Effect of drying conditions on adhesion strength of a pressure-sensitive adhesive

ABSTRACT Regular surface roughness could be created on the surface of a pressure-sensitive adhesive (PSA) film through Marangoni flow during the drying process. Therefore, tailoring of both surface and bulk characteristics of the PSA and consequently its adhesion strength could be expected by controlling the drying conditions. Herein, surface and bulk properties of a water-based PSA dried at various temperatures and humidities were scrutinized. Increase in the drying temperature improved the adhesion strength of the PSA to poly(ethylene terephthalate) due to enhanced surface nanoroughness of its film. At constant humidity, the higher the Péclet number, the higher the Marangoni number and the rougher the PSA film surface. Drying humidity rise, however, improved the adhesion strength due to more uniform distribution of copolymers constituting the PSA, better interdiffusion of chains through the interface of polymer particles in a prolonged drying process, and increased surface free energy of the film. The adhesion strength of the PSA, similar to the here-defined viscoelastic dissipation ability, demonstrated a power dependence on the film surface nanoroughness. This newly-defined parameter considers taking advantage of the real viscoelastic dissipation of the PSA regarding its potentiality through thorough wetting of the substrate surface with the PSA.


Introduction
Pressure-sensitive adhesives (PSAs) adhere to nearly any surface under low contact pressure and short contact times without any chemical reaction. [1,2]mong all polymers, acrylics are used extensively to synthesize PSAs. [3]This wide application is due to their specifications, such as adjustable glass transition temperature (T g ), molecular architecture, crosslink density, low plateau modulus, competitive price compared to other classes of polymers, compatibility with additives, and processability. [3]Nonetheless, the base acrylic polymers which are used for the synthesis of PSAs, namely poly(n-butyl acrylate) (PBA) or poly (2-ethylhexyl acrylate), require some structural changes to provide them with the desired adhesion properties. [3,4]This requirement results from their high entanglement molecular weight, low T g , and low-tomedium molecular weight chains. [3]Therefore, high-T g components, such as methyl methacrylate (MMA) or acrylic acid, are added to improve the adhesion properties. [5,6]SAs are prepared through various polymerization techniques such as solution, bulk, and emulsion polymerization.Despite the possibility of preparing solvent-based PSAs, [7] there is an ever-increasing tendency toward water-based formulations due to their advantages over the former, including the absence of volatile organic compounds (VOC), reduced odor, inflammability, non-toxicity, and reduced costs. [8]In the preparation of a PSA through emulsion polymerization, the synthesis methodology (semi-batch or batch) and water solubility and reactivity ratio of the monomers are effective in the final properties of the PSA. [6]For instance, in the batch emulsion polymerization, the reactivity ratio of monomers controls the distribution of their sequence in the chain. [6]Consequently, formation of chains rich in each of the monomers leads to their phase separation or dissolution according to phase diagrams of the formed chains. [5,9]The formation of butyl acrylaterich (BA-rich) and MMA-rich chains during the preparation of an acrylic PSA using BA and MMA monomers through batch emulsion polymerization has been reported previously. [6]This was due to the greater tendency of MMA macroradicals to react with MMA rather than BA, having a reactivity ratio of higher than unity. [6]Besides, the BA-rich and MMA-rich chains phase separated due to dissimilarity of their solubility parameter [10] and stiffness.
In addition to the synthesis methodology, in the preparation of water-based PSAs, film formation process is a critical step. [11]The drying process of the water-based PSAs goes through three main stages of latex film formation: (1)  evaporation of water to concentrate the particles, (2) deformation of particles into space-filling polyhedra, and (3) interdiffusion of polymer chains among particles and subsequent improved cohesive strength. [8]The drying rate and softness of polymer particles are the two most essential system parameters governing the film formation. [12]At the onset of latex drying, hydrodynamic and elastocapillary stresses play roles in film formation. [13]The hydrodynamic stress generated from the directional water flow transports particles to the front.It then compresses the deposits, named the coffee-ring effect. [14]The elastocapillary stress, resulting from the tendency of central fluid domain for energy minimization, pulls on the packed outer ring, effectively stretching the packed bed of particles inwards and potentially causing surface wrinkling. [13,15]Particles deform after evaporation of the bulk of water, and the mechanical properties develop.
Routh and Russel [16,17] shed light on the film formation mechanisms by introducing a generalized map, which considers the critical process parameters of Péclet number,Pe, and deformability, � λ.Being a criterion for vertical inhomogeneity of particles distribution (normal to the substrate), Pe is determined using the ratio of the characteristic time for diffusion of polymer particles from top to bottom of the film, t diff , to the drying time, t dry , Equation (1).� λ is defined as the ratio of the characteristic time needed for complete particle deformation, t def to t dry , Equation (2).
