Hydrogel nanocomposites as pressure-sensitive adhesives for skin-contact applications

Soft Matter

C

Dynamic Article Links <

Cite this: Soft Matter, 2011, 7, 2025

PAPER

www.rsc.org/softmatter

Hydrogel nanocomposites as pressure-sensitive adhesives for skin-contact applications Nadia Ba€ıt,abc Bruno Grassl,*c Christophe Derailc and Ahmed Benabouraa Received 7th October 2010, Accepted 29th November 2010 DOI: 10.1039/c0sm01123a This study investigates the effects of monodisperse polystyrene nanoparticle fillers on the network formation, rheological properties and adhesion performance of hydrogel nanocomposites based on polyacrylamide and poly(acrylamide-hydroxyethyl methacrylate). We demonstrated a simultaneous increase in elasticity and tack of these humid composite materials. A 1H-NMR kinetic study showed quasi-total conversion of these monomers during the polymerization–reticulation process and the formation of inhomogeneities within the hydrogel network structure due to the difference in reactivity ratios of the comonomers: acrylamide (AM) and hydroxyethyl methacrylate (HEMA)(rAM ¼ 0.41  0.01 and rHEMA ¼ 7.4  0.3). The rheological properties of these materials were found to be affected by their chemical composition (HEMA content, presence of nanoparticles and heterogeneities). We investigated the adhesion properties of our materials using a probe test tack. Measurements were carried out on a human skin substitute to compare with metal and investigate the potential use of these hydrogel nanocomposites as dermatological patches. The adhesion energy was found to be related to the chemical composition and rheological properties of the hydrogels, as well as to the surface properties of both the adhesive and the substrate.

1 Introduction Hydrogels have been used as potential carriers in drug delivery systems.1–5 Their use as adhesives for dermatological patches, such as in cataplasm and wound healing dressings, has also been recognized as effective.6 The adhesion properties, as well as the mechanical strength of hydrogels, are important characteristics to consider in the development of dermatological patches.7,8 Generally, hydrogels exhibit poor mechanical properties.9–11 To increase the range of their application, two types of approaches are necessary: (i) controlling the microstructure to restrict inhomogeneities and (ii) improving the mechanical strength while maintaining the capacity for water absorption.12,13 Several methods have been reported to produce hydrogels with relatively high mechanical strength combined with high water content: topological (TP) gel,14 nanocomposite (NC) gel,15 interpenetrating polymer networks (IPNs)16–20 and snake-cage (SC) gel. TP gels have figure-eight cross-linkers that can slide a Laboratoire de Synth ese Macromol eculaire et Thio-Organique Macromol eculaire, Facult e de Chimie, USTHB, Bp 32, El-Alia, Bab-Ezzouar, 16111, Algeria b Centre de Recherche Scientifique et Technique en Analyses PhysicoChimiques (CRAPC), Bp 248, RP 16004 Alger, Algeria c Universit e de Pau et des Pays de l’Adour, Institut Pluridisciplinaire de Recherche sur l’Environnement et les Mat eriaux (IPREM)-Equipe de Physique et Chimie des Polym eres-UMR CNRS/UPPA 5254, H elioparc, 2 Avenue du Pr esident Angot, 64053 Pau, Cedex 9, France. E-mail: [email protected]

This journal is ª The Royal Society of Chemistry 2011

along the polymer chains. With this flexible cross-linker, TP gels absorb large amounts of water and can be highly stretched without fracture. In NC gels, polymer chains can be cross-linked by inorganic clay slabs, instead of organic cross-linking agents, on the scale of several tens of nanometres.21,22 NC gels are also highly stretchable and possess other favorable physical properties such as excellent optical transparency. IPN gels consist of two interpenetrating polymer networks: the first is made of highly cross-linked rigid polymers, and the second is made of loosely cross-linked flexible polymers. In SC gels, a linear hydrophobic23 or hydrophilic polymer24,25 (snake) with good mechanical properties is entrapped in a cross-linked hydrophilic polymer (cage). Gels with such high mechanical strength have led to breakthrough applications in industrial and biomedical fields and also raise new fundamental problems in gel science. In particular, the development of hydrogels filled with nanoparticles as pressure-sensitive adhesives would be a major advance for skin-contact applications. In this work, the use of conventional hydrogels filled with nanoparticles as pressure-sensitive adhesives (PSAs) for skincontact applications is presented for the first time; we demonstrate a simultaneous increase in elasticity and tack of these humid composite materials. For that, we studied classical hydrogels containing polymer nanoparticles—hydrogel nanocomposites—which belong to the family of snake-cage (SC) gels in which monodisperse, hydrophobic and non-ionic nanoparticles of polystyrene (PS)-based polymers take the place of the Soft Matter, 2011, 7, 2025–2032 | 2025

