Abstract
Abstract
Ultraviolet germicidal irradiation (UVGI) is traditionally based on low-pressure mercury lamps emitting at 254 nm, whereas UV–C light-emitting diodes (LEDs) have emerged as promising mercury-free alternatives. However, direct comparisons between these technologies are frequently confounded by differences in irradiance, exposure geometry, and accumulated UV–C dose, making it difficult to isolate wavelength-dependent effects from dosimetric bias. In this study, two closed irradiation chambers equipped with a low-pressure Hg lamp (254 nm) and an InGaN UV–C LED array (280 nm) were comprehensively evaluated in terms of emission spectrum, temporal irradiance stability, and spatial dose distribution. Radiometric measurements established matched irradiation conditions corresponding to source-to-sample distances of 41.0 cm for the Hg lamp and 5.5 cm for the LED system. For microbiological testing, both sources were operated for 270 s to reach stable irradiance, while the plates remained shielded from UV–C until the timed exposure began. Under matched accumulated UV–C doses, Staphylococcus aureus ATCC 25923 exhibited comparable microbial inactivation kinetics. A single-exponential decay model with a residual offset yielded similar model-estimated characteristic doses (D90, D99, and D99.9), while the survivor curves agreed in order of magnitude with historical and contemporary susceptibility data. Pairwise comparisons identified a statistically significant difference at only one exposure condition, whereas ANCOVA did not detect a statistically significant dose × technology interaction. These findings demonstrate that, when rigorous radiometric equivalence is established, the apparent performance differences between 254 and 280 nm become substantially smaller than generally inferred from nominal wavelength alone.