Thermoresponsive antimicrobial wound dressings via simultaneous thiol-ene polymerization and in situ generation of silver nanoparticles

Abstract

Thiol-ene polymerization and a one electron transfer reaction were simultaneously utilized in the present study for the preparation of semi-IPNs composed of a thermoplastic polyurethane elastomer, crosslinked poly(N-isopropylacrylamide) and silver nanoparticles (AgNPs). Application of these materials as thermoresponsive and antibacterial wound dressings with proper mechanical properties, efficient handling of wound exudates and easy peeling from the wounded area were examined. The thermoresponsivity of the membranes was elucidated via differential scanning calorimetry and measuring water absorption at different temperatures. Ease of removal of the designed dressings from the wound bed was confirmed based on qualitative examination of adhered cells to the dressings at temperatures lower and higher than the lower critical solution temperature of the prepared membranes. The proper bulk hydrophilicity and water vapour transmission rate of designed dressings showed their ability for managing of wound exudates. The potential ability of prepared dressings for protection of the wound bed from external forces over the entire period of healing was confirmed by their excellent tensile properties even at a fully hydrated state. In situ generation and dispersion of AgNPs into the matrix of the dressings, as well as the size of these particles were elucidated by EDX and TEM methods. An MTT assay against human dermal fibroblast cells performed on dressings with and without AgNPs approved their appropriate cytocompatibility. And finally, the measured antimicrobial activity of the dressings against different Gram positive and Gram negative bacteria as well as a fungal strain showed promising efficiency of impregnated AgNPs for combating microorganisms.

---

Introduction

Wound dressings are an important section of the medical and pharmaceutical wound care market worldwide. Wound dressings are designed to facilitate the natural healing process. In fact, wound healing is a dynamic process consisting of several cellular and biochemical activities and overlapping phases including inflammation, new tissue formation, and remodeling.^1,2 Unfortunately, the human body cannot heal dermal injuries completely. Since skin forms a protective barrier around the body, damage to the dermis may cause several problems like severe dehydration as well as infection. Therefore, using wound dressings is mandatory for the promotion of rapid wound healing with the best functional and cosmetic results.³ For this purpose, wound dressings should have proper cytocompatibility, appropriate flexibility, gas permeability, durability, and more importantly the ability to manage exudates to prevent scab and scar formation on wounded area. Meanwhile, non-adherence to wound bed and prevention from bacterial/fungal infection of damaged tissues are two key features of versatile modern wound dressings.⁴

To prepare wound dressings having the most of aforementioned features, selection of materials used for the preparation of dressings and mode of their fabrication have prime importance. Polyurethane framework was selected in the present study as the basic ingredient of target dressings. The primary reason for this selection is long and established history for application of polyurethane for the preparation of dressing membranes. This interest to polyurethane is attributed to excellent physico-mechanical and biological properties of polyurethanes.^4–6

During healing process the volume of wound exudates reduces gradually, therefore, the tendency for adherence of dressing to wound increases, consequently, removal of dressing from wound bed is painful and may cause secondary injury.^7,8 One accepted methodology for imparting cell adhesion control and easy peeling characteristic in wound dressings is utilization of thermoresponsive materials. Poly(N-isopropylacrylamide) (PNIPAAm) is the material commonly used for the preparation of thermoresponsive wound dressing. PNIPAAm exhibits a sharp lower critical-solution temperature (LCST) at about 32 °C. At temperatures lower than LCST, this polymer is highly hydrophilic, however, it possesses hydrophobic nature at temperature above than LSCT.^9,10 When dressings containing optimized amount of PNIPAAm apply on wound bed, they are at body temperature, which is higher than LCST. Just before removal by external cooling to temperature lower than LCST, these dressings transform to fully hydrophilic state with the least tendency to adhere to damaged tissue.¹¹ Unfortunately, such systems containing PNIPAAm moieties suffer from the disadvantage of poor flexibility in both dry and hydrated states. Therefore, several strategies were followed to maintain desired level of flexibility for these materials. Preparation of thermoresponsive semi-interpenetrating polymer networks (semi-IPNs) composed of crosslinked PNIPAAm and linear polyurethane via thermal initiated free radical polymerization is one of the best methods followed for solving this problem.^12,13

Preserving the moist environment over wounded area is known as the best condition for self healing of wounds. However, this condition can simultaneously increase the chance of pathogenic microorganism proliferation and colonization. Infection of wound bed can prolong inflammatory phase of healing and postpone the natural recovery of damaged tissue. To reduce the likelihood of infections, antimicrobial property is becoming an indispensable quality for a medically favoured dressing nowadays.¹⁴ In this regard, different types of silver nanoparticles (AgNPs) were prepared and examined for production of highly versatile antimicrobial wound dressings.¹⁵ It was reported that utilization of such disinfectant in dressing membranes not only decreased both local and systemic inflammation, but also led to faster healing.^16,17 Therefore, AgNPs based disinfectant was used in this study. Unfortunately, adding preformed AgNPs to dressing membranes is not an easy task and many problems regarding improper loading and dispersion of such nanoparticles may be faced. Finding an easy procedure to overcome this problem was another concern of the present study. To this end, interesting method for in situ generation of AgNPs from impregnated silver salts has been utilized.^18,19

To prepare thermoresponsive, antimicrobial semi-IPN dressing membranes consisting AgNPs, the simultaneous thiol-ene type click polymerization and single electron transfer reaction were performed on mixtures consisting a linear thermoplastic polyurethane elastomer, NIPAAm and different amounts of silver salt. All of the prepared materials were fully characterized by spectroscopic methods. Preventing adhesion and controlling of infection were two key elements of designed dressings that fully rationalized based on standard methods. Physical and mechanical properties as well as cytocompatibility of these materials were also evaluated to find better perspective about their practical applicability as wound dressings. To the best of our knowledge there is no similar study for the preparation of semi-IPN type wound dressings with combination of antimicrobial activity and thermoresponsivity through application of thiol-ene click polymerization reaction.