linear polymer. The effect of nanoparticle incorporation on the rheological and adhesive properties of these hydrogels is the primary focus of this work. We chose to study hydrogels based on acrylamide (AM) and 2-hydroxyethyl methacrylate (HEMA) that were cross-linked with N,N0 -methylene-bis-acrylamide and filled with monodisperse PS particles. Evidence of enhancements in the mechanical strength and adhesion properties on artificial human skin in comparison to metallic surfaces is presented, and the effects of hydrogel nanocomposites composition on these properties are described.

2 Experimental 2.1 Materials Nanoparticles were synthesized from styrene purified by being passed through a basic activated alumina column before use. Ammonium persulfate [APS, (NH4)2S2O8] and divinylbenzene were purchased from Aldrich and used without further purification. Two surfactants were used as received for the preparation of nanoparticles: Igepal CO 897ª (called NP40, hydrophilic– lypophilic balance HLB ¼ 17.8), obtained from Rhodiaª, and Ninolª (HLB ¼ 11.4), obtained from Stepanª. For hydrogel synthesis, acrylamide (AM) (purity >99%), 2-hydroxyethyl methacrylate (HEMA) (purity >99%) and N,N0 -methylene-bisacrylamide (Bis) (purity >99%) were purchased from Aldrich. Potassium persulfate (KPS) was used as the initiator for hydrogel polymerization. Salt-free Milli-Q water, with an electrical conductivity of 18.3 MU cm at 25  C, was filtered through a 0.22 mm Millipore filter prior to use. Reaction mixtures were prepared with the filtered water. All 1H-NMR spectra were recorded with D2O as the solvent (purchased from SDS, purity >99.9%). Four main products were used as received for synthesis of the human skin equivalent: gelatine PS 30 Bloom 250, purchased from Rousselot; glycerol, obtained from Baker; castoryl maleate (called ceraphyl RMT), marketed by the ISP company; and formaldehyde, purchased from Acros Organics. 2.2 Latex particle synthesis Latex particles were synthesized according to the procedure used by Kohut-Svelko et al.26 The reaction vessel was purged with nitrogen to remove traces of oxygen. In a typical polymerization, a reaction mixture of non-ionic surfactants (NP40 62 wt%, Ninol 38 wt%) was dissolved in distilled water (200 mL) in a threenecked round-bottom flask under a nitrogen atmosphere. The purified monomers styrene (87 g, 0.836 mol) and divinylbenzene (3.7 wt% vs. styrene), used as the cross-linking agent, were added to the aqueous solution and emulsified under vigorous stirring. This emulsion was then heated under mechanical stirring up to 70  C. An aqueous solution of APS (0.3 g in 5 mL of water) was then added drop by drop. Polymerization was allowed to proceed for 24 h at 70  C under mechanical stirring. 2.3 Nanoparticle analysis The nanoparticle size distribution in the emulsion was characterized on an environmental scanning electron microscope (ESEM), and analysis was carried out using an ElectroScan instrument operating at 20 kV. A multi-acquisition analysis 2026 | Soft Matter, 2011, 7, 2025–2032

Fig. 1 Nanoparticle size distribution achieved by environmental scanning electron microscope (ESEM) analysis (inset: ESEM images of PS latex particles).

