Impact of material properties for improved Pseudomonas aeruginosa biofilm inactivation with 280 nm UV LEDs

Inactivation rates on different materials

Inactivation potential for 280 nm UV LEDs on pure culture P. aeruginosa biofilm was assessed across eight materials: extruded Polytetrafluoroethylene (PTFE), Acrylonitrile Butadiene Styrene (ABS), Viton®, Silicone, High Density Poly Ethylene (HDPE), Stainless Steel, Porex (expanded PTFE), and Polycarbonate. Biofilm growth was quantified using standard plate counts, and data were analyzed using a non-linear model to determine the dose–response curve for each material tested. The UV LED irradiation reduced cell viability across all tested materials. As illustrated in Fig. 2, the dose–response curves have two regions of interest: slope and tailing.

The sloped (log-linear) region represents the rate of inactivation, whereas the tailing region demonstrates that increasing fluence does not result in additional log reduction for that test condition. Tailing may occur in instances where there is a resistant subpopulation of a microbe or if there is shielding/shadowing of a subset of microbes, which would prevent them from being exposed to germicidal light. Geeraerd’s Model can quantify this effect in the N_res term, which can provide insights into the impact of material surface in addition to microbial effects. Silicone, polycarbonate, and ABS demonstrated rapid inactivation with steep initial increases in log reduction values at lower fluence levels, achieving over 1.5 LRV at ~ 10 mJ/cm2. PTFE and ABS required higher fluences, ~ 12 mJ/cm² and ~ 35 mJ/cm², respectively, to achieve 1.5 LRV. Conversely, stainless steel, HDPE and Viton® did not achieve 1.5 LRV before exhibiting tailing. These observations suggest that surface characteristics play a crucial role in the susceptibility of biofilms to UV LED inactivation. Material surface impacts biofilm attachment and accumulation, which can account for differences in inactivation potential.

The fluence required to reach the tailing region of log reduction varied among the materials. Silicone and polycarbonate reached their maximum log reduction at around 10 mJ/cm², while Porex, stainless steel and Viton® required 15 mJ/cm², PTFE required approximately 25 mJ/cm², and ABS and HDPE required upwards of 35 mJ/cm². These results indicate that the effectiveness of 280 nm UV LED biofilm inactivation is highly dependent on the material’s properties and the applied fluence. The variability in LRV across tested materials demonstrates the importance of further surface characterization for results to be fully contextualized and analyzed.

The calculated inactivation rate constants (k value), residual standard errors and the peak LRV achieved for each material are provided in Table 1. These values allow for a comparison of UV LED effectiveness across material types and further illustrate the impacts and variability that material surface can have on experimental results.

Inactivation rate constants provide a clear picture of the inactivation effectiveness across the materials tested. A one-way analysis of variance  (ANOVA) revealed a statistically significant difference in k values among various material types (p = 3.92e−9). To further investigate these differences, a Tukey’s Honestly Significant Difference (HSD) test was run, which identified several significant comparisons. The most notable significant differences were observed between Porex and ABS, with a mean difference of 0.1527 (p = 0.000126), indicating that Porex is significantly more effective than ABS. Similarly, significant differences were found between Silicone and ABS (p = 0.00014), and PTFE and ABS (p = 0.0284), suggesting that both Silicone and PTFE also demonstrate higher effectiveness compared to ABS. In contrast, the comparisons between HDPE and ABS (p = 1.00), Polycarbonate and HDPE (p = 0.0821), as well as those between Silicone and PTFE (p = 0.535), showed no significant differences. These findings highlight the differences between materials and emphasize the importance of material choice when considering inactivation effects within a system.

Potential role of surface properties for UV inactivation of biofilms

With the limited number of published studies which employ UV LEDs for the inactivation of biofilm on material surfaces, the comparison of collected k values to published values is difficult. Pousty et al.² employed 270 nm emitting LEDs to inactivate pure culture P. aeruginosa biofilm on three material types (PTFE, polycarbonate and PVC) and observed k-values of 0.133 ± 0.023, 0.416 ± 0.089 and 0.416 ± 0.089, respectively. The k values for PTFE and polycarbonate used in both studies fall within the confidence interval, indicating similar inactivation effectiveness between the experiments.

Chemicals used for surface inactivation, such as chlorine (Cl₂), are commonplace in many homes and industries alike; however, these chemicals may have limited use cases due to surface or human sensitivity and the inability for their inactivation to be augmented using additive effects. Surface reflectance has been shown to impact UV irradiation, and it is hypothesized that increased surface reflection may present an additive inactivation impact for surface inactivation^50,51. The diffuse reflectance for each material, measured in triplicate in increments of 0.5 nm, from 200 to 400 nm, is illustrated in Fig. 3 provides the reflectance percentage at 280 nm for each material.

Diffuse reflectance measurements of materials (a–h) used for biofilm formation in CDC biofilm reactors for inactivation by 280 nm UV LED irradiation (n = 3). Shaded areas represent 95% confidence intervals around the mean.

