Effect of the Strength of Stickers on Rheology and Adhesion of

Article Cite This: Langmuir XXXX, XXX, XXX−XXX

pubs.acs.org/Langmuir

Effect of the Strength of Stickers on Rheology and Adhesion of Supramolecular Center-Functionalized Polyisobutenes X. Callies,*,†,‡,∥ E. Ressouche,§ C. Fonteneau,§ G. Ducouret,†,‡ S. Pensec,§ L. Bouteiller,§ and C. Creton*,†,‡ †

Downloaded via UNIV OF SUNDERLAND on October 16, 2018 at 21:56:00 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Laboratoire de Sciences et Ingénierie de la Matière Molle, CNRS, ESPCI Paris, PSL Research University, 10 rue Vauquelin, 75005 Paris, France ‡ Laboratoire Sciences et Ingénierie de la Matière Molle, Sorbonne-Université, 10 rue Vauquelin, 75005 Paris, France § Sorbonne Université, CNRS, IPCM, Chimie des Polymères, F-75005 Paris, France S Supporting Information *

ABSTRACT: In order to systematically investigate the effect of the strength of the supramolecular interactions on the debonding properties of associative polymers, a series of model systems have been characterized by probe-tack tests. These model materials, composed of linear and low dispersity poly(isobutylene) chains (Mn ≈ 3 kg/mol) center-functionalized by a single bis-urea sticker, are able to self-assemble by four hydrogen bonds. Three types of stickers are used in the present study: a bis-urea with a methylene diphenyl (MDI) spacer, a bis-urea with a tolyl (TOL) spacer, and a bis-urea with a xylyl (XYL) spacer. In order to investigate the influence of stickers in depth, both the nanostructure of the materials and the linear rheology were investigated by small-angle Xray scattering (SAXS) and oscillatory shear, respectively. For two types of stickers (TOL and XYL), the association of polymers via hydrogen bonds induces the formation of bundles of rodlike aggregates at room temperature and the behavior of a soft elastic material was observed. For bis-urea MDI, no structure is detected by SAXS and a Newtonian behavior is observed at room temperature. In probe-tack experiments, all these materials show a cohesive mode of failure, a signature of flowing materials as previously observed for tri-urea center-functionalized poly(butylacrylate) (PnBA3U). However, XYL center-functionalized polyisobutene shows much higher debonding energies than PnBA3U, revealing the importance of the strength of noncovalent bonds in the scission/recombination dynamics. On the basis of the analysis of the debonding images, this effect is discussed via the mechanical behavior at large deformation.

■

formation of dimers, clusters, or aggregates.14,15 These physical cross-links strongly increase the relaxation time of the polymer chains and affect their rheological properties in the linear and nonlinear regimes,16,17 particularly at low molecular weights.18 The effect of stickers on rheology includes different molecular processes, such as formation of clusters and phase segregation, that are strongly dependent on the chemical architecture of the polymer chains. As previously reported in the literature, the functionalization of unentangled or lightly entangled polymer chains could induce a dissipative behavior in the melt state,19,20 particularly interesting for soft adhesives or pressure-sensitive adhesives (PSA21). In order to have an optimized combination of resistance to shear and “tacky” character, PSA require very well-defined viscoelastic properties, in particular to reach large bulk deformations and thus high debonding energies.22

INTRODUCTION

The interest in developing innovative and smart materials by applying supramolecular chemistry concepts to polymers1−3 has considerably grown in the last few years. These materials can be composed of polymer chains functionalized by strongly interacting moieties, often called “stickers”. Among the numerous studies focusing on the properties of these new materials, some of them reported interesting results for nonsticky but yet self-healing materials4,5 or for stimuliresponsive adhesives.6,7 In the field of adhesion,8 diverse applications were reported in the melt state,9 in wet conditions,10,11 and even underwater.12 These studies raised fundamental questions about the effect of dynamic sticker groups on the adhesive properties of supramolecular materials, either through changes in the surface energy13 or through changes in the rheological properties of these materials. Since the first investigation on ionomers several decades ago, it is well-known that the noncovalent association of stickers in the melt state can lead to a temporary network via the © XXXX American Chemical Society