where μ, R, H, _ E, k, T, η 0 , and γ are viscosity of the continuous phase, original particles radius, thickness of the wet film, evaporation rate, Boltzmann constant, absolute temperature, low-shear viscosity of the polymer, and latex surface tension. [18]The film formation mechanism of a soft polymer latex is wet sintering accompanied by skin formation, [16,17] happening in the range of Pe > 1 and � λ < 1.This is because of the fact that, on the one hand, in Pe > 1, faster evaporation than diffusion results in particles accumulation near the top interface of latex with the air, and, on the other hand, in � λ < 1, due to the shorter time required for particles deformation compared with the time needed for water evaporation, particles are susceptible to coalescence and skin formation. [19]Ma et al. [20] investigated the accumulation of particles on top of the drying latex film using cryogenic electron microscopy.Routh and Zimmerman [21] showed that at higher Pe, a steeper vertical concentration gradient would be expected, meaning that there would be a more sharply defined boundary between the packed particles at the top of the film and a dilute dispersion near the bottom. [22]Therefore, process conditions, namely temperature, relative humidity (RH), air flow rate, and film thickness, influence the film formation by affecting water evaporation rate, particles arrangement, interdiffusion of polymer chains among particles, and the final properties of the film. [23]n order to understand the effect of the mentioned events occurring during the synthesis and film formation processes on adhesion energy of a PSA, let us recall the fundamental equation of Gent and Schultz [24] and Andrews and Kinloch [25] for the adhesion strength, G, Equation (3): G 0 and ϕ are the intrinsic adhesion energy and bulk viscoelastic dissipation, respectively. [5]When the fracture is cohesive inside the adhesive, the C-C fracture energy should be considered in calculation of G 0 .However, when the fracture occurs at the interface, G 0 is equal to the surface chemistry-dependent thermodynamic work of adhesion, W 12 .Although the adhesion strength of a PSA to a substrate is several orders of magnitude higher than their W 12 , indicating the more pronounced role of viscoelastic dissipation, the thermodynamic tendency of the PSA and the substrate is a prerequisite for a proper adhesion. [26]Formation of various components during the polymerization process, their arrangement during the film formation process, and also the film surface nanoroughness will affect the adhesion energy of the PSA.
The idea of the research originates from the above-mentioned vertical concentration gradient of components in the film formation process of water-based PSAs which was thought to create regular surface roughness through Marangoni flow [27] due to vertical surface tension gradients driven by heat and/or mass transfer. [28]Marangoni flow is a process with some similarities to wrinkling, though substantially different from that. [29]Guo et al. [30] reported formation of a wrinkled surface pattern during drying of poly(methacrylic acid)-b-poly(butyl acrylate) copolymer dispersion in methanol.They concluded that more restriction in mobility of PBA through higher drying temperature or higher T g of the copolymer led to the formation of clearer surface wrinkles with a shorter wavelength.Crosby and coworkers [31] controlled the adhesion strength of a crosslinked PBA via wrinkle patterning created by curing the surface of the PBA elastomer film swelled with BA monomer.Sufficiently small wavelength of wrinkles led to improved adhesion strength due to an increase in the total contact line per area.Also, the positive influence of substrate surface roughness on the adhesion strength of various polymers through mechanical interlocking has been reported in the literature. [32,33]evertheless, to the best knowledge of the authors, the effect of surface nanoroughness of a water-based PSA on its adhesion strength has been overlooked.
Therefore, this work aims to elucidate the relationship between the most crucial process variables, i.e., temperature and humidity, on the surface and bulk properties of a water-based acrylic PSA and its adhesion strength.To achieve the goal, the nanoroughness of the film containing BA-rich and MMA-rich copolymers, which formed during batch emulsion polymerization of BA and MMA, has been controlled through the drying process of the soft polymer latex at three different temperatures and two relative humidities.Changes in the surface nanoroughness of the dried PSAs were followed by atomic force microscopy and confirmed by calculating the Marangoni number.Furthermore, the effect of drying process parameters on mechanical and viscoelastic properties of the PSAs was investigated.The dependence of the adhesion strengths of the PSAs on their surface nanoroughness was scrutinized and evaluated by a novel parameter, named viscoelastic dissipation ability.

Synthesis
Poly(methyl methacrylate-co-butyl acrylate) containing 30 wt. % methyl methacrylate (MBC30) was synthesized through batch emulsion copolymerization based on the prescribed recipe [34] (Table S1).After preparation of separate water phases containing KOH, KPS, and SDS, they were poured, together with the monomers and TDM, into a one-liter glass reactor.The emulsion copolymerization was conducted at 80°C and 250 rpm for four hours using an anchor impeller.The formulation produced monodisperse particles with a diameter of 135 nm and particle size distribution of 0.011, determined by dynamic light scattering technique using Zetasizer Nano Series (Malvern Instruments, UK).The total number-average and weight-average molecular weights of the resulting copolymers were about 9.33 × 10 4 and 2.53 × 10 5 g/ mol, respectively (Figure S1), measured by gel permeation chromatography using Agilent 1100 (Agilent Instruments, USA).

