Adhesion Depends on Interfacial Strength: Time and Temperature Effects Authors: Shi-Qing Wang, Zehao Fan, Tianji Pang, Zhe Cui Published: 2026 ABSTRACT In the conventional description, peeling of an adhesive from a solid substrate is expressed in terms of adhesion energies at adhesive-substrate interfaces. However, this energy-based perspective faced considerable difficulties in explaining why the adhesion energy, also known as peel strength Γp, is many orders of magnitude greater than Dupré’s thermodynamic work of adhesion, Γ0. In this study, we present experimental evidence to demonstrate that polymer adhesion is governed by interfacial strength σinterf, defined by the pair of adhesive and substrate. A tensile adhesion test is performed to estimate σinterf from adhesion strength σadh equal to the engineering stress at detachment. This test stretches, at various rates under different temperatures, one end of a ribbon-like specimen with the other end adhered to the same substrate used in peeling tests until adhesive detachment. The dependence of σadh on applied rate and temperature is found to be the same as that of Γp from conventional peeling tests on peeling speed v p and temperature. Our stress perspective shows that Γp explicitly correlates with σadh through a characteristic length scale P instead of Γ0. Here, P is the distance from the peeling front, beyond which the adhesive undergoes little deformation. Since P is a much larger length scale than a molecular scale and σadh directly depends on polymer–substrate interfacial interactions, Γp explicitly depends on σinterf. As activated processes, in both peeling and tensile adhesion tests, interfacial debonding on shorter time scales and lower temperature required higher interfacial stress. Introduction Polymer adhesion is commonly interpreted within the Dupré work-of-adhesion (Γ0) framework, − where peel strength Γp is regarded to reveal adhesion fracture energy. On the other hand, Γp is orders of magnitude greater than Γ0, whose magnitude can be comparable to the surface energy of the adhesive. Specifically, peeling tests − are often characterized by Γp in the following form as a function of peeling speed v p and temperature T: 1 Γp=Γ0[1+fadh(vp,T)],⁣withfadh≫1 where the second component f adh is supposed to arise from viscoelastic dissipation associated with bulk deformation of the adhesive. , In the present study, we suggest, based on new experiments, that polymer adhesion should be understood in a different way, i.e., in terms of interfacial strength σinterf instead of Γ0. Peel strength, Γp, as a measure of mechanical resistance against interfacial debonding of the adhesive off the substrate, directly manifests interfacial strength. We suggest that a tensile adhesion test can be carried out to estimate interfacial strength σinterf on different time scales at different temperatures. We will investigate whether Γp depends on temperature and characteristic time scales, associated with peeling speed v p and stretching rate λ̇ in peeling tensile adhesion tests, as σinterf does. Tensile adhesion and peeling experiments can be carried out to measure adhesion strength σadh and peel strength Γp. These two quantities are related through a length scale P that measures an effective region over which peeling force acts on the interface at the peeling front, beyond which adhesive experiences little deformation. This characteristic length P largely depends on interfacial strength, along with thickness B and the stretchability of adhesives. Our description of adhesion in terms of Γp/P ≈ σadh offers a stress perspective to the current paradigm that treats adhesion from Griffith’s energy perspective, , i.e., in terms of Γp. In this alternative framework Γp is expressed in terms of normalized force (σadh) required to exceed the interfacial strength, which is time and temperature dependent, along with P. This paper is organized as follows. In Section , we not only describe a rarely adopted experimental protocol to characterize interfacial strength but also present our analysis and methodology. Specifically, we discuss the relationship between probe tests and peeling tests. We show various results from different experiments in Section . Specifically, Section describes an experimental estimate of interfacial strength in terms of adhesion strength σadh based on both types of probe tests (tensile adhesion and poker chip), defined as the engineering stress at adhesive failure as a function of probe speed and temperature. Results from peeling tests at different peeling speeds and temperatures are presented in Section . Using adhesives of different thicknesses (0.02 vs 1 mm), we show that peeling tests are characterized by both peel strength, Γp, and characteristic length scale, P. Interfacial strength prescribes peel strength. Thus, peel strength increases with increasing peeling speed and decreasing temperature because overcoming interfacial strength is an activated process, i.e., both time-dependent and temperature-sensitive. When peeling only involves adhesive failure, we conclude in Section that (a) time- and temperature-dependent interfacial debonding explains the