(i ¼ 671), as shown in Fig. 1, provided the number, surface, weight-averaged particle diameter and particle-diameter dispersity: dn ¼ 410, ds ¼ 470, dw ¼ 505 nm and dw/dn ¼ 1.15, respectively. The following relationships were used: dn ¼ Snidi/ Sni, ds ¼ Snidi3/Snidi2, dw ¼ Snidi4/Snidi3 and amplitude (in Fig. 1) proportional to nidi.6 Dynamic light scattering measurements, carried out on a VASCO apparatus developed by Cordouan Technologies, show the absence of aggregates in the emulsion. The technique works on a very thin suspension layer to avoid multiscattering phenomena and can be successfully applied to concentrated, dense and opaque solutions (laser wavelength: 658 nm). 2.4 Gel samples Gels were prepared by classical radical polymerization with Bis as the cross-linker and KPS as the initiator at a reaction temperature Tr ¼ 60  C. Reaction mixtures were prepared a few hours before polymerization and thoroughly mixed for 2 h prior to use. The sample solutions were degassed with nitrogen. Hydrogels were prepared by adding comonomers (AM and HEMA), cross-linker (Bis) and initiator (KPS) to D2O and water for NMR and rheological/adhesion experiments, respectively. The hydrogel nanocomposites containing hydrophobic nanoparticles were prepared by the addition of latex into the above mentioned solutions. 2.5 Gel compositions Let us introduce the following parameters to classify the hydrogel compositions: - f represents the volume fraction of polymer particles in the total volume of the reaction mixture, - [I2], [AM], [HEMA] and [Bis] represent molar concentrations of initiator and vinylic units of acrylamide, HEMA and Bis, respectively, taking into account the volume of the aqueous phase without the contribution of hydrophobic nanoparticles, i.e. in 1  f. - [I2] and the molar fraction of the cross-linking agent, [Bis]/ ([AM] + [HEMA]), are constant and equal to 4.6  103 mol L1 This journal is ª The Royal Society of Chemistry 2011

and 0.47 mol%, respectively. The weight fraction of comonomers in the aqueous phase (1  f) is constant and equal to 7 wt%. Hydrogels without nanoparticles are termed HX, where ‘‘X’’ indicates the hydrogel molar HEMA content ([HEMA]/[AM] + [HEMA]). Hydrogel nanocomposites were named HX-Y-Z, where ‘‘Y’’ refers to the volume fraction of nanoparticles f and Z to the type of nanoparticles; in this study, Z ¼ PS. For example, H20-26-PS contains 20 mol% HEMA and 26% polystyrene nanoparticles. 2.6

1

H-NMR monitoring

The NMR apparatus used was a 400 MHz Bruker Advanced AM400 spectrometer. Once the reaction mixtures were prepared and mixed, they were poured into NMR glass tubes (5 mm interior diameter), purged with nitrogen and sealed with paraffin film. Around 150 measurements were taken continuously during the polymerization in order to follow the monomer unit conversion as a function of time. Fig. 2 shows an example of the advance of the polymerization reaction of the precursory mixture of the H10 hydrogel. To follow-up on the reaction kinetics, we considered four integrations of 1H-NMR spectra: the first two integrations, A (at 6.4–6.1 ppm) and B (at 6.1–5.8 ppm), correspond to vinylic protons of acrylamide and (2-hydroxyethyl methacrylate) monomers, respectively; the third integration, C, attributed to –CH2– monomeric HEMA and polymeric PHEMA; and the last, D, was assigned to aliphatic protons of PAM, HEMA and PHEMA. In this way, we obtained the overall polymer conversion versus time. 2.7 Skin substitute synthesis and characterization Human skin substitute was synthesized and characterized according to a procedure used recently by our team.27 In summary, the human skin substitute was developed from fat components added to natural reticulated protein to mimic the mechanical and interfacial properties of real skin. The typical procedure proposed by Renvoise et al.27 and Lir et al.28 to obtain

the final blend was as follows. First, 10 grams of gelatine (PS30, Rousselot) was dissolved in 80 g of water at T ¼ 45  C and in the presence of 4.5 mL of 1 M NaOH to adjust the pH to 9.0. After 20 min, 10 g of glycerol (Baker) and 3 g of fat components (Ceraphyl RMT, ISP) were added and stirred for 10 min at the same temperature. Finally, a solution of formaldehyde (37% in water, Acros Organics) was added. At the end, this blend was deposited on an aluminium surface modified by chemical abrasion to mimic the surface rugosity of human skin (79.3 mm of mean depth of the crevasses; compared to 91 mm and 36 mm for old and young men, respectively) and was kept at T ¼ 33  C for 5 hours. The final film was used directly for tack experiments. Values of the Young’s modulus of the final film are very close to those reported in the literature for human skin and range from 2  105 to 108 Pa. The surface properties (surface tension and its components) are reported here in Section 2.10. 2.8 Rheological measurements Rheological measurements were made using a dynamic rotational rheometer (DSR, Rheometrics) operating under controlled stress. For uncharged and hydrogel nanocomposites, we measured the storage G0 and loss G00 moduli as functions of the circular frequency in the linear domain detected by a stress sweep at the same temperature. A stainless steel parallel-plate geometry with a 25 mm diameter was used, and measurements were carried out at a temperature of T ¼ 25  C controlled by a Peltier device. Samples were tested in the form of discs immediately after their synthesis, as mentioned above. In order to keep the hydrogels’ moisture constant, a water trap was used to minimize evaporation. To verify that the rheological properties remained constant with time, we performed time sweep measurements for 2000 s at a constant angular frequency (10 rad s1). These experiments showed that G0 and G00 remain constant during this time, demonstrating the stability of our samples. This time allows us to perform stress sweeps and frequency sweeps without evolution of the samples. 2.9 Test tack

Fig. 2 Spectral evolution of reaction mixture for in situ H10 hydrogel formation in an 1H-NMR probe at t ¼ 370, 1180, 400, 4030 and 4580 s for spectra 1 to 5, respectively.

