Adhesion and Contact Aging of Acrylic Pressure-Sensitive Adhesives to Swollen Elastomers

Fluid-infused (or swollen) elastomers are known for their antiadhesive properties. The presence of excess fluid at their surface is the main contributor to limiting contact formation and minimizing adhesion. Despite their potential, the mechanisms for adhesion and contact aging to fluid-infused elastomers are poorly understood beyond contact with a few materials (ice, biofilms, glass). This study reports on adhesion to a model fluid-infused elastomer, poly(dimethylsiloxane) (PDMS), swollen with silicone oil. The effects of oil saturation, contact time, and the opposing surface are investigated. Specifically, adhesion to two different adherents with comparable surface energies but drastically different mechanical properties is investigated: a glass surface and a soft viscoelastic acrylic pressure-sensitive adhesive film (PSA, modulus ∼25 kPa). Adhesion between the PSA and swollen PDMS [with 23% (w/w) silicone oil] retains up to 60% of its value compared to contact with unswollen (dry) PDMS. In contrast, adhesion to glass nearly vanishes in contact with the same swollen elastomer. Adhesion to the PSA also displays stronger contact aging than adhesion to glass. Contact aging with the PSA is comparable for dry and unsaturated PDMS. Moreover, load relaxation when the PSA is in contact with the PDMS does not correlate with contact aging for contact with the dry or unsaturated elastomer, suggesting that contact aging is likely caused by chain interpenetration and polymer reorganization within the contact region. Closer to full saturation of the PDMS with oil, adhesion to the PSA decreases significantly and shows a delay in the onset of contact aging that is weakly correlated to the poroelastic relaxation of the elastomer. Additional confocal imaging suggests that the presence of a layer of fluid trapped at the interface between the two solids could explain the delayed (and limited) contact aging to the oil-saturated PDMS.


Relaxation during constant indentation (effect of viscoelasticity)
During constant indentation, the load  in the PSA -PDMS system relaxes over time for all  � .The relaxation curves can be fit to a poro-viscoelastic model to obtain the overall viscoelastic and poroelastic timescales in the system.
Where  0 is maximum compressive load on the system that corresponds to  = 0.   and   are the loads after viscoelastic and poroelastic relaxation.Eq. (S1) is valid for a system where the poroelastic and viscoelastic timescales are sufficiently different i.e.,   ≪   which is indeed the case for our constant indentation experiments.(See Table S1).For dry PDMS Eq. (S1) reduces to a stretched exponential relaxation Eq. ( S2) is an alternate mathematical expression for the Prony series. 1 A dispersion factor  < 1 indicates that the system cannot be described using a single exponential relaxation time, but rather a distribution of timescales 0 <   < ∞.The contribution of each timescale in the distribution depends on .The mean characteristic time for viscoelasticity 〈  〉 can be calculated for the stretched exponential distribution using Eq.(S3).In Fig. S3 we see that most of the load on the dry PDMS has relaxed after 100 , which is similar to the relaxation of PDMS on glass. 2 Therefore, the bulk viscoelastic relaxation in the PDMS-PSA system is governed by the relaxation of the PDMS.It can be seen in Fig. S4 that for the same nominal stress, the PSA undergoes an indentation of ~6  with a glass probe.The presence of PDMS in the system would further suppress bulk deformation in the PSA as the load will be distributed between the PSA and PDMS. 3

Capillary adhesion model
Fig. S6.Capillary adhesion model for swollen PDMS adhesion with PSA.Side view image of fully saturated PDMS in contact with PSA (a) before jump-off -both JKR and capillary forces act on the PSA-PDMS system.(b) After jump-off -liquid capillary bridge formed at the interface contributes to total attractive force.(c) Individual contributions of JKR adhesive force   and capillary force   and total contribution   calculated using Eq.S4-S6.

S-6
For swollen PDMS in contact with PSA (Fig. S5), we find that the effective diffusivity obtained (see Table S1) is very similar to that for glass-PDMS 2 indicating that oil diffusion into the PSA is has either negligible contribution to relaxation and/or happens over a much different timescale.Estimating extent of relaxation: The extent of relaxation is defined as the fraction of total relaxation at a given contact time   , where  0 is the initial load at time  = 0. (  ) is the load at contact time  =   , and  ∞ is the equilibrium load at time  → ∞.
We obtain (  ) as the average of the load for the last n seconds where the slope of the force time curve becomes constant (within 1%).Typically, n~ 20% of the total contact time.
Since we do not know  ∞ for an indented sample a priori, we assume that all samples at the same oil content will reach the same state of equilibrium and estimate  ∞ to be the minimum obtained value of (  ).To correct for variations across experiments, we rescale with the initial load to obtain  ∞ / 0 .
(Starred value in Table S2 below).The extent of relaxation can therefore be rewritten as

Fig. S1 .
Fig. S1.Confocal microscopy set-up.(a) Locating the top surface of the PSA using reflected light.The beam splitter (BS) splits the output signal into fluorescence(fluor.)and reflectance(refl.).Intensity of reflected light is measured as a function of depth in z-direction.The first two intensity peaks from the top are ~25  apart, which is around the thickness of the PSA.(b) Schematic of swollen PDMS in contact with PSA.(c) Normalized intensity vs. emission wavelength of swollen

Fig. S2 .
Fig. S2.Fluorescence intensity as a function of depth for a swollen PDMS ( � = 1) in contact with glass.

Fig. S3 .
Fig. S3.Debonding curves for PDMS-PSA adhesion after contact at (a) constant load F = 10 mN for 100 s, (b-d) at constant indentation depth  = 37 for three different contact times contact time tc=100 s, 1000 s and 10000 s.

Fig. S5 .
Fig. S5.Load relaxation during indentation of PSA in contact with a glass indenter.The nominal stress on the material  = / is maintained at 91 .

1 −.Fig. S10 .
Fig. S10.Strain energy release rate as a function of time with PSA and glass for constant load F=10 mN experiments.(a)  � = 0, (b)  � = 0.25, (c)  � = 0.6, and (d)  � = 1.Filled symbols denote adhesion with PSA and open symbols denote adhesion with glass.Dashed lines represent power-law fit with contact time.Dotted dashed green line in (d) represents capillary adhesion model with contact aging.

Table S1 .
Poro-viscoelastic relaxation parameters obtained by fitting Eq.S4 and Eq.S5 to load relaxation data.

Table S3 .
Power law exponents for strain energy release rate   vs contact time data shown in Fig. S9.