Preparation of PSA films
In order to prepare MBC30 PSA films under controlled drying conditions, while recording the latex drying rate, a homemade 27-liter temperature and humidity control chamber was used (Figure S2).The chamber was equipped with a 588 × 10 −2 N bending load cell (AB120C, Sewhacnm Instruments, South Korea) with an accuracy of 98 × 10 −6 N, a continuous weight logging system, and temperature and humidity controllers.To prepare free-standing or backed PSA films, 17 g MBC30 latex was cast each time in uncovered or poly(ethylene terephthalate) (PET)-covered Teflon molds, 8 cm in diameter.The drying was then conducted at 25°C, 35°C, and 40°C and 15% and 75% RH until the weight is fixed.After the film formation, surface of the PSA film was rinsed with distilled water for surfactant removal and redried for 15 minutes.The initial thickness (H) of all cast latexes was 2.3 mm, and the final thickness of the films was measured by ultrasonic thickness gauge.The temperature of the surface of the drying latex films was measured using an infrared thermometer (Testo 830-T1, Testo Instruments, Germany).

Fourier transform infrared (FTIR) spectroscopy
During the synthesis of MBC30, due to the difference in the reactivity ratio of BA and MMA radicals (r MMA = 2.24, and r BA = 0.414 [6] ), MMA-rich and BArich chains were formed.Although the formation of these chains has been mentioned in our several previous works, [5,6,9,35] FTIR test was performed to reconfirm the formation of these components.For this purpose, by taking advantage of the difference in the molecular weight of these two components, [6] they were first separated by the sedimentation process.To do so, first, 17 g MBC30 latex or PBA latex (for comparison) was dried in a Teflon mold, 8 cm in diameter, at 25°C and 15% RH.The film was then rinsed with distilled water for surfactant removal and redried.MBC30 film was dissolved in toluene and precipitated in methanol.After 24 hours fractionation of MBC30 chains in a separatory funnel, the first drop (MBC30-1) and the thirteenth drop (MBC30-2) were dried under the above-mentioned condition.The whole polymer was fractionated into sixteen drops.To identify the chemical groups in the free-standing MBC30-1, MBC30-2, and PBA films, their FTIR spectra were obtained using Fourier transform infrared spectrometer (PerkinElmer, Frontier, USA).

Differential scanning calorimetry (DSC)
To determine the glass transition temperature of the free-standing MBC30 film dried at 25°C and 15% RH, DSC thermograms were obtained using differential scanning calorimeter (DSC 1, Mettler Toledo Instruments, Switzerland) at a heating rate of 5°C/min over a temperature range of −70°C to 150°C under nitrogen purging rate of 80 mL/min.

Rheomechanical spectroscopy (RMS)
Rheological characterization of the thick free-standing MBC30 film dried at various temperatures and humidities was performed using a rheometric mechanical spectrometer (Anton Paar, Physica, MCR 501, Austria).The tests were conducted using a parallel plate fixture 25 mm in diameter at a constant 1 mm gap.To obtain linear rheological properties of MBC30 films, frequency-sweep experiments were done in oscillatory shear deformation mode at a strain amplitude of 0.5%.To obtain zero shear viscosity for calculation of � λ, the shear rate sweep in the range of 0.0001-0.001(1/s) was performed at 25°C, 35°C, and 40°C (Figure S3).

Atomic force microscopy (AFM) and spectroscopy
Surface of the PET substrate and the PSA films was investigated by AFM (No. 0101/A, Ara Pajohesh Instruments, Iran) in non-contact mode at room temperature.Imaging was performed using a silicon cantilever with a spring constant in the range of 0.18 N/m, a resonance frequency of 190 kHz, and a tip with a curvature radius of 10 nm.Several images of 10 μm × 10 μm were collected to obtain the topographical height and corresponding phase images.The reported PSA surface nanoroughness calculated by the AFM software was an average of the root mean square roughness data obtained from three measurements.The lateral wavelength of wrinkles in the x-direction regarding the distance between two valleys was also averaged at four different y coordinates of each sample.Nanoscale adhesion test was performed in contact mode on at least ten points of the sample surface.The nanoscale adhesion energy was calculated using Johnson-Kendal-Roberts (JKR) theory [36] : where F adhesion is the adhesion force recorded during retraction of the AFM tip from the surface of the PSA and R is the radius of curvature of the AFM tip.

Field-emission scanning electron microscopy (FESEM)
Surface and cross-section of the PSA films were studied by (MIRA3, Tescan Instruments, Czech Republic) at an accelerating voltage of 20 kV under vacuum at room temperature.For taking cross-sectional images, the PSA films were fractured in liquid nitrogen.All samples were coated with a thin layer of gold.

