Sticky degradable bioelastomers

Article pubs.acs.org/cm

Sticky Degradable Bioelastomers Richard Vendamme* and Walter Eevers Flemish Institute for Technological Research (Vito N.V.), Boeretang 200, Mol 2400, Belgium ABSTRACT: This Article investigates the important, and often overlooked, structure−property relationships underlying the complex viscoelastic and adhesion behaviors of soft polyester elastomers, an emerging class of degradable biomaterials with promising applications in industry, biotechnology, and medicine. A family of poly(isosorbide fatty alkylates) with different molecular architectures and physical aspects (viscous, sticky, rubbery, solid) is reported. We demonstrate that the adhesiveness of these materials can be mainly understood in terms of bulk viscoelasic factors, in contrast to alternative ideas reported in the literature. These results shed some light on the intimate structure of bioelastomers, and notably on the decisive role of a well-adjusted macroscopic cluster of percolated polyester chains for tailoring key biomaterial functions such as elasticity, stickiness, fibrillation, and biodegradation. By establishing a bridge between polyester biomaterials and the material science of sticky things, this Article provides robust design principles for diverse functional biomaterials with tailored dissipative characters such as adhesives with tunable stickiness and degradation profiles, or scaffolds mimicking the nonlinear elasticity of supersoft biological tissues.

■

INTRODUCTION The biological function of natural soft biomaterials is often related to their unique biomechanical properties. For instance, sandcastle worms in the sea and snails on land produce viscous hydrophobic secretions that strongly stick underwater even in turbulent environment,1 the adhesiveness of the glue droplets produced by orb-weaving spiders for coating their silk and catching prays can be explained by the theory of viscoelastic solid,2 whereas the wing supports of beetles endure millions of deformation cycles thanks to rubber-like elements that are mechanically very similar to man-made car tires. Examples also abound in the human body: mucus secretions are viscoelastic fluids that aid protecting the lungs by sticking foreign particles that could enter them during breathing,3 ligaments are viscoelastic solids that return to their original dimension up to a certain level of deformation before showing permanent deformation (creep),4 and the extracellular matrix (ECM) is an elastic cross-linked network of various proteins displaying complex nonlinear viscoelastic behaviors including strainhardening at large deformations.5 Synthetic soft biomaterials that mimic some of the elaborated features observed in nature are attracting the attention of (bio)chemical engineers due to their promising applications in industry, biotechnology, and medicine.6 Among the various classes of materials that have been used to synthesize such substances, polyesters have recently emerged as a versatile and sustainable platform for various reasons: polyesters are readily synthesized in bulk without organic solvents,7 they tend to degrade in various environments due to the presence of hydrolytically degradable ester bonds,8 and they can be synthesized from monomers derived from renewable resources or even found in the human metabolism (endogenous).9 In © 2017 American Chemical Society

2002, Langer introduced poly(glycerol sebacate) (PGS) as a tough biodegradable elastomer combining remarkable properties such as rubber-like mechanical properties, in vitro and in vivo biocompatibility, and complete adsorption under the skin.10 The significance of that discovery triggered the development of a wide variety of polyesters networks derived from low molecular weight multifunctional alcohols and carboxylic acids monomers.11,12 Up to now, the scientific community has mainly paid attention to the chemical properties and biological evaluation of polyester bioelastomers. It has also been shown that the mechanical properties of these biorubbers could be tailored to a large extent by tuning the polymerization degree and the monomers’ molecular structure and molar ratio.12,13 Several polyester network structures displaying stronger mechanical properties have been introduced, such as super tough elastomer for load-bearing applications14 or stiff degradable elastomers.15 However, the other side of the mechanical spectra (i.e., soft and ultrasoft elastomers) has not been fully explored yet. Intriguingly, the Langer team noted that PGS networks with low degrees of cross-linking were found to be exceedingly sticky on a variety of substrates, including human skin and even hardto-stick substrates like Teflon.16 Chung et al. recently mapped the properties of PGS as a function of the degree of esterification (DE) and confirmed that substances with DE of 60−75% were indeed sticky, a property that these authors attributed to the adhesive character of the remaining hydroxyl groups.17 Considering the huge body of literature dedicated to Received: April 21, 2017 Revised: June 1, 2017 Published: June 1, 2017 5353

DOI: 10.1021/acs.chemmater.7b01635 Chem. Mater. 2017, 29, 5353−5363

Article

Chemistry of Materials

Figure 1. Molecular structures of poly(isosorbide fatty alkylates). P1, poly(IS C36 dimerate); P2 and P3, poly(IS C36 dimerate-co-sebacate); P4 and P5, poly(IS C36 dimerate-co-succinate). Branched and network structures fully based upon ester linkages were achieved by cross-linking the carboxyterminated polymers with epoxidized linseed oil (ELO), an epoxy-functional triglyceride where the three fatty acid elements are linked to a central glycerol core via ester bonds.

establishing for the first time a bridge between polyester biomaterials and the material science of sticky things, this Article provides robust design principles for diverse functional biomaterials with tailored dissipative characters such as adhesives with tunable stickiness and degradation profiles, or scaffolds mimicking the nonlinear elasticity of ultrasoft biological tissues.

polyester biomaterials, it is surprising that sticky biodegradable elastomers have never been the subject of more in depth investigations and are still a largely overlooked topic to date, despite their obvious practical applications as functional glues, sealants, or scaffolds in various domains ranging from industry to surgery. Herein, we investigate the delicate structure−property relationships underlying the rich viscoelastic and adhesion behaviors of soft polyester biomaterials. We demonstrate that the adhesiveness of soft polyester networks can be mainly understood in terms of bulk rheological factors, putting to rest the idea that it stems from unreacted chemical groups. To that end, we designed a family of poly(isosorbide fatty alkylates) with different molecular architectures and physical aspects (viscous, sticky, rubbery, solid) as model polyester elastomers. Mechanical properties at both small and large strains and adhesion performances have been investigated by various techniques and complemented with theoretical frameworks initially developed for pressure-sensitive adhesives, resulting in the establishment of a new viscoelastic phase diagram depicting generic adhesive archetypes. These results shed some light on the intimate structure of soft bioelastomers, and notably on the decisive role of a well-adjusted macroscopic cluster of percolated polyester chains for tailoring key biomaterial functions such as elasticity, stickiness, and fibrillation. Early stage biodegradation has been investigated in specific controlled conditions for selected specimens, providing insights into the degradation pathways of linear and branched polyester structures. The conclusions obtained herein with model isosorbidebased networks are more generally applicable to the wide variety of soft polyester biomaterials developed so far. By

■

RESULTS AND DISCUSSION Synthesis of Poly(isosorbide fatty alkylates) with Various Compositions and Topologies. Our approach to designing amorphous 3D network structures fully based upon ester linkages is to end-link carboxy-terminated linear chains with epoxy-functionalized triglycerides. We anticipated that this synthetic approach will lead to more reproducible network structures than elastomers directly synthesized by reacting multifunctional alcohols and carboxylic acid, which often suffer from high stoichiometric variability because of the partial evaporation of monomers during synthesis (i.e., glycerol in the case of PGS synthesis).17 We have recently shown that renewable polyesters made from a combination of sugar-based isosorbide (IS) and lipidbased dimerized fatty acid (C36 dimer) combine the intrinsic properties of their precursors, such as flexibility, softness, and functionality, and form the basis of a new class of self-adhesive materials.18 In the chemical section of this Article, we expand this polyester platform by investigating additional comonomers and different form factors such as macroscopic gels, adhesive tapes, and fibers. The chosen monomers all have relevance for both industrial and biotechnological applications. For instance, C36 dimers (obtained by dimerization of C18 fatty acids such as oleic or linoleic acids) are well-known building blocks of hot5354


