Luminescent Electro‐Spun Nanofibers Crosslinked with Boronic Esters Exhibiting Controlled Release of Carbon Dots for Detection of Wound pHs and Enhanced Antimicrobial

ABSTRACT

Timely and accurate assessment of wounds during the healing process is crucial for proper diagnosis and treatment. Conventional wound dressings lack both real‐time monitoring capabilities and active therapeutic functionalities, limiting their effectiveness in dynamic wound environments. Herein, we report our proof‐of‐concept approach exploring the unique emission properties and antimicrobial activities of carbon nanodots (CNDs) for simultaneous detection and treatment of bacteria. This approach centers on the fabrication of well‐defined CND‐embedded poly(vinyl alcohol) (PVA) e‐spun nanofibrous mats, which are crosslinked with degradable boronic ester (BE) crosslinks. The BE‐CND/PVA mats exhibit stimuli‐responsive degradation to pHs and hydrogen peroxide as well as pH‐responsive release of CNDs. Promisingly, the mats turn out to be hemocompatible with blood and biocompatible with skin cells. Furthermore, they exhibit notable antimicrobial activity against Gram‐negative bacteria and demonstrate great potential for real‐time monitoring of wound pH to assess the wound status. These results suggest that BE‐CND/PVA mats could significantly enhance wound healing by providing localized therapeutic action, reducing the risk of bacterial infections, and enabling non‐invasive monitoring of wound progress.

Fabrication of well‐defined CND‐embedded e‐spun nanofibrous mats crosslinked with degradable boronic ester crosslinks in response to pHs and hydrogen peroxide found in wounds, and capable of pH‐responsive release of CNDs, for simultaneous detection and treatment of bacteria.

1. Introduction

Wound healing is a complex biological process involving a series of coordinated events, including hemostasis, inflammation, proliferation, and tissue remodeling [1]. The choice of appropriate wound dressing is essential to facilitate wound healing, because it provides a protective barrier, maintains moisture balance, prevents infection, and supports cellular migration and proliferation [2]. Compared with traditional dressings such as gauze, bandages, and cotton pads, modern dressings have been designed to provide an ideal microenvironment for cellular activity, promote faster healing, and in some cases deliver therapeutic agents such as antimicrobial compounds or growth factors [3]. Among various dressings, electrospun (e‐spun) nanofibrous scaffolds offer unique structural and functional advantages. E‐spun nanofibers mimic the extracellular matrix of human tissue, providing a highly porous, interconnected structure that enhances cell adhesion, infiltration, and regeneration [4]. Additionally, their high surface area to volume ratio allows for efficient drug loading and release of therapeutic agents [5].

Most e‐spun nanofibers have been fabricated by e‐spinning process for polymer solutions because the process offers a wide range of configurations that provide great flexibility in material selection, process parameters variability and instrument setup adaptability [6]. Among various polymers used for electrospinning, poly(vinyl alcohol) (PVA) is widely studied owing to its biocompatibility, non‐cytotoxicity, hydrophilicity, and film‐forming ability [7]. PVA is a water‐soluble synthetic polymer that possesses great mechanical strength, making it suitable for fabricating wound dressings. Additionally, its hydrophilic nature helps in maintaining the moisture balance necessary for efficient wound healing [8].

Bare PVA nanofibers are required to be crosslinked to retain their structural integrity and high mechanical strength in an aqueous environment, with rapid degradation for controlled/enhanced release of encapsulated antimicrobials when exposed to wound environments [9]. Chemical crosslinking through the formation of ester [10], carbamate [11], and acetal [12] crosslinks has been explored; however, this approach presents critical drawbacks, including concerns about cytotoxicity or uncontrolled release of encapsulated antimicrobials [13]. Boronic ester (BE) bonds, formed by condensation reaction of boronic acid and diols, can be cleaved in response to acidic and alkaline pHs as well as ROS and glucose, which are all found in wounds [14]. Our proof‐of‐concept investigation demonstrates the versatility of our approach exploring degradable BE chemistry toward the fabrication, pH‐responsive degradation, and controlled release of well‐defined PVA e‐spun fibrous mats crosslinked with BE crosslinks. Given their antimicrobial activities, our approach opens up new avenues for stimuli‐responsive wound dressings with enhanced biocompatibility, stability, and controlled drug release [15].