Contact angle measurement
Surface free energies (SFEs) of PET and polystyrene (PS) as substrates and also the dried PSAs were determined using the sessile drop method and distilled water and formamide as probe liquids. [37]Nonpolar liquids such as diiodomethane or dibromonaphthalene could not be used, since they removed the PSAs surface roughness.After calculating surface free energies of the substrates by analyzing the drop profiles photographed by a CCD camera using ImageJ software (version a1.52 and contact angle plugin) and Owens -Wendt-Rabel -Kaelble (OWRK) equation, [38] the latex contact angle was measured on PET and PS substrates at 25°C, 35°C, and 40°C.The reported contact angles were averages of the contact angle data obtained from at least ten drops per sample.The maximum contact angle error was ±2°.The thermodynamic work of adhesion of the PSAs on PET, W 12 , was also calculated using Equation ( 5). [39]ere γ 1 and γ 2 are the SFEs of the materials in contact, and d and p stand for the dispersive and polar components of the SFEs, respectively.

Adhesion test
To measure the adhesion strength of the PSAs, T-peel test was performed using a homemade universal testing machine equipped with a 1 kN load cell at room temperature and atmospheric pressure.The PSA film was bonded to a PET substrate 20 μm in thickness and the joint was rolled over with a 170 g or 2000 g roller.The sample length and width for the peel test were 60 and 12 mm, respectively.The nip-to-nip distance and applied strain rate were 20 mm and 300 mm/min, respectively.The peel strength was calculated using 2F=w, where F the average peel force recorded during the test after passing a maximum and w stands for the sample width. [40]The adhesion strength of each PSA was obtained by averaging the results of two repetitions for each case.

Tensile test
Tensile test was performed on 5 mm × 30 mm free-standing PSA films dried at various conditions using a homemade universal testing machine with a 100 N load cell at room temperature and atmospheric temperature.The nip-to-nip distance and applied strain rate were 10 mm and 300 mm/min, respectively.

Characterization of MBC30 copolymer
][42] It has been shown that the MMA-rich chains have a lower molecular weight than the BA-rich chains. [5,6]This difference in the molecular weight of the components was attributed to the difference in their type of termination mechanism and radical activity.The characteristics of high-molecular weight BA-rich and lowmolecular weight MMA-rich copolymers in MBC30 determined in our previous studies have been tabulated in Table S2. [35]Herein, to re-verify the formation of these copolymers, they were separated from each other using the sedimentation process.FTIR spectroscopy was performed on the separated components and also on pure PBA.During the sedimentation process, chains with higher molecular weight (BA-rich) precipitated first, and chains with lower molecular weight (MMA-rich) precipitated at the end.The FTIR spectra of the two precipitated drops of the MBC30 (MBC30-1 and MBC30-2, respectively) were compared with that of the pure PBA (Figure 1).The characteristic band of the stretching vibration of C=O group appeared at 1732 cm −1 for all samples. [3]The stretching vibration of C-H appeared at 2954 and 2874 cm −1 .Oh et al. [41] showed that the intensity of C-H stretching band absorption at 2874 cm −1 is more significant than 2960 cm −1 in polymethyl methacrylate (PMMA).However, in PBA, the intensity of C-H stretching band absorption at 2874 cm −1 is lower than 2960 cm −1. [43]erein, the area under the deconvoluted peaks at 2960 and 2875 cm −1 was calculated.The ratio of the area under the peak at 2954 cm −1 to that at 2874 cm −1 was 2.47, 1.87, and 1.35 in the spectrum of PBA, MBC30-1, and MBC30-2, respectively.As this ratio was highest in PBA, in accordance with the results reported by Lacerda et al., [43] the higher obtained value for MBC30-1 compared to MBC30-2 implied the BA-rich composition of chains in the first precipitated drop of MBC30 and the MMA-rich composition of chains in the drop closer to the last precipitated drop of MBC30.
Therefore, MBC30 copolymer consists of the BA-rich and MMA-rich chains.The dried synthesized MBC30 copolymer showed an upper critical solution temperature phase diagram (between its components, i.e., BA-rich and MMA-rich chains), being located in the two-phase region at room temperature. [6,9]o investigate the location of these two components in the dried PSA film, DSC analysis was performed on samples from the center and front of the PSA film dried at 25°C and 15% RH (Figure 2).The existence of the BA-rich copolymers with low T g (about −26°C) was observed in the first derivative of the heat flow curve for both the center and front of the film.However, the presence of three strong high-temperature peaks in the thermogram of the front of the film compared to a relatively weak high-temperature peak in the thermogram of the center of the film indicated a more significant accumulation of the MMA-rich copolymers at the front compared to the center of the film (Figure 2, and Figure S4).
After confirming the formation of MMA-rich and BA-rich chains during the synthesis processes, in the following sections, we will focus on the effect of film formation process on the PSAs properties.