Article


Table 1. Co- and Ter-polyesters Synthesized from Isosorbide (IS), a C36 Dimer Fatty Acid (Radia 0975), and a Second Dicarboxylic Acid (Sebacic or Succinic) characterization of the linear polyestersa

monomer mixture in wt % (mol %) IS P1 P2 P3 P4 P5

19.63 30.12 35.17 32.48 41.75

(48.6) (48.7) (48.6) (49.0) (48.8)

Radia 0975 80.37 40.16 20.63 50.64 29.13

(51.4) (16.6) (7.3) (19.5) (8.8)

second diacid none sebacic sebacic succinic succinic

29.72 44.20 16.88 29.13

(34.7) (44.1) (31.5) (42.3)

Tα (°C)

Mw (g/mol)

PDI

AV (mg-KOH/g)

OHV (mg-KOH/g)

−18.2 −7.8 −2.2 13.7 44.6

27032 19195 25672 32114 26064

2.57 3.20 2.85 5.08 4.30

13.52 23.28 27.03 11.33 14.73

2.89 3.74 4.06 4.45 3.88

a

Tα was determined by rheology in the linear regime and represents the characteristic temperature corresponding to the segmental relaxation time τseg (cf., Figure 2a); Mw, weight-average molecular weight; PDI, polydispersity index; AV, acid value; and OHV, hydroxyl value.

Figure 2. (a) Temperature dependence of the shear modulus (G′) and of the loss factor (tan δ) for P1-α%X bioelastomers. τseg and τstruc indicate the temperature windows corresponding to the segmental and structural relaxations, respectively, while Tα is the characteristic temperature corresponding to τseg. (b) Influence of the concentration of isosorbide [IS] in the polymer (in mmol IS/g of polymer) on Tα for the five linear polyesters (solid lines are guides for the eye). (c) Creep test (strain as a function of time at 25 °C) for P1-α%X bioelastomers.

Because the reaction is a step-growth polymerization, a statistical distribution of chain lengths is obtained, with average molecular weight and chain end-groups controlled by monomer stoichiometry.25 As we targeted carboxyl-functional polymers, a well-defined excess of diacid was introduced in the reaction vessel, and control over polymers molecular weights has been achieved by tuning the monomers stoichiometry. The extent of polymerization and conversion of acid groups were monitored via both GPC and end-groups titration (Table 1). Despite the excess of carboxylic acid groups in the monomer mixture, the OH values of the obtained polyesters are always non-negligible. Koning et al. made similar observations for a series of acidterminated poly(IS succinates) and attributed the residual OH values to the difference in reactivity between the two hydroxyl functionalities of IS, the endo OH being significantly less reactive than its exo counterpart because of intramolecular hydrogen bonding.21 Besides linear polyesters, elastomeric biomaterials with various cross-linking density and gel contents were also prepared by curing the linear polyesters with epoxidized linseed oil (ELO). The detailed mechanism of this curing reaction is highlighted in Figure 1, and has been fully elucidated

melt adhesives, their noncrystallization behavior has been used to design supramolecular rubbers with striking self-healing capacities,19 and their nontoxic character makes them appealing for medical devices.20 Isosorbide has already been utilized in various applications ranging from industrial powder coating resins,21 adhesives,22 and as a bioactive monomer for tissue engineering scaffolds.23 Epoxydised plant oil (soybean, linseed, etc.) is commonly used in coatings, and has also been introduced as a base material for smart printable biomedical scaffolds.24 Linear polyesters with varying molecular weight have been prepared via bulk polycondensation of isosorbide as the only diol, in combination with a C36 dimer, and an optional second diacid (sebacic or succinic acid), leading to the generic polyester structure displayed in Figure 1 and according to the monomer compositions given in Table 1. The proton NMR spectra of polymers P1, P3, and P5 confirm the polymer structures, giving an account of all groups contained in the repeating units, with signals exhibiting the expected area and multiplicity, whereas no trace of any other signal could be detected (Figure 1). 5355


Article

Chemistry of Materials elsewhere.18 Cross-linked formulations are always designated by the codes P1-α%X, where X refers to the wt % of ELO relative to the weight of polymer P1. Influence of Molecular Parameters on Thermomechanical Properties. Mechanical properties of the polyester biomaterials have been studied by monitoring the frequency dependencies of the shear elastic modulus (G′) and loss modulus (G″) in dynamic deformation experiments with sinusoidal signals in the linear viscoelastic regime. Figure 2a reports the evolution of the real shear modulus component (G′) and of the loss factor tan δ (tan δ = G″/G′) as a function of temperature for the P1-α%X series. The maximum of tan δ around −20° appearing for both linear and cross-linked polymers corresponds to a segmental relaxation (τseg) of polymeric units near the glass transition temperature and is therefore a short-range phenomena involving only a few monomers.26,27 Here, we note Tα is the characteristic temperature corresponding to the relaxation time τseg (Figure 2a). In the molten state, the tan δ of P1 increases very rapidly as the temperature is raised (tan δ > 10 at 50 °C), indicating a rapid global flow (G″ > G′) associated with chain disentanglement, but no other relaxation mode could be detected. In view of designing self-adhesive elastomers, the glass transition temperature of the base polymer must be carefully considered. It is generally accepted that an adhesive substance should possess a glass transition temperature at least 30 °C below its application temperature to be tacky (i.e., sticky to the touch).28 Figure 2b illustrates the influence of the IS content on Tα for several ter-polyesters prepared with sebacic and succinic acid. For each series of polyesters, Tα increases almost linearly with the IS content, in accordance with the Fox−Flory theorem. On the basis of Figure 2b, we conclude that only P1 could be considered as a viable candidate for designing elastomers displaying self-adhesive properties at room temperature, whereas the sebacate copolymers (P2 and P3) could be useful for designing pressure-sensitive adhesives operating around 30−40 °C (or around body temperature), and P4 and P5 could play the role of hot-melt adhesives (i.e., solventfree thermoplastic adhesives that are melted by heating before use). Although all of the polyester synthesized herein could be cross-linked with ELO, we therefore decided to select P1 as the base polyester for designing our model sticky bioelastomers in the rest of this study. The viscoelastic profiles of P1-α%X systems tend to superimpose with that of P1 around Tα because stress relaxation at this temperature is mainly controlled by shortrange motions involving only a few monomer units, without depending on topological factors.26 Nevertheless, a slight increase of Tα with cross-linking density could be noticed in Figure 2a, highlighting the loss of mobility of the chains due to the cross-linking process. A limited amount of branching is sufficient to considerably increase the molecular weight of the polyesters. P1-4%X, which displays a tan δ of 1.25 at ambient temperature, is still predominantly liquid (i.e., tan δ > 1). However, a second relaxation mode characteristic of the structural relaxation (τstruc) of polymer chains is now detected around 40 °C. This relaxation is very broad because of the statistical nature of the branching reaction that results in a wide distribution of molecular weight and branched structures. The more crosslinked networks P1-5%X, P1-7.5%X, and P1-10%X display tan δ values below 1 over the whole temperature range and are