Carbon nanodots (CNDs) are photoluminescent, carbon‐based, nanoscale materials that exhibit unique optical properties, high biocompatibility, excellent water solubility, and facile functionalization. Owing to these features, CNDs have been widely explored in biomedical engineering and biotechnology for bioimaging, biosensing, and drug delivery [16]. Further, CNDs have shown antibacterial properties, making them a promising alternative to antimicrobials against antibiotic‐resistant bacteria [17]. Given their versatile optical properties, mostly optical and emission properties, CNDs have been incorporated into the design of PVA e‐spun nanofibers for the development of effective CND/PVA hybrid materials with enhanced functionalities suitable for various applications, mostly biosensing and environmental remediation. For example, reports describe CND/PVA nanofibers crosslinked with glutaraldehyde for the detection of doxorubicin [18] and thermally‐activated delayed fluorescence [19] as well as PVA nanofibers embedded with graphene quantum dots for sensing hydrogen peroxide and glucose [19]. In addition to biosensing, reports describe PVA nanofibers co‐doped with AuNPs and CNDs for the detection of cyanide [20], CND/PVA fibers functionalized with dipyrrole‐labeled boronic acid to remove ionic pollutants (fluoride ions) [21], and CND‐embedded PVA/chitosan co‐nanofibers for the detection and removal of dye pollutants in wastewater [22]. Despite these advances, only a few reports describe CND/PVA fibers for wound dressing and healing to the best of our knowledge [23].

We have hypothesized a new strategy aiming at the controlled release of CNDs as therapeutic agents from chemically crosslinked PVA e‐spun nanofibrous mats for simultaneous infection detection and treatment. Given that wound infections often lead to an alkaline microenvironment and an elevated level of reactive oxygen species (ROS), we anticipate that our stimuli‐responsive CND/PVA hybrid nanofibers could exhibit the selective release of CNDs at wound sites, thereby providing a dual‐functional wound dressing capable of both sensing wound pH to give real time information and combating infection in a controlled manner.

In this work, as a proof‐of‐concept to investigate our hypothesis for simultaneous detection and treatment of bacteria owing to unique emission properties and antimicrobial activities of CNDs, we fabricated well‐defined CND/PVA e‐spun nanofibrous mats crosslinked with degradable BE crosslinks (BE‐CND/PVA mats) (Scheme 1). A commercially available, benzen‐1,4‐diboronic acid (DBA) was chosen as a phenyldiboronic acid crosslinker to crosslink PVA fibrous mats in an organic solvent. The fabricated BE‐CND/PVA mats were characterized for stimuli‐responsive degradation to pHs and hydrogen peroxide, followed by pH‐responsive release of CNDs. Further, their biocompatibility with cells, hemocompatibility with blood and antimicrobial activities against both Gram‐positive and Gram‐negative bacteria, as well as their potential for real‐time monitoring of wound pH to assess the wound status, were examined.

SCHEME 1.

Well‐defined CND/PVA e‐spun nanofibrous mats crosslinked with degradable BE crosslinks (BE‐CND/PVA mats) for simultaneous detection and treatment of bacteria owing to unique emission properties and antimicrobial activities of CNDs.

2. Experimental

2.1. Lab‐Made Electrospinning Setup

The lab‐made electrospinning setup includes a Spellman CZE1000R high‐voltage direct current power supply (5–30 kV) and a spinneret with a conventional stationary collector consisting of an aluminum foil target in a Fisher model 630D incubator (see Figure S1). The spinneret includes a syringe pump (LEGATO110 KD Scientific) equipped with a 5 mL syringe and a blunt‐ended single‐nozzle metal needle of 25 gauge and 1.5 inches long. In the incubator, a small muffin fan (75 mm, 6 W) was used to circulate air. The collector was set to a negative high voltage (e.g., −25 kV), while the spinneret and metal‐body incubator were grounded.

2.2. Instrumentation and Characterization

Scanning Electron Microscopy (SEM) was conducted using a Phenom ProX with a resolution of 8 nm or less at an acceleration voltage of 10 kV. Nanofiber samples were dried in a vacuum oven for 24 h before being mounted on a stub using double‐sided carbon tape and then coated with a 5 nm thick gold layer using a Cressington 108 Auto Sputter Coater. The average diameter of the fiber was determined using ImageJ software on over 200 randomly selected nanofibers.

UV/Visible spectra were recorded using a Cary 60 UV–vis Agilent Spectrophotometer (Agilent Technologies) over a range of 200–800 nm using a 1 cm quartz cuvette with a scan speed of 400 nm s⁻¹. Data was processed using Cary Eclipse software, and the background was corrected with solvents.