Experimental

Materials

Thermoplastic polyurethane elastomer (PU, Desmopan 5377 A) was obtained from Bayer Company. TPU was freed from moisture by heating at 90 °C under vacuum. N,N-Dimethylacetamide (DMAc), N-isopropyl acrylamide (NIPAAm), N,N′-methylenebisacrylamide (MBA), pentaerythritol tetra (3-mercaptopropionate) (PETMP) were supplied from Aldrich and used as received. Azobisisobutyronitrile (AIBN) was purchased from Aldrich and purified by recrystallization from ethanol and dried at room temperature under vacuum prior to use. Silver nitrate was obtained from Merck and purified by recrystallization from water. Human Dermal Fibroblasts (HDF-C646) were supplied by Pasteur Institute of Iran. S. aureus (ATCC 6538), and P. aeruginosa (ATCC 15449) bacteria and C. albicans (ATCC 10231) were purchased from Iranian Research Organization for Science and Technology (IROST).

Preparation of thermoresponsive dressing membranes

In a 50 ml beaker equipped with magnetic stirrer and heating mantel was placed TPU and DMAc. The content was stirred at 40 °C until all granules were dissolved. Then, other ingredients including NIPAAm, MBA, PETMP, and AIBN were added and after complete dissolution the total volume of the beaker was adjusted to 20 ml by adding more DMAc. The content of flask was transferred to a Teflon mould and placed in to a vacuum oven. The oven temperature was increased to 90 °C and polymerization reaction was continued under controlled vacuum for 12 h. The prepared membranes were soaked in distilled water for 24 h to remove any residual solvent and unreacted monomers. After complete drying at 60 °C under vacuum, the membranes were used for further characterization.

For preparation of membranes containing silver nanoparticles, appropriate amount of AgNO₃ was added to beaker along with other ingredient. Then, the resulting solution was subjected to ultrasound irradiation using probe ke 76 for 5 min. Then the reaction was continued according to the synthetic procedure described above.

The compositions of all prepared membranes were tabulated in Table 1.

Table 1 Feed compositions and gel contents of prepared dressing membranes^a

Sample	TPU (g)	NIPAAm (g)	MBA (g)	AIBN (g)	PETMP (g)	Silver nitrate (g)	Gel content (%)
a According to analysis of variances p-values of <0.05 were considered significant. The difference between quantities with similar superscripts is not significant (p ≥ 0.05).b Due to improper mechanical property this sample was not evaluated further.
PU	1.000	—	—	—	—	—	—
TRPU-25	1.000	0.333	0.067	0.008	0.463	—	98.7 ± 0.2^a
TRPU-50	1.000	1.000	0.200	0.024	1.390	—	98.7 ± 0.8^a
TRPU-75^b	0.333	1.000	0.200	0.020	1.390	—	—
TRPU-25/Ag-0.5	1.000	0.330	0.067	0.008	0.463	0.009	97.1 ± 1.0^a
TRPU-50/Ag-0.5	1.000	1.000	0.200	0.024	1.390	0.018	96.8 ± 0.9^b
TRPU-50/Ag-2.5	1.000	1.000	0.200	0.024	1.390	0.09	96.5 ± 1.0^c

---

Evaluation of wound exudates management and oxygen permeability of dressings

The moisture and oxygen permeability of the membranes were determined according to the procedure described in Section S1.†

Thermoresponsivity study

The thermoresponsive behaviour of samples were studied by using differential scanning calorimetry. To have a better insight about thermoresponsive behaviour of the membranes, they were examined through determination of their surface and bulk hydrophilicity. The related procedures were described in Section S2.†

Quantification of the amount of stratum corneum removed

The procedure followed for evaluation of adhered stratum corneum to dressing membranes were fully described in Section S3.†

In vitro cytocompatibility assays

Biocompatibility of prepared membranes was evaluated against human dermal fibroblast according to the procedure described in Section S4.†

Cell attachment and detachment

The possible attachment and subsequent detachment of human dermal fibroblasts to dressings via decreasing temperature of culture medium were studied according to the procedure described in Section S5.†

Antibacterial study

Antimicrobial activity of prepared membranes was evaluated based on “colony forming count” method as described in Section S6.†

Measurements

ATR-FTIR spectra were recorded on Bruker Instrument (model Aquinox 55, Germany) in the range of 4000–400 cm⁻¹ at a resolution of 4 cm⁻¹ and signal averaged over 8 scans. Dynamic mechanical thermal analysis (DMTA) was carried out on a Triton instrument (model Tritec 2000, England) at temperature range of −80 to 100 °C, heating rate of 5 °C min⁻¹ and frequency of 1 Hz. The values of storage modulus, loss modulus, and tan δ versus temperature were recorded for each sample. Mechanical properties including tensile strength, modulus, and elongation at break were determined from stress–strain curves with MTS tensile tester model 10/M at a strain rate of 10 mm min⁻¹. The measurements were performed at 25 °C with a film thickness of about 0.5 mm and stamped out with an ASTM D638 Die. Reported data were an average of at least 3 measurements.

The gel content of prepared membranes was evaluated. For this purpose the membranes were dried under vacuum for 24 h at room temperature and weighed. Then, the samples were immersed in deionized water for 24 h. Insoluble sample was dried at 50 °C and weighed. The gel content was defined as follows:

where W_d and W_i designated the weight of dried membrane after immersing in water and the initial weight of the membrane, respectively.

Silver atoms content and their distribution map were measured by EDX analysis using a scanning electron microscope (SEM, Tescan, Vega II, Czech) equipped with an energy dispersive X-ray analyzer system (EDX, Oxford Instrument, INCA, England). The amount of elemental silver in the polymer structure was measured with an X-ray fluorescence spectrometer (Philips X'unique II, Holland) operating X-ray source voltage 30 kV and current 30 mA with 1.0 ppm resolution.

The presence of silver nanoparticles in dressing membranes were also investigated by transmission electron microscopy (TEM) technique using a Zeiss-EM10C microscope operating at an accelerating voltage of 80 kV. Ultra-thin sections (about 50 nm thick) were prepared by using a cryo-ultramicrotome (Reichert OMU3 Vienna, Austria), equipped with a diamond knife keeping the sample at −100 °C.