therefore mainly elastic (G″ < G′). In that case, the broad structural relaxation does not solely correspond to the relaxation of free chains, but to the combined relaxation of all constitutive elements of the gel (percolated and dandling chains, networks defects, and sol fraction). The fact that τstruc is shifted to lower temperatures as the cross-linking degree increases indicates that the average molecular weight between cross-linked (known as Mc) is lower for the network P1-10%X than for the lightly cross-linked network P1-5%X, because molecular conformations relax with faster modes as the macromolecular size decreases.29 In addition to the property of adhesiveness, useful PSA must also possess an elastic cohesiveness and a resistance to flow under shear (known as resistance to creep). In Figure 2c, a model creep test has been done for the system P1-α%X with various cross-linking levels. The strain percentage of recovery after creep is related to the density of elastically active chain in the bioelastomer (and therefore to G′) whether the lost percentage is related to the energy dissipated by the system (in the form of heat) during the creep phase. The extent of strain after applying a stress for 10 min is the lowest for the system P1-10%X, confirming the more solid-like character of this bioelastomer. An interesting difference is observed between the two loose networks P1-4%X and P1-5%X. Although P1-5%X displays a good percentage of recovery after creep typical of viscoelastic polymers, P1-4%X shows a more viscous character with a high strain and very low recovery. Physical Appearances, Form Factors, and Adhesion Profiles. The conclusions drawn from the above rheological analysis are nicely complemented by carefully observing the physical aspects of the various polymers (Figure 3). Starting with the linear polyesters, we note that P1 is a viscous honeylike and sticky material lacking macroscopic cohesion and ultimately flowing like a liquid. P2 and P3 display an increased cohesion at room temperature as compared to P1, together with a pasty-like appearance and lower tackiness levels. P4 is a nonsticky flexible thermoplastic, while P5 is a hard and glassy material. Peculiar characteristics of all linear poly(isosorbide fattyalkylates) in their molten state include a strong adhesion on various substrates (metal, plastics, human skin, etc.), and an inherent ability to form long legs (fibrillar structure) upon debonding of the adhesive join. To illustrate this surprising fiber-forming character, we created long and well-defined fibers from P5 (Figure 3) as follows. P5 was first heated to 85 °C on a heated plate. A fiber thread was then initiated by placing the tip of a needle into the melt and by sticking the initiated thread onto the edge of a rotating cylinder (150 rpm). With this simple setup, we could obtain fibers up to 40 m long and with a diameter of 30 μm. These spaghetti-like fibers were nonsticky at room temperature and could be used to tie knots similar to sutures. To some extent, one can consider that these P5 fibers correspond to the fibrils observed during the debonding of branched structure such as P1-2%X, but in a frozen state (i.e., below the polyester Tg). Cross-linking density has a profound impact on the physical properties of P1-α%X elastomers, and especially on their adhesion behaviors (see examples in Figure 3). Adhesion strength was investigated by laminating the elastomers precursors on a PET film, curing this bilayer stack in an oven, and then measuring the adhesion properties of the resulting tape using a standard 180° peel test. Special attention was given to the failure mode of each specimens by monitoring 5356


Article


residues on the substrates, or interfacial failure mode associated with a clean removal), as reported in Table 2. From 0% to 3.5% of cross-linking, the adhesion strength increases gradually from 110 to 1376 cN/20 mm, respectively, while the failure mode is characterized by large fibrils that fail in a cohesive manner at large extension. Similar observations are made for P1-4%X, although the adhesion strength is already slightly decreasing. P1-5%X displays an interesting behavior, combining rather high adhesion strength (712 cN/20 mm), a clear fibrillation process, and a clean removal where the feet of the extended fibrils ultimately detached interfacially from the substrate without leaving residues. Finally, the adhesion strength decreases from 7.5% to 15% of cross-linking to reach the lowest adhesion level of 143 cN/20 mm. The more cross-linked networks do not display any observable fibrils during the peeling process and always fail cohesively. Viscoelastic Phase Diagram and Adhesion Archetypes. In this section, we performed frequency sweep rheological measurements in the linear viscoelastic regime, analyzed the results with frameworks initially developed for pressure-sensitive adhesives, and constructed a generic viscoelastic phase diagram that could depict the main adhesive archetypes by simply using results obtained from rheological measurements in the linear regime. To that end, a set of three criteria has been determined. (i) At first, the so-called Dahlquist criteria stipulates that the shear elastic modulus G′ at the bonding frequency should be lower than 0.1 MPa (or 105 Pa) for the adhesive layer to be able to form a good contact with the substrate.30 This means that at the bonding frequency, all materials showing stickiness at room temperature should display viscoelastic properties on the left side of the vertical curve corresponding to G′ = 105 Pa (see Figure 4). Below this threshold, the debonding process will result from the complex interplay between the bulk and the interfacial properties of the biomaterials layer. (ii) Second, the bonding strength of a soft biomaterial could be largely influenced by the progressive development of a fibril structure arising from the vertical expansion of cavitation bubbles.31 To define a relevant fibrillation criterion, let us quickly get back to some basics of adhesion theory. The growth of a defect initially present at the interface is governed by two competitive parameters: the interfacial growth of a crack, which is controlled by the critical energy-release rate (noted GC), and the bulk deformation of the glue layer, which is controlled by the average stress within the glue layer and which is mainly determined by the elastic modulus of the adhesive (noted E).

Figure 3. Physical appearance of bioelastomers. P1 is a viscous honeylike liquid. P1-2%X is a very tacky substance that lacks macroscopic cohesion. P1-5%X is a sticky material with good cohesion: a rectangular piece can be folded in its middle, stuck into the shape of a farfalle, and unfolded again in a reversible manner. P1-7.5%X is a rubbery material with limited tack: it can be twisted but does not display self-adhesion. P1-10%X behaves like a rubber: because of its nontacky character, a formulation cured around a metallic wire creates a material that can subsequently slip along the wire to give a hollowtube (Øinternal = 1 mm). P5 is a hard, glassy material at rt and can be processed at elevated temperature in the form of nonsticky fibers with adjustable diameter (here Ø = 30 μm) and sufficient flexibility to tie knots. The color of the circles on each micrograph corresponds to the adhesive archetypes of Figure 4.

the occurrence of fibrils during the peeling process as well as the failure mode of the glue layer (cohesive failure leaving

Table 2. Physical Characterization of P1-α%X Bioelastomers and Biogels gel structure

adhesion properties and failure scenarioa

P1-α%X

gel (%)

peel strength (N/m)

fibril

failure mode

0%X 2%X 3.5%X 4%X 5%X 7.5%X 10%X 15%X

0 1.5 1.8 5.6 48 76.5 86.6 81.6

∼55 131 688 611 356 178 135 71.5

yes yes yes yes yes no no no

COH COH COH COH ITF ITF ITF ITF

fitting parameters for the large strain modelc

rheology at small strainb −5

G′ (Pa) ×10

−1

tan δ/G′ (Pa ) ×10

0.01

959.98

0.13 0.13 0.37 0.85 2.44 2.48

13.01 11.15 2.31 0.65 0.10 0.11

5

Gv (kPa)

Ge (kPa)

Jm

De

49.9 55.6 94.8 122

1.1 6.3 15.8 63

3170 1500 195 71

0.23 0.21 0.13 0.29

a Adhesive strength was measured by a 180 °C peel test from BA Steel substrates (the bioelastomeric glue layer had a thickness of 20 μm). COH and ITF refer to cohesive and interfacial failure modes, respectively. bValues measured at 0.10 Hz (see the full rheological profiles in Figure 4). cSee model description in eq 4 and the resulting best fits in Figure 7.

