are you an electrochemist or battery engineer by any chance?
 

ps56k

Enlightened
Joined
Oct 1, 2004
Messages
215
great info - just stumbled across it as our son bought some IMR 10440's for his ITP EOS flashlight that I got him years ago,
and I had no idea what he was talking about .... battery wise.
 

gmac

Newly Enlightened
Joined
Oct 12, 2014
Messages
1
Thanks, great info. I have not seen these cell types listed with any kind of protected circuit, is that not required? Does anyone know whether the Pila Charger can handle these cell types?
 

david57strat

Enlightened
Joined
May 29, 2012
Messages
485
Some great info, here. I have to get ready for work, but I'll look forward to reading this, when I get home!
 

SubLGT

Flashlight Enthusiast
Joined
Nov 18, 2013
Messages
1,170
Location
Idaho, USA
Isn't it time to let the IMR and ICR terminology fade away into oblivion? It simply does not provide that much information, IMO. Or it provides misleading information.

If IMR = LiMn2O4
If ICR = LiCoO2

Then to what group does the Panasonic NCR18650B (LiNiCoAlO2) belong? It is commonly called an ICR battery, but it is actually a "hybrid". Even calling a battery a hybrid does not provide much information.

 
Top