phenomenology summarized by eq and (b) characteristics such as time- and temperature-dependent peel strength Γp from peeling tests and debonding stress σadh from probe tests merely reflect how interfacial interactions between adhesive and substrate vary on different time scales at different temperatures. Experimental and Technical Backgrounds New MethodTensile Adhesion for Probe/Tack Tests To clarify the nature of peeling that involves only adhesive failure, we perform tensile adhesion tests, as depicted in Figure a. Upon placing a freshly fractured VHB strip on various substrates including bare glass and steel and after a given amount of time for adsorption, the other end of the strip is fixed onto an Instron clamp while the substrate is held stationary to act like the second clamp in such a tensile test. At different values of stretching rate λ̇ = V/L 0, engineering stress σengr = F/BW is measured as a function of extension until termination by adhesive failure at the substrate. Taking σadh to be given by σengr at debonding in such a tensile adhesion test, we measure σadh as a function of the time elapsed to the point of interfacial detachment, i.e., t db = (λadh – 1)/λ̇. (a) Tensile adhesion test is defined as shown, where the substrate is in contact with adhesive of area B × W. (b) Poker chip-like configuration, where the layer of adhesive may match the probe size or be a larger sheet. The probe is round-shaped with a diameter a. (c) Peeling test for adhesive with thickness B fixed on the backing. A force arises from an area of size P × W, which is maximum at the peeling front. Materials and Their Preparation Two commercially available adhesives are used to describe pertinent physics involved in polymer adhesion. They are VHB4910 and Scotch tape, both purchased from Amazon. VHB is in the form of sheets with a thickness of B = 1.0 mm. Scotch tape refers to a commercial transparent tape (Scotch Transparent Tape, Cat. 600, 3M) and has a thickness of 0.02 mm for its adhesive layer. Prior to testing, substrates were carefully cleaned following the sequence of hexane, alcohol, and water, and then dried by compressed air. All specimens were handled using tweezers to avoid contamination. Both probe/tack and peeling tests are based on the free-spinning wheel device shown in Figure a, along with an actual photograph of the apparatus in Figure b, which was earlier used by Kaelble. The substrate for tensile adhesion tests is either the surface of the wheel, made of either a borosilicate glass tube with a diameter of ca. 50 mm or steel tubes with a diameter of 48 mm, or steep tube wrapped with plastic films (Mylar and Kapton). (a) Schematic of the wheel device for 90° peeling that is fixed on the base of the Instron 5969 universal testing instrument equipped with a temperature-controlled environmental chamber. The upper clamp of the Instron holds the specimen to travel upward for all three types of tests depicted in Figure a–c. (b) Photo of the actual peeling device. To prepare specimens for tensile adhesion tests, VHB sheets were cut into rectangular specimens with approximate dimensions of 15 mm × 12.5 mm × 1 mm (length × width × thickness). The strips were sufficiently stretched in uniaxial extension so that a small notch introduced with scissor cutting triggers instant rupture, creating two smooth and flat fracture surfaces to be introduced into contact with the substrate for tensile adhesion tests. For all poker chip tests, the probe’s surface is made of a circular glass disk with diameters a = 3 mm that is fixed onto the upward traveling clamp at various speeds after being brought into contact with a flat stationary VHB sheet with typical dimensions of 110 mm × 70 mm × 1 mm, naturally achievable because VHB is a double-sided adhesive. For tests involving Scotch tape, the tape was cut into 40 mm × 19 mm strips and fixed on a flat surface using superglue, with the adhesive side facing upward. For peel tests, VHB sheets were cut into long strips with a width ranging from 10 to 12.5 mm and a length of around 140 mm. Scotch tape has a width of 19 mm, and a length of around 140 mm is wrapped around either glass or steel tube. Two videos of peel tests, along with the corresponding load vs time curves in Figure S1a,b (Section I of the Supporting Information (SI)), respectively, show peeling of Scotch tape and VHB as Movies 1a and 1b at respective v p = 10 and 0.2 mm/min. Adhesion between VHB or Scotch and the glass surface is established by bringing the adhesive into contact with the substrate under various conditions. For tensile adhesion and peeling tests presented in Figures a–c and , VHB was in contact with the glass tube for 5 min at 90 °C before testing. For the data in Figure , the contact time for both the tensile adhesion test and the poker chip was 10 min at 20 °C. For the data in Figure , peeling data involved a contact time of 30 min at 90 °C, while poker chip tests had both VHB and Scotch tape in contact with the glass tube for 10 min at 20 °C. For the tensile adhesion data in Figure , both substrates were in contact with VHB for 30 min at 90 °C. The strip length L 0 for tensile adhesion tests varies from 10 to 