5357


Article


(iii) Finally, a third and last criterion must be defined for finalizing our adhesion phase diagram. Once the fibrils are formed, two options could still be considered: a clean interfacial failure mode in which the foot of the fibrils will ultimately peel off from the substrate in a clean manner (i.e., without leaving residues), and a cohesive failure mode where the fibrils will break due to thinning of their central section.28,31,34 A key parameter here is the amount of external work energy that could be stored elastically within the fibrils. We assume that a biomaterial should possess a sufficient level of elasticity for detaching in the interfacial mode, and therefore cannot be liquid. We thus use tan δ to assess its relative elasticity. Below the tan δ = 1 line of Figure 4, the material is more elastic (G″ < G′) and is expected to form fibrils that are sufficiently robust to ultimately peel off interfacially, while above this line, more liquid (G″ > G′) gels will lead to global flow associated with cohesive failure. On the basis of the three criteria described above, we established the adhesion phase diagram of Figure 4 by plotting tan δ/G′ against G′. Five distinct areas could be distinguished on that diagram. On the adhesive side of the graph (i.e., on the left of the Dahlquist line), three adhesive archetypes could be distinguished from top to bottom: sticky liquid with uncohesive fibrils, adhesive with fibrils and clean removal, and removable adhesive without fibrils, respectively. It is interesting to note at this point that the failure modes experimentally observed for the linear polymers and the P1-α%X networks and reported in Table 2 correlate nicely with the adhesive archetypes proposed in Figure 4 in the frequency range 0.01−1 Hz, confirming the relevance of this simple, yet practical, representation (the lower frequencies corresponds to the establishment of the adhesive bond, while the higher frequency range relates to the more dynamic debonding phase). However, despite the good match obtained between the viscoelastic diagram of Figure 4 and our experimental observations, it is important to keep in mind that the approach consisting of using small-strain parameters (such as tan δ) to predict large-strain phenomena (like the detachment/breakup of fibrils) does not always work and could present some limitations. In the present case, it works because the bioelastomers are homogeneous and only weakly nonlinear. To overcome these shortcomings, Creton and Ciccotti recently introduced a new adhesion diagram built upon two large-strain parameters, Chard and Csoft, both obtained from the Mooney representation of the stress−strain curves and corresponding to the cross-linking and entanglements densities, respectively.35 Although this technique may require a slightly more elaborated data analysis as compared to easily obtainable rheological data, it is expected to be more widely applicable because the resulting adhesion diagram incorporates nonlinear elements.35 Micromechanical Effects at Small Strain. The establishment and development of an adhesive bond between a soft biomaterial and a substrate is a complicated phenomenon involving various parameters such as the rheological profile of the elastomer and the increased interfacial energy resulting from macromolecular rearrangements and the development of a 2D network of noncovalent bonds near the interface.36 Because many substrates can be quite rough at the mesoscale, it is essential for the material to be sufficiently flexible and compliant to establish an intimate bond with the substrate (cf., the Dahlquist criterion).28,30 This partly explained why an elastic and rubbery sample (P1-10%X) displays a much lower adhesion than a more viscoelastic specimen (P1-4%X), because

Figure 4. Generic viscoelastic diagram of states and key adhesion archetypes obtained by plotting the evolution of tan δ/G′ versus G′ for a selection of linear and cross-linked polyester biomaterials over the frequency range 0.01−1 Hz (the frequency increases from left to right for each measurement). The three criteria used for defining the diagram are indicated, as well as the resulting material phases (see color code for reference) and a scheme of the failure mode for each adhesive archetype.

For elastic layers, the ratio GC/E could be used as a fairly accurate predictor for the formation of fibrils before final detachments.32 For simple elastomers showing relatively weak adhesion mainly attributed to van der Waals forces, GC/E could be linked to routinely obtained viscoelastic parameters by using eq 1:33 ⎡ G″ ⎤ ⎛ GC ⎞ ⎛ tan δ ⎞ ⎟ = k·⎢ ⎜ ⎟ ≈ k·⎜ ⎥ ⎝ G′ ⎠ ⎝ E ⎠ ⎣ (G′)2 ⎦

(1)

Following this insightful approach, one can now establish a quantitative criterion for the formation, or not, of fibrils using only results obtained from rheological measurements in the linear regime. Typically, the lower is tan δ/G′, the more crack propagation is favored and the sooner the detachment is expected to occur. On the other hand, higher values of tan δ/G′ will indicate the increasing tendency of the polyester biomaterial to form fibrils. Creton et al. established experimentally that for general PSA, a threshold value of tan δ/G′ = 105 Pa−1 could be taken as a good approximation for predicting the formation of fibrils.33 Practically, one can consider that adhesives showing a value of tan δ/G′ above the horizontal line tan δ/G′ = 105 Pa−1 in Figure 4 will display fibrillation upon debonding, while below this limit, adhesives will detach without forming fibrils. 5358


Article


was prepared by spin-coating a toluene solution on a glass substrate, dried at room temperature, and then transferred to the surface of a water bath by the flotation technique. Small pieces of elastomers were cut, heated at 80 °C for 30 min, and the floating PS film was then stuck on the bioelastomer surface. After the system was dried and relaxed at room temperature for several hours, wrinkled patterns such as that displayed in Figure 5b eventually appeared. Measurements of the wrinkles wavelengths λ were performed by analyzing optical micrographs. Buckling patterns were observed for all elastomers above 5%X, which indicates that the PS film is sensing a predominantly elastic (solid-like) response from these elastomers. The highest λ (i.e., the lowest Es) was obtained for P1-5%X, whereas the lowest λ (i.e., the highest Es) was obtained for P1-10%X (Figure 5c). Quantitative analysis of the results using eq 2 is however hazardous, because our networks are viscoelastic and not purely elastic, and consequently λ may also be influenced by other relaxation processes such as creep. Surprisingly, no wrinkles were observed for cross-linking degrees below 5%, even if P1-4%X and P1-5%X have quite similar shear modulus. This observation clearly denotes that radically different relaxation modes (such as diffusion or reptation) should operate for viscous liquid-like polyesters displaying tan δ > 1,40 and indirectly supports the viscoelastic phase diagram of Figure 4 (i.e., viscoelastic liquids fail cohesively due to their insufficient elastic component). The micromechanical properties of soft bioelastomers near interfaces are not only important in the framework of adhesion science but also for biotechnological applications: it is, for instance, known that the viscoelasticity of a biomaterial has a direct impact on its interaction with cells and on its biological activity in vivo and in vitro.41 Nonlinear Viscoelasticity at Large Strain. Although the viscoelastic diagram obtained at small strain could qualitatively describe the debonding scenario of the elastomers (including the occurrence of a fibrillar structure), it does not provide quantitative information on their large-strain (nonlinear) mechanical behavior occurring during the debonding phase.42 Therefore, tensile experiments are useful to provide this information. The tensile curves of P1-α%X elastomers and of linear succinate copolyesters (P4 and P5) are shown in Figure 6. This figure highlights the remarkable diversity of tensile behaviors that could be obtained by combining a rather limited set of monomers.

of its inability to sufficiently interlock with the substrate (Figure 5a).