The diffuse reflectance at 280 nm demonstrated a broad range of values, reflecting their varying interactions with UV light. Viton®, a firm black rubber, exhibited the lowest reflectance at 5.2%, followed by Silicone, a soft, transparent rubber, with a reflectance of 6.83%. Polycarbonate, a frosted glass-like material, showed slightly higher reflectance at 7.4%, closely followed by ABS, a hard black plastic, at 8.4%. HDPE, an opaque white plastic, reflected 12.8% of UV light at 280 nm, while Stainless Steel (316L), known for its shiny silver surface, had a reflectance of 16.7%. The two most reflective materials were PTFE, a hard white plastic, and Porex, a flexible white plastic sheet, which demonstrated reflectance of 44.2% and 84.4% respectively, at 280 nm. This variation in reflectance at 280 nm highlights the importance of material properties in UV light interaction and their potential for effective biofilm inactivation.

A one-way ANOVA was conducted on surface reflectance across the materials and revealed a significant effect of material type (p < 2e−16). Significant differences in reflectance between material pairs were also noted. HDPE demonstrated a substantially higher reflectance compared to ABS, with a difference of 16.54% (p < 0.0001), while the difference between Porex and ABS was even more pronounced at 75.71% (p < 0.0001). A two-way ANOVA examining the interaction between material and wavelength also yielded significant results for material type (p < 2e−16), wavelength (p < 2e−16), and the interaction of material with wavelength (p < 2e−16). This indicates that both the type of material and the wavelength significantly influence reflectance, and the interaction suggests the effect of material on reflectance varies with wavelength. These findings highlight the complex relationship and indicate that surface reflection can contribute to an increase in inactivation capacity. However, a greater dataset with increased material types would have to be tested to identify the extent of the impacts.

Surfaces with increased roughness or variation can contribute to increased biofilm attachment^52,53,54. The surface roughness values are shown in Table 2, where R_a represents the average roughness of the material above the mean line, R_z is the average maximum height of the profile, and R_sm is the mean width of the profile elements.

A correlation analysis revealed a significant correlation between surface reflectance and surface roughness (p = 0.04). These findings indicate that higher surface roughness is associated with increased diffuse reflectance. To understand the impacts on inactivation due to surface roughness, a further correlation analysis was run between k value and surface roughness (p = 0.202) and surface roughness and N_res (p = 0.484), suggesting neither correlation is statistically significant for the materials tested in this study. This indicated that surface roughness alone does not govern the potential for a material to be inactivated.

Additionally, a linear regression was run to understand the impacts of surface roughness versus k values (p = 0.202, R² = 0.255) and N_res (p = 0.484, R² = 0.0848). However, the regressions were not significant. While there appears to be a small relationship between roughness and k-values, in this study, the lack of statistical significance suggests that roughness does not provide predictive power for k values or N_res.

The materials which measured with higher R_a and R_z values, such as silicone rubber (R_a = 13.7 µm, R_z = 108.7 µm) and Porex (R_a = 19.84 µm, R_z = 100.01 µm), provide more surface area and microenvironments for biofilm attachment and formation. In contrast, materials like ABS (R_a = 3.26 µm, R_z = 22.8 µm) and stainless steel (R_a = 3.45 µm, R_z = 33.3 µm) have smoother surfaces, which can inhibit biofilm attachment due to the minimal surface texture. This is represented in this study as materials with higher roughness also resulted in higher maximum LRVs. Silicone (R_a = 13.69 µm) and Porex (R_a = 19.86 µm) were measured as the two roughest materials, but also demonstrated the first and third largest LRVs, 2.05 and 1.75 CFU/cm², respectively. Furthermore, the modelled N_res for both silicone and porex, 1.86 and 1.75, were amongst the largest of the tested materials. This demonstrates that these materials have the potential to provide favourable conditions for microbial shielding, allowing for a larger resistant sub-population than the other tested materials.

However, this observation does not hold for all materials, for example, HDPE, which measured comparatively low on surface roughness (R_a = 4.63 µm, R_z = 45.89 µm) but exhibited a higher LRV (LRV max. = 1.17). This change could be attributed to additional surface characteristics of HDPE, such as zeta potential, less electrostatic repulsion or hydrophobicity^55,56.

Other approaches to enhance biofilm inactivation with UV LEDs

Optimization of UV LED parameters, including wavelength, intensity, and exposure time, is crucial for maximizing biofilm inactivation across material types. Identifying surface characteristics such as roughness and reflectivity allows for the tailoring of UV LED inactivation processes to specific use cases. For instance, ABS required a fluence of 20 mJ/cm² to achieve maximum inactivation, while PTFE and Viton® reached peak inactivation at 10 mJ/cm². Understanding the differences between material types and their respective inactivation needs allows for only the required fluences to be applied.

These findings have practical implications for the selection of materials in applications requiring biofilm inactivation using UV LEDs. Materials like ABS, PTFE, and Porex, which show higher and rapid inactivation, are recommended for applications where quick and effective biofilm control is essential. Conversely, materials like HDPE and Polycarbonate, which exhibit lower inactivation rates, may require higher fluence or prolonged exposure for effective biofilm control. Further research is recommended to optimize UV LED parameters for each material to enhance inactivation efficiency. Ultimately, future work should investigate additional material types and the potential synergistic effects with other variables (e.g., light wavelength) to advance the possibilities of UV LED biofilm inactivation.