12.5 mm on the glass tube and 11.5 mm on the steel tube. Relationship between Tensile Adhesion and Poker Chip Tests For adhesives such as Scotch tape with an adhesive layer as thin as 0.02 mm, only poker chip tests like that shown in Figure b are feasible for an approximate estimate of σadh. In literature, a poker chip configuration with a thin layer of adhesive in contact with a probe is often employed for tack tests. The complicated fibrillation or cavitation , makes it challenging to extract useful and pertinent information. Poker chip tests are typically characterized in terms of the work to debond (adhesive thickness multiplying the area under the stress vs strain curve), , which has the same dimension as Γ0. To clarify the information revealed by poker chip tests, it is instructive for us to compare them with tensile adhesion tests since the interpretation of the latter is more straightforward and better-defined when debonding leaves nothing on the substrate, i.e., when the interfacial failure is entirely adhesive. In both tensile adhesion and poker chip tests, shown in Figure a,b, rheological properties of a viscoelastic adhesive merely influence how fast the force F increases with time until detachment at F c, whereas debonding of adhesive from the substrate solely depends on interfacial interactions. Stretch rate λ̇ or speed V can be varied to evaluate adhesion strength σadh on different time scales at debonding in tensile adhesion tests and the peak nominal stress in poker chip tests. When adhesive failure prevails, poker chip tests reveal similar information. According to the analysis provided in Appendix A, we can estimate interfacial strength σinterf using such probe tests to measure adhesion strength as 2 σadh=Fc/BW(tensile adhesion)⁣or⁣Fmax/πa2(poker chip) As shown in the following section, peeling tests reveal the same information on σadh. It is conceivable and perhaps even plausible that chain dynamics at the interface between adhesive and substrate can influence the interfacial strength: rapid chain movements might enable readsorption to offset any force-induced desorption. If such an effect existed or was dominant, interfacial strength could show a complicated temperature dependence. Relationship between Peel Strength and Interfacial Strength For 90° peeling, shown in Figure c, we bring adhesives onto a freely spinning wheel as sketched in Figure a, along with a photo of the actual apparatus in Figure b. For the present analysis, we take the fluctuating v f to be around the mean value given by the probe (crosshead) speed v p. Peeling produces a force F p acting on an area given by WP, where P can be comparable to adhesive thickness B in the presence of strong interfacial interactionsvanishing interfacial interactions are expected to result in diminishing P, decreasing all of the way to a monolayer scale, independent of B. The normalized force, often known as peel strength Γp, 3 Γp=Fp/W depends on σinterf and P, as shown in eq in Appendix B. Consequently, we further normalize Γp to introduce a potentially material-intrinsic quantity, 4 σpeel=Fp/WP=Γp/P and show it to be proportional to σinterf in eq . We call σpeel peeling stress. In Section , we will present experimental results involving pressure-sensitive adhesives (VHB and Scotch tape) for probe/tack and peeling tests to explore the framework for polymer adhesion. At the foundation, we assert that Γp reflects the interfacial strength. In other words, as shown in Appendix B, σpeel of eq from peeling tests and σadh of eq from tensile adhesion tests have the same origin, i.e., both related to σinterf. As shown in Figure c, peeling tests involve a characteristic length P whose magnitude largely depends on the strength of interfacial interactions and can be larger than adhesive thickness B. Thus, Γp/P , not Γp, may be more intrinsic to the pair of adhesive and substrate. The new dimensional analysis reveals (Γp/P) ∝ σinterf in place of Γp ∝ Γ0 in eq . Experiments in the following Section support the basic concept that σpeel defined by eq is another measure of interfacial strength (cf. Appendix B). In other words, the peel strength of eq can be expressed as 5 Γp∼σadhP Equation indicates that any peeling speed v p and temperature dependencies of Γp stem from those of adhesion strength σadh, where P may be a weak function of speed and temperature. This conclusion differs from the prevailing view in the field of polymer adhesion, summarized in eq . When Γp involves a characteristic length scale P that is much greater than a monolayer scale, it far exceeds the magnitude of Γ0. An instructive analogy exists between elastomeric fracture and adhesion, specifically, between tearing and peeling, both involving crack propagation and both operationally characterized in terms of a normalized force that shares the dimensions with the energy release rate or surface energy. In both cases, the force is higher at a higher speed and lower temperature. In both tearing and peeling, time and temperature effects arise, e.g., higher stress emergent on shorter time scales, because the crack propagation is an activated process. Moreover, a characteristic length scale