Figure 5. Mesoscale interlocking and buckling phenomena at soft biomaterial interfaces. (a) Representation of the adhesion interfaces between bioelastomers P1-4%X and P1-10%X with a solid substrate (the black areas represent voids). (b) Schematic cross-section of the buckling experiment and typical optical micrograph obtained after depositing a PS nanofilm deposited on a heated piece of P1-10%X and cooling to room temperature (black triangles indicate the strain direction). (c) Influence of the bioelastomer cross-linking density on λ.

To further characterize the interfacial mechanical properties of our bioelastomers, we have used an elegant technique introduced by Stafford and called strain-induced elastic buckling instability for mechanical measurements (SIEBIMM).37 Its basic principle is that an ultrathin and stiff film deposited on an elastomeric substrate of much lower modulus will undergo a buckling instability when subjected to planar compressive forces to relieve and minimize the strain energy of the system.38 This technique could be used to measure the modulus of a thin film using a calibrated elastomer, or for measuring the elastomer properties using a film of known modulus. In that later case, buckling takes place at a wavelength λ, given by eq 2, where β is the nanofilm thickness, and Ef, νf, Es, and νs are the Young’s moduli and Poisson’s ratios of the film and substrate, respectively.39 Es =

Ef ·(1 − νgel 2) ⎡ 2·π ·β ⎤3 ·⎢ ⎥ 3·(1 − νf 2) ⎣ λ ⎦

(2)

We used this inverse SIEBIMM concept to qualitatively explore the surface micromechanical properties of polyester bioelastomers. To do so, an ultrathin film of polystyrene (PS)

Figure 6. Tensile results of cross-linked polymers P1-α%X and of the linear polymers P4 and P5. The ultimate tensile strength (UTS) of P5 exceeds 14 MPa. 5359


Article


intermediate strain softening effect combined with a strain hardening at high extension is a property often observed for conventional pressure-sensitive adhesives (acrylic-based or natural rubber-based).28,43 The tensile curve of a bioelastomer represents its specific fingerprint and encapsulates precious information on its intimate molecular structure. To extract such information, we applied a model of uniaxial stress−strain curve recently introduced by Deplace.44 This constitutive model combined in parallel the Upper Convected Maxwell (UCM) model generally used to describe viscoelastic fluids with the Gent strain-hardening model used to describe the purely elastic deformation of rubbery networks (including the finite extensibility of polymer chains in the network). Prediction of the stress−strain can be obtained using eq 3:

The tensile behavior of P5 has a high tensile modulus (high initial slope of the tensile curve) and a very low strain, which are both expected from brittle and glassy polymers, while P4 displays a more complex thermoplastic behavior. The stress− strain curve of the more cross-linked bioelastomer (P1-10%X) displays a nonlinear shape typical of elastomers, but is also found in vulcanized rubber and even in biological materials such as ligaments.4,5 It is notable that the elongation at break and the ultimate tensile strength of P1-10%X (4.1 and 0.67, respectively) are quite similar to the values reported for PGS (∼3.1 and >0.5, respectively).10 The lightly cross-linked elastomers P1-5%X and P1-4%X both display quite similar tensile moduli at small strain, a very large elongation, and a pronounced strain softening at intermediate elongations (see enlarged views of the tensile curves in Figure 7a). However, one notable difference between

σN(λ) = {σN,v(λ)} + {σN,e(λ)}

(3)

Equation 3 can be further developed into eq 4, where De is the Deborah number (related to the relaxation time of the viscous component), λ is the extension ratio, Jm describes the finite extensibility of the network, and Gv and Ge are the initial shear moduli of the viscoelastic part and of the elastic part of the elastomer, respectively. ⎧⎡⎛ G ·De ⎞ ⎛ ⎛ (1 + De) ⎞⎞ ⎟ · ⎜1 − exp⎜ − ⎢⎜ v σN,v(λ) = ⎨ · (λ − 1)⎟⎟+ ⎪ ⎝ ⎠ ⎝ ⎠⎠ De De + 1 ⎝ ⎩⎣ ⎪

⎛ (1 − 2·De) ⎞⎞⎤ 1 ⎫ ⎛ 2·Gv ·De ⎞ ⎛ ⎟ · ⎜1 − exp⎜ − ·(λ − 1)⎟⎟⎥ · ⎬ ⎝ 1 − 2·De ⎠ ⎝ ⎝ ⎠⎠⎦ λ ⎪ De ⎭ ⎪

⎜

⎧⎡⎛ ⎪ ⎪⎢⎜ + ⎨⎢⎜ ⎪⎢⎜⎜ 1 − ⎪⎣⎢⎝ ⎩

Ge

(

λ 2 + 2·λ−1 − 3 Jm

)

⎤ ⎫ ⎞ ⎥ ⎪ ⎟⎛ ⎞ 1 1⎪ ⎟ ·⎜λ 2 − ⎟⎥ · ⎬ ⎝ ⎠ ⎥ λ λ⎪ ⎟⎟ ⎠ ⎦⎥ ⎪ ⎭ (4)

The tensile curves for the three adhesive archetypes identified in Figure 4, sticky liquid with cohesive fibrils (P14%X), adhesive with interfacial fibrils (P1-5%X), and removable adhesive without fibrils (P1-7.5%X), have been fitted with eq 4 using the set of adjustable parameters given in Table 2. Very satisfying fits could be obtained for these three tensile curves (Figure 7). The sum (Ge + Gv) increases with the cross-linking density, in line with the assumption that more cross-linked networks will have a higher tensile modulus. The tensile modulus of the bioelastomer is linked to Gv and Ge as follows: E ≈ 3 × [Ge + Gv]. The molecular weight between cross-links could thus be obtained using eq 5:26,27

Figure 7. (a) Enlarged views of the tensile curves of the three adhesive archetypes (P1-4%X, P1-5%X, and P1-7.5%X), and best fits obtained using the nonlinear model of eq 4 (parameters of the model are given in Table 2). The color of the circles corresponds to the adhesive archetypes described in Figure 4. (b) Results of (a) represented as Mooney stress σR as a function of 1/λ.