This journal is ª The Royal Society of Chemistry 2011

The probe tack test is typically used to evaluate the adhesion performances of pressure-sensitive adhesives (PSAs).29 We carried out this test to explore the adhesion properties of samples stuck onto different substrates based on their chemical composition. We used metallic (substrate-probe) surfaces and viscoelastic (substrate-probe) surfaces (artificial skin) specially formulated to mimic human skin. The characteristics of both surfaces are discussed later in Section 2.10. Tack experiments were performed using a TA-XT2i texture analyzer. Samples were tested in the form of discs with 32 to 34 mm diameters at T ¼ 25  C. The method consists of contacting a flat punch with the sample stuck on the surface. As indicated in Fig. 3, we controlled the compressive force (Fc) during the contact time (tc ¼ 120 s). After tc, the flat punch was removed at a constant rate (10 mm s1), during which we measured the force versus time and displacement. During this debonding stage, one can plot a nominal stress (s ¼ F/A0) versus strain (3 ¼ h/h0  1) curve from the force-displacement curve. A0 and h0 are, respectively, the initial area and thickness of the tested hydrogel disc. Soft Matter, 2011, 7, 2025–2032 | 2027

Fig. 3 (a) Probe tack procedure. Phase 1: approach; phase 2: contact between probe and sample; phase 3: debonding step. (b) Typical stressstrain curves obtained during the debonding phase with (1) soft and (2) hard substrates.

The main parameters from the stress–strain curve are the maximum values of the stress smax and the strain 3max, which are reported in Fig. 3. The total work of adhesion Wadh performed to detach the adhesive sample from the surface is experimentally obtained Ð from the integral under the stress–strain curve (Wadh ¼ sd3).29 We have not directly observed the surface during the debonding phase because the model skin is not sufficiently transparent. 2.10 Contact angle and surface tensions measurements We used a Tracker Dynamic Tensiometer and the well-known direct method measurement of the angle between a suitable liquid drop and the examined material surface. On the basis of the Owens–Wendt method,30 we evaluated the surface tension of the substrates using three liquids with surface tensions gL, dispersive gLd and polar gLP components as follows: water (gL ¼ 72.8 mN m1, gLd ¼ 21.8 mN m1, gLP ¼ 51.0 mN m1), diiodomethane (gL ¼ 50.8 mN m1, gLd ¼ 50.8 mN m1, gLP ¼ 0 mN m1) and glycerol (gL ¼ 63.4 mN m1, gLd ¼ 37.0 mN m1, gLP ¼ 26.4 mN m1). The surface tensions measured were as follows: for aluminium substrate, gL ¼ 43 mN m1, gLd ¼ 32 mN m1, gLP ¼ 26 mN m1 according to data in the literature;31 for artificial human skin substrate, gL ¼ 29 mN m1, gLd ¼ 27 mN m1 and gLP ¼ 2 mN m1 according to the previous work.27

3 Results and discussion 3.1 Hydrogel synthesis The effects of monomer, initiator and cross-linker concentrations, polymerization temperature and volume fraction of 2028 | Soft Matter, 2011, 7, 2025–2032