emerges, either describing a stress saturation zone in bulk fracture or a region of size P at the peel front in interfacial fracture. Interfacial Strength vs Substrate Surface Energy Since Γp as a dimensionalized force has the dimension of work of adhesion, eq is widely used to describe speed and temperature effects in peeling tests. There also exists an impression from the literature that substrates with higher surface energy γs should produce higher Γp because Γ0 varies with γs. The phenomenological description based on eq warrants several comments. First, the dimensional argument leading to eq is likely inspired by the energy perspective in fracture mechanics, taking Γp to represent the energy released per unit area during peeling. But this proposal of Γp ∼ Γ0 lacks first principle-based reasoning. Second, Γ0 is much smaller than Γp and does not capture the fact that peeling involves stretching of chain network in the adhesive. In other words, peeling involves a destructive, non-equilibrium activation process. Third, peeling is a phenomenon of interfacial strength σinterf being overcome by emergent stress. Thus, Γp reflects mechanical resistance in the form of eq (force per length) or eq (or force per area). Fourth, while using organic substrates assures that Γ0 remains at the level of γa ≈ γs on the order of tens of mJ/m2, even a clean (unoxidized) metallic substrate with γs ≈ 1 J/m2 does not usually result in Γ0 higher than ca. 0.1 J/m2. Since the relationship between γs and σinterf is unclear, we can conduct tensile adhesion tests on different substrates (cf. Section ) to shed some light. Interfacial Debonding in Peeling and Probe/Tack Tests as an Activated Process When adhesive–substrate interfacial interactions are established by physical adsorption alone, adhesion with higher cohesive strength involves only interfacial debonding. Adhesive failure is also an activated process because the interfacial detachment involves collective desorption or dissociation of noncovalent van der Waals bonds at multiple adsorption sites. Since debonding adhesives from substrate is an activated process, interfacial strength characterized by the degree and intensity of segmental adsorption on the substrate is both time- and temperature-dependent. We introduce a concept of lifetime t intfS for interfacial interactions (adhesion) that is dependent on the state of adsorption. In the absence of any explicit discussion of the molecular picture, we assert that a polymer–substrate interface has finite lifetimes, and adhesion has finite strength as detachment amounts to complete removal of polymer–substrate contacts. It is stress, not the energy release rate, that dictates the lifetime. In the case of adhesion, the bonds are van der Waals in origin, and in the case of elastomeric fracture, carbon–carbon covalent bonds are typically involved. To the leading order, the dependence of t intfS on interfacial stress and temperature may be of the Arrhenius form, 6 tintfS=tintfS0⁡exp[EintfS(σadh)kBT] where the energy barrier height E intfS can be lowered during peeling and probe/tack tests. Inverting this equation, approximating E intfS as E 0 – vσadh, like the treatment of solid fracture, , we may find σadh to change logarithmically or very weakly with the characteristic time involved in the measurement of σadh. Moreover, on the same time scale, eq predicts that adhesion strength σadh and therefore peel strength Γp per eq are both larger at lower temperatures. Like our theory for elastomeric fracture, , we theorize here that adhesive failure through interfacial debonding occurs when t intfS decreases due to increasing interfacial stress until it meets the experimental time scale, characterized by the elapsed time at debonding, which is dictated by the applied rate, as shown below in eq . In treating rubber friction against smooth solid surfaces, Schallamach advanced the concept of rate-dependent detachment of polymer chains using reaction-rate theory of Eyring. Peeling strength was suggested to be a function both of the energy of interfacial bonds and of bulk energy loss in a viscoelastic adhesive. When an interfacial fracture involves chemical bonding of adhesives to a substrate, breaking of covalent bonds is involved. Chaudhury has treated such a case , within the framework of Kauzmann–Eyring–Bell , to explain rate and temperature dependencies. To our knowledge, peeling speed and temperature effects on peeling strength for adhesion with physical contacts have not yet been interpreted in terms of an activated debonding process. On the other hand, Bell was among the first to treat cell adhesion as activated processes. Results and Discussion Given the current consensus , regarding the nature of polymer adhesion, e.g., characteristic speed and temperature dependencies of peel strength (cf. eq ), we propose focusing on the essential case where peeling involves only adhesive failure. In the absence of cohesive failure, we show that tensile adhesion and poker chip tests, sketched in Figure a,b, disclose similar, equivalent information about interfacial strength. Therefore, we will first present in Section results from tensile adhesion tests