Effect of drying condition on film formation mechanism
Before investigating the effect of film formation conditions on the PSA nanoroughness and morphology, it is necessary to express the reason for adopting the film formation conditions briefly.The drying temperature was investigated in the range of 25°C to 40°C.At temperatures below 25°C, the drying time increased a lot due to approaching the freezing temperature of water; therefore, choosing lower temperatures was not practical.On the other hand, at a temperature higher than 40°C, for example 50°C, drying was accompanied by blistering of the final film, Figure S5.Therefore, herein, the maximum latex drying temperature for having a defect-free MBC30 copolymer film was 40°C.Also, to evaluate the humidity effect on film formation and adhesion of the PSA, low and high-enough RH values of 15% and 75% were used.Indeed, RH value of 15% was the humidity of the laboratory and RH values above 75% caused supersaturated conditions which were inefficient.
The drying kinetics of MBC30 latex was studied to determine the main mechanism of film formation and the vertical distribution of components during the drying process at various temperatures and humidities using Pe and � λ. Figure 3 shows the evolution of the mass of MBC30 latexes over time at the studied drying conditions.The drying rate was calculated using where S, ρ and A were the slope of the mass versus time curve (in g/s), the density of water (in g/cm 3 ), and the mold surface area (in cm 2 ). [19]ncrease in the temperature and decrease in the humidity of the film formation process increased the latex drying rate (Table 1), which was comparable to the drying rate of soft polymer latexes reported in the literature. [8]The drying time is considered as the crossover time of the tangent lines on the observed steep and flat sections of the mass versus time curve. [44]The trend of change in the mass of the cast film at 15% RH was consistently declining, indicative of the continuous evaporation of water in the course of latex drying (Figure 3a).
However, at 75% RH, the mass change curve exhibited an upward trend at the onset of drying and decreased after representing a maximum (Figure 3b), which was a result of the absorption of water vapor in the cast latex under high humidity condition. [45]In other words, at 75% RH, water absorption competed with water evaporation, and reduction in the mass of the MBC30 latex appeared when the latter surpassed the former.The observed plateau region after about 68% mass loss of the latex dried at 25°C and 75% RH was attributed to the severe entrapment of water underneath the skin layer.Such a plateau region is a characteristic feature in the very slow drying of low-T g latexes. [46]o calculate � λ and Pe, the latex surface tension was determined at 15% RH and various temperatures.To make more proper estimations for the 75% RH case, the low-humidity data were shifted according to the change in water surface tension with humidity in Ref. [47] Pe and � λ were calculated using Equations ( 1) and (2), respectively (Table 1).It is worth mentioning that although the lowshear viscosity of the films dried at high humidity might be slightly different from those dried at low humidity, the minor change in this quantity could not significantly alter the � λ magnitude to an amount higher than unity.To be sure, however, the low-shear viscosity of the dried films at 40°C and two humidities of 15 and 75% were measured, Figure S3.The results showed that increase in the humidity had a minimal effect on the low-shear viscosity, which was considered in the calculations.The obtained Pe > 1 and � λ < 1 for all drying conditions were indicative of the non-uniform vertical distribution of water and particles during the drying process and the mechanism of film formation of wet sintering along with the skin formation. [17,19]As increase in Pe intensifies the skinning in the wet sintering regime due to the more accessible transport of particles to the top of the film, [19] the film being dried at 40°C and 15% RH was photographed for an explicit representation of the mechanism (Figure S6).There was such sufficient mechanical integrity in the skin layer that it could be peeled off, and a wet colloidal film was revealed underneath (Figure S6).Increase in Pe with the increase in the drying temperature and decrease in the drying humidity was attributed to the more decrease in the drying time than the decrease in the characteristic time for particles diffusion in the latex medium.Chen et al. [48] also observed a similar trend of change in Pe with temperature and humidity.The trend of change of � λ at both drying humidities was inclining with the increase in the drying temperature.η 0 decreased with increase in the temperature, which tends to decrease � λ.On the other hand, increase in _ E and decrease in γ latex tend to increase � λ with increase in the temperature.The overall increase in � λ with temperature indicates the predominance of the roles of over the role of η 0 .