nanoparticles on the network structure and hydrogel mechanical properties were studied for acrylamide gels only using rheological32 or NMR12 monitoring. Generally, the structure was changed by modulating the molar ratio of the acrylamide monomer and the cross-linker. For a given acrylamide monomer concentration, there exists an optimal cross-linker concentration and an optimal polymerization temperature which give rise to an ‘‘ideal’’ hydrogel, i.e., exhibiting a maximal elasticity. That can be explained by the heterogeneity of the network due to the difference of reactivity ratios of the acrylamide and the cross-linking agent (rAM ¼ 0.52 and rBis ¼ 5.2 for AM and Bis, respectively).12 In this work, we show the effect of HEMA comonomer content on the elasticity of poly(acrylamide-hydroxyethyl methacrylate) hydrogels due to the difference in reactivity ratios between AM and HEMA. The determination of rAM and rHEMA using the ratio of AM and HEMA rate constants from NMR data was carried out by a convenient and accurate method described in the previous work.12 The reactivity ratios can be determined from the ratio of the apparent rate constants of the propagation reactions of AM and HEMA, respectively (aAM, aHEMA), determined for initial decay rates and for separate experiments in which the relative initial concentrations [AM]0 and [HEMA]0 are varied and the same initiator concentration is used. This yields the following:   1 þ rAM ½AM0 =½HEMA0 aAM  ¼ (1) aHEMA ½AM0 =½HEMA0 þ rHEMA where, [AM] ¼ [AM]0eaAMt; [HEMA] ¼ [HEMA]0eaHEMAt

(2)

Fig. 4a shows the determination of rAM and rHEMA using the ratio of AM and HEMA rate constants from eqn (1). The experimental values of rAM and rHEMA obtained are 0.41  0.02 and 7.4  0.3, respectively. Our results showed that a non-ideal copolymerization occurs with a certain tendency to form larger blocks of HEMA because of the preference of HEMA for homopropagation and a nonnegligible weight fraction of polyacrylamide chain in the network, as show in Fig. 4b. This led to the formation of inhomogeneous networks. The weight fraction of the network to include a copolymer chain with 10–90 mol% of HEMA, W10,90, can be estimated from Fig. 4b: W10,90 ¼ 6.3%, 34%, 75% and 28% for H10, H20, H50 and H80, respectively. These values show that the heterogeneity is maximal for 50–60 mol% of HEMA in the initial feed. The effect of these heterogeneities on the mechanical properties of hydrogels is discussed in the following section. 3.2 Rheological behavior The evolution of W10,90 and the elastic moduli of different, uncharged P(AM-HEMA) hydrogels is reported in Fig. 5. These hydrogels have been obtained in pure water with a total conversion of the monomers AM and HEMA and the same molar ratio of cross-linker used to perform the above-mentioned 1 H-NMR kinetics. From 0% to 50% of HEMA, G0 decreases from 1300 Pa to 180 Pa. For higher HEMA content, G0 increases to reach its maximum value of 1700 Pa for pure HEMA. This complex evolution of G0 correlated with W10,90 clearly shows that This journal is ª The Royal Society of Chemistry 2011

Fig. 4 (a) Determination of rAM and rHEMA using the ratio of AM and HEMA rate constants from eqn (2). The values of rAM and rHEMA are 0.41 and 7.4, respectively. (b) Instantaneous HEMA composition in the copolymer vs. total monomer conversion in weight%.

 Gb ¼

Fig. 5 Evolution of W10,90 and the elastic modulus (G0 ) of different uncharged P(AM-HEMA) hydrogels.

lower elasticity and firmness are obtained for the most heterogeneous network, i.e. H50 hydrogel. This result highlights the effect of the hydrogel structure on its rheological properties. To enhance the mechanical properties of these hydrogels, monodisperse nanoparticles of PS were introduced in H00, H10 and H20 with a constant weight fraction of 0.07 of comonomers in the aqueous phase (1  f) and a 0.26 volume fraction of polymer particles. Thus, we obtained the hydrogel nanocomposites denoted H00-26-PS, H10-26-PS and H20-26-PS. Note that 1H-NMR kinetics showed a total conversion of these monomers during the polymerization–reticulation process. The values of the storage modulus, G0 , for each sample are reported in Table 1 and show that there is a decrease in the elasticity of uncharged hydrogels with an increase in the HEMA hydrogel composition. This can be explained by the increase of heterogeneities in the network structures, as mentioned above for uncharged gels. The presence of PS nanoparticles within the networks gives rise to the elastic modulus G0 of the hydrogels according to rheological modeling of a composite of two elastic bodies without interfacial tension.33 The elastic modulus of the blend Gb, in our case, a PS nanoparticle-filled hydrogel, can be related to the elastic modulus of the matrix Gm and the dispersed phase Gd as follows: This journal is ª The Royal Society of Chemistry 2011

ð2Gd þ 3Gm Þ þ 3fPS ðGd  Gm Þ ð2Gd þ 3Gm Þ  2fPS ðGd  Gm Þ

 (3)

where Gd is close to 6  109 Pa34 and Gm is the elastic modulus of the matrix as measured for the hydrogel without PS nanoparticles. The values of G0 listed in Table 1 show that the experimental elasticity of each PS nanoparticle composite hydrogel is close to that predicted by the rheological model. Moreover, note that the hydrogel nanocomposites elasticity changes in the same way as was observed for the uncharged gels with the increase of HEMA molar content. The data suggest that PS nanoparticles have a very small influence on network formation during the polymerization process. In other words, it can be assumed that the presence of the nanoparticles in the mixture has no considerable effects on the heterogeneity of the networks generated by the difference in reactivity ratios between the comonomers AM and HEMA.

