Abstract
This study synthesized PS-S⁺, PS-D⁺, and PS-T⁺, which contain carbon-carbon (C-C), carbon-carbon double bond (C=C), and carbon-carbon triple bond (C≡C) linkers between the electron donor and acceptor parts, respectively. Unlike PS-S⁺ and PS-T⁺, PS-D⁺, which contains a flexible carbon-carbon double bond linker, can absorb energy under white light irradiation, enhancing the oscillation of the connected receptor groups and leading to the formation of a non-coplanar structure. The non-coplanar structure formed by PS-D⁺ can reduce the energy gap (ΔEₛ₋ₜ) between the singlet and triplet states, thereby facilitating intersystem crossing, increasing the production of reactive oxygen species (ROS), and shifting the generation mechanism from Type II to Type I. In in vitro (E. coli and S. aureus) and in vivo (infected mice) wound dressing experiments, the developed fibers exhibited long-term good antibacterial performance, demonstrating their potential as sustainable and biosafe wound dressings to accelerate healing.
Design and Photophysical Properties of PS-S⁺, PS-D⁺, and PS-T⁺PS-S⁺, PS-D⁺, and PS-T⁺ were synthesized using the strong electron donor dimethoxy-substituted triphenylamine (TPA) and the electron acceptors benzothiadiazole and pyridine. The carbon-carbon (C-C), carbon-carbon double bond (C=C), and carbon-carbon triple bond (C≡C) linkers were inserted between the electron donor and acceptor to obtain PS-S, PS-D, and PS-T, respectively. They were then reacted with iodomethane to generate the positively charged PS-S⁺, PS-D⁺, and PS-T⁺. Triphenylamine, as a rotatable electron donor, possesses aggregation-induced emission (AIE) characteristics, which can suppress non-radiative decay, activate radiative emission, and enhance intersystem crossing.
When excited at the maximum absorption wavelengths of 457, 486, and 486 nm, PS-S, PS-D, and PS-T exhibited aggregation-induced emission effects. These derivatives do not emit light in tetrahydrofuran but activate their fluorescence when the proportion of the poor solvent water reaches 70%. The maximum absorption wavelengths of the positively charged photosensitizers PS-S⁺, PS-D⁺, and PS-T⁺ redshift to approximately 520, 532, and 530 nm in tetrahydrofuran, respectively. However, these photosensitizers do not produce fluorescence in both solution and solid states. In the aggregated state, the previously dissipated excited state energy (S₁) is released through enhanced radiative transitions and intersystem crossing pathways. This dual pathway provides significant opportunities to simultaneously enhance fluorescence intensity and excite to the triplet state (T₁). However, these processes compete with each other, and the increase in fluorescence intensity may hinder the efficiency of reactive oxygen species production.
Production of Reactive Oxygen Species by PS-S⁺, PS-D⁺, and PS-T⁺
We used the common reactive oxygen species sensor 2,7-dichlorodihydrofluorescein diacetate (DCFH-DA) to evaluate the overall reactive oxygen species production capabilities of PS-S, PS-D, PS-T, and PS-S⁺, PS-D⁺, and PS-T⁺. When the non-fluorescent DCFH solution containing each photosensitizer was continuously irradiated under white light (10 mW cm⁻²), the fluorescence intensity of 2,7-dichlorofluorescein (DCF) at 525 nm rapidly increased. All synthesized photosensitizers exhibited superior reactive oxygen species production capabilities compared to the commercial photosensitizer deuteroporphyrin e6 (Ce6). However, the reactive oxygen species production capabilities of PS-S⁺, PS-D⁺, and PS-T⁺ were significantly higher than their precursors PS-S, PS-D, and PS-T. The reactive oxygen species production capabilities of PS-D⁺ and PS-T⁺, which connect the electron acceptor pyridine through conjugated carbon-carbon double bonds and carbon-carbon triple bonds, were higher than that of PS-S⁺, which only contains carbon-carbon bonds. This trend was also observed in their precursors PS-S, PS-D, and PS-T.
We studied the types of reactive oxygen species produced by PS-S⁺, PS-D⁺, and PS-T⁺, which exhibited superior reactive oxygen species production capabilities. The dihydrorhodamine 123 (DHR123) probe can detect superoxide anions (O₂⁻), typically used to detect Type I reactive oxygen species. First, to exclude the influence of Type II reactive oxygen species on the DHR123 detection results, we conducted a control experiment using the typical Type II photosensitizer rose bengal (RB). Under white light irradiation, the DHR123 solution underwent slight photodegradation over time, resulting in a slight increase in fluorescence intensity at 530 nm. After adding RB under the same conditions, there was almost no change in fluorescence intensity at 530 nm. This result indicates that the singlet oxygen (¹O₂) produced by RB did not significantly affect the fluorescence of DHR123. Under continuous white light irradiation, the fluorescence intensity of DHR123 at 530 nm gradually increased in the presence of the three photosensitizers, indicating the production of Type I reactive oxygen species (O₂⁻).
