All 5 electrode thicknesses looked fine until the charge rate went up
Argonne National Laboratory and BMW built 5 versions of the identical lithium nickel manganese cobalt oxide battery chemistry, changing only the electrode thickness, from 2.2 to 6.6 milliamp hours per square centimetre, the measure of active material on the electrode. A thin electrode is what most laboratory papers test, because it is easy to make and recovers well. A thick electrode is what a real car battery pack needs, because it carries less inert packaging weight per stored energy. At a gentle discharge, all 5 delivered between 92% and 97% of their own rated capacity, so thickness alone showed nothing wrong.
All 5 rows are the identical lithium nickel manganese cobalt oxide chemistry against graphite, differing only in electrode thickness. Figures measured at a slow, C tenth discharge rate, before the charge rate was raised in later testing described in the body text.
Show the numbers
| 2.2 milliamp hours per square centimetre | 93 |
| 3.3 milliamp hours per square centimetre | 93 |
| 4.4 milliamp hours per square centimetre | 97 |
| 5.5 milliamp hours per square centimetre | 95 |
| 6.6 milliamp hours per square centimetre | 92 |
The difference showed up only once the charge rate went up. All 5 cycled fine for 285 cycles at a moderate rate. On cycle 286 the rate rose. The thinnest electrode showed no measurable extra wear. The 2 thickest suffered a large, irreversible loss of capacity from lithium plating, metallic lithium building up on the electrode surface instead of moving back inside it. The paper states a working safety guideline directly, a graphite electrode should avoid charge current densities near or above 4 milliamps per square centimetre.
Cutting the electrolyte a cell carries raises the same number
A Spanish research institute, CIDETEC, found the same pattern in the electrolyte a cell carries. The team built a lithium sulfur pouch cell, a flat, foil wrapped cell using sulfur instead of the nickel and cobalt in many lithium ion cells today, first with a low sulfur loading and plenty of electrolyte, the typical conditions for a coin cell, the small disc shaped cell many laboratories use. That version reached 327 watt hours per kilogram. Raising the sulfur loading and cutting the electrolyte to capacity ratio, the weight of electrolyte carried for every amp hour stored, moved the same pouch cell to 436 watt hours per kilogram, a gain of 33% calculated here from the 2 figures CIDETEC reported.
The first 2 bars are the identical lithium sulfur pouch cell built by CIDETEC, source 2, with only the sulfur loading and the electrolyte to capacity ratio changed between them. The third bar is a separate, already finished 27 layer nickel rich lithium metal pouch cell built by a different team, source 4, shown to demonstrate the same mechanism reaches a working production cell, not to claim the 2 chemistries are directly comparable to each other.
Show the numbers
| Coin cell style loading | 327 |
| Production style loading | 436 |
| 27 layer finished cell | 489.7 |
Stacking more layers dilutes the weight that earns nothing
A team publishing in Nature Communications found stacking layers works the same way. Building a solid state lithium battery, using a solid electrolyte instead of a liquid one, the team found a pouch cell needs more than 12 stacked positive electrode layers before it clears 250 watt hours per kilogram, because more layers dilute the passive weight, the foil tabs, the outer casing and the current collectors that store no energy themselves, across more active material. A single layer laboratory cell, the format most papers test, carries that same passive weight across far less material, so a 1 layer result overstates a many layer version of the same chemistry. A separate team in China proved the fix already works at scale, building a 27 layer, 6.86 amp hour pouch cell with a nickel rich cathode and a lithium metal anode that reached 489.7 watt hours per kilogram finished, still holding 91.1% of its capacity after 150 cycles.
Labs keep testing the easy way because a factory line is not available
Fang Dai and Mei Cai, battery researchers at the General Motors Research and Development Center, explain why the gap keeps appearing.
Ideally, commercially production line-made cell, no matter pouch, cylindrical, or prismatic format, provide better reproducible result.
Fang Dai and Mei Cai, battery researchers at the General Motors Research and Development Center. Source 5.
A production line cell is more consistent than the small, hand built cell most university laboratories can afford, but a production line is not something a university laboratory can build. That is why a coin cell or a single layer laboratory pouch cell keeps producing headline numbers that a 27 layer, factory style cell, thick electrodes, real electrolyte amounts and fast charging included, does not repeat.