Fluorescence spectra were recorded using a Cary Ellipse Fluorescence Spectrophotometer (Agilent Technologies) using a 1 cm quartz cuvette. All emission data were collected by setting λ_ex at 350 nm over a range of 355–600 nm. The excitation and emission slits were set to a width of 5 nm with a medium PMT voltage and a scan rate of 600 nm min⁻¹. All data were processed using Cary Eclipse software. Background corrections were made to account for solvents.

Fourier‐Transform Infrared (FTIR) spectra were recorded using a Thermo Scientific Nicolet iS5 equipped with an iD5 attenuated total‐reflection (ATR) accessory. The spectra were collected with 64 scans at a resolution of 0.4 cm⁻¹, gain of 1, optical velocity of 0.4747, and an aperture setting of 100. Data was processed using Omnic 9 software.

Differential Scanning Calorimetry (DSC) measurements were performed using a DSC Q20 (TA Instruments) to evaluate the thermal properties of nanofibers. All nanofiber mats were measured under a nitrogen atmosphere with a flow of 50 mL/min. The samples were tightly packed in a Tzero aluminum pan and equilibrated at −80°C for analysis. For every sample, three cycles were performed, with each cycle involving heating to 250°C at a rate of 10°C min⁻¹ and then cooling to −80°C at a rate of 5°C min⁻¹. Thermal properties, including glass transition temperature (T_g) and melting temperature (T_m), were analyzed from the second heating run.

2.3. Materials

Poly(vinyl alcohol) (PVA, 89–98 kDa, 99% hydrolyzed), benzene‐1,4‐diboronic acid (DBA, >95%), phosphate buffered saline (PBS) tablets, Mueller Hinton Broth 2 (MHB) microbiology culture medium and agar powder (quality level 100) were purchased from Millipore Sigma and used without any further purification unless otherwise mentioned. N,N‐dimethylformamide (DMF, ACS grade) was treated with 3 Å molecular sieves (8–12 mesh) to remove residual water.

2.4. Synthesis of Well‐Defined CNDs

CNDs were synthesized using a CEM Discover SP microwave reactor. Citric acid (0.384 g, 500 mm) and ethylenediamine (375 mm) were mixed with distilled water (4 mL) in a microwave reaction tube. The resulting mixture was subjected to sonication for 15 min to form a homogeneous solution. The formed solution was heated to 210°C for 10 min and then dialyzed over Milli‐Q water using a cellulose dialysis tubing (MWCO = 3.5–5.0 kDa) over 5 days. Outer water was changed twice per day. After dialysis, the solution was washed twice with acetone and ethanol, and the precipitate was collected by centrifugation (10 000 rpm × 10 min) after each wash. Precipitate was dried in an oven at 80°C for 14 h and then crushed to a fine powder.

2.5. Fabrication of CND/PVA Nanofibers

PVA powder (5.0 g) was dissolved in deionized water (50 mL) to prepare a 10 w/v % aqueous PVA solution under magnetic stirring at 80°C for 3 h. Aliquots of the formed PVA solution (10 mL) were mixed with the various amounts of the purified CNDs at 0.5, 1, 2, and 5 wt% based on PVA under magnetic stirring at room temperature for another 1 h. A series of the formed aqueous CND/PVA solutions was degassed and e‐spun under the optimal conditions.

2.6. Crosslinking of CND/PVA Nanofibers with DBA

Pieces of CND/PVA (1 wt%) mats (≈ 50 mg, containing 1 wt% CND), after being removed from aluminum foils, were immersed in DMF (20 mL) containing different amounts of DBA for 48 h. The amount of DBA was varied as the mol equivalent ratio of BA/2OH to be 0.25/1, 0.5/1, 1/1, and 2/1. The mats were dried in a vacuum oven preset at 60°C for 24 h. DBA incorporation efficiency (%) and OH reacted (%) were calculated.

$$\begin{array}{ccc}& & \mathrm{DBA}\mathrm{incorporation}\mathrm{efficiency}(\%)\hfill \\ & & =\frac{(W_{\mathrm{f}}-W_{\mathrm{i}})/\mathrm{M}{\mathrm{W}}_{\mathrm{DBA}}}{\mathrm{mole}\mathrm{of}\mathrm{DBA}\mathrm{in}\mathrm{recipe}}\times 100\%\hfill \end{array}$$ (1)

$$\mathrm{OH}\mathrm{reacted}(\%)=\frac{4\times \left(W_{\mathrm{f}}-W_{\mathrm{i}})/\mathrm{M}{\mathrm{W}}_{\mathrm{DBA}}\right)}{\mathrm{Initial}\mathrm{mol}\mathrm{of}\mathrm{OH}\mathrm{in}\mathrm{PVA}}\times 100\%$$ (2)

where, W _f and W _i are the weight of dried CND/PVA mat before and after treatment with DBA, and MW_DBA is the molecular weight of DBA (165.75 g/mol).