3.3 Adhesion performance of hydrogels Before discussing the qualitative and quantitative analyses of our results, it is important to report that our samples always debonded from the two substrates used without leaving a residue. This interfacial debonding is a very interesting characteristic of these hydrogels relevant to the application for which they are intended: namely, skin-contact application. Fig. 6 shows the stress–strain curves obtained from the tack tests for PAM and P(AM-HEMA) uncharged hydrogels (H0, H10 and H20) and PS hydrogel nanocomposites PAM and P(AM-HEMA) hydrogels (H0-26-PS, H10-26-PS and H20-26-PS), using metallic and artificial skin surfaces. The values of smax, 3max and Wadh are listed in Table 1. As shown, there is a very marked change in the shape of the stress–strain curves going from uncharged to charged hydrogels and from a metallic substrate to a viscoelastic one, suggesting the existence of different mechanisms during debonding. Indeed, one can note the following: (i) The maximum stress smax is affected by the substrate nature. Indeed there is a significant decrease in smax values when going from metallic substrate to the skin equivalent. Nevertheless, based on our results one cannot differentiate the roughness and the viscoelasticity effects. Soft Matter, 2011, 7, 2025–2032 | 2029

Table 1 Rheological and adhesive characteristics of charged and uncharged P(AM-HEMA) hydrogels f H00 H00-26 H10 H10-26 H20 H20-26 a

0 26 0 26 0 26

G0 (Gb)a/Pa

tan d

Probeb

smax/kPa

3max

Wadh/J m2

Contact angle ( )

1280

0.11 0.10

950

0.31

1800 (1830)

0.11

640

0.38

1200 (1230)

0.09

39 28 41 29 37 32 35 18 33 21 49 21

1.5 1.7 1.7 2.6 1.4 1.2 1.2 2.0 1.1 1.2 1.1 2.0

46 38 58 176 39 27 44 92 36 22 53 101

44

2400 (2460)

m–m s–s m–m s–s m–m s–s m–m s–s m–m s–s m–m s–s

69 58 65 61 67

Elastic modulus of the blend Gb calculated according to rheological modeling.33 b m: metal, s: skin.

(ii) The stress decreases regularly in the case of metallic substrates. (iii) Contrarily, when using the synthetic skin substrate, the stress decreases to a plateau value that remains nearly constant then decreases again to zero. It is also worth noticing that the plateau is much larger in the case of PAM and P(AM-HEMA) charged hydrogels. (iv) The maximum deformation 3max of the hydrogels obtained during the tack experiments increases between metal–metal and skin–skin substrates. The low value of smax can be explained by the fact that the detachment takes place without cavitation on these systems, as is known to occur in the conventional pressure sensitive adhesive.35 Therefore, the work Wadh required for debonding was used to characterize the tack of the hydrogel nanocomposites. The values of Wadh reported in Table 1 are the same order of magnitude as commercially available classes of PSAs used for different conventional applications35 or dedicated for skin-contact applications (acrylics, polyisobutylenes, silicones).7,35 As noted previously, according to the adhesion requirements for skin-contact applications, hydrogel nanocomposites adhere to synthetic skin without leaving any residue on the skin upon removal, even for samples kept in wet conditions for one month. The values of Wadh obtained, combined with the interfacial debonding, highlight the great potential of our materials as adhesives for dermatological patches. A thorough examination of Wadh values showed the complexity of these systems. For both the metallic and viscoelastic substrates, there is a decrease in the adhesion energy with increasing hydrogel HEMA composition in the case of uncharged hydrogels. Further, the values of the elastic modulus decrease as a function of the HEMA composition yet tan d (¼G00 / G0 ) increases at the same time, indicating the dissipation increase within the hydrogels. These contradictory results can be explained by the fact that the debonding occurred before the hydrogels had totally deformed themselves. In the same way, for the range of HEMA hydrogel compositions considered in this section, the presence of PS nanoparticles greatly influenced work required for debonding. In particular, for the same type of surface, there was an improvement in the adhesion properties of the above-cited hydrogels when filled with PS nanoparticles. This 2030 | Soft Matter, 2011, 7, 2025–2032