⎡ ρ·R·T ⎤ Mc = ⎢ ⎥ ⎣ (Gv + Ge) ⎦

(5)

where R is the universal gas constant, T is the temperature, and ρ is the density. We obtained Mc values of 50 135, 28 059, and 16 080 for P1-4%X, P1-5%X, and P1-7.5%X, respectively. As a comparison, a Mc value of 18 300 was reported for PGS,10 which indicates that sticky elastomers should possess much larger Mc than purely rubbery networks. An important parameter in this model is the ratio Gv/Ge, which describes how dissipative is the bioelastomer. Typically, Gv/Ge increases while the cross-linking density decreases (from 1.9 for 7.5%X, to 6.0 for 5%X and 8.8 for 4%X), which indicates that the softening at intermediate strain during a tensile test is

these two networks is the strain hardening effect observed with P1-5%X, but not with P1-4%X. This nonlinear strain-hardening behavior of P1-5%X appears more clearly using the Mooney stress representation of Figure 7b, where the measured stress is normalized by the expected behavior of a neo-Hookean rubber in uniaxial extension (with σR = σN/[λ − (1/λ2)]). In fact, the strain hardening is now indicated by a clear minimum in the 1/ λ representation at high extension (i.e., on the left side of the graph). In contrast, viscoelastic liquids such as P1-4%X are characterized by the absence of a well-defined minimum. An 5360


Article


test item was determined, the percentage of biodegradation can be calculated as the percentage of solid carbon, which has been converted to gaseous, mineral carbon under the form of CO2. Finally, the relative digestion (noted RD%) of the test item was obtained by comparing the degradation of the sample with the degradation of a cellulose standard. Figure 8 displays the RD% of linear polyesters P1, P2, and P3 after 25 and 45 days of incubation. Despite the rather similar molecular weights of these specimens, clear compositional effects could be seen after only 45 days of incubation: P3 was almost completely (>95%) digested, one-half of P2 (51,9%) was composted, whereas P1 only showed a moderate RD% (14.5%). P1 has a higher hydrophobic character than P3 due to its higher concentration of C36 fatty acids and its lower concentrations of isosorbide and ester bonds. Consequently, P1 displays much slower degradation kinetics than P3, whereas P2 has an intermediate behavior. At this stage, Figure 8 does not allow us to draw definitive conclusions regarding the final compostability of these polyesters (i.e., the norm EN 13432 states that >90% of the test item should be degrade within a period of 180 days); however, it clearly shows that the biodegradation kinetics could be tailored to a large extent by tuning the monomer composition of the polyester chains.45,46 Topological effects also play a role in the biodegradation process. P1-10%X shows RD% of 2.8% and 8.3% after 25 and 45 days, respectively. These values correspond to 37% and 57% of the RD% of the corresponding linear chain P1 for similar incubation periods. This delayed degradation could be explained by the higher molecular weight of P1-10%X (chainextended chains, branched structures, and macroscopic percolated cluster of infinite molecular weight) as compared to P1. Overall, these results demonstrate that biodegradability could be integrated as an additional feature to sticky bioelastomers, and that the degradation kinetics could be largely tailored by modulating molecular parameters such as the monomer composition and the cross-linking structure of the bioelastomer.

more pronounced for lightly cross-linked and gives a sign of increasingly good adhesion. Another parameter is Jm, which describes the finite extensibility of the network. A low enough value of Jm is necessary to have a well cross-linked network and thus a clean debonding without residues during a tack test. The cross-linking affects significantly the value of Jm (from 71 for 7.5%X, to 1500 for 4%X, and even 3170 for 3.5%X). Finally, the model of eq 4 also attributes a value to De. This parameter does not drastically influence the fitting, and its physical interpretation is quite difficult because real polymeric networks have several relaxation times and not only one (as indicated by the large width of τstruc on Figure 2). Influence of Compositional and Topological Effects on Biodegradation. Natural adhesives produced by animals or plants are intrinsically sticky and biobased, but also biodegradable. To design synthetic adhesives that also have the faculty to biodegrade, one needs to anticipate the expected degradation environments: aqueous solutions (buffer, effluents, etc.), human body, soil, etc. Up to now, the degradability of polyester bioelastomers has been mainly investigated in the framework of biotechnological applications.10−16 Triggered by the potential industrial applications of sticky degradable bioelastomers, we have evaluated here the degradation of our bioelastomers in compost environment. The controlled composting biodegradation test is an optimized simulation of an intensive composting process where the biodegradability is determined under aerobic conditions. The polymer was finely mixed with the inoculum and introduced into static reactor vessels where it was extensively composted under optimum oxygen, temperature (∼60 °C), and moisture conditions.45 The inoculum consisted of stabilized mature compost derived from the organic fraction of municipal waste. During the aerobic biodegradation of organic materials, a mixture of gases (mainly CO2 and H2O) is the final decomposition product, while part of the organic material will be assimilated for cell growth (Figure 8). The CO2 production is monitored and integrated to determine the cumulative CO2 production. After the carbon content of the

■

CONCLUSION In essence, we have demonstrated by using model biopolyesters of various molecular composition and structure that the stickiness of amorphous, low Tg, bioelastomers is mainly determined by bulk rheological factors and not by the adhesive character of residual functional groups. All of the molecular parameters (compositional and topological) that have been shown to influence the viscoelacticity and adhesion profiles of a bioelastomer are also the ones that will determine its biodegradation, which implies that the design of sticky degradable bioelastomer with tailored adhesiveness (high or low, with or without fibrils) and biodegradation (fast or slow, complete or partial) should be conducted in an integrative manner. Beyond isosorbide-based polyesters, the conclusions drawn in this Article could be generalized to a wide variety of biodegradable elastomers synthesized so far (including PGS and its analogues). It is important to study how the viscoelastic nature of apparently soft and weak bioelastomeric substances is allowing them to establish surprisingly strong adhesive bonds via efficient energy dissipation mechanisms such as cavitation and fibrillation. The possibility to synergistically control the viscoelastic, adhesive, and biodegradability profiles of soft polyester biomaterials by fine-tuning their molecular structures

Figure 8. Early stage aerobic digestion by compost microorganisms. Compost biodegradation stems from the hydrolytic depolymerization of the chains followed by the aerobic digestion of the monomers by various microorganisms (bacteria, fungi, etc.). The charts report the relative digestion (RD%) for P1, P2, P3, and P1-10%X, after incubation periods of 25 and 45 days, respectively. 5361


Article


Rheology at Small Strain. Dynamic mechanical analysis (DMA) was performed using a mechanical spectrometer AR 2000 rheometer (TA Instruments) in parallel plate geometry (8 mm diameter and 0.5 mm gap) in dynamic mode with a strain of 0.01. Creep tests were obtained in a similar geometry. A shear stress of 1000 Pa was instantaneously applied to the gel, and the resultant strain was monitored as a function of time. After 10 min of creep, the shear stress was removed, while the strain was still monitored during a relaxation period of 10 min. Strain-Induced Elastic Buckling Instability for Mechanical Measurements. An ultrathin (160 nm thick) polystyrene film was prepared by spin-coating (2000 rpm for 15 s) on a clean glass substrate. The thickness was monitored by scratching the film with a scalpel on several locations and measuring the profile by atomic force microscopy. The modulus of the PS film was calibrated using a poly(dimethylsiloxane) (PDMS) reference elastomer (Sylgard 184 from Dow Corning, prepared in a ratio of 10:1 by mass of base to curing agent, and with a uniaxial compression modulus of 1.2 MPa). PS films were transferred in a water bath by the flotation technique before being stuck to the surface of a preheated bioelastomer. Because of the adhesive nature of our elastomers, the PS films were always firmly attached to the surfaces, and no delamination or cracks occurred. The system was then dried dry at room temperature for several hours, and the patterns were analyzed with images acquired on an optical microscope equipped with a digital camera. Pressure-Sensitive Adhesion. Reactive formulations were diluted with ethyl acetate (solid base 70 wt %) and roll-coated manually on a 50 μm thick polyethylene terephtalate (PET) base film, and the bilayer stack was thermally cured. Dry thickness of glue layers was 20 ± 2 μm. After curing, the adhesive surfaces were covered with an antiadhesive siliconized paper. The adhesive strength has been investigated with a 180° peel test as follows: 2 cm wide, 10 cm long stripes were cut and placed on a table (adhesive face on top). One-half of the tape length is covered with a second 50 μm-thick PET film, and this assembly is turned upside down. The adhesive area is then applied on the reference surface, and a 2 kg cylinder is rolled twice on the tape. Peel tests were conducted on a Zwick Z005 testing machine at a constant peeling speed of 300 mm/min after a dwell time of 15 min. Tensile Testing. Tensile tests were conducted on a Zwick Z005 testing machine with a speed of 300 mm/min. The initial distance between grips was 10 mm, and the initial cross-section of the cured adhesive sheets was 10 × 0.8 mm. Measurements were done in triplicate. Biodegradation. Degradation studies in controlled compost conditions were performed by the company Organic Waste Systems N.V. (Ghent, Belgium) according to the norm EN 13432.