and compare them with those from conventional poker chip tests to confirm the meaning of poker chip tests and illustrate the origin of peel strength. Probe/Tack Tests Tensile Adhesion Test Availability of VHB sheets of finite thickness makes it possible to carry out a new test that we call “tensile adhesion”, as shown in Figure a. Movie 2 shows such a test at V = 200 mm/min with L 0 = 14 mm, with its stress vs time curve in Figure S2 in Section II of the SI, indicating (a) significant partial interfacial detachment along the edge of width (W = 12.5 mm) occurs 33 s after the startup and complete detachment ensues 3 s later, (b) during this period of 3 s, engineering stress changes from 0.43 to 0.52 MPa. Visual inspection confirms that the tensile adhesion test involves only interfacial debonding. Given the time- and temperature-dependent nature of σadh described in Section , we perform tensile adhesion tests at various temperatures using different speeds so that debonding occurs on different time scales given by 7 tdb=(λdb−1)/λ̇,⁣withλ̇=V/L0 where λdb is the stretch ratio of the adhesive strip at debonding and only weakly increases with λ̇. Therefore, we can control t db (∼1/λ̇) by choice of V and L 0. Such tests establish how interfacial interactions between the adhesive and substrate depend on time and temperature. Conversely, regarding t db as a measure of t intfS of eq , we explore how the lifetime of interfacial interactions depends on the applied interfacial stress. For example, Figure a summarizes the raw data (tensile stress vs time curves), presented in Figure S3a–d in Section III of the SI, in terms of stress σadh according to eq versus the time t db at debonding per eq 6 . It is clear from Figure a that σadh is systematically lower on longer time scales and increases with decreasing temperature. Like the phenomenon of elastic rupture, which also involves an activated process associated with covalent bond dissociation, there is time–temperature equivalence (TTE), where the lifetime t intfS of polymer–substrate interactions is the hidden internal clock. It is conceivable that polymer dynamics can influence lifetime t intfS if the kinetics of polymer adsorption and stress-induced desorption can affect the barrier height E 0 in eq . Since adhesive failure takes place when t intfS = t db, tensile adhesion tests provide a measure of t intfS as a function of applied stress. While the concept of TTE per eq is well-defined for a fixed interfacial stress, TTE cannot be used to perform a horizontal shift along the time axis in Figure aeach of the four sets of σadh vs t db involves a range of interfacial stress. On the other hand, we can forcefully shift the time scale for the four sets of data in Figure a to form an approximate master curve in Figure b. The shift factor A T (circles), loosely representing t intfS(T)/t intfS(T ref), in Figure c still encompasses some profound information. It approximately reveals how the lifetime t intfS increases with decreasing temperature. Since higher interfacial stress (σadh) is involved at lower temperatures, we can expect the temperature dependence of t intfS to be stronger than that shown in Figure c, i.e., changing more than three decades from 22 to 80 °C. Also plotted in Figure c are data points from a second shift factor A T(peel) in squares that is used to shift σpeel data along the time axis for peeling tests, as presented below in Section . Taking t db of eq as representing the lifetime t intfS, the concept of TTE allows us to explore the dependence of t intfS on interfacial stress σadh over a much wider range, as shown in Figure b. Approximately, t db (regarded as t intfS) shows an exponential dependence on σadh in agreement with eq , where the stress dependence of the energy barrier E intf may be linearized as E intf = E 0 – vσadh. (a) Engineering stress at debonding from tensile adhesion tests, read from Figure S3a–d in the SI, where the two open circles indicate cohesive failure. (b) “Master curve”, constructed from (a) by shifting along the time axis. (c) Shift factors A T as a function of temperature, used in the construction of (b) and in Figure . Poker Chip Test Results from tensile adhesion tests allow us to clarify the meaning of poker chip tests. Figure a shows that both poker chip (open symbols) and tensile adhesion tests (filled symbols) produce similar adhesive failure, although tensile adhesion produces a cleaner interpretation of the nature of interfacial detachment. The comparison largely shows the same trend: On longer time scales, σadh of eq , is lower. In poker chip tests, debonding first takes place at the weakest site/region of adsorption. Approximately, the peak stress signals the onset of detachment. The agreement between tensile adhesion and poker chip tests is shown in filled and open circles in Figure b, indicating that the poker chip test is valid as an approximate setup to characterize adhesion. (a) Engineering stress vs time from both tensile adhesion tests (boxed in three filled symbols) and poker chip tests (four open symbols), involving crosshead speed V from 500 to 0.1 mm/min, where elapsed times at stress peak are denoted by