improvement was much more marked when using the artificial skin substrate. Contrary to the results obtained from uncharged hydrogels, the work for debonding did not change monotonically with increasing HEMA composition. There was first a decrease and then an increase in Wadh with increasing HEMA composite network composition. All of these results can be interpreted by considering important differences at the interface between the hydrogels and synthetic skin or hydrogels and metal (roughness, surface tensions, etc.). This interface has not been completely studied here and will provide interesting future work, in addition to the previous study carried on this type of substrate by using peeling experiments.36 Particularly the effects of roughness and viscoelasticity will have to be differentiated. With the aim of supporting our tack test results, we carried out contact angle measurements on hydrogel surfaces using diiodomethane, which is a commonly used liquid to evaluate its hydrophilicity. The contact angle values recorded and given in Table 1 show that for uncharged hydrogels, the contact angle increases with increasing HEMA content. This result is related to the hydrophobic character of poly(hydroxyethyl methacrylate). The introduction of PS nanoparticles within the PAM and P(AM-HEMA) networks has two distinct effects. (i) For a given HEMA hydrogel composition, when going from an uncharged hydrogel to PS nanoparticle-filled one, there is an increase in the contact angle that results from the hydrophobic character of polystyrene. This evolution is similarly observed for the values of Wadh obtained for the two surfaces tested. (ii) By varying the HEMA composition in the hydrogel nanocomposites, we note that contact angles change in the same way as does the Wadh of the same materials obtained for the two different substrates: namely a decrease followed by an increase of their values. Through these results, we have shown the influence of the surface properties of hydrogel nanocomposites on their adhesive character. We cannot formally conclude about the complex mechanism of adhesion between the hydrogel nanocomposites and skin substitute, but we have shown that the incorporation of hydrophobic nanoparticles in conventional hydrogels increases the elasticity and hydrophobicity of the hydrogel surface sufficiently to attain pressure-sensitive adhesives (PSAs) requirements for skin application. In addition, the composition and network This journal is ª The Royal Society of Chemistry 2011

demonstrate a simultaneous increase in elasticity and tack of these humid composite materials. The most important shortcoming in currently available adhesives is their inability to adhere during periods of strenuous exercise and under humid conditions. We have demonstrated that hydrogel nanocomposites may be the solution to this problem because the composites are elastic and contain more than 60% of water, making them compatible with excipients (i.e., liquids that are sometimes used to enhance skin transport of certain drugs). In this context, the advantage of using hydrogel nanocomposites is the ability of the adhesives to absorb a significant amount of the excipient without losing their adhesive character or elastic modulus.

Acknowledgements The ‘‘Programme National Exceptionnel’’ is gratefully acknowledged for a PhD grant to N. Ba€ıt to perform experimental work in IPREM. We thank Abdel Khoukh for assistance with NMR experiments and Virginie Pellerin for help with the ESEM experiments.

References

Fig. 6 Effect of the substrate-probe surfaces and the chemical composition of the hydrogels on the stress–strain curves (m–m for metal–metal and s–s for skin–skin).

heterogeneity in the hydrogel matrix do not affect their use since it is the nanoparticle fillings that mainly govern the adhesive properties.

4 Conclusion In this work, the use of conventional hydrogels filled with nanoparticles as pressure-sensitive adhesives (PSAs) for skincontact applications is presented for the first time; we This journal is ª The Royal Society of Chemistry 2011