Received: July 26, 2018 Revised: September 23, 2018 Published: September 27, 2018 A

DOI: 10.1021/acs.langmuir.8b02533 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir Although weakly interacting groups are usually used in the elaboration of acrylate-based PSA (such as acrylic acid23), few systematic studies were carried out on well-defined model systems with other types of polar associating groups.20,24 The general objective of our project is to identify the key elements, in terms of chemical architecture and strength of stickers, to design the macromolecular structure of soft supramolecular adhesives. Our work is based on a systematic study of well-defined chemical systems and consists in exploring the link between the supramolecular chemistry of polymer chains, their self-assembly properties, and finally their rheological and adhesive properties. The chemistry of the model materials is inspired from an extensive study of supramolecular self-assembly of small aromatic molecules bearing urea moieties in solution.25,26 In a preliminary work,20 some of the authors investigated the bulk properties of a bis-urea functionalized low-molecular-weight polyisobutene [PIB tolyl (TOL), Mn ≈ 3 kg/mol]. They reported that the self-assembly of the polymer chains by hydrogen bonding interactions induces an organized structure detected by smallangle X-ray scattering (SAXS) at room temperature. The supramolecular nanostructure forms a physical gel and prevents the material from flowing at small deformation. At large deformation, PIB TOL flows with a high elongational viscosity, much higher than expected from its molecular weight, revealing interesting dissipative properties for the elaboration of new adhesives. Then, in order to understand the physico-chemistry of these new systems in depth, linear and monodisperse poly(butylacrylate) chains functionalized by tri-urea stickers (PnBA3U) in the middle were synthesized with functional initiators.27 The systematic study of these supramolecular poly(butylacrylate) systems (PnBA3U) revealed a pronounced effect of stickers on the nanostructure28 and the resulting rheological properties.29 Although PnBA3U is elastic and nanostructured at a high density of stickers (i.e., at low molecular weight of the PnBA side chains), it is unstructured and behaves like a viscoelastic fluid at a low density of stickers (i.e., high molecular weight). However, the effect of stickers on the debonding properties was surprisingly low;30 both high and low molecular weights show cohesive failure at low strains comparable to the behavior of a low-molecular-weight polymer. The mechanical analysis of the adhesive films during the adhesion test itself suggested that these unexpected results may be induced by the specific behavior of these materials at large deformations, characterized by a strong softening of the supramolecular nanostructure and the absence of strain hardening.30 In order to investigate the effect of the binding strength of stickers in a more apolar matrix, two new supramolecular systems are investigated in the present study. Like PIB TOL, these two materials are composed of linear and low dispersity poly(isobutylene) chains (Mn ≈ 3 kg/mol) but centerfunctionalized by two different stickers: either a xylyl (XYL) or a methylene diphenyl spacer (MDI) (see Figure 1 and Table 1). The strength of hydrogen bonding interactions between the urea groups depends on the molecular architecture of the sticker and thus may be easily tuned according to the desired properties of the final supramolecular material.25 The XYL sticker differs from the TOL sticker by a second methyl group on the aromatic ring, changing the orientation of the ortho urea group. As this urea group is preferentially aligned out of the aromatic plane, the urea−urea

Figure 1. Chemical structure of the stickers of the three investigated supramolecular materials [R = poly(isobutylene)].

