UVA 365nm LED vs Mercury Lamp: Industrial Curing Compared

UVA 365nm LEDs now match medium-pressure mercury lamps for most adhesive, coating and screen-printing ink curing, with instant switching, no warm-up and roughly 50-70% lower energy consumption. The comparison is system-level: spectrum, dose delivery, thermal load and lifecycle cost. The numbers below are what matter on a production line.

Spectral Match to Photoinitiators

A mercury lamp emits a line spectrum with strong peaks at 254, 313, 365, 405 and 436 nm plus a broad continuum that mostly becomes heat. A 365nm LED concentrates its output into one band with FWHM of 10-15 nm, aligned to the absorption peaks of the photoinitiators used in free-radical and cationic systems. Cure is defined by dose in J/cm² at the substrate, and a matched spectrum delivers that dose without UVB or UVC side reactions, which reduces yellowing, surface tack and substrate heating.

One discipline transfers from lamp lines. A recipe qualified under mercury was really qualified against a mix of 313, 365 and 405 nm lines, so verify photoinitiator absorption against the LED band before switching sources.

Irradiance, Power and Heat

Commercial 365nm chips run wall plug efficiency of roughly 15-30%, below the 50%+ typical at 395-405 nm, so head design carries the load. Focused COB heads deliver 8-20 W/cm² at the window and dense arrays exceed 20 W/cm², at or above a lamp-reflector system at working distance. LEDs emit almost no IR, so substrate temperature rise comes from absorbed UV and head radiation, not from a radiant IR load. At equal optical dose, the lamp dumps most of its input power as heat that the HVAC system then removes; this is the core of the 50-70% energy saving reported in retrofit audits.

Mercury systems need several minutes to stabilize, and output drifts as electrodes age. LED heads switch in microseconds, strobe with conveyor speed and log dose per part, which enables closed-loop cure control instead of fixed line speed.

Limits and Where Lamps Persist

Penetration depth at 365 nm is shallower than at 395-405 nm because shorter wavelengths are absorbed nearer the surface. Thick, heavily pigmented or highly filled layers may need higher irradiance, longer exposure or a dual-wavelength head. Kilowatt-class web lines at extreme speeds still sometimes justify lamp systems, although LED efficiency gains keep narrowing that gap.

Irradiance also falls with distance, so plan the optics so the working gap is fixed by the fixture rather than adjusted by operators. Selection sequence: match wavelength to the photoinitiator, set the dose window from the formulation datasheet, verify irradiance at working distance with a radiometer calibrated in the 365 nm band, and budget cooling so Tj stays at or below the level behind the manufacturer’s lifetime data.

FAQ

Q: What dose do typical adhesives need?

Formulation dependent, commonly 0.5-3 J/cm² for structural and laminating adhesives. Take the value from the supplier datasheet and confirm it with a dose meter at the fixture.

Q: How do 365nm LEDs fail compared with lamps?

Lamps wear through electrode erosion and arc drift. LEDs depreciate instead: watch silicone encapsulant yellowing under high UV dose, solder fatigue from thermal cycling, and irradiance loss tracked against the L70 curve.

Q: Is 365nm always the right choice?

No. Deep-section cure or low photoinitiator loading often favors 385-405 nm for penetration and efficiency, while 365 nm suits thin coatings and fast surface-cure lines.