1 A. S. Hoffman, J. Controlled Release, 1987, 6(1), 297–305. 2 N. A. Peppas, P. Bures, W. Leobandung and H. Ichikawa, Eur. J. Pharm. Biopharm., 2000, 50(1), 27–46. 3 N. A. Peppas, Y. Huang, M. Torres-Lugo, J. H. Ward and J. Zhang, Annu. Rev. Biomed. Eng., 2000, 2, 9–29. 4 X. Z. Zhang, D. Q. Wu and C. C. Chu, Biomaterials, 2004, 25(17), 3793–3805. 5 A. Guiseppi-Elie, Biomaterials, 2010, 31(10), 2701–2716. 6 Y. Onuki, M. Hoshi, H. Okabe, M. Fujikawa, M. Morishita and K. Takayama, J. Controlled Release, 2005, 108(2–3), 331–340. 7 S. Venkatraman and R. Gale, Biomaterials, 1998, 19(13), 1119–1136. 8 I. Strehin, Z. Nahas, K. Arora, T. Nguyen and J. Elisseeff, Biomaterials, 2010, 31(10), 2788–2797. 9 J. A. Johnson, N. J. Turro, J. T. Koberstein and J. E. Mark, Prog. Polym. Sci., 2010, 35(3), 332–337. 10 G. M. Kavanagh and S. B. Ross-Murphy, Prog. Polym. Sci., 1998, 23(3), 533–562. 11 J. Kopecek, Biomaterials, 2007, 28(34), 5185–5192. 12 C. Thevenot, A. Khoukh, S. Reynaud, J. Desbrieres and B. Grassl, Soft Matter, 2007, 3(4), 437–447. 13 W. C. Lin, W. Fan, A. Marcellan, D. Hourdet and C. Creton, Macromolecules, 2010, 43(5), 2554–2563. 14 Y. Okumura and K. Ito, Adv. Mater., 2001, 13(7), 485–487. 15 S. G. Starodoubtsev, A. A. Ryabova, A. T. Dembo, K. A. Dembo, I. I. Aliev, A. M. Wasserman and A. R. Khokhlov, Macromolecules, 2002, 35(16), 6362–6369. 16 J. P. Gong, Y. Katsuyama, T. Kurokawa and Y. Osada, Adv. Mater., 2003, 15(14), 1155–1158. 17 Y.-H. Na, T. Kurokawa, Y. Katsuyama, H. Tsukeshiba, J. P. Gong, Y. Osada, S. Okabe, T. Karino and M. Shibayama, Macromolecules, 2004, 37(14), 5370–5374. 18 T. Nakajima, H. Furukawa, Y. Tanaka, T. Kurokawa, Y. Osada and J. P. Gong, Macromolecules, 2009, 42(6), 2184–2189. 19 B. B. Mandal, S. Kapoor and S. C. Kundu, Biomaterials, 2009, 30(14), 2826–2836. 20 Y. Liu and M. B. Chan-Park, Biomaterials, 2009, 30(2), 196–207. 21 K. Haraguchi, T. Takehisa and S. Fan, Macromolecules, 2002, 35(27), 10162–10171. 22 K. Haraguchi, R. Farnworth, A. Ohbayashi and T. Takehisa, Macromolecules, 2003, 36(15), 5732–5741. 23 S. Nagaoka, Polym. J. (Tokyo, Jpn.), 1989, 21(10), 847–850. 24 M. J. Hatch, J. A. Dillon and H. B. Smith, Ind. Eng. Chem., 1957, 49(11), 1812–1819. 25 H. Bodugoz-Senturk, C. E. Macias, J. H. Kung and O. K. Muratoglu, Biomaterials, 2009, 30(4), 589–596.

Soft Matter, 2011, 7, 2025–2032 | 2031

26 N. Kohut-Svelko, S. Reynaud, R. Dedryvere, H. Martinez, D. Gonbeau and J. Francois, Langmuir, 2005, 21(4), 1575–1583. 27 J. Renvoise, D. Burlot, G. Marin and C. Derail, Int. J. Pharm., 2009, 368(1–2), 83–88. 28 I. Lir, M. Haber and H. Dodiuk-Kenig, J. Adhes. Sci. Technol., 2007, 21(15), 1497–1512. 29 C. Creton and P. Fabre, in Adhesion Science and Engineering: the Mechanics of Adhesion, Elsevier, New York, 2002. 30 D. K. Owens and R. C. Wendt, J. Appl. Polym. Sci., 1969, 13(8), 1741–1747.

2032 | Soft Matter, 2011, 7, 2025–2032

31 A. Rudawska and E. Jacniacka, Int. J. Adhes. Adhes., 2009, 29(4), 451–457. 32 D. Calvet, J. Y. Wong and S. Giasson, Macromolecules, 2004, 37(20), 7762–7771. 33 M. Bousmina, Rheol. Acta, 1999, 38(3), 251–254. 34 F. Leonardi, J. C. Majeste, A. Allal and G. Marin, J. Rheol. (Melville, NY, U. S.), 2000, 44(4), 675–692. 35 H. Lakrout, P. Sergot and C. Creton, J. Adhes., 1999, 69(3–4), 307–359. 36 J. Renvoise, D. Burlot, G. Marin and C. Derail, J. Adhes., 2007, 83, 403–416.

This journal is ª The Royal Society of Chemistry 2011