Table 1. Molar Mass of the Materials samples PIB PIB PIB PIB

TOL MDI XYL REF

Mn,RMNa (g/mol)

Mn,SEC tripleb (g/mol)

D̵ b

2900 3600 2450

2600 3190 2600 2200

1.3 1.24 1.3 1.7

a

Calculated by 1H NMR. bEvaluated by size exclusion chromatography triple detection in tetrahydrofuran.

interactions between the XYL stickers are facilitated, promoting the self-association of the polymer chains.25 As for the MDI sticker, the two urea groups are linked by a methylene group that allows free rotation, preventing the cooperative association of both urea groups and thus the formation of stable aggregates over long time scales.27 In the melt state, the self-assembly of stickers in the three supramolecular model systems has been investigated by SAXS, whereas the linear rheology was investigated by small amplitude oscillatory shear (SAOS) tests and the adhesive properties were characterized by probe-tack tests. Compared to the more common peel test, the force measurements in the probe-tack test are less influenced by the elastic stretching of the substrates and the geometry of the interface is simpler, making the test more sensitive to differences in material properties.31 After briefly describing the chemistry of the studied materials and the experimental methods, the influence of the strength of stickers on the polymer self-assembly and on the macroscopic properties will be shown and discussed.

■

EXPERIMENTAL SECTION

Materials. The synthesis is described in the Supporting Information. As shown in Table 1, the molecular weight of the polymers is close to 3000 g/mol for the three systems and the polydispersity is close to 1.3. Fourier transform infrared spectroscopy (FTIR) analysis shows that the three polymers are strongly hydrogenbonded because no free NH vibration is detected in these materials (see Figure S1). A reference nonfunctional PIB (PIB REF) was purchased from Acros. Scattering. The SAXS experiments were conducted on a device equipped with a copper rotating anode (λ = 1.54 Å) (Rigaku Corporation, Tokyo, Japan), a Gobel’s mirrors collimation system (ELEXIENCE, Verrières le Buisson, France), and a two-dimensional detector (Princeton Instrument SCX2D, Trenton, NJ, USA). For the SAXS experiments, the distance between the sample and the detector was equal to 550 mm. The exposure time was set at 300 s. All samples were placed between two thin kapton films to prevent flow. Because no attempt was made to ensure a constant mass of sample in the X-ray beam, the absolute intensity of the patterns for different samples will not be discussed. For all patterns, the kapton signal was subtracted. For all materials, samples were first diluted in toluene and deposited on a thin kapton film; the solvent was slowly evaporated overnight and the sample was then placed in partial vacuum during a week. B

DOI: 10.1021/acs.langmuir.8b02533 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 2. Characterization of the nanostructure (left) of the PIB functionalized by XYL (blue), TOL (green), MDI (red) spacers and unfunctionalized PIB (black) by SAXS at room temperature and frequency dependence of the viscoelastic moduli of the same materials (right) at 25 °C. analyzed by using the methods developed by Tanguy and coworkers34 in order to characterize the in situ elongation of the material during the debonding step. This method was previously explained in detail.30 Briefly, the true stress σe in the walls between cavities (component of the stress parallel to the tensile direction) can be estimated before the equilibration of pressure as well as in the remaining fibrils after the equilibration of pressure

Rheology. Dynamical rheological measurements were performed on a strain-controlled rheometer ARES LS1 (TA Instruments). The frequency dependence of the viscoelastic moduli G′ and G″ was characterized with a parallel plate geometry (diameter 8 and 25 mm) at 25 °C. The ARES LS1 rheometer was equipped with a Peltier device to regulate the temperature of the sample. As all samples flow under stress, the final gap was selected by compressing the sample with the upper plate for each experiment. The gel state characterized by SAOS tests was confirmed by classic qualitative tests based on upturned vials. Adhesion Tests. Adhesion tests were carried out on thin (100 ± 10 μm) adhesive layers deposited on glass slides (2.6 × 10 × 0.2 cm3, purchased from Preciver) by slow evaporation from solution. Solutions were prepared by dissolving 300 mg of the polymer in 2 mL of toluene. Solutions were then deposited on glass slides, left for 2 days under a glass cover at room temperature, and then for 2 more days at 70 °C under reduced pressure (∼200 mbar) to remove the residual solvent. The thickness of dry films was measured by a white light scanning technique with an optical profilometer (Microsurf 3D, Fogale nanotech). The home-made probe-tack setup is described in detail elsewhere.30,32 Adhesion tests were carried out at room temperature (between 22 and 24 °C) and in the same range of relative humidity (between 20 and 30%). The probe-tack test33 that was carried out consists in bringing the surface of a solid probe into contact with the thin adhesive layer coated on a rigid substrate and in measuring the force FT required to detach it at a constant debonding speed Vdeb. During the debonding step, the displacement d(t) of the probe relative to the adhesive layer is measured in order to calculate the thickness of the adhesive layer h(t) h(t ) = d(t ) − K × FT(t ) + h0