2.7. Stimuli‐Responsive Degradation of BE‐CND/PVA Mats

For pH responses, pieces of BE‐CND/PVA (1 wt%) mats (≈12 mg) fabricated with BA/2OH = 2/1 were immersed in PBS solutions (50 mL) at pH = 5.4, 7.4, and 8.4 for 48 h. Either 1 m NaOH or 1 m HCl was used to adjust pHs. The mats were taken and dried in a vacuum oven preset at 40°C for 24 h. Their weights were recorded for gravimetric analysis. For the hydrogen peroxide response, pieces of BE‐CND/PVA (1 wt%) mats (≈12 mg) fabricated with BA/2OH = 2/1 were immersed in aqueous hydrogen peroxide solution (1.0 mm, 50 mL) for 72 h. Mats were taken and dried in a vacuum oven preset at 40°C for 24 h. Their weights were recorded for gravimetric analysis.

2.8. pH‐Responsive Release of CNDs from BE‐CND/PVA Mats using Fluorescence Spectroscopy

To construct correlation curves of CNDs at pH = 5.4, 7.4, and 8.4, a series of aqueous solutions of CNDs at various concentrations at 0.5–15 µg/mL were prepared. Their emission spectra were recorded at λ_ex = 350 nm. To study pH‐responsive release of CNDs, pieces of BE‐CND/PVA (1 wt%) mats (≈12 mg) fabricated at BA/2OH = 2/1 were immersed in PBS solutions (15 mL) at pH = 5.4, 7.4, and 8.4. Aliquots (3 mL) were taken to record their emission spectra for given time intervals. The same volumes of fresh buffer solutions were backfilled to keep the sink condition.

2.9. In Vitro Hemocompatibility

Sprague Dawley rat RBCs (Fisher Scientific) were used to evaluate the hemocompatibility for BE‐PVA nanofibers with and without 1 and 5 wt% CNDs. RBCs as received were diluted in sterile PBS solution (2 wt% RBC suspension) and subjected to centrifugation (1000 g × 5 min × 4°C) to obtain transparent, pink‐colored supernatant (indicating that RBCs were not hemolyzed and were ready for tests). Sterilized disk‐shaped BE‐PVA nanofibers (≈ 0.5 mg, 0.8 mm in diameter, 0.05 mm in thickness) were incubated in the RBCs suspension (0.7 mL) for 4 h at 37 ± 2°C under gentle stirring. The mats were then removed, and the suspension of RBCs was subjected to centrifugation (1000 g × 5 min × 4°C). The absorbance of the supernatant was measured at λ = 577 nm and 620 nm as references. RBCs suspended in deionized water as a positive control and PBS as a negative control were included.

2.10. Cell Viability using Resazurin Reduction Assay

Human Foreskin Fibroblast HFF‐1 cells (ATCC, SCRC‐1041) were cultured at 37°C in a humidified atmosphere of 5% CO₂ in air, in standard tissue culture plates containing 5 mL Dulbecco's‐modified eagle medium (DMEM, Multicell, Wisent Inc., Canada) supplemented with 10% fetal bovine serum and 1% penicillin–streptomycin. Disk‐shaped fibrous mats with and without CNDs (≈35 µg and 5 mm diameter) were sterilized under UV for 10 min, and placed in a 96‐well plate (Thermo Fisher Scientific, USA). HFF‐1 cells (100 µL) with a density of 5 × 10⁵ cells/mL were seeded on the fibrous mats and co‐cultured for 24, 48, and 72 h. Media and cells only were included as controls. The resazurin reduction assay was used to determine the proliferation rate of cells. After the given incubation times, the mats were removed, and 10% resazurin solution (0.15 mg/mL, 20 µL) was added to each well and incubated at 37°C for 4 h. The resazurin solution was sterilized prior to use by filtration through a poly(ethene sulfone) membrane with a 0.2 mm pore size. Absorbance of the culture media was measured using a plate reader (BioTEK Synergy H1) at λ = 570 nm and 600 nm. The experiment was performed in 10 replicates (n = 10) for each condition. The obtained mean values were analyzed relative to those obtained by the control group.