Statistical analysis

Statistical analyses were performed via PASW Statistics program package, version 18 (SPSS Inc., Chicago, IL, USA). Comparison of obtained data for different samples was performed with One-Way ANOVA with Tukey posthoc test. The significance level was set at p < 0.05.

Results and discussions

Synthesis and spectroscopic characterization of dressing membranes

Thiol-ene reaction represents one of the best newly developed systems for the polymerization of unsaturated monomers with attractive click characteristics.²⁰ Utilization of this reaction for the preparation of desired dressing membranes was considered. The sequences of reactions led to the formation of final products are outlined in Scheme 1. The radicals generated through thermal decomposition of AIBN acted as a source of donor species which enabled rapid reduction of impregnated Ag⁺ ions to metallic silver (Ag⁰). Simultaneously, the thiol radicals formed from PETMP handled the thiol-ene step-growth polymerization reaction led to formation of crosslinked network from unsaturated monomers.^21,22

Scheme 1 The schematic rout of thiol-ene network preparation.

The feed composition of the prepared membranes and their gel fractions are collected in Table 1. The high value of gel fraction (more than 98%) for the prepared samples confirmed the high efficiency of the thiol-ene reaction.²³

FTIR spectroscopy was used for characterization of the prepared membranes (Fig. 1). For better discrimination of peaks related to the thiol-ene moieties, the FTIR spectrum of PU component was also recorded. According to FTIR spectrum of pure PU sample (Fig. 1a), peaks appeared at 1703 and 1730 cm⁻¹ were attributed to the stretching vibration of C=O bonds of ester and urethane groups. The peak at 3300 cm⁻¹ was related to urethane N–H stretching vibration. The combination of N–H out of plane bending and C–N stretching vibrations was detected at 1534 cm⁻¹. After performing thiol-ene polymerization reaction and introducing thermoresponsive phase composed of NIPAAm and MBA, the characteristic peaks of PU segment were preserved and new peaks were also detected (Fig. 1b). The peak observed at 1645 cm⁻¹ was related to stretching vibration of C=O bond of amide groups. The broad peak centred at 3274 cm⁻¹ was attributed to combination of stretching vibration of urethane N–H bonds and new amide N–H groups originate from NIPAAm and MBA moieties.²⁴ The peaks at 1456 and 1385 cm⁻¹ were related to the bending vibration of CH₂ and CH₃ groups of NIPAAm and MBA, respectively. Also, the peaks related to stretching vibration of aliphatic C–H bonds in the range of 2825–2939 cm⁻¹ were broadened after incorporation of crosslinked thermoresponsive segment, which contains new aliphatic C–H groups.

Fig. 1 FTIR spectra of (a) PU, (b) TRPU-50 and, (c) TRPU-50/Ag-2.5.

The dressing membranes containing AgNPs were simply prepared through simultaneous polymerization and one-electron transfer reaction to impregnated silver ions.

During polymerization of systems containing silver ion, intense orange to brown colour was developed (Fig. 2). This phenomenon was the first evidence for the production of metallic silver nanoparticles. Intensity of colour was related to the concentration of silver ions. Darker colour was obtained for samples containing higher concentration of silver ions. This phenomenon was related to surface plasmon resonance effect, caused by a collective excitation of the free electrons in the silver nanoparticles.

Fig. 2 Photographs of (a) TRPU-50, (b) TRPU-50/Ag-0.5, (c) TRPU-50/Ag-2.5, and (d) TRPU without thiol-ene reaction.

Despite a partial decrease in measured gel contents for these samples, the network formation was conducted well in this case. Consumption of free radicals generated in this system by silver ion was responsible for this happening. However, high functional group conversion and appropriate gel content of final networks was indebted to high efficacy of thiol-ene reaction. It is worth to mention that the same phenomenon was detected by Phillips et al. during thiol-ene polymerization/photo reduction of systems containing gold ions.²⁵ Samples containing silver nanoparticles were also subjected to the FTIR study.

A representative example is shown in Fig. 1c. No considerable difference was detected and characteristic peaks of PU and thiol-ene segments were recognized for these samples.

Characterization of silver nano particles

The silver distribution map within polymeric matrix was acquired by EDX technique. Fig. 3 shows the standard EDX spectrum recorded on the examined sample. According to presented spectrum, the peaks located between 2 kV and 4 kV were related to the silver characteristic lines K and L that confirms the presence of silver within the membrane.²⁶ The rest of the lines of the EDX spectrum correspond to other elements present in the membrane (mainly S, O, and C). In addition, the silver atomic map recorded by EDX showed a uniform distribution of silver metal in membrane matrix owing to in situ generation and immediate distribution of nanoparticles upon their formation with least possible agglomeration.²¹

Fig. 3 (a) Elemental analysis, (b) map of silver atoms on the surface, and (c) SEM as well as distribution of silver atoms on cross-section of TRPU-50/Ag-2.5.

The EDX analysis together with the XRF results makes possible to confirm the reduction of silver ions to elemental silver inside the composite membranes. Amounts of measured Ag⁰ for TRPU-50/Ag-2.5 and TRPU-50/Ag-0.5 membranes detected by XRF are collected in Table 2. The data of this table confirmed the presence of Ag⁰ within the network.

Table 2 Amount of silver detected by XRF

Sample	Metallic silver (%w)
TRPU-50/Ag-0.5	0.02
TRPU-50/Ag-2.5	0.05

---

To find a better insight regarding size and size distribution of the silver particles, TEM micrograph was recorded for the representative sample (Fig. 4a). The presence of well dispersed AgNPs with average size around 30–50 nm was clearly detected (Fig. 4b). The narrow size distribution and highly dispersion of AgNPs was attributed to the fact that silver ions were loaded before polymerization and AgNPs formed during formation of crosslinked network.²⁷

Fig. 4 (a) TEM micrographs at different magnification of thiol-ene semi-IPN system in the presence of 2.5 wt% of silver precursor and (b) distribution of AgNPs size calculated from the TEM images using Image J software.