σe =

∫

(3)

with the force Fm required to stretch fibrils and the load-bearing area Ae, that is, the projected area of fibrils obtained from the images of the contact. In the presence of bubbles, Fm is calculated by subtracting the force Fp because of the work against the atmospheric pressure from the measured force FT. The force Fp ≈ AcPatm is estimated by considering the projected area Ac of a convex envelope around bubbles and the atmospheric pressure Patm ≈ 105 Pa. After the coalescence of all bubbles, Fp = 0. In the case of viscoelastic fluids, the elongational flow in the fibrils can be characterized by an approximate local transient elongational viscosity η+ in the fibril walls via the Hencky strain rate ϵ̇H σ η+ = e ϵ̇H (4) The Hencky strain rate ϵ̇H is calculated by considering an effective elongation ⟨λ⟩ along the tensile direction in the thinnest cross section of the walls by assuming a deformation of the wall at constant volume. ϵ̇H =

■

(1)

with h0 the initial thickness of the adhesive film and K = 5.5 μm/N the compliance of the apparatus (measured with a blank test). Probe-tack curves conventionally plot the nominal stress σ0(t) = FT(t)/A0 versus the nominal strain ε0(t) = (h(t) − h0)/h0 or the stretching ratio λ0(t) = 1 + ε0(t), where A0 is the contact area between the probe and the adhesive layer at the maximum compression stage. The area under the σ0 = f(ε0) curve is used to calculate the debonding energy of the soft layer, Wadh (J/m2), that is, the energy necessary to detach the probe from the layer for the given traction speed Wadh = h0 σ0 dε0

Fm Ae

d(ln⟨λ⟩) dt

with ⟨λ⟩ =

A0 Ae

(5)

RESULTS SAXS experiments and SAOS tests shed light on the influence of the strength of stickers on the degree of self-assembly of polymer chains. PIB MDI, PIB XYL, and PIB REF have been investigated by SAXS, and their scattering spectra are compared with PIB TOL35 in Figure 2. As previously observed for PIB TOL, a sharp peak is detected for PIB XYL, whereas a broad and low peak is noticed for PIB MDI and no peak is detected for PIB REF. A characteristic distance can be determined from the Bragg relation for the two first materials: d = 4.3 ± 0.2 nm for PIB TOL and d = 4.1 ± 0.2 nm for PIB XYL. The molecular organization at large length scales is associated with a solidlike behavior characterized with a plateau of the storage modulus at low frequency in SAOS experiments (see Figure 2). At high ω, the strong frequency dependence of the viscoelastic moduli reveals a dissipative behavior for the PIB TOL and PIB XYL materials. On the

(2)

The experimental procedure of the alignment between the probe and the glass slide was recently described in detail.30 In order to get a perfectly smooth and reflective surface, the stainless steel probes (diameter Φ = 5.95 ± 0.02 mm) were mechanically polished. In our home-made setup,32 the thin adhesive layer in contact with the probe is observed via an optical microscope connected to a charge-coupled device camera. In the case of the formation of a fibrillar structure during the debonding process, images can be C

DOI: 10.1021/acs.langmuir.8b02533 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 3. Stress−strain curve obtained by probe-tack tests for PIB MDI (A), PIB TOL (B), and PIB XYL (C) at different debonding speeds and at room temperature. The dashed lines are theoretical curves for Newtonian liquids and yield stress fluids. The pictures (D) illustrate the debonding processes at different debonding speeds [Vdeb = 1 (α, γ, δ) and 10 μm/s (β)] for the three investigated supramolecular materials: PIB MDI (α, β), PIB TOL (γ), and PIB XYL (δ).