Effect of drying condition on nanoroughness and morphology
After determination of the film formation mechanism, AFM analysis was performed to study the surface morphology of the PSA films dried at various temperatures and humidities.The phase and height images of the PSAs are shown in Figure 4a-l and S7, respectively.Because of the horizontal center-tofront gradient of the particles concentration arising from the stress exerted by water to the particles and the presence of surfactant, according to the DSC thermogram and the literature, [49] both the center and front of the surface of the PSA films were studied by AFM.The existence of distinct components, namely the BA-rich and MMA-rich copolymers, was also verified by the observed dispersion of higher-modulus (brighter) aggregates in lowermodulus (darker) counterparts.
Moreover, comparing the corresponding images revealed that increase in the drying temperature intensified the appearance of more stiffnessdifferentiated components on the surface (images in Figure 4 from left to right).Furthermore, the difference in the stiffness of components at the front of PSA films was more than that at the center, confirming the horizontal gradient of components concentration with the intensified appearance of MMA-rich copolymers at the front, in accordance with the DSC result.The AFM height images of the PSA films showed an increased PSA surface nanoroughness of the film with increase in the drying temperature at both humidities (Figure 5a,b).These changes could be clarified by the vertical and horizontal distribution of BA-rich and MMA-rich components in the PSA film.In other words, more intensified accumulation of the components at the surface of the PSA film, which was a consequence of enhanced Pe, led to the formation of more significant phase-separated domains, rich in either stiffer or softer components.This consequently increased the surface nanoroughness.Besides, the fronts of the films were of higher surface nanoroughness than the centers, which could be interpreted by more accumulation of stiffer higher-T g (MMA-rich) components at the front of the film, being unable to interdiffuse for closing all the voids and forming a smooth film.Similarly, the lateral wavelength of wrinkles was longer at the drying humidity of 75% rather than 15% and at the center rather than the front (Table S3).Paying detailed attention to the relevance of surface topography of the PSA films to their drying kinetics unveiled a decrease in the surface nanoroughness (Figure 5a,b) with the increase in the drying time of the PSA films (Figure 3a,b).The increase in the film surface nanoroughness with temperature rise could be described by formation of regular surface corrugations due to the Marangoni phenomenon [28] by calculating the Marangoni number [50,51] : where ΔT, α, and η are the temperature difference between the bottom and the top of the surface of latex film, water diffusivity, [51] and latex viscosity, respectively.dγ dT was calculated based on the slope of the line tangent to the latex surface tension versus temperature curve at any required temperature. [52]t is worth mentioning that in drying complex polymer systems, the Marangoni flow could be triggered by both temperature and concentration gradients interconnectedly. [53]In the current study also the drying polymer system was complex, and since the thermal or solutal origins of the Marangoni flow could not be separated precisely, their effect was investigated by change in the vertical temperature gradient and therefore in the surface tension.Indeed, the increased accumulation of components at the surface of the polymer film, which was equivalent to the increased vertical concentration gradient, could lead to an elevated vertical temperature gradient owing to low thermal conductivity of polymers.An increase in the Marangoni number with temperature rise or humidity reduction was in accordance with the increase in the surface nanoroughness of the films (Figure 6a,b).The increase in the average Marangoni number also was in agreement with the increase in Pe as the origin of intensified accumulation of components at the surface of the film, which in turn raised the vertical temperature gradient in the drying film.Interestingly, the Marangoni number for drying of the front of the films was higher than that of the center, consistent with the higher surface nanoroughness at the front of the films (Figure S8).
To scrutinize the vertical and horizontal gradient of components concentrations and confirm the morphology observed in the AFM images, FESEM analysis was performed on cross-section and surface of the center and front of the PSA films dried at 25°C and 40°C at both low and high humidities (Figure 7).The presence of agglomerates in the cross-sectional images of the PSA films (Figure 7a-h) proved the claim of more irregular arrangement of components and hence the presence of more voids near the top of the film, especially at the front of the films, similar to the AFM results.The agglomeration was intensified with increase in the temperature or decrease in the humidity.Indeed, acceleration of the drying process provoked an increase in bottom-to-top and center-to-front gradient of components concentrations on the one hand and failed to provide sufficient time for deformation of particles to be able to close all the voids on the other hand. [54]Besides, more resemblance of the center images to the front images of the films dried at lower temperatures or higher humidity meant gentler gradient of components concentrations and more homogeneity of the PSA film in the horizontal direction.
FESEM images of the surface of the PSA films dried at the above-mentioned conditions were shown in Figure S9.The PSA film dried at 40°C demonstrated more accumulation of stiffer components at the top of the film for both center and front samples compared to those dried at 25°C at both humidities.In other words, increased temperature, reduced humidity, or proximity to the front of the film led to a more irregular arrangement of components due to the  accelerated drying especially at the top of the film.Indeed, the simultaneous existence of all three factors caused agglomeration of components at the top of the film due to the intensified phase separation of the MMA-rich and BA-rich copolymers.