Study of thermoresponsive behaviour

To have an insight regarding the LCST of the membranes, the phase transition of hydrated TRPU-50/Ag-2.5 as representative sample was analyzed by DSC considering water as reference (Fig. 5a). Thermal behaviour of this sample at dry state was also examined (Fig. 5b). No transition was detected in temperature range of 25–45 °C in dry state. However, as shown in Fig. 7a, a transition was detected centred at 32.5 °C which was close to the LCST of pure PNIPAAm hydrogel reported by Zhang et al.²⁸

Fig. 5 DSC heating thermograms of TRPU-50/Ag-2.5 in the (a) hydrated and (b) dry state.

The endothermic peak observed in DSC thermogram was indicated possible shrinking of the polymeric chains with heating. Therefore, the LCST was the point at which the hydrophobic interactions of the isopropyl pendant groups of PNIPAAm compensated the hydrophilic nature of the amide pendant groups. This is the reason for network structure collapsing and corresponding expelling of water from structure.^29,30

It is well-known that thermoresponsive polymers have different bulk and surface hydrophilicity at temperatures lower and higher than LCST. Thus, the surface and bulk hydrophilicity of the prepared membranes were measured by evaluation of their water contact angle and the amounts of absorbed water at different temperatures below and above LCST. In Fig. 6a, the temperature dependency of equilibrium swelling of samples is shown. Due to an abrupt conformational rearrangement, all samples showed temperature-sensitive swelling behaviour around 32 °C, which is within physiologically important interval of human body. Close inspection of equilibrium swelling data for prepared samples revealed less significant thermoresponsive behaviour for samples containing lower wt% of PNIPAAm moieties (TRPU-25 series) than those had higher wt% of PNIPAAm moieties (TRPU-50 series).³¹ Furthermore, based on Fig. 6a, the samples containing in situ generated AgNPs revealed higher swelling capacity than those without AgNPs. This phenomenon can be related to lower crosslink density of networks obtained in the presence of silver ions. This issue will be further discussed.

Fig. 6 Thermoresponsivity behaviour evaluated by (a) bulk and (b) surface hydrophilicity versus different temperatures (according to analysis of variances the difference between quantities with similar superscripts is not significant (p > 0.05)).

The contact angle of water droplets on prepared samples as a means of surface hydrophilicity was also measured at two temperatures below and above LCST (25 and 37 °C) (Fig. 6b). The temperature dependency of surface wettability for all samples clearly detected. The lower contact angle at 25 °C comparing to 37 °C was strong indication for decreasing of surface hydrophilicity. This phenomenon was mainly due to intra and inter molecular hydrogen bonding between polymer and water molecules that was disrupted upon heating.³² The effect of silver particles on the surface wettability of membranes was also evaluated and no significant difference was found in their contact angle values with those without AgNPs.

It is well-known that surface hydrophilicity of biomaterials has determining effect on proteins adhesion and cells interaction. Usually, the contact angle of about 60 to 70° is an optimum range for cell adhesion and proliferation.³³ Fortunately, average water contact angle measured for samples with higher wt% of PNIPAAm was in the range of 60 to 70° at 37 °C, but at lower temperature (25 °C) this value shifted to lower range (40 to 50°). Therefore, these samples have proper wettability for cell supporting at body temperature. High wettability at lower temperature is beneficial for easy removal of dressing. Therefore, based on bulk and surface hydrophilicity data, samples with higher content of PNIPAAm are appropriate absorptive materials for wound drainage and they can prevent adherence at the wound site. As well, easier peeling of dressing and consequently decreasing the secondary injuries and promotion of healing process is expected for these samples.

To find a practical evidence for this claim, the amounts of stratum corneum removed from the surface of skin by dressing membranes at two different temperatures (below and above LCST) were examined. Fig. 7a–d shows the stained and dichotomized images of the stratum corneum printed at two different temperatures on the adhesive tape from wound dressings (TRPU-50/Ag-0.5 and TRPU-50/Ag-2.5 as representative examples). Both the stained and dichotomized images were almost the same; accordingly, assessment of the areas of the stratum corneum removed from healthy skin was correct. Fig. 7e shows the desquamated area (%) of the stratum corneum for each wound dressings. There was a significant difference in the percent of desquamated area at two studied temperatures (p < 0.05), with lower percentage of desquamated area removed from skin after cooling procedure. This result showed appropriate function of the prepared dressings for easy peeling, therefore, a secure condition for newly formed epithelium and facilitation of wound healing was guaranteed.³⁴

Fig. 7 Stained images of the stratum corneum printed on the adhesive tape from each wound dressing and the dichotomized images of TRPU-50/Ag-2.5 at (a) lower and (b) higher temperature than LCST and TRPU-5-/Ag-0.5 at (c) lower and (d) higher temperature than LCST. (e) Percentage of desquamated area of the stratum corneum for each wound dressing.

Viscoelastic and mechanical properties of dressings

DMTA study was performed for the evaluation of viscoelastic behaviour of dressing membranes containing different wt% of silver and PNIPAAm. For comparison, the same study was extended to pure PU elastomer before fitting into the thiol-ene generated PNIPAAm network.

