Is it possible for you to elaborate more on the internal chemistry breaking down?
How is time involved in the breakdown of the internal chemistry? I've always understood heat but never understood the time portion of it.
The problem with lithium ion cells that use a cathode based on cobalt (LiCoO2) is that they oxidize over time. So no matter how careful you are, they will lose capacity over time from the moment they leave the factory. I built my own electric vehicle (electric motorcycle) using lithium cobalt batteries, and I've lost about 40% capacity over 2.4 years and this was while I was babying them. Cobalt-cathode lithium ion and lithium polymer rechargeable are the ones used in most consumer electronics applications (laptops, cell phones, MP3 players, etc).
There's a discussion of this on Wikipedia here:
http://en.wikipedia.org/wiki/L...onal_Li-ion_technology
As Bobsmith mentioned, to maximize the life of lithium cobalt (ie. laptop batteries) and lithium polymer batteries (ie. iPod batteries), you want to keep them about 50% state of charge or as close to 50% charged as possible (so when you use them charge to 70%, discharge to 30%). You want to keep them cool at all times - storage around 30F/-1C works well and you want to avoid deep discharges or high current output (eg. >5C, although this depends on the chemistry).
As was mentioned the problem of "memory" in nickel batteries is not really a problem with the batteries, but rather with the charging systems. A good charger can cycle out the "memory effect" (
http://www.duracell.com/oem/re...le/Nickel/voltdep.asp). Since the car manufacturers have any quality battery engineers involved in the design of their battery systems, memory effects on NiCd/NiMH in automotive applications is not an issue. In fact, if battery manufacturers use more expensive components in their rechargeable NiMH battery chargers, home users shouldn't see the "memory" effect either.
Lithium ion (LiCoO2), lithium polymer and lithium ion phosphate and actually most other battery major rechargeable chemistries (like lead-acid) use a constant currrent style charging system that basically charges at a fixed current until a voltage is reached and then taper off the current to hold at that voltage and then when the current reaches a low-enough rate, they cut off.
With lithium ion(LiCoO2) you want to be very careful not to each the maximum voltage (else they can explode), you also don't want to discharge them below their minimum voltage (else they explode, or if you are lucky, they stop working), you want to avoid getting them too hot (else they explode), or charging at too high a rate (this reduces the longevity of the battery, and causes heat... which could lead to explosion).
A newer chemistry called lithium ion iron-phosphate (LiFePo) doesn't have any of these explosion issues, and oxidizes very slowly - in fact it's more robust and safer to use than NiMH (which can vent hydrogen at high temperatures), but presently it's energy capacity per weight is closer to NiMH than it is to traditional LiCoO2. An even newer chemistry - as mentioned by MarkR - is lithium ion titanate (LiTiO3), but it's not in high volume manufacturing yet.
The high-volume mainstream plug-in electric hybrid cars scheduled for release in ~2010 seem to be falling into two main chemistries: NiMH (Toyota Prius, among others), and LiFePo (Chevy Volt, Saturn Vue, among others). Since LiFePo doesn't have a huge advantage in terms of energy capacity per weight over NiMH, it's biggest advantages lie in it's temperature range and it's tolerance of high discharge current. But the advantage of NiMH over LiFePo is that NiMH's characteristics are extremely well understood and NiMH manufacturing is less complex. One huge advantage LiFePo has over NiMH, however, is that the patents for automative applications of LiFePo batteries are not controlled by an oil company - Chevron - like NiMH automotive batteries are and so broad licensing of the patents involves makes LiFePo development easier.