2.11. Measurement of Antimicrobial Activity

Tests were performed according to guidelines published by the American Association of Textile Chemists and Colorists: AATCC‐TM100‐2019 [24]. This standard protocol was conducted against Escherichia coli (E. coli, ATCC 25922) and Staphylococcus aureus (S. aureus, ATCC 29213) at wound pH = 8.4 and physiological pH = 7.4. BE‐PVA mats with and without CNDs were used for the experiment. A stack of 10 mats (1 cm × 1 cm) was incubated in an overnight bacterial culture (400 µL) for 48 h. The mats were transferred to separate sterile flasks, each containing MHB (50 mL) whose pHs were adjusted to 7.4 and 8.4. The resultant mixtures were vigorously agitated and swirled for 30 s to release any bacteria that were adhered to the mats. To quantify bacterial survival, the bacterial suspensions were serially diluted in pH‐adjusted MHB at ratios of 1:10, 1:100, 1:1000, and 1:10 000. A 10 µL volume of each dilution was plated onto pH‐adjusted MHA. The plates were then incubated at 37°C for 18 h after which the bacterial colonies were counted. All samples were plated in triplicate to ensure reproducibility.

2.12. pH Sensing of Wet BE‐CND/PVA Mats

Pieces of BE‐CND/PVA (1 wt%) mats (≈12 mg) fabricated at BA/2OH = 2/1 were immersed in PBS solutions at pH = 4–12 for 1 h at room temperature. The mats were taken and exposed to UV light (λ_ex = 365 nm). Digital images were analyzed for their emission intensity using the Image‐J software.

3. Results and Discussion

3.1. Synthesis of CNDs and Fabrication of CND‐Embedded e‐Spun Nanofibers

Well‐defined CNDs were synthesized by a microwave‐assisted pyrolysis process from citric acid and ethylenediamine precursors in water (Figure S2a), as described elsewhere [25]. They were spherical with an average diameter of 1.4 ± 0.2 nm by TEM analysis (Figure S2b). They had adsorption at λ = 300–400 nm with the maximum wavelength at λ = 355 nm. In addition, they had emission in the blue region (thus, blue emission) with the maximum wavelength at λ = 450 nm when being excited at λ = 355 nm (Figure S2c).

An aqueous 10 wt% PVA solution was mixed with various amounts of CNDs at 0.5, 1.0, 2.0, and 5.0 wt% based on PVA. As seen in Figure 1a, the resultant CND/PVA solutions were homogeneous and transparent. They were then subjected to e‐spinning using our lab‐made setup under the optimized conditions, e.g., 25 G 1 ½ needle, flow rate of 1 mL/h, nozzle‐collector distance of 9 cm, 25 kV voltage at 35°C to fabricate CND/PVA e‐spun nanofibers on aluminum substrates. It can be expected that all CNDs in the solutions could be transferred to the corresponding e‐spun nanofibrous mats during our e‐spinning process.

FIGURE 1.

Digital images of (a) aqueous CND/PVA solutions and e‐spun mats, (b) SEM images, and (c) size distributions of e‐spun CND/PVA nanofibers embedded with various amounts of CNDs.

They were first characterized for their morphologies by SEM analysis. As seen in Figure 1b, all CND/PVA nanofibers appeared to be well defined, spaghetti‐shaped, and droplet‐free. Data analysis using ImageJ suggests that their average diameters ranged from 98 to 119 nm and did not appear to rely on the amount of CNDs. One plausible explanation could be due to the fact that the incorporation of CNDs up to 5% is too little to influence the entanglement of high molecular weight PVA chains in e‐spun nanofibers.

CND/PVA nanofibers were then characterized for their thermal properties as glass transition (T_g) and melting (T_m) temperatures, by DSC analysis. All DSC thermograms show a single glass transition (Figure S3). Their T_g increased from 71.2°C to 84.5°C with increasing amounts of CNDs in the fibers (Figure S4). Interestingly, their T_m remained unchanged regardless of the amount of CNDs (Figure S5), suggesting that the presence and amount of CNDs could not affect the crystallinity of linear PVA chains in the fibers. Further, the e‐spun CND/PVA nanofibers were characterized by FT‐IR spectroscopy. As compared in Figure S6, bare PVA and CND/PVA nanofibers show vibration modes at 3340, 2928, 1434, and 1091 cm⁻¹ attributed to O─H stretching, C─H stretching, C─H bending, and C─O stretching, respectively. Their vibration modes did not appear to be altered in the presence of CNDs up to 5 wt%.

3.2. Studies of BE‐Induced Crosslinking on e‐Spun PVA Nanofibers

E‐spun PVA fibers are required for crosslinking to retain their structural integrity in aqueous environments because of the high solubility of PVA chains. As illustrated in Figure 2a, BE chemistry was studied to fabricate CND/PVA nanofibers crosslinked with BE bonds (called BE‐CND/PVA nanofibers). DBA, a commercially available diboronic acid crosslinker, was examined in an organic solvent. Pieces of dried fibrous mats were immersed in DMF containing DBA. After 24 h, the mats were dried in a vacuum oven at 60°C to remove residual DMF.

