High-voltage lithium-cobalt-oxide is the quiet cathode upgrade that reshaped premium wearables over the last three years. Pushing the charge cut-off from 4.35 V to 4.45 V, and now to 4.48 V, buys roughly 8–12% more volumetric energy density at the cell. Here’s what that looks like in real products.
The short version
At the cell level:
- 4.40 V LCO: typical 690–720 Wh/L, 500–700 cycles to 80% SOH.
- 4.45 V LCO: typical 720–750 Wh/L, 450–600 cycles.
- 4.48 V LCO: typical 750–780 Wh/L, 400–500 cycles.
The numbers vary 5–10% between suppliers and formulations, but the trade-off is consistent: each step up in voltage buys density at the cost of cycle life.
Why it works
A lithium-cobalt-oxide cathode holds more lithium at higher states of charge. Pushing the upper cut-off from 4.35 V to 4.48 V extracts more of that lithium per charge. The underlying electrochemistry is the same; the practical difference is in the coating composition and in the electrolyte additives that stabilise the cathode at the higher voltage.
The cycle-life penalty comes from two mechanisms. First, higher voltage accelerates electrolyte oxidation at the cathode interface, consuming active lithium over cycles. Second, the cathode itself experiences more structural stress during the larger lithium swing, accelerating particle cracking. Additive chemistry mitigates both effects but doesn’t eliminate them.
The pack-level reality is smaller
The 8–12% cell-level gain does not translate 1:1 into runtime. Several factors compress the benefit by the time it reaches the user:
- Charger IC losses are proportional to voltage. Charging to 4.48 V instead of 4.40 V slightly increases the IR² losses in the charge path. Maybe 0.5–1% of the gain disappears here.
- Upper end of charge is less accessible. Above 4.4 V the cell can only absorb charge at progressively lower C-rates without stress. In practice, most fast-charge implementations taper at 4.35–4.40 V and complete the last few percent slowly. On real usage patterns this effectively caps users at 98% rather than 100%.
- Voltage sag under load is worse at high SOC. The usable energy between 4.48 V and 3.0 V is higher than between 4.40 V and 3.0 V, but the usable energy above a load-induced minimum of 3.4 V is closer.
Net real-world gain: 5–8% instead of the headline 8–12%.
Who’s shipping it in 2026
4.45 V is now the default for premium smartwatches and AR glasses. 4.48 V is still flagship-only — mostly because the electrolyte-additive packages are more expensive and because warranty claims from early-generation 4.48 V cells scared some OEMs off.
The picture by category in mid-2026:
| Category | Typical cut-off |
|---|---|
| Mid-range TWS earbuds | 4.35 V |
| Premium TWS earbuds | 4.40–4.45 V |
| Premium smartwatches | 4.45 V |
| Mid-range AR glasses | 4.45 V |
| Flagship AR glasses | 4.45–4.48 V |
| Medical wearables | 4.35 V (conservative) |
Qualification gotchas
Three things that commonly surprise programs qualifying their first HV LCO cells:
- Swelling is worse. The cathode stress shows up as incremental volumetric expansion. Expect 1–2% more cell swelling at full charge compared to 4.40 V equivalents.
- Calendar aging at 100% SoC is harsher. Storing a 4.48 V cell at full charge and elevated temperature loses capacity 30–50% faster than the 4.40 V equivalent under the same conditions. Design SoC limits accordingly.
- UN 38.3 overcharge test applies 2× rated voltage. For a 4.48 V cell that’s 8.96 V to the cell terminals, which is a harder condition than the 4.40 V case (8.80 V). Some early 4.48 V cells failed T7 where the 4.40 V equivalent passed.
Where it’s going next
The cathode research community has shown 4.55 V and even 4.60 V platforms in lab samples. Commercial viability at those voltages depends on electrolyte stability that doesn’t yet exist at mass-production cost. Our rough expectation: 4.50 V becomes commercially viable in 2027 for flagship products; higher voltages remain lab curiosities through 2028.
In parallel, semi-solid chemistry (see the separate article) offers a different path to density gain without the voltage risk — so the industry ends up with two parallel premium tiers, one voltage-driven and one electrolyte-driven.
What OEMs should consider
- Don’t chase 4.48 V for a mid-tier product. The cost premium and cycle-life penalty aren’t worth a single percentage point of runtime on a commodity device.
- Qualify 4.45 V for your premium tier, 4.48 V only for flagships with bulletproof thermal design.
- Test calendar aging at full charge and 40 °C. Most HV LCO failures in the field are calendar, not cycle.
- Don’t mix voltages across SKUs. One cell, one cut-off — mixing creates firmware and warranty confusion.