either t detch or t db. In tensile adhesion tests (filled symbols), specimen length L 0 = 10.5, 10.6, and 5 mm, respectively, for V = 5, 50, and 500 mm/min. The arrow shows the trend: σadh decreases with decreasing V. (b) Adhesion strength σadh, read from the peak stresses in (a) at either t detach (open circles) in poker chip tests or at t db (filled circles) for debonding in tensile adhesion tests. Also plotted is σadh from poker chip tests on Scotch tape (open squares) for comparison with open circles (VHB). Upon such validation of poker chip tests, we can demonstrate the essence of peeling tests in Section by comparing a thick adhesive, such as VHB, with a thin one, such as Scotch tape that is only amenable to characterization by poker chip tests. The present Scotch tape happens to show interfacial strength comparable to that of VHB when bare glass is the common substrate. Specifically, reading from the poker chip tests on Scotch tape, presented in Figure S4a,b in Section IV of the SI, open squares in Figure b show only slightly higher magnitudes than open circles of VHB. Peeling Tests Peeling behavior of adhesives is described by both the emergent force, normalized as Γp per eq , and a characteristic length P shown in Figure c. According to our analysis in Section , Γp is not a material function because it varies with adhesive thickness B. Using the two adhesives with remarkably different thicknesses in VHB and Scotch tapes but similar interfacial strength, we make a comparison in Figure a between them based on the raw data, i.e., peel strength vs time in Figure S5a,b in Section V of the SI. At each speed, we measure in “steady state” the characteristic length P and peel strength Γp, as shown in Figure a, based on video recording of peeling, e.g., images in Figure S6a,b in Section VI of the SI, whereas Figure S6c,d compares the front geometries at two speeds. Here, higher Γp at higher V occurs without greater stretching of VHB because VHB shows stronger stress response at higher rates. Although Figure b shows comparable adhesion strength σadh, Γp of these two adhesives are markedly different. Notably, the much higher Γp of VHB also has proportionally higher P due to higher thicknes. The much higher Γp of VHB stems from the fact that adhesion occurs over a much larger area of WP, with P of VHB being ten times P of Scotch tape, as shown in Figure a. The similar interfacial strength of VHB and Scotch tape on the glass surface should reveal similar intrinsic characteristics. The peeling stress σpeel of eq instead of the Γp of eq is the quantity to describe interfacial interactions in peeling tests. Figure b shows that both VHB and Scotch tape show comparable peeling stress (filled symbols) and adhesion strength (open symbols) from poker chip tests over a wide range of comparable time scales, illustrating the origin of peel strength, i.e., validating eq . The difference in adhesion conditions for peeling and poker chip tests evidently did not result in any significant difference in interfacial strength. Figure b also verified that the magnitudes of σpeel and σadh for VHB are comparable to those of Scotch tape. Here for the comparison between σpeel and σadh, at peeling speed v p, taken to equal the crosshead speed V, a pertinent time scale is identified as P/v p, so that the two stresses can be evaluated on the same footing in Figure b. Since these time scales originate from bond lifetime t intfS of eq , we replot Figure b in Figure c, showing indeed that they undergo exponential decrease as a function of stress. The data should not be extrapolated to zero stress on this semilogarithmic plot since a stress threshold is expected: Detachment on any realistic time scale would require nonvanishing stress. Peeling may be regarded as fracture-mechanical, requiring a characteristic stress and length to describe, as shown in eq . Consequently, peeling of the two different adhesives (VHB and Scotch tape) is described by the peel strength, given by the product of comparable adhesion strength but very different length P. The approximately 1 order-of-magnitude difference in Γp in Figure a stems from the corresponding difference in P. (a) Characteristic length P vs peel strength Γp defined by eq . (b) Plotting the data in (a) in terms of the intrinsic quantity–peeling stress per eq , along with adhesion strength from Figure b, as a function of the characteristic time, given by either P/v p or debonding time t db. (c) Characteristic times as a function of adhesion and peel strengths. To examine the temperature effect on σpeel, we carry out 16 peeling tests at four temperatures with four rates at each temperature. Figure summarizes the raw data and preliminary analysis presented in Figures S7a–d, S8a–d and S9 in the SI (respectively, in Sections VII, VIII, and IX of the SI). Like Figure b, the master curve in open symbols in Figure involves a horizontal shift of data in Figure S9 involving a shift factor A T(peel), i.e., squares in Figure c. Also plotted in Figure are the same data points from Figure b, in filled symbols. Since the two straight lines are drawn to have the same slope, Figure shows σpeel(P/v p = t db) = σadh(t