FIGURE 2.

Schematic illustration of crosslinking CND/PVA mats through the formation of BE bonds in DMF for 48 h. Digital images show (a) the mats before and after being crosslinked; (b) SEM image, (c) diameter distribution, and (d) FT‐IR spectrum of BE‐PVA fibrous mat crosslinked with BA/2OH = 2/1.

The resulting mats appeared to be opaque, confirming that their nanofiber structure remained unchanged upon crosslinking. With our choice of CND/PVA mat (containing 1 wt% CND), along with bare PVA mat as a control, the amount of DBA was varied as mol equivalent ratio of BA/2OH ranging at 0.25/1 – 2.0/1. Note that BA/2OH implicates one BA group in DBA reacting with two OH groups in PVA chains to form a BE bond. To get an insight into the efficiency of crosslinking through the formation of BE bonds, gravimetric analysis was used to determine % efficiency of DBA incorporated into fibrous mats compared with that in the recipe (%DBA incorporation efficiency) and %OH groups reacted with DBA in fibrous mats (%OH reacted). As compared in Figure S7 a,b, %DBA incorporation efficiency increased from 1.0% to 5.3% for PVA/CND mat and 1.5 to 5.9% for bare PVA mats with an increasing amount of DBA (e.g., BA/2OH). Similarly, when the amount of DBA increased, %OH reacted increased from 0.3% to 10.6% for PVA/CND (1wt%) mat and 0.4% to 11.8% for bare PVA mats. Interestingly, no significant difference in crosslinking efficiency was observed for PVA/CND fibers, compared with bare PVA fibers.

Particularly, CND/PVA (1 wt%) mats crosslinked with BA/2OH = 2/1 were characterized by SEM and FT‐IR techniques. As seen in Figure 2b,c, SEM image shows well‐defined, droplet‐free, spaghetti‐shaped fibers with an average diameter to be 178.1 ± 19.6 nm upon crosslinking with BA/2OH = 2.0/1, which is larger than the corresponding uncrosslinked ones (113.2 ± 18.9 nm). FT‐IR spectrum in Figure 2d shows the appearance of vibrational modes at 1300 and 662 cm⁻¹ corresponding to B‐O‐C bending and O‐B‐O stretching frequencies, respectively [26], confirming the formation of new BE crosslinks. These results confirm the fabrication of BE‐CND/PVA mats.

To get an insight into the possibility of losing CNDs during the crosslinking process, DMF solutions were analyzed for UV/Vis spectroscopy. No noticeable absorbance corresponding CNDs was observed, suggesting that the significant amounts of CNDs are retained in BE‐CND/PVA mats.

3.3. Stimuli‐Responsive Degradation of BE‐CND/PVA Nanofibers

As illustrated in Figure 3a, BE‐CND/PVA mats could degrade in acidic and alkaline pHs as well as in the presence of hydrogen peroxide through the cleavage of BE bonds. With the choice of BE‐PVA/CND (1 wt%) mats, their stimuli‐responsive degradation was investigated in aqueous solution, with %degradation determined by gravimetric analysis based on mass loss as well as change in morphologies of fibers by SEM analysis.

FIGURE 3.

Schematic illustration of degradation of BE‐CND/PVA in (a) the presence of acidic and alkaline pH and hydrogen peroxide, (b,e) %degradation, (c,f) digital images, and (d,g) SEM images of BE‐CND/PVA fibrous mats after being incubated in (b–d) pH = 5.4 (acidic), 7.4 (neutral), and 8.4 (alkaline) for 48 h and (e–g) in response to hydrogen peroxide for 72 h.

For pH response, pieces were incubated in buffer solutions at pH = 5.4 (acidic) and 8.4 (alkaline), along with pH = 7.4 (neutral) as a control. As seen in Figure 3b, %degradation was 9.4% at pH = 5.4 and 5.8% at pH = 8.4, which appeared to be greater than that (1.7%) at pH = 7.4. The plausible reason could be due to the hydrolysis of BE bonds under alkaline and acidic conditions. Digital and SEM analyses confirm the significant loss of well‐defined fibrous forms upon degradation at pH = 5.4 and 8.4, while the preservation of more fibrous forms at pH = 7.4 (Figure 3c,d). For the hydrogen peroxide response, fiber mats were incubated in 1.0 mm aqueous hydrogen peroxide solution. %Degradation was determined to be 7.6%, which could be attributed to the cleavage of BE bonds (Figure 3e). Digital and SEM analyses also exhibit the significant loss of fibrous forms upon degradation (Figure 3f,g).