The variations in storage modulus (E′), loss modulus (E′′), and loss tangent (tan δ) as a function of temperature for all studied samples are collected in Fig. 8. Neat PU elastomer showed a single phase structure in studied temperature range since just one peak was observed at about −18 °C in tan δ and loss modulus curves; meanwhile a decrease in storage modulus was detected. This transition was associated to glass transition (T_g) of soft segment of PU composed of polyols used for its synthesis. However, a two-phase structure was detected for all thermoresponsive dressings, since two α-transitions were distinguished in their DMTA curves. The first transition at lower temperature was related to T_g of PU soft segment (T_g,pu) and that appeared at higher temperature was attributed to T_g of crosslinked network obtained via thiol-ene polymerization of NIPAAm, MBA, and PETMP (T_g,tr). Upon introduction of PNIPAAm network, the T_g,pu was shifted to higher temperature due to more restriction for segmental mobility of PU arose from intermolecular interaction of PU and PNIPAAm segments. Comparison of T_g values for two different series of dressings containing different wt% of PNIPAAm moieties, showed a reverse relationship between PNIPAAm content and T_g of both segments. For TRPU-50 series lower T_g,pu and T_g,tr were detected in comparison to TRPU-25 series. This phenomenon can be attributed to the presence of flexible thioether bonds produced during thiol-ene polymerization reaction.³⁵ For system containing higher NIPAAm content (TRPU-50 series), higher corresponding concentration of PETMP was used in initial formulations that led to formation of more flexible thioether bonds in final network. Interestingly, lower modulus at rubbery plateau region and higher maximum height of tan δ curve were detected for these samples as a result of formation of networks containing more flexible thioether bonds (Fig. 8a). It is worth to mention to tremendous advantage of thiol-ene system for formation of nearly ideal, uniform polymeric networks in comparison to the common networks produced from classical free radical polymerization. In contrast to conventional radical polymerization reaction, thiol-ene system can lead to formation of networks with lower heterogeneity.³⁶ It was remarkable in Fig. 8c that the width of T_g,tr peak was narrow and entire glass transitions of PNIPAAm segments occurred in narrow interval. The homogeneity of network produced in this system may be one of the reasons for desirable mechanical behaviour of these samples which is beneficial for wound dressing application.

Fig. 8 (a) Storage modulus, (b) loss modulus, and (c) tan δ versus temperature curves for different membranes.

The effect of in situ generated AgNPs on thermo-mechanical property of produced networks was also studied. It can be seen from the Fig. 8c that for samples containing 0.5 wt% of AgNPs, the T_g,pu peak was shifted to higher temperature. The same phenomenon was reported in literature^37,38 upon introduction of inorganic particles like Ag. The increased hindrance for motion of polymeric chains was responsible for the increment of T_g. The T_g,pu peak was almost disappeared for samples containing 2.5 wt% of AgNPs. This phenomenon was also attributed to restriction of segmental motion as a result of high interaction with AgNPs. Also, upon introduction of AgNPs, the T_g,tr value decreased and the maximum height of corresponding tan δ peak becomes higher.^25,39 The reason suggested for this happening was retardation effect of silver ions on polymerization reaction. In fact, the silver ions competed with NIPAAm and MBA monomers for reaction with free radicals generated from initiator. Eventually, this phenomenon led to networks with lower crosslink density. The looser networks with higher degree of freedom of chain segments could interact more efficiently with penetrated water molecules and higher water absorption was resulted for these samples as mentioned at Fig. 6a.

Also, it is worthwhile to mention that due to the extension of rubbery region of the prepared materials to 100 °C (Fig. 8a) they can withstand the thermal condition needed for autoclave sterilization procedure.

Ability of preserving wounds from external pressure and mechanical shocks are the essential characteristics of ideal wound dressings. Therefore, wound dressings need to be elastic and unlikely to rip. As well, maintaining flexibility and strength during healing period is a crucial property for ideal dressing membranes.⁴ To find insight regarding these important issues, tensile properties (tensile strength, modulus, and elongation at beak) of dressings were evaluated from their stress–strain curves recorded on both dry and fully hydrated samples (Fig. 9a and b and Table 3).

Fig. 9 Stress–strain curves for the prepared membranes in the (a) dry and (b) hydrated states.

Table 3 Tensile properties of semi-IPN membranes^a

Sample	Modulus (MPa)		Tensile strength (MPa)		Elongation at break (%)
	Dry	Wet	Dry	Wet	Dry	Wet
a According to analysis of variances p-values of <0.05 were considered significant. The difference between quantities with similar superscripts is not significant (p ≥ 0.05) for data of each column.
PU	1.08 ± 0.11^a	1.06 ± 0.11^a	15.60 ± 1.24^a	12.26 ± 0.60^a	1300.79 ± 68.50^a	913.10 ± 56.58^a
TRPU-25	2.49 ± 0.37^b	2.12 ± 0.17^b	9.21 ± 1.16^b	7.49 ± 1.01^b	307.62 ± 21.70^b	308.82 ± 19.41^b
TRPU-50	1.30 ± 0.16^c	1.26 ± 0.13^c	14.14 ± 1.30^a	8.90 ± 0.28^b	940.56 ± 26.12^c	639.5 ± 21.70^c
TRPU-50/Ag-2.5	0.30 ± 0.08^d	0.22 ± 0.09^d	2.36 ± 1.41^c	1.50 ± 0.10^c	755.43 ± 32.4^d	600.81 ± 31.50^d

---

Membranes with different compositions showed different mechanical behaviour. The synthesized semi-IPNs were highly flexible, with elongation at beak from 300 to 950% in the dry state. Fortunately, very good flexibility of dressings (300 to 650% elongation at break) was preserved at fully hydrated state with maximum value for TRPU-50 sample. This property was related to special structural characteristics of PNIPAAm network prepared through thiol-ene polymerization reaction.

It is worth to mention that the samples with the same compositions prepared through classical radical polymerization procedure without adding PETMP component, led to formation of very brittle membranes which could not be folded at all (Fig. 2d).

Also the tensile strengths from 2 to 14 MPa at dry state and 1 to 9 MPa in hydrated state were recorded for the prepared samples, which were comparable to tensile strength of skin (2.5–16 MPa). It is obvious that the reduction of tensile strength of membranes in hydrated state is related to reduced intermolecular adhesion of chain segments due to plasticizing effect of water molecules.⁴⁰

Interestingly, the reduction of tensile strength of dressings at hydrated state was much lower than some of related systems made through classical radical polymerization.¹¹ 20–35% reduction in tensile strength of hydrated samples was recorded in our materials containing different wt% of PNIPAAm, however, up to 65% reduction was recorded for semi-IPNs samples obtained from classical radical polymerization procedure.¹¹ The values for modulus of dressings at dry and wet conditions were about 0.3 to 2.5 MPa and 0.2 to 2.5 MPa, respectively. Also, the semi-IPN membranes had higher modulus than parent polymer. In addition, the modulus for dressing membrane with 50% PNIPAAm was lower than the sample with 25% PNIPAAm. This phenomenon was related to more flexibility of TRPU-50 network. In fact, higher amount of flexible thio-ether bonds was formed during thiol-ene polymerization reaction as a result of higher concentration of NIPAAm and PETMP.^41,42 The findings of tensile properties were in accordance to DMTA results.