PS-D⁺, which contains carbon-carbon double bonds, exhibited the highest superoxide anion (O₂⁻) production capability, although its overall reactive oxygen species production capability was comparable to that of PS-T⁺. Additionally, after adding PS-S⁺, PS-D⁺, and PS-T⁺ and irradiating for 3 minutes, a significant enhancement in fluorescence was observed. Importantly, when the oxygen radical scavenger vitamin C (Vc) was added simultaneously, the fluorescence intensity of DHR123 was nearly consistent with the unirradiated state, indicating that this signal originated from Type I reactive oxygen species (O₂⁻). These results strongly confirm that PS-S⁺, PS-D⁺, and PS-T⁺ primarily produce Type I reactive oxygen species (O₂⁻), classifying them as Type I photosensitizers.
Furthermore, we used 9,10-anthracenediyl-bis(methylene) dipropionic acid (ABDA) probe to verify the production of singlet oxygen (¹O₂), which is typically used to detect Type II reactive oxygen species. When singlet oxygen is produced, the absorbance of ABDA at 378 nm decreases with increasing irradiation time. Similarly, after irradiating for 6 minutes, the absorbance of ABDA at 378 nm decreased in a dimethyl sulfoxide/water mixed solution (water volume fraction of 90%) in the presence of the same concentration of PS-S⁺, PS-D⁺, and PS-T⁺. When the singlet oxygen output produced by RB was set to 1, the relative yields of PS-S⁺, PS-D⁺, and PS-T⁺ were calculated to be 2.33, 1.71, and 1.75, respectively. The trend of singlet oxygen production was: PS-S⁺ > PS-T⁺ > PS-D⁺. When using green light (495-570 nm, matching the absorption range of the photosensitizers) as the irradiation light source for reactive oxygen species performance testing, it was found that under the same testing conditions, green light irradiation exhibited superior reactive oxygen species production performance. These results indicate that all three photosensitizers can produce both Type I and Type II reactive oxygen species. PS-S⁺ has the lowest overall reactive oxygen species production capability but the highest Type II reactive oxygen species production capability. PS-D⁺ and PS-T⁺ have nearly the same overall reactive oxygen species production capability, while PS-D⁺ has the highest Type I reactive oxygen species production capability.
To further provide conclusive evidence for the generation of radical-type reactive oxygen species by PS-S⁺, PS-D⁺, and PS-T⁺, we employed electron spin resonance (ESR) spectroscopy. Under dark conditions, no signals were produced in the ESR spectrum when 5,5-dimethyl-1-pyrroline N-oxide (DMPO) and 2,2,6,6-tetramethyl-4-piperidone (TEMP) were mixed with the photosensitizer solution. When PS-S⁺, PS-D⁺, and PS-T⁺ were irradiated with TEMP, significant singlet oxygen signals were detected. Under light conditions, superoxide anion signals could be detected when the spin trapping agent DMPO was mixed with the photosensitizer in a methanol solution. Additionally, under light conditions, hydroxyl radical signals (1:2:2:1) were observed in the water solution mixed with DMPO and the three photosensitizers, which are related to the generation of paramagnetic radicals. Therefore, we confirmed that PS-S⁺, PS-D⁺, and PS-T⁺ can simultaneously produce Type I (O₂⁻ and ・OH) and Type II (¹O₂) reactive oxygen species, with PS-D⁺ exhibiting the highest Type I reactive oxygen species production efficiency.
Antibacterial Performance of PS-S⁺, PS-D⁺, and PS-T⁺
To evaluate the photodynamic antibacterial performance of PS-S⁺, PS-D⁺, and PS-T⁺, Escherichia coli (E. coli) and Staphylococcus aureus (S. aureus) were selected as representative Gram-negative and Gram-positive bacteria, respectively. Solutions containing different concentrations of PS-S⁺, PS-D⁺, and PS-T⁺ were prepared in dimethyl sulfoxide and dissolved in phosphate-buffered saline, co-cultured with bacterial suspensions. White light with an intensity of 50 mW cm⁻² was chosen as the light source, which was confirmed to have no antibacterial activity itself. First, antibacterial experiments were conducted at a photosensitizer concentration of 1.0 μM. Control plates without PS-S⁺, PS-D⁺, and PS-T⁺ showed abundant colony growth of E. coli and S. aureus under both light and dark conditions. However, E. coli and S. aureus treated with 1.0 μM PS-S⁺, PS-D⁺, and PS-T⁺ exhibited almost no colony growth under white light conditions, while numerous colonies were present on the dark group plates. Antibacterial experiments were conducted using different concentrations of photosensitizers. The antibacterial performance of PS-S⁺, PS-D⁺, and PS-T⁺ against E. coli and S. aureus increased with the increase in photosensitizer concentration. At a concentration of 0.6 μM, all groups exhibited almost no colony growth under white light irradiation. The survival rates of E. coli at 0.6 μM PS-S⁺, PS-D⁺, and PS-T⁺ were 7.46%, 1.60%, and 3.64%, respectively. In contrast, the survival rates of S. aureus at 0.6 μM PS-D⁺ and PS-T⁺ were both < 0.1%. These in vitro experiments collectively confirm that under white light conditions, PS-S⁺, PS-D⁺, and PS-T⁺ exhibit powerful antibacterial effects, with PS-D⁺ demonstrating the highest antibacterial activity.