3.4. pH‐Responsive Release of CNDs from BE‐CND/PVA Mats

Emission spectroscopy was utilized to investigate the release kinetics of CNDs from BE‐CND/PVA mats upon degradation in pH = 5.4 (acidic) and pH = 8.4 (alkaline), compared with pH = 7.4 (neutral). From emission spectra of CNDs (Figure S8), their correlation curves were constructed with emission intensity at λ = 450 nm over concentration of CNDs at different pHs (Figure S9). Then, the pieces of BE‐CND/PVA mats were incubated at the pHs and the emission spectra of supernatants were recorded for a given time interval (Figure S10). Given the correlation curves, along with the amounts of CNDs embedded in the fibers, their emission intensity at λ = 450 nm was followed to calculate %release of CNDs. As seen in Figure 4a, %release increased within 24 h and gradually increased to the range of 65%–75% at all the pHs. It appeared to be greater at pH = 8.4 than pH = 5.4 and further pH = 7.4.

FIGURE 4.

%Release of CNDs from BE‐CND/PVA (1 wt%) fibrous mats incubated in aqueous buffer solutions at pH = 5.4 (acidic) and 8.4 (alkaline), (a) compared with pH = 7.4 (neutral) as a control; viability of HFF‐1 cells incubated with BE‐CND/PVA fiber mats containing 1 and 5 wt% CNDs for 24, 48, and 72 h, (b) determined by resazurin reduction assay; and (c) growth of E. coli ATCC 25922 and (d) S. aureus ATCC 29213 incubated with BE‐CND/PVA mats, compared with BE‐PVA mats as a control, at wound pH = 8.4 and physiological pH = 7.4, determined by AATCC‐100 assay.

The pH responsiveness of the fiber mat system is particularly advantageous for wound dressing applications, where dynamic pH variations occur during the healing process. Under normal physiological conditions, skin maintains a slightly acidic pH (≈5.5); however, upon injury, the wound environment becomes more alkaline (pH = 7–9) due to bacterial colonization, inflammation, and exudate accumulation. This alkaline shift can impair wound healing by disrupting enzymatic activity, promoting bacterial colonization, and inhibiting tissue regeneration [27]. The observed pH‐responsive release of antibacterial CNDs aligns well with this shift, ensuring that the therapeutic CNDs are released in response to the wound state. This targeted release mechanism precisely enhances the antimicrobial efficacy when needed, potentially reducing bacterial proliferation and promoting a favorable healing environment. This highlights the potential of the developed system as a stimuli‐responsive wound dressing capable of delivering antibacterial agents in a controlled and pH‐dependent manner.

3.5. Studies of Hemocompatibility and Cytocompatibility

BE‐CND/PVA nanofibrous mats were evaluated for the hemocompatibility with and without 1 and 5 wt% CNDs with Sprague Dawley rat RBCs using an in vitro hemolysis assay. %Hemolysis was determined based on the amount of hemoglobin. As summarized in Table S1, %hemolysis was as low as 0.7%–1.4% for BE‐CND/PVA and CND‐free BE‐PVA mats, suggesting that both mats are non‐hemolytic.

Further, they were evaluated for biocompatibility as viability with Human Foreskin Fibroblast HFF‐1 cells up to 72 h, using an Alamar Blue assay. As seen in Figure 4b, the viability was greater than 90% for all mats with and without 1 and 5 wt% CNDs. Note that 5 wt% CNDs is equivalent to 2 µg/mL. This result is promising in that the mats are biocompatible and have great potential for biomedical applications. Regardless, the CNDs themselves used in our experiments were reported to be non‐toxic to both HFF‐1 and HeLa cells up to 10 000 µg/mL [25].

3.6. Antimicrobial Activity

Guidelines published by the AATCC TM100‐2019 assay, a widely recognized quantitative method to assess the antibacterial efficacy of fabrics, were used to evaluate the antibacterial efficacy of BE‐CND/PVA mats against E. coli (ATCC 25922) and S. aureus (ATCC 29213) in wound pH = 8.4 and physiological pH = 7.4. Figure 4c,d shows their bacterial growth as a measure of inhibition efficiency in log(CFU/mL), compared with BE‐PVA mats (no CNDs) as a control. As seen in Figure 4c, the growth of E. coli was not inhibited by either mat at pH = 7.4. Promisingly, at pH = 8.4, the growth of the bacteria was significantly inhibited with BE‐CND/PVA. Such inhibition could be attributed to the enhanced release of embedded CNDs due to pH‐sensitive cleavage of BE crosslinks in alkaline pH. The BE‐PVA mats also inhibited the growth of E. coli at pH = 8.4, though not to the same extent. This could be due to the intrinsic antimicrobial property of boronic esters and boronic acids against bacteria [28]. In contrast, the growth of S. aureus was not inhibited by either mat at pH = 7.4 and minimal inhibition was observed for BE‐CND/PVA mats at pH = 8.4 (Figure 4d).