from the macroscopic level down to the molecular level is opening some appealing perspectives in various fields. At first, the industrial use of polyester derived from renewable resources is an attractive way to improve the carbon footprint of the chemical and adhesive industries by decoupling polymer production from depleting fossil feedstock.47 Additionally, biodegradation and/or composting of industrial adhesives and rubbers could contribute to a circular economy by offering more sustainable end-of-life alternatives to various consumer goods. It is thus not a surprise that the promotion of renewable resources and that of biodegradable polymers are two key recommendations mentioned in the latest report from the Ellen Macarthur Foundation “The New Plastics Economy: Rethinking the Future of Plastics”. In the medical arena, there are still unmet needs for skinfriendly adhesives that strongly attach to the skin but also allow for a gentle and trauma-free removal, especially for patients with sensitive skins such as new-borns.48 Polyester chemistry may also provide an alternative for patients suffering from allergy to the widely used poly(acrylates) skin adhesives. In the field of surgery, a recent study of Vakalopoulos demonstrated that the viscoelasticity of commercial surgical glues in their cured states could be correlated to some extent with their adhesive strength. 49 Del Nido and Karp developed a photocurable and biodegradable surgical glue based on acrylated PGS, and demonstrated that physical interlocking with the substrate (such a beating heart) combined with an optimal degree of cross-linking were essential to optimize the adhesion level.50 Despite these advances, all of the surgical glues developed so far are mostly solid-like in their cured state, showing only a very limited dissipative character. We anticipate that introducing more pronounced viscous elements in surgical glues could lead to stronger adhesive bonds between soft tissues thanks to the establishment of energy dissipations mechanisms promoting fibrillation instead of crack propagation. Sticky degradable bioelastomers could thus pave the way to solutions and products that stick to the triple bottom line of sustainable development: social (people), ecological (planet), and economic (profit).

■

■

EXPERIMENTAL SECTION

Materials. High-purity isosorbide (99.5%+) was obtained from Roquette Frères (Lestrem, France). The fully hydrogenated dimer fatty acid (Radia 0975) was supplied from Oleon (Ertvelde, Belgium). Epoxidized linseed oil (ELO) was obtained from Cognis (Düsseldorf, Germany) under the trade name Dehysol B376. The catalyst for the epoxy-carboxy curing reaction (Nacure XC-9206) was supplied from King Industries, Inc. (Norwalk, Connecticut). All chemicals were used as received without further purification. Preparation of Poly(IS fatty alkylates). Linear polyesters were prepared according to the polycondensation procedure described earlier,18 according to monomer feed composition given in Table 1. Branched and cross-linked specimens were prepared as follows: predetermined amounts of P1, ELO, and catalyst (Nacure XC-9206) were homogenized, and the resulting mixture was poured into Teflon molds and then cured at 155 °C for 45 min. Adhesive tapes were also prepared in view of the adhesion measurements. Analytics. The average molecular weights of the fatty acid-based polymers were determined by a gel permeation chromatography (GPC) instrument (Waters model pump 515 and Waters 2414 refractive index detector) with styragel columns relative to polystyrene (PS) standards using tetrahydrofuran (THF) as eluent. The acid and hydroxyl values of the polymers were monitored via standard titration methods as previously described.18

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Richard Vendamme: 0000-0002-8931-0851 Notes

The authors declare no competing financial interest.

■

REFERENCES

(1) Zhao, Q.; Lee, D. W.; Ahn, B. K.; Seo, S.; Kaufman, Y.; Israelachvili, J. N.; Waite, J. H. Underwater contact adhesion and microarchitecture in polyelectrolyte complexes actuated by solvent exchange. Nat. Mater. 2016, 15, 407−412. (2) Sahni, V.; Blackledge, T. A.; Dhinojwala, A. Viscoelastic solids explain spider web stickiness. Nat. Commun. 2010, 1, 19. (3) King, M. Viscoelastic properties of airway mucus. Fed. Proc. 1980, 39, 3080−3085. (4) Pioletti, D. P.; Rakotomanana, L. R. On the independence of time and strain effects in the stress relaxation of ligaments and tendons. J. Biomech. 2000, 33, 1729−1732.

5362


Article


(28) Satas, D. Handbook of Pressure Sensitive Adhesive Technology; Springer: New York, 1989. (29) Urayama, K.; Miki, T.; Takigawa, T.; Kohjiya, S. Damping Elastomer Based on Model Irregular Networks of End-Linked Poly(Dimethylsiloxane). Chem. Mater. 2004, 16, 173−178. (30) Dahlquist, C. A. An investigation into the nature of tack. Adhes. Age. 1959, 2, 25. (31) Villey, R.; Cortet, P. P.; Creton, C.; Ciccotti, M. In-situ measurement of the large strain response of the fibrillar debonding region during the steady peeling of pressure sensitive adhesives. Int. J. Fract. 2017, 204, 175. (32) Crosby, A. J.; Shull, K. R.; Lakrout, H.; Creton, C. Deformation and failure modes of adhesively bonded elastic layers. J. Appl. Phys. 2000, 88, 2956−2966. (33) Deplace, F.; Carelli, C.; Mariot, S.; Retsos, H.; Chateauminois, A.; Ouzineb, K.; Creton, C. Fine tuning the adhesive properties of a soft nanostructured adhesive with rheological measurements. J. Adhes. 2009, 85, 18−54. (34) Vendamme, R.; Olaerts, K.; Gomes, M.; Degens, M.; Shigematsu, T.; Eevers, W. Interplay between viscoelastic and chemical tunings in fatty-acid-based polyester adhesives: engineering biomass toward functionalized step-growth polymers and soft networks. Biomacromolecules 2012, 13, 1933−1944. (35) Creton, C.; Ciccotti, M. Fracture and adhesion of soft materials: a review. Rep. Prog. Phys. 2016, 79, 046601. (36) Feldstein, M. M.; Siegel, R. A. Molecular and nanoscale factors governing pressure-sensitive adhesion strength of viscoelastic polymers. J. Polym. Sci., Part B: Polym. Phys. 2012, 50, 739−772. (37) Stafford, C. M.; et al. A buckling-based metrology for measuring the elastic moduli of polymeric thin films. Nat. Mater. 2004, 3, 545− 550. (38) Vendamme, R.; Ohzono, T.; Nakao, A.; Shimomura, M.; Kunitake, T. Synthesis and Micromechanical Properties of Flexible, Self-Supporting Polymer-SiO2 Nanofilms. Langmuir 2007, 23, 2792− 2799. (39) Wilder, E.; Guo, S.; Lin-Gibson, S.; Fasolka, M. J.; Stafford, C. M. Measuring the modulus of soft polymer networks via a bucklingbased metrology. Macromolecules 2006, 39, 4138−4143. (40) Ferry, J. D. Viscoelastic Properties of Polymers, 3rd ed.; John Wiley & Sons: New York, 1980. (41) Gobaa, S.; Hoehnel, S.; Lutolf, M. P. Substrate elasticity modulates the responsiveness of mesenchymal stem cells to commitment cues. Integr. Biol. 2015, 7, 1135−1142. (42) Zosel, A. Effect of Cross-Linking on Tack and Peel Strength of Polymers. J. Adhes. 1991, 34, 201−209. (43) Creton, C. Pressure-Sensitive Adhesives: An Introductory Course. MRS Bull. 2003, 28, 434−439. (44) Deplace, F.; Rabjohns, M. A.; Yamaguchi, T.; Foster, A. B.; Carelli, C.; Lei, C. H.; Ouzineb, K.; Keddie, J. L.; Lovell, P. A.; Creton, C. Deformation and adhesion of a periodic soft-soft nanocomposite designed with structured polymer colloid particles. Soft Matter 2009, 5, 1440−1447. (45) Doi, Y. F.; Fukuda, K. Biodegradable Plastics and Polymers; Elsevier: Netherlands, 1994. (46) Tuominen, J.; Kylma, J.; Kapanen, A.; Venelampi, O.; Itävaara, M.; Seppäla, J. Biodegradation of Lactic Acid Based Polymers under Controlled Composting Conditions and Evaluation of the Ecotoxicological Impact. Biomacromolecules 2002, 3, 445−455. (47) Belgacem, M. N.; Gandini, A. Monomers, Polymers and Composites from Renewable Resources; Elsevier: Oxford, 2008. (48) Laulicht, B.; Langer, R.; Karp, J. M. Quick-release medical tape. Proc. Natl. Acad. Sci. U. S. A. 2012, 109, 18803−18808. (49) Vakalopoulos, K. A.; Wu, Z.; Kroese, L.; Kleinrensink, G. J.; Jeekel, J.; Vendamme, R.; Dodou, D.; Lange, J. F. Mechanical strength and rheological properties of tissue adhesives with regard to colorectal anastomosis: an ex vivo study. Ann. Surg. 2015, 261, 323−331. (50) Lang, N.; et al. A blood-resistant surgical glue for minimally invasive repair of vessels and heart defects. Sci. Transl. Med. 2014, 6, 218.