Evaluation of wound exudates management and oxygen permeability of dressings

An ideal dressing should control evaporative water loss at an optimal rate from the wound bed to prevent scab formation due to excessive dehydration and desiccation of wounds; therefore, a proper WVTR is required.^43,44 On the other hand, accumulation of exudates can cause wound maceration and damage surrounding tissues; therefore, the dressings should absorb the extra exudates and have absorptive property. The data of WVTR and EWA are collected in Table 4 and Fig. 6a, respectively. The parent PU elastomer showed the lowest values of WVTR and EWA which associated to hydrophobic behaviour of this material. However, upon introduction of PNIPAAm through thiol-ene reaction, both of these factors improved due to higher hydrophilicity of PNIPAAm arose from their amide linkages. WVTR values for the prepared membranes were in the range of 300–500 g per m² per day. Further, the amount of permeability and permeance were measured for the dressing membranes. Permeation of water vapour through the membrane consists two steps including adsorption of water vapour on the surface (solubility) and then diffusion of water through the membrane (diffusion).⁴⁵ According to Fig. 6, membranes with higher PNIPAAm content exhibited higher solubility and surface wettability. In addition, diffusion coefficient of water (D) through TRPU membranes was calculated based on Fick's law and data are reported in Table 4. D increased with increasing PNIPAAm content (PU < TRPU-25 < TRPU-50). It was perceptible that both solubility and diffusion factors were favour for the improvement of WVTR, P, and permeance of dressing membranes with higher content of PNIPAAm.

Table 4 WVTR, permeability, permeance, and diffusion coefficient data for semi-IPN membranes dressing

Sample	Water vapour transmission (WVTR) (g per day per m^2)	Water vapour permeability (P) (kg m^−1 s^−1 Pa) × 10^12	Water vapour permeance (kg m^−2 s^−1 Pa) × 10^9	Diffusion coefficient (D) (mm^2 h^−1) × 10^3	O2 penetration (ml Pa^−1 s^−1 cm^2) × 10^3
a Data based on (ref. 47).
PU	281.9 ± 16.5^a	0.31 ± 0.02	0.89 ± 0.05	47.7 ± 1.2^a	0.30 ± 0.02^a
TRPU-25	312.7 ± 18.3^b	0.34 ± 0.02	0.99 ± 0.05	49.3 ± 1.0^b	0.50 ± 0.05^b
TRPU-50	336.3 ± 43.2^b	0.37 ± 0.04	1.07 ± 0.13	52.1 ± 2.6^c	2.11 ± 0.21^c
TRPU-25/Ag-0.5	381.4 ± 28.5^c	0.42 ± 0.03	1.21 ± 0.09	53.1 ± 2.1^d	—
TRPU-50/Ag-0.5	428.0 ± 35.2^d	0.47 ± 0.04	1.36 ± 0.11	55.8 ± 1.8^d	—
TRPU-50/Ag-2.5	519.7 ± 30.5^e	0.58 ± 0.03	1.65 ± 0.09	60.4 ± 3.6^e	2.50 ± 0.14^d
Bioclusive^a	394.0 ± 12.0	0.44 ± 0.01	1.26 ± 0.04	—
IntraSite^a	354.0 ± 42.0	0.39 ± 0.03	1.13 ± 0.09	—
Comfeel^a	308.0 ± 16.0	0.32 ± 0.01	0.91 ± 0.02	—
Dry human skin^a	215	0.24	0.68	—
Wet human skin^a	350	0.39	1.12	—

---

It was noticed that samples with AgNPs showed higher ability for permeation of water vapour. This was related to lower crosslink density of their corresponding networks.

The average values of WVTR, P, and permeance of membranes developed in this study (Table 4) were comparable to the commercial products suitable for low to moderated exuding wounds.

In addition to water vapour transmission, continuous supply of oxygen to the tissue is vital for the healing process and resistance to infection. Oxygen is involved in many different processes like oxidative killing of bacteria, re-epithelialization, angiogenesis, and collagen synthesis that eventually lead to wound healing. Therefore, oxygen has been explored as a therapeutic modality to aid wound healing.⁴⁶ Therefore, oxygen permeation through membranes was evaluated (Table 4). From the collected data, it was obvious that, sample with highest PNIPAAm and AgNPs possessed better performance for transmitting oxygen. Again, this feature was attributed to the lower crosslink density of this sample.

Biocompatibility study

Wound dressings should not liberate any cytotoxic materials. As well, having proper cytocompatibility is indispensable property of ideal dressings. Evaluation of biocompatibility assay for dressings containing biocidal additives is more critical, since most of these compounds may show toxic effects on cells in wound bed. AgNPs with broad-spectrum antimicrobial activity was used in the present study as antibacterial agent. Recent studies showed potential inflammatory and cytotoxic effects of AgNPs, however, with proper tuning of AgNPs concentration, reaching to appropriate biocompatibility index was possible.^47,48

To find an insight regarding this issue, human dermal fibroblast cells were exposed to membranes and leachates extracted from them. MTT results for leachates extracted from dressings after 24 and 48 h into PBS solution were collected in Fig. 10a.

Fig. 10 Cell viability of human dermal fibroblast cells (a) in contact with leachates extracted from membranes in 24 h and 48 h and (b) cultured directly on the surface of the films for 48 h (according to analysis of variances the difference between quantities with similar superscripts is not significant (p > 0.05)).

Fortunately, acceptable viability of cells was recorded for all of the samples with and without embedded AgNPs. In addition, no significant difference was detected between these samples (p = 0.627). This observation was confirmed the proper concentration of in situ generated AgNPs. Therefore, it was concluded that the concentration of the released biocidal agent was lower than toxic level for dermal fibroblast cells. This observation indirectly confirmed high efficiency of thiol-ene reaction for generation of PNIPAAm network, since the leachates were free from toxic monomeric NIPAAm residues. Cells behaviour on direct contact with dressing materials was also studied via either evaluation of cells morphology in vicinity of samples or by measuring their viability by MTT assay (Fig. 10b and 11). The healthy condition of cells with spindle shape morphology spread all over the tissue culture medium confirmed the absence of released toxic materials from samples during 48 h of incubation period. Also, the MTT assay on cells cultured directly on samples revealed very good cytocompatibility (>70%).