Effect of drying condition on mechanical and rheological properties
To affirm the above interpretations on the observed morphologies, the mechanical properties of the free-standing PSA films were evaluated by tensile test (Figure 8).Ultimate strength of the PSA films reduced with increase in the drying temperature or decrease in the drying humidity.This was attributed to the existence of voids and weak interfaces in fast-dried films due to lessordered particles and less opportunity for chains to interdiffuse across the particles interface.Under such circumstances, weaker mechanical properties of more inhomogeneous films with numerous small cracks were conceivable.Moreover, the samples prepared from the front of the PSA films showed higher ultimate strength than those from the center, which was in accordance with more accumulation of hard components (MMA-rich) at the front than at the center in microscopic images.
The curves of storage and loss modulus of PSAs prepared at different temperatures are shown in Figure 9.The lower storage modulus of all samples than the Dahlquist criterion (3.3 × 10 5 Pa [6] ) at low frequencies (0.1-1 s −1 ) proved that they could be regarded as PSAs, Figure 9a.At constant humidity, with increase in the drying temperature, the storage and loss moduli of the PSAs decreased in the whole studied frequency range, which were in accordance with their tensile behavior.These results, on the one hand, were attributed to the weaker mechanical properties of the films due to the incomplete interdiffusion of the polymer chains across the particles, and, on the other hand, in agreement with AFM and FESEM results, may be assigned to the intensified phase separation of the MMA-rich and BA-rich components with increase in the drying temperature.
At a constant film formation temperature, for instance, at 40°C, increase in the humidity led to an increase in the storage and loss moduli of the PSA.The observed results followed the increase in the tensile strength of the PSA with increased humidity.This was also attributed to the better interdiffusion of the chains across the particles and better dispersion of MMA-rich components in the BA-rich matrix, with increase in the humidity.Therefore, the results showed that the change in both drying temperature and humidity significantly affected the PSAs surface nanoroughness and bulk properties.