contrary, the PIB REF sample (Mn ≈ 2,2 kg/mol ≪ Me ≈ 8 kg/mol36,37) exhibits a Newtonian fluid behavior with a viscosity η = 28 Pa·s (G″ = ηω and G″ ≫ G′) and PIB MDI shows a viscoelastic fluid behavior: a flowing zone is observed at low frequency, whereas a dissipative zone is detected at high frequency (G″ ≈ G′). The adhesive properties of the center-functionalized PIB were characterized by probe-tack tests. The experiments were carried out at room temperature, and four debonding velocities Vdeb = 1, 10, 50, and 100 μm/s were investigated. For all materials, the debonding process is cohesive over the studied Vdeb range: some macroscopic residues of the adhesive layer remain on the steel probe (see Figure S5). The stress−strain curves for all three materials are shown in Figure 3 along with the representative debonding picture at Vdeb = 1 and 10 μm/s. Three experimental parameters were defined to characterize the adhesive behavior:30 the maximal stress at the peak σ0,max, the critical strain λ0,c, and the debonding energy Wadh. The critical strain λ0,c is estimated when all cavities are coalesced, that is, when the stress drops (see the following paragraphs for more details). For PIB MDI, two distinct shapes of stress−strain curves are observed, respectively, at low and high debonding speeds (see Figure 3A). For Vdeb = 1 μm/s, stress−strain curves show a stress peak at small strain and a continuous decrease of stress following a power law σ0 ≈ λ0−6. For higher debonding velocities, a sharp decrease from σ0 ≈ 105 Pa to σ0 ≈ 104 Pa is also noticed at intermediate strains (λ0 ≈ 2−4). At larger deformations (λ0 > 4), the stress again decreases in a progressive way (σ0 ≈ λ0α with −4 < α < −5). As illustrated by the debonding pictures α and β in Figure 3D, the variation

of the shape of the stress−strain curves reveals a change in deformation mechanisms. For Vdeb = 1 μm/s, only fingerlike instabilities growing from the circular interface (between the ambient atmosphere and the bulk) toward the center of the probe are observed (see picture α). At higher debonding velocities, cavities also grow near the center of the probe (see picture β). For adhesion tests carried out on thin films of PIB MDI (h0 ≈ 100 μm), the Deborah number De, defined as the ratio of the relaxation time τc ≈ 2 s (see Figure S6) and the characteristic time scale τexp = h0/Vdeb of the debonding test, is lower than 1 for low debonding velocities (Vdeb ≤ 10 μm/s). Within these experimental conditions, PIB MDI can be considered as a Newtonian fluid. The mechanical response of viscous liquids in the probe-tack geometry has been well described in the literature, in particular by Poivet and coworkers.38 In this model system,39 the detachment of the probe induces a drop of pressure in the liquid layer and the flow of the liquid from the periphery of the disk to its center. At very low debonding speed, this flow can be modeled by a Poiseuille flow and depends only on the geometry parameters and the fluid viscosity η. For large confinement ratios R = a0/h0 and in the absence of cavities, the relationship between the nominal stress σ0 and the fluid viscosity η is given by the following expression: σ0 =

3 ̇ 2 5 ηλ 0(R /λ 0 ) 2

(6)

Equations 1 and 6 can be combined to fit the experimental data measured at Vdeb = 1 μm/s, as shown by the dashed line in Figure 3A. The theoretical curve was calculated with the following parameters: η = 6500 Pa·s, a0 = 3.1 mm, h0 = 125 D

DOI: 10.1021/acs.langmuir.8b02533 Langmuir XXXX, XXX, XXX−XXX

Article

Langmuir

Figure 4. Maximal stress (A), debonding energy (B), and maximal strain before the equilibration of pressure (C) for adhesion tests carried out on PIB MDI, PIB TOL, and PIB XYL films.