Fig. 11 Optical microscopy images of human dermal fibroblast cells in direct contact with washed membranes after 48 h incubation: (a) TRPU-50, (b) TRPU-50/Ag-0.5, (c) TRPU-50/Ag-0.5, (d) TRPU-50/Ag-2.5.

Close inspection of MTT results showed a significant (p < 0.05) predictable decrease of cellular viability for samples containing AgNPs, however, the viability of cells on direct contact with these samples was still in acceptable range.

Cell attachment and detachment

Evaluation of biocompatibility assay showed proper behaviour of the prepared dressing against fibroblast cells. Therefore, during practical application of these dressings, migration and adhesion of cells from wounded area on the surface of these dressing membranes would be probable. Therefore, removal of these dressing may cause secondary injury on wound bed. The significant increase of bulk and surface hydrophilicity of these membranes upon lowering temperature from LCST is a good means for repelling of cells attached on the surface of dressings. To examine this possibility, dermal fibroblast cells were cultured on the surface of both TRPU-50/Ag-2.5 as representative example for thermoresponsive dressings and parent PU elastomer as non-thermoresponsive sample. The cells attached to sample surface at 37 °C and remained attached after cooling to 20 °C were stained and visualized by fluorescence microscopy. Results are collected in Fig. 12. No considerable difference was detected for number of attached cells to PU sample at both temperatures. However, significant reduction of adhered cells was recorded for TRPU-50/Ag-2.5 sample after cooling procedure. Therefore, lower secondary damage during removal of this dressing was expected.

Fig. 12 Fluorescent microscopy images of human dermal fibroblasts remained attached to the surface of (a) PU and (b) TRPU-50/Ag-2.5 at 37 °C and 20 °C.

Antibacterial study

The aim of the present study was to design a biocompatible and antimicrobial wound dressing that could be effective for a broad spectrum of microorganisms. Among different possibilities, AgNPs was chosen in the present work. Silver is a powerful broad spectrum antimicrobial agent that has been used to control infection of different wounds. Close inspection of antibacterial dressing materials available in the market shows popularity of using silver based dressing to prevent wound infections. Although different mechanisms were suggested for biocidal activity of silver based biocidal,^49–51 the most of recent studies have proved the size dependency of biocidal activity of AgNPs. More activity was observed for smaller particles.⁵² The method developed in the present study enabled us to produce AgNPs with proper distribution and size of about 30–50 nm. AgNPs of this size range are suitable for exhibiting promising antibacterial activity.⁵³

It has been shown that there is a difference in toxic concentration range of AgNPs for mammalian and microbial cells. This difference provides an excellent opportunity for tuning the concentration of AgNPs in wound environment. Therefore, having a practical ability for controlling the amounts of impregnated silver particles in the dressing membrane is an important issue. Generation of AgNPs through simultaneous thiol-ene polymerization and single electron transfer to silver salt was selected method utilized in the present work. The concentration of generated AgNPs could be controlled by initial concentration of silver salt.

As it can be seen in Fig. 13, the results of antimicrobial activities demonstrated complete killing ability of microorganisms by TRPU-50/Ag-2.5 membrane.

Fig. 13 Antibacterial activity of the prepared membranes against (a) P. aeruginosa, (b) S. aureus, and (c) C. albicans.

Conclusion

Thermoresponsive semi-IPN membranes containing fully dispersed nano sized silver particles (30–50 nm) were prepared successfully through thiol-ene radical initiated polymerization reaction. The resultant membranes exhibited thermoresponsivity in appropriate temperature range as determined by measuring the water absorption at different temperatures and through DSC data. This property of prepared dressings made possible their easy removal from wound bed. Protection of wounds from external forces made also possible due to appropriate tensile properties and flexibility of designed dressings.

Evaluation of biological activity of prepared dressings by studying interaction of these materials against human dermal fibroblasts and three different microbial strains proved excellent cytocompatibility for fibroblast cells and efficient antimicrobial activity against S. aureus, P. aeruginosa, and C. albicans.