Antibacterial Performance of PS-S⁺, PS-D⁺, and PS-T⁺ Fibers
Qualitative analysis of the antibacterial fibers was performed using the agar plate diffusion method to determine the inhibition zones. Pure cellulose fibers were placed alongside PS-S⁺, PS-D⁺, and PS-T⁺ fibers on the same plate and cultured under both light and dark conditions. After 18 hours of cultivation, the inhibition zone widths were measured. Under light conditions, no bacterial growth was observed beneath the antibacterial fibers, and a clear inhibition zone appeared around them. However, almost no inhibition zones were observed on the plates under dark conditions. After placing the agar plates in natural indoor light for three months, the areas surrounding the antibacterial fibers showed no bacterial or fungal growth, while pure cellulose fibers were covered with bacteria and fungi. In contrast, when the agar plates were placed in a dark environment indoors, both the antibacterial fibers and pure cellulose fibers were covered with bacteria and fungi. These results indicate that the antibacterial fibers possess long-term antibacterial capabilities under sunlight conditions. Subsequently, the antibacterial effects of PS-S⁺, PS-D⁺, and PS-T⁺ fibers at different concentrations were evaluated using an oscillation co-culture method. By observing the growth of bacterial colonies under different mass fractions of the photosensitizer antibacterial fibers, the impact of photosensitizer content on antibacterial efficacy was assessed. Under white light irradiation (50 mW cm⁻²), when the PS concentration > 0.5 wt%, the antibacterial efficacy of PS-S⁺, PS-D⁺, and PS-T⁺ fibers against E. coli and S. aureus reached 99%. Even after washing the fibers with deionized water 15 times, their antibacterial activity remained above 90%. Notably, as the number of washes increased, the photosensitizer dissolved into the water, leading to a decrease in the generation of antibacterial reactive oxygen species, thereby reducing the antibacterial effect of the fibers.
Antibacterial Fibers for Wound Photodynamic Therapy
The antibacterial fibers obtained through wet spinning were woven in a crisscross pattern to create small wound dressing fabrics approximately 1.5×1.5 cm2 . Furthermore, we established a wound model infected with Staphylococcus aureus in mice and evaluated the antibacterial capabilities of PS-S⁺, PS-D⁺, and PS-T⁺ dressings. Mice were randomly divided into 5 groups: infection control group, cellulose fiber treatment group, and 3 experimental groups treated with different antibacterial fibers. As the treatment time progressed, the skin wound area of the mice gradually decreased. On day 8, the wound healing of the PS-D⁺ and PS-T⁺ fiber-treated groups was superior to that of the other treatment groups. The wound closure rates of the control group were 44.02±14.63%, while the wound closure rates of the PS-D⁺ and PS-T⁺ fiber groups were 80.41±12.35% and 79.51±8.54%, respectively, nearly double that of the control group. Mice treated with PS-D⁺ (P<0.0001) and PS-T⁺ (P<0.0001) fibers exhibited significantly higher wound healing rates than the control group, but there was no significant difference between the two groups. This indicates that PS-D⁺ and PS-T⁺ fibers can effectively promote the healing of infected wounds. To further observe the role of antibacterial fibers in the inflammatory and proliferative phases of wound healing, we collected mouse skin tissues at different time points for hematoxylin-eosin (H&E) staining and Masson staining experiments. On days 5 and 8, hair follicle growth was observed in the PS-D⁺ and PS-T⁺ fiber groups. On day 2, the hair follicle growth in the PS-D⁺ (P<0.05) and PS-T⁺ (P<0.05) fiber groups showed significant differences compared to the control group. By day 5, this difference became more pronounced, with hair follicle growth levels showing extremely significant differences (P<0.0001). Similarly, on day 8, the PS-D⁺ (P<0.0001) and PS-T⁺ (P<0.0001) fiber groups exhibited abundant and evenly distributed collagen deposition, showing significant differences compared to the control group. This confirms that PS-D⁺ and PS-T⁺ fibers can promote wound healing and hair follicle growth in mice.
Conclusion
This study designed and synthesized three photosensitizers, namely PS-S, PS-D, and PS-T, which contain carbon-carbon single bonds (C−C), carbon-carbon double bonds (C=C), and carbon-carbon triple bonds (C≡C) linkers between the electron donor and acceptor molecules, respectively. Reactive oxygen species (ROS) generation experiments and density functional theory (DFT) calculations indicate that using π bridges to separate the donor and acceptor can significantly enhance the overall generation efficiency of reactive oxygen species. In particular, the flexible carbon-carbon double bond in PS-D undergoes trans-cis photoisomerization under UV light irradiation, forming a non-coplanar structure. During this process, the singlet-triplet energy gap (ΔEₛ₋ₜ) decreases from 0.968 eV to 0.8733 eV, enhancing intersystem crossing (ISC) and reactive oxygen species generation, and shifting the generation mechanism of reactive oxygen species from Type II to Type I.
References
https://doi.org/10.1021/jacs.5c07262