Furthermore, different inhibition efficiencies were observed for the two bacterial strains. Such observed strain‐dependent antibacterial activity could be attributed to differences in bacterial cell wall structures and electrostatic interactions between the CNDs and bacterial membranes. E. coli, a Gram‐negative bacterium, possesses a thinner peptidoglycan layer surrounded by an outer membrane composed of lipopolysaccharides and hence is negatively charged due to the presence of phosphate and carboxyl groups [29, 30]. This structure allows easier penetration of released CNDs through strong electrostatic attraction between positively charged CNDs and negatively charged bacterial membranes [29]. This interaction promotes adhesion, disrupts membrane integrity, leading to bacterial cell death. In contrast, S. aureus, a Gram‐positive bacterium, has a thicker peptidoglycan layer, which provides greater structural rigidity and resistance to nanoparticle penetration, thereby reducing the effectiveness of CNDs [31].

Overall, these findings demonstrate that pH‐responsive CND release plays a key role in antibacterial activity, with greater inhibition observed at alkaline pH due to enhanced degradation. Additionally, the differential inhibition between Gram‐negative and Gram‐positive bacteria highlights the importance of bacterial cell wall structure in determining susceptibility to CND‐based antimicrobial systems. These results further support the potential of BE‐CND/PVA nanofibers as stimuli‐responsive antibacterial wound dressings, where the controlled release of CNDs in alkaline wound environments could selectively inhibit bacterial proliferation.

3.7. Biological Sensing of Wound pH Changes

The potential of BE‐CND/PVA mats was investigated for sensing pH changes in wounds using fluorescence imaging, as described elsewhere [32]. Pieces of fiber mats were incubated in a broad range of pHs at 4–12 for 48 h. Digital images of wet mats were captured upon irradiation of UV light at λ_em = 365 nm (Figure 5a), and emission intensity was quantified using ImageJ software. As seen in Figure 5b, emission intensity linearly increased from pH = 4 to 6.5, followed by a gradual rise to pH = 12. This result indicates a greater dependence on emission intensity within the pH range of 4–6.5, suggesting its potential as a pH sensor in this pH range.

FIGURE 5.

Digital images of wet mats upon (a) the irradiation of UV light at λ = 365 nm and (b) emission intensity at λ_em = 450 nm at pH = 4–12.

4. Conclusion

Well‐defined, droplet‐free CND/PVA e‐spun nanofibers were fabricated via e‐spinning of aqueous PVA solution containing various amounts of CNDs up to 5 wt%. Their diameters remained independent of the concentrations of CNDs, while their glass transition increased with an increasing amount of CNDs inside the mats. In the presence of DBA, the fabricated CND/PVA mats were crosslinked, and importantly, their fibrous morphology was retained during the crosslinking process, confirmed by gravimetric, FT‐IR, SEM analysis. The amount of DBA turned out to be the important parameter that ensures control over the efficiency of BE‐induced crosslinking reaction. The fabricated BE‐CND/PVA mats degraded rapidly in both acidic and alkaline pHs as well as in the presence of hydrogen peroxide, which leads to enhanced release of CNDs in alkaline pH. Promisingly, they were non‐hemolytic to blood and non‐toxic to skin cells. They also suppressed bacterial growth, particularly against Gram‐negative E. coli at alkaline pH, suggesting potential benefits to wound healing. These findings underscore the potential of BE‐CND/PVA nanofibrous mats for advanced dual‐functional wound dressings, offering both antibacterial activity and real‐time pH‐responsive detection.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Supporting File 1: mabi70068‐sup‐0001‐SuppMat.

Lokuge N. D., Casillas‐Popova S. N., Singh P., et al. “Luminescent Electro‐Spun Nanofibers Crosslinked with Boronic Esters Exhibiting Controlled Release of Carbon Dots for Detection of Wound pHs and Enhanced Antimicrobial.” Macromolecular Bioscience 25, no. 12 (2025): e00352. 10.1002/mabi.202500352

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.