α < −1). This variability can be attributed to the morphology of the fibrillar structure, which may strongly change from one test to another. At higher debonding speeds (Vdeb ≥ 10 μm/s), the nominal stress σ0 unexpectedly increases after the pressure equilibrates, before finally decreasing at larger deformations. Such stress increase, which is not observed for PIB MDI and PIB XYL, does not arise from the morphology of the fibrillar structure and seems to be the consequence of the dynamic character of PIB TOL, favoring a fast molecular reorganization when the hydrostatic stress drops (see the Analysis and Discussion). For PIB XYL, stress−strain curves are similar to those observed for PIB TOL: they are all characterized by a stress plateau and, after the coalescence of cavities, residual stresses are detected, revealing the presence of remaining fibrils between the probe and the glass slides (see Figures 3C and S5). For both materials, these fibrils break in the bulk at large deformation and the mode of fracture is cohesive. A clear improvement of the adhesion properties is observed at a low debonding speed where PIB XYL is characterized by the highest stress maxima σ0,max, highest strain maxima λ0,c, and highest debonding energy Wadh (see Figure 4). However, the three parameters σ0,max, λ0,c, and Wadh show a less pronounced rate dependence for PIB XYL than for PIB TOL and, at high debonding speeds, the parameters σ0,max and λ0,c are lower for PIB XYL than for PIB TOL. The variation of the stress−strain curves with the strength of stickers seems to reveal the competition of different effects of the supramolecular associations on the debonding mechanisms, as revealed by the analysis of the debonding images in the following section.

μm, and K = 106 N/m. The viscosity η is relatively close to the complex viscosity η* = 5400 Pa·s measured in shear experiments at low frequency. At high λ0, the decrease of stress observed in the experimental data is stronger than that in the theoretical one (which follows a power law σ0 ≈ λ0−5). The discrepancy with the prediction of Poivet and coworkers may be due to Saffman−Taylor instabilities, which are not taken into account in this analytical model.38 At higher debonding velocities (Vdeb ≥ 10 μm/s), the drop of pressure in the liquid induces the macroscopic growth of gas bubbles initially entrapped in the liquid layer or at the interface between the probe and the liquid. When the pressure p around these bubbles drops under a certain pressure threshold pc, the Laplace pressure cannot overcome the expansion of the gas and these internal cavities grow rapidly. At the transition between the low-velocity regime and the high-velocity regime (see picture β in Figure 3D), the macroscopic growth of gas bubbles occurs in the center of the liquid disc, where the pressure is minimal. One of the most important features of the bulk cavitation process is the emergence of a plateau in the stress−strain curves (see Figure 3A). The plateau ends by a more dramatic decrease of force when the cavities meet the fingers coming from the periphery. The drop of the nominal stress is roughly equal to the pressure of the surrounding atmosphere.40 For PIB TOL, finger-like instabilities and cavitation are observed for all debonding speeds and all stress−strain curves show a stress plateau because of the growth of cavities (see Figure 3B and picture γ in Figure 3D). In comparison with PIB MDI, higher stresses are required to stretch fibrils and fibrils are more stretched when the instabilities break the walls of bubbles. This is confirmed in Figure 4: for all debonding velocities, PIB TOL is characterized by higher stress maxima σ0,max, higher strain maxima λ0,c, and higher debonding energy Wadh than for PIB MDI. Although stress−strain curves are extremely similar to those observed for PIB MDI, the flow through the debonding layer is different because PIB TOL is a yield stress fluid. Among numerous works in the literature, Derks and coworkers41 investigated the mechanical response of yield stress fluids in the probe-tack geometry in the low debonding speed regime (i.e., in the absence of cavitation and fibrillation) and predicted a much more progressive decrease of stress (σ0 ≈ λ0−5/2) at large strains than for Newtonian fluids. Although cavities are present for PIB TOL, a similar progressive drop of stress is interestingly observed for Vdeb = 1 μm/s at large nominal strains (see Figure 3B). The experimental slopes vary much from one debonding test to another (σ0 ≈ λ0α with −2