db), thus revealing the meaning of peeling tests and demonstrating that time and temperature dependencies of peel strength stem from those of interfacial interactions, not from the bulk viscoelasticity of adhesives: Since tensile adhesion tests involve only adhesive failure yet reveal the same dependencies, we conclude that the same physics is at play, i.e., the speed and temperature effects in peeling tests stem from the dependence of interfacial strength on temperature that also varies on different time scales. Comparison between the master curve of peeling stress σpeel (right-hand-side Y-axis from 1.2 × 10–3 to 1.2) against (P/v p)A T(peel) and that of adhesion strength σadh (left-hand-side Y-axis from 1.2× 10–2 to 12) against t db A T at different temperatures, where A T(peel) and A T are, respectively, the squares and circles in Figure c. Double Y-axes are employed to displace one master curve from the other by one decade to avoid overlap. The agreement between σadh and Γp/P in Figures c and for either VHB or Scotch tape configurations can be understood by comparing eqs and , both proportional to interfacial strength with prefactors on the order of unity. Surface Energy of Substrate vs Adhesion Strength Section indicates that interfacial strength has no direct relationship with the substrate’s surface energy. Although the surface energies of Mylar as well as Kapton, glass, and steel may differ significantly, from 30 to 1000 mJ/m2, there is no theory to describe how their interfacial interactions with VHB vary. Figure , read from Figure S10 in Section X of the SI, shows that the adhesion strengths are within a factor of 3 over two decades of time scale. Similarly, peel strength is only lower on the steel tube than on the glass tube at room temperature, as shown in Figure S11 in Section XI of the SI. Since appreciable pressure is applied to establish contact between VHB and the steel tube, we have assumed that VHB achieved comparable access to the steel surface due to adequate pressure. Thus, the surface roughness of the steel tube may not be the reason for the observed lower adhesion strengththe oxidative layer on the steel tube may be the origin. In contrast, Mylar and Kapton films should be very smooth down to submicrons. Thus, the results in Figure indicate a certain correlation between interfacial strength and surface energy of the substrate. Further exploration of the relationship between the surface energy of substrate and adhesion strength is an interesting topic for future study, but beyond the scope of the current work. Because of the correlation between σadh and Γp, demonstrated in Figure and eq , it is reasonable to conclude that peel strength follows the same trend as described by Figure . In fact, since P decreases with adhesion strength, peeling strength on these plastic surfaces can be expected to be significantly lower than indicated by these tension adhesiont tests. Debonding time from tensile adhesion tests at room temperature (26 °C) as a function of adhesion strength σadh on four different substrates, made of glass, steel, Kapton (PI), and Mylar (PET), respectively. The horizonal line shows the spread in σadh of ca. 3 on a given time scale. Conclusions Departing from the conventional description of polymer adhesion, we show that time and temperature effects in peeling tests can be understood in terms of interfacial strength and a characteristic length scale that defines the area of active stress at the peeling front. Like adhesive failure in tensile adhesion tests (cf. Figure a), peeling (cf. Figures c and a,b) amounts to stress-induced interfacial debonding of van der Waals bonds between adhesive and solid substrate. Since debonding is an activated interfacial process, both probe and peeling tests show similar time and temperature dependencies. Specifically, higher peel strength at higher peeling speeds and lower temperatures stems from the fact that only higher interfacial stress can cause interfacial debonding to take place on shorter time scales and at lower temperatures. In other words, the time- and temperature-dependent phenomenology of peel strength Γp has a different origin from that proposed based on energy dissipation associated with viscoelastic bulk deformation of adhesives. In the literature, the influence of interfacial interactions on Γp has been mentioned within the phenomenological framework of eq . But the present viewpoint contrasts with the prevailing interpretation of the nature of polymer adhesion, which regards peel strength Γp in eq as energy per unit area. Making a connection between tensile adhesion and peeling tests has permitted us to conclude that Γp is merely a mechanical quantity, measuring the magnitude of a partially normalized force associated with interfacial interactions on a given time scale and at a temperature of choice. Probe/tack tests (e.g., tensile adhesion tests) and peeling tests describe the same physics, characterized by interfacial strength. Explicitly, our experiments show that interfacial failure in tensile adhesion tests is the same process of detachment of adhesive from solid substrate as that involved in peeling. In both cases, the