(5) Storm, C.; Pastore, J. J.; MancKintosh, F. C.; Lubensky, T. C.; Janmey, P. A. Nonlinear elasticity in biological gels. Nature 2005, 435, 191−194. (6) Khademhosseini, A.; Vacanti, J. P.; Langer, R. Progress in tissue engineering. Sci. Am. 2009, 300, 64−71. (7) Flory, P. J. Principles of Polymer Chemistry; Cornell University Press: Ithaca, NY, 1953. (8) Albertsson, A. C. Degradable Aliphatic Polyesters; Springer: Berlin, 2002. (9) Vendamme, R.; Schüwer, N.; Eevers, W. Recent synthetic approaches and emerging bio-inspired strategies for the development of sustainable pressure-sensitive adhesives derived from renewable building blocks. J. Appl. Polym. Sci. 2014, 131, 40669. (10) Wang, Y.; Ameer, G. A.; Sheppard, B. J.; Langer, R. A tough biodegradable elastomer. Nat. Biotechnol. 2002, 20, 602−606. (11) Bettinger, C. J. Synthetic biodegradable elastomers for drug delivery and tissue engineering. Pure Appl. Chem. 2011, 83, 9−24. (12) Serrano, M. C.; Chung, E. J.; Ameer, G. A. Advances and Applications of Biodegradable Elastomers in Regenerative Medicine. Adv. Funct. Mater. 2010, 20, 192−208. (13) Bruggeman, J. P.; de Bruin, B. J.; Bettinger, C. J.; Langer, R. Biodegradable poly(polyol sebacate) polymers. Biomaterials 2008, 29, 4726−4735. (14) Pereira, M. J. N.; et al. A highly tunable biocompatible and multifunctional biodegradable elastomer. Adv. Mater. 2013, 25, 1209− 1215. (15) Mizrahi, B.; Shankarappa, S. A.; Hickey, J. M.; Dohlman, J. C.; Timko, B. P.; Whitehead, K. A.; Lee, J. J.; Langer, R.; Anderson, D. G.; Kohane, D. S. A Stiff Injectable Biodegradable Elastomer. Adv. Funct. Mater. 2013, 23, 1527−1533. (16) Wang, Y.; Ameer, G.; Langer, R. S. Biodegradable polymer. WO 2003064496 A3 assigned to Massachusetts Institute of Technology, 2003. (17) Li, X.; Hong, A. T. L.; Naskar, N.; Chung, H. J. Criteria for Quick and Consistent Synthesis of Poly(glycerol sebacate) for Tailored Mechanical Properties. Biomacromolecules 2015, 16, 1525− 1533. (18) Vendamme, R.; Eevers, W. Sweet solution for sticky problems: chemoreological design of self-adhesive gel materials derived from lipid biofeedstocks and adhesion tailoring via incorporation of isosorbide. Macromolecules 2013, 46, 3395−3405. (19) Cordier, P.; Tournilhac, F.; Soulié-Ziakovic, C.; Leibler, L. Selfhealing and thermoreversible rubber from supramolecular assembly. Nature 2008, 451, 977−980. (20) Domb, A. J.; Maniar, M. Absorbable Biopolymers Derived from Dimer Fatty Acids. J. Polym. Sci., Part A: Polym. Chem. 1993, 31, 1275−1285. (21) Noordover, B. A. J.; van Staalduinen, V. G.; Duchateau, R.; Koning, C. E.; van Benthem, R. T. A.; Mak, M.; Heise, A.; Frissen, A. E.; van Haveren, J. Co- and Terpolyesters Based on Isosorbide and Succinic Acid for Coating Applications: Synthesis and Characterization. Biomacromolecules 2006, 7, 3406−3416. (22) Gallagher, J. J.; Hillmyer, M. A.; Reineke, T. M. Acrylic Triblock Copolymers Incorporating Isosorbide for Pressure Sensitive Adhesives. ACS Sustainable Chem. Eng. 2016, 4, 3379−3387. (23) Gorna, K.; Gogolewski, S. Biodegradable porous polyurethane scaffolds for tissue repair and regeneration. J. Biomed. Mater. Res., Part A 2006, 79, 128−138. (24) Miao, S.; Zhu, W.; Castro, N. J.; Nowicki, M.; Zhou, X.; Cui, H.; Fisher, J. P.; Zhang, L. G. 4D printing smart biomedical scaffolds with novel soybean oil epoxidized acrylate. Sci. Rep. 2016, 6, 27226. (25) Carothers, W. H. Studies on polymerization and ring formation. I. An introduction to the general theory of condensation polymers. J. Am. Chem. Soc. 1929, 51, 2548−2559. (26) Dusek, K.; Prins, W. Structure and Elasticity of Non-Crystalline Polymer Networks. Adv. Polym. Sci. 1969, 6, 1−102. (27) Erman, B.; Mark, J. E. Structures and Properties of Rubberlike Networks; Oxford University Press: New York, 1997. 5363


Sticky degradable bioelastomers

Recommend Documents