References

1. A. D. Metcalfe and M. W. J. Ferguson, Biomaterials, 2007, 28, 5100–5113.
2. M. B. Dreifke, A. A. Jayasuriya and A. C. Jayasuriya, Mater. Sci. Eng., C, 2015, 48, 651–662.
3. J. Boateng and K. Matthews, J. Pharm. Sci., 2008, 97, 2892–2923.
4. A. Yari, H. Yeganeh, H. Bakhshi and R. Gharibi, J. Biomed. Mater. Res., Part A, 2014, 102, 84–96.
5. H. Bakhshi, H. Yeganeh, A. Yari and S. K. Nezhad, J. Mater. Sci., 2014, 49, 5365–5377.
6. A. Yari, H. Yeganeh and H. Bakhshi, J. Mater. Sci.: Mater. Med., 2012, 23, 2187–2202.
7. S. Y. Lin, K. S. Chen and L. Run-Chu, Biomaterials, 2001, 22, 2999–3004.
8. J. M. Yang, S. J. Yang, H. T. Lin and J. K. Chen, J. Biomed. Mater. Res., Part B, 2006, 80, 43–51.
9. L. Klouda and A. G. Mikos, Eur. J. Pharm. Biopharm., 2008, 68, 34–45.
10. B. Jiang, J. C. Larson, P. W. Drapala, V. H. Pérez-Luna, J. J. Kang-Mieler and E. M. Brey, J. Biomed. Mater. Res., Part B, 2012, 100, 668–676.
11. T. T. Reddy, A. Kano, A. Maruyama, M. Hadano and A. Takahara, Biomacromolecules, 2008, 9, 1313–1321.
12. T. Thimma Reddy and A. Takahara, Polymer, 2009, 50, 3537–3546.
13. T. T. Reddy, A. Kano, A. Maruyama, M. Hadano and A. Takahara, J. Biomed. Mater. Res., Part B, 2009, 88, 32–40.
14. A. Drosou, A. Falabella and R. S. Kirsner, Wounds, 2003, 15, 149–166.
15. Z. Aziz, S. F. Abu and N. J. Chong, Burns, 2012, 38, 307–318.
16. J. Tian, K. K. Y. Wong, C. M. Ho, C. N. Lok, W. Y. Yu, C. M. Che, J. F. Chiu and P. K. H. Tam, ChemMedChem, 2007, 2, 129–136.
17. X. Liu, P. Lee, C. Ho, V. C. H. Lui, Y. Chen, C. Che, P. K. H. Tam and K. K. Y. Wong, ChemMedChem, 2010, 5, 468–475.
18. O. Eksik, M. A. Tasdelen, A. T. Erciyes and Y. Yagci, Compos. Interfaces, 2010, 17, 357–369.
19. M. R. Reithofer, A. Lakshmanan, A. T. K. Ping, J. M. Chin and C. A. E. Hauser, Biomaterials, 2014, 35, 7535–7542.
20. A. B. Lowe, Polym. Chem., 2014, 5, 4820.
21. G. Colucci, E. Celasco, C. Mollea, F. Bosco, L. Conzatti and M. Sangermano, Macromol. Mater. Eng., 2011, 296, 921–928.
22. S. Beigi Burujeny, M. Atai and H. Yeganeh, Dent. Mater., 2015, 31, 244–261.
23. T. Yang, M. Malkoch and A. Hult, J. Polym. Sci., Part A: Polym. Chem., 2013, 51, 363–371.
24. B. Sun, Y. Lin and P. Wu, Appl. Spectrosc., 2007, 61, 765–771.
25. J. P. Phillips, N. M. Mackey, B. S. Confait, D. T. Heaps, X. Deng, M. L. Todd, S. Stevenson, H. Zhou and C. E. Hoyle, Chem. Mater., 2008, 20, 5240–5245.
26. M. Puchalski, P. Dąbrowski, W. Olejniczak, P. Krukowski, P. Kowalczyk and K. Polański, Mater. Sci., 2007, 25, 473–478.
27. A. Chandra Babu, M. N. Prabhakar, A. Suresh Babu, B. Mallikarjuna, M. C. S. Subha and K. Chowdoji Rao, Int. J. Carbohydr. Chem., 2013, 2013, 1–8.
28. X. Z. Zhang, D. Q. Wu and C. C. Chu, J. Polym. Sci., Part B: Polym. Phys., 2003, 41, 582–593.
29. M. S. Kim and H. J. Chun, Tissue Eng. Regener. Med., 2011, 8, 117–123.
30. B. Liu, J. Hu and Q. Meng, J. Biomed. Mater. Res., Part B, 2009, 89, 1–8.
31. M. Y. Jen, T. L. Hao and J. Y. Shu, J. Membr. Sci., 2005, 258, 97–105.
32. D. Roy, W. L. Brooks and B. S. Sumerlin, Chem. Soc. Rev., 2013, 42, 7214–7243.
33. K. L. Menzies and L. Jones, Optom. Vis. Sci., 2010, 87, 387–399.
34. J.-P. P. Chen, C.-Y. Y. Kuo and W.-L. L. Lee, Appl. Surf. Sci., 2012, 262, 95–101.
35. M. Black and J. W. Rawlins, Eur. Polym. J., 2009, 45, 1433–1441.
36. C. E. Hoyle and C. N. Bowman, Angew. Chem., Int. Ed., 2010, 49, 1540–1573.
37. S. H. Hsu, C. M. Tang and H. J. Tseng, Acta Biomater., 2008, 4, 1797–1808.
38. D. Paul, S. Paul, N. Roohpour, M. Wilks and P. Vadgama, J. Funct. Biomater., 2013, 4, 358–375.
39. C. C. Chang and C. H. Chang, Polym. Int., 2010, 59, 910–916.
40. R. Gharibi, H. Yeganeh, H. Gholami and Z. M. Hassan, RSC Adv., 2014, 4, 62046–62060.
41. L. Lecamp, F. Houllier, B. Youssef and C. Bunel, Polymer, 2001, 42, 2727–2736.
42. S. Beigi, H. Yeganeh and M. Atai, Dent. Mater., 2013, 29, 777–787.
43. P. Wu, A. C. Fisher, P. P. Foo, D. Queen and J. D. S. Gaylor, Biomaterials, 1995, 16, 171–175.
44. S. M. Bishop, M. Walker, A. A. Rogers and W. Y. Chen, J. Wound Care, 2003, 12, 125–128.
45. J. M. Yang, S. J. Yang, H. T. Lin, T. H. Wu and H. J. Chen, Mater. Sci. Eng., C, 2008, 28, 150–156.
46. P. G. Rodriguez, F. N. Felix, D. T. Woodley and E. K. Shim, Dermatol. Surg., 2008, 34, 1159–1169.
47. P. V AshaRani, G. Low Kah Mun, M. P. Hande and S. Valiyaveettil, ACS Nano, 2009, 3, 279–290.
48. R. de Lima, A. B. Seabra and N. Durán, J. Appl. Toxicol., 2012, 32, 867–879.
49. I. Sondi and B. Salopek-Sondi, J. Colloid Interface Sci., 2004, 275, 177–182.
50. W. R. Li, X. B. Xie, Q. S. Shi, H. Y. Zeng, Y. S. Ou-Yang and Y. Ben Chen, Appl. Microbiol. Biotechnol., 2010, 85, 1115–1122.
51. K. C. Chaw, M. Manimaran and F. E. H. Tay, Antimicrob. Agents Chemother., 2005, 49, 4853–4859.
52. G. a. Martinez-Castanon, N. Niño-Martínez, F. Martínez-Gutierrez, J. R. Martínez-Mendoza and F. Ruiz, J. Nanopart. Res., 2008, 10, 1343–1348.
53. M. Guzman, J. Dille and S. Godet, Nanomedicine, 2012, 8, 37–45.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ra11618j

---