strength varies with the temperature on different time scales. Since the interfacial detachment is adhesive rather than cohesive, it is straightforward to conclude that any observed rate and temperature dependencies reflect interfacial bonding as an activated process. Viscoelastic properties of adhesives in such a test merely dictate how quickly tensile stress builds up and generates interfacial stress all of the way up until interfacial strength is reached. It is beyond the scope of the current study to explore any relationship between the network structure of adhesives and their interfacial strength with solid substrates. To reiterate, when peeling involves only adhesive failure, the same physics is involved in peeling as in the tensile adhesion test. Specifically, both normalized peel strength, i.e., peeling stress, and adhesion strength, are comparable on the same time scales at various temperatures. The strength of interfacial interactions between adhesive and solid substrates varies with the surface energy of the substrate, based on our measurement of adhesion strength on four different substrates. The generality of this conclusion remains to be further explored. Since interfacial interactions are different for different substrates, it is already known that surface treatment can produce considerable changes in Γp. We plan to describe the effect of surface treatment in our next paper. Like our treatment of polymer fracture, where we take the stress perspective and recognize how local stress and characteristic length combine to describe fracture, we have shown here that polymer adhesion in the case of peeling can be similarly understood in terms of interfacial strength σinterf and pertinent length scale P, in contrast to the conventional description based on work of adhesion, which is a fracture mechanics-motivated, energy-based perspective. Here P, crudely related to Γp according to eq , is not the elasto-adhesive length , defined by Γp/E. In both elastomeric fracture and adhesion, the energy perspective can only interpret the rate and temperature effects in terms of viscous energy dissipation when materials only undergo predominantly elastic deformation before fracture and detachment. Our stress perspective naturally encompasses cases where Γp ≫ Γ0. We have not only explained why peel strength may vary with peeling speed and temperature but also indicated that interfacial strength dictates its magnitude along with a characteristic length P. By explaining time and temperature effects on polymer adhesion in terms of the lifetime of interfacial interactions, this work revives previous concepts ,, of kinetics of interfacial interactions to elucidate the nature of polymer adhesion. The findings have direct implications for the design of adhesive interfaces, soft coatings, and polymer–solid contacts through physical adsorption, where interfacial debonding leads to adhesive failure. The present alternative framework explored may be equally relevant and plausibly applicable to the adhesion of all soft materials whose cohesive strength is higher than the adhesion strength. The present conclusions also have implications regarding the adhesion theory based on the JKR , model for a spherical body (of radius R) in contact with a flat substrate. When contact between a rubber sphere and a hard solid surface involves strong physical absorption, detachment may involve a magnitude of the pull-off force F c much greater than F JKR ∼ Γ0 R. In this case, comparison with peeling tests should produce insightful results. In the case of adhesive detachment, time and temperature effects on F c are plausibly interfacial in origin, not related to Dupré work of adhesion and/or bulk viscoelastic properties. Such a viewpoint offers a complementary perspective to the standard energy perspective, in which Persson and co-workers have attempted to generalize the JKR approach by associating adhesion energy (peel strength) with viscous energy dissipation. In the case of peeling, we show in the present work that the origin of such time and temperature effects is activation of interfacial interactions, not viscoelastic dissipation. The separation of two elastic bodies in physical contact also amounts to overcoming interfacial strength, involving F c/R naturally much greater than Γ0. Thus, future studies of time and temperature effects ,, on F c (≫F JKR) may benefit from our stress perspective. In summary, although our experiments utilize two common pressure-sensitive adhesives and four substrates, they reveal a framework that may be universal when the substrate is a smooth solid surface. Whenever peeling results in adhesive failure, the process is fundamentally interfacial; the increase in the peel strength observed at higher speeds or lower temperatures stems from the thermally activated dissociation of interfacial bonding. When the substrate is also made of a polymer network, as often studied, the nature of interfacial interactions can be more complicated. Time and temperature effects may arise from polymer dynamics at the interface, where entanglement due to chain uncrossability can influence interfacial strength. Supplementary Material