Effect of drying condition on adhesion properties
After elucidation of the PSA film formation mechanism and the effect of drying process parameters on the surface nanoroughness and bulk properties of the dried PSA films, their adhesion strengths were investigated.To do so, their thermodynamic work of adhesion to the PET substrate was calculated after the determination of the SFE of the PSAs and the substrate using the sessile drop method and Equation (5).W 12 of the adhesive on the substrate increased with the increase in the drying temperature or humidity, following increased SFE of the PSA films (Figure 10a,b).The highest and the lowest W 12 belonged to the PSAs dried at 40°C and 75% RH and 25°C and 15% RH, respectively.The effect of temperature rise on W 12 could be described by the enhanced Pe-driven accumulation of components from the bottom to the top at the film/air interface and the increased surface nanoroughness of the film could be described.In other words, an increase in the surface nanoroughness involved the presence of a higher number of components at the surface, which increased the SFE of the film. [55]Moreover, the intensified accumulation of the shorter MMA-rich copolymers with higher SFE than longer BA-rich copolymers at the surface could be considered as another cause for the increased SFE with drying temperature rise.Such surface migration of short, though high SFE components to the surface of polymeric films has been known to be entropically favorable. [56]Besides, W 12 and SFE were of higher amounts at the fronts rather than the centers in all drying conditions, which could be assigned to the intensified already stated events at the fronts compared to the centers.The higher SFE of the PSAs dried at increased humidity was attributed to the facilitated transfer of more components to the top surface through more favored migration of the surfactants to the film/humid air interface, which was confirmed by measuring a 10-fold mass of the surfactants on top of the surface of the films dried at 40°C and 75% RH compared to the one dried at the same temperature but 15% RH.
In addition to the thermodynamic work of adhesion, the microscopic adhesion of the PSAs was investigated using atomic force spectroscopy.The adhesion force (F adhesion ) recorded during retraction of the AFM tip from the PSA surface enhanced with increase in the drying temperature and humidity (Figure 11a-f).Thus, the highest and the lowest adhesion force belonged to the PSAs dried at 40°C and 75% RH, and at 25°C and 15% RH, respectively.The adhesion energy of the PSAs calculated using JKR theory (Equation 4) enhanced with increase in the surface nanoroughness of the PSA films dried at both low and high humidities (Figure 12a,b).This was assigned to increase  in the thermodynamic work of adhesion of the PSAs on the one side, and improvement in their bulk viscoelastic dissipation (due to the variation in the arrangement of the components and, as a consequence, the rheological properties) during retraction of the AFM tip on the other side (Equation 3).
Macroscopic adhesion of the PSAs was also investigated using T-peel test on PET substrate.The height image of the PET surface in Figure S10 revealed a surface roughness of 22 nm.In the standard version of this test (ASTM D1876), the film is adhered to the substrate using a 2000 g roller.However, to fully analyze the film surface nanoroughness effect in the current study, each film was adhered to the substrate using two types of 2000 g and 170 g rollers in independent tests.At both drying humidities, increase in the drying temperature enhanced the adhesion strength of the films rolled by the 2000 g roller (Figure 13a,c), in accordance with the AFM results (Figure 11).The reduced adhesion strength of the films rolled by the lighter roller (Figure 13b,d) was assigned to the insufficient wetting of the substrate by the film under slight pressure.Indeed, the reverse effect of drying temperature rise on the trend of change in the adhesion strength using two kinds of rollers implied the influence of film surface nanoroughness on the adhesion strength.It could be attributed to elimination of the weak boundary layer caused by voids [57] due to the film surface nanoroughness and improved wetting of the film on the substrate by applying sufficient contact pressure during bonding.A similar argument has been raised in the literature for the adverse effect of highly rough substrates on the adhesion strength of a soft polymer film. [1,58]Furthermore, the adhesion strength of the films was higher at the center compared to the front in any drying conditions.It could be assigned to the presence of soft BArich components at the center rather than the front and also the rougher film surface at the front, which consequently resulted in less wettability of the adhesive adhered to the substrate using the same roller for bonding the center of the film to the substrate.Increase in the drying humidity augmented the adhesion strength of the films rolled by either roller.It was attributed to the previously mentioned increase in the SFE of the film on the one side and improved dispersion and distribution of the components in the system on the other side, as a result of increased drying humidity.
To clarify the relationship between the adhesion strength and the surface nanoroughness, the viscoelastic dissipation factor, ϕ, of the films was calculated using W 12 and Equation (3) (Table 2).The viscoelastic dissipation factor of the film rolled with 2000 g roller was higher than that rolled with 170 g roller, which confirmed the unsuitable bonding of the film to the substrate under low contact pressure.Furthermore, comparing the films rolled with the standard roller, i.e., 2000 g, uncovered the higher viscoelastic dissipation factor for the films dried at 75% RH than those dried at 15% RH.It could be assigned to the higher interdiffusion of the chains across the particles interface at higher humidity because of the longer film formation time compared to the drying at lower humidity and also better distribution of the MMA-rich component in the BA-rich matrix.The inclining trend of the viscoelastic dissipation factor of the films dried at 15% RH with drying temperature rise could be referred to the monotonic increase in the PSA surface nanoroughness (Figure 5a) and thus increase in the effective contact area of the PSA with the surface of the substrate after being bonded under sufficient pressure.As a result, a larger volume of the PSA contributes in energy dissipation during the peel test (Table 2).The drying temperature rise, however, first reduced and then slightly improved the viscoelastic dissipation factor of the film at 75% RH.The declining trend of the viscoelastic dissipation factor of the films dried at 75% RH with drying temperature rise from 25°C to 35°C could be attributed to the decreased interpenetration as a consequence of shorter drying time and also to the appearance of voids in the PSA film.However, the increase in the viscoelastic dissipation factor with temperature rise from 35°C to 40°C was owing to the steep increase in the PSA film surface nanoroughness in this temperature range (Figure 5b).
To thoroughly investigate the effect of film surface nanoroughness on the adhesion strength, the ratio of the viscoelastic dissipation factors obtained for the films adhered to the substrate using 2000 g and 170 g rollers, named viscoelastic dissipation ability ( ), was calculated (Table 2).The term was plotted as a function of the film surface nanoroughness (Figure 14a) and fitted with a power function.This quantity was defined to indicate the ratio of the film viscoelastic dissipation factor after the complete wetting of the substrate to that with incomplete wetting due to its surface nanoroughness.It obtains the viscoelastic dissipation that could be represented by the film, provided that it is adhered to the substrate completely using a standard roller.
Increased film surface nanoroughness resulted in the enhancement of viscoelastic dissipation ability.In other words, by increase in the surface nanoroughness, the film would be more capable of utilizing its potentiality to dissipate the energy applied for debonding the joint.It could be deduced from the fact that the rougher the film surface, the better its wettability on the substrate under standard pressure, and the more its viscoelastic dissipation.The enhancement of the viscoelastic dissipation ability of the films with the increase in the surface nanoroughness was similar to the enhancement of the adhesion strength (Figure 14b).Indeed, the adhesion strength of the films dried at both low and high humidities and adhered to the substrate using 2000 g roller, also, followed power functions with respect to the film surface nanoroughness.Not only are the results applicable to designing PSAs, but also they could be valuable for understanding the nature of bio-inspired soft adhesion [59] such as gecko adhesion.Although it has been mentioned that viscoelasticity might play an essential role in bio-inspired soft adhesion, [60] it has been less considered, being still believed to originate mainly from intermolecular forces. [61]

Conclusions
In the current study, the effect of the most crucial process parameters, temperature, and humidity, was scrutinized on the film formation and adhesion strength of a water-based soft acrylic film.The intensified vertical concentration gradient of components in the drying latex blend film as a result of the increased drying temperature and the decreased drying humidity was under the increased Péclet number in the wet sintering mechanism along with skin formation.The drying temperature rise increased the vertical concentration gradient of components in the soft film and surface nanoroughness, surface free energy, thermodynamic work of adhesion, and adhesion strength of the soft films.However, drying humidity rise, despite reducing the surface nanoroughness and vertical concentration gradient of components, enhanced the adhesion strength and viscoelastic dissipation factor of the films as a result of the increase in the homogeneity of the components in the system, more uniform distribution of MMA-rich copolymers in the whole system, better interdiffusion of chains through the interface of particles, and increased surface free energy of the films.The adhesion strength and the viscoelastic dissipation ability of the films showed power dependence on the film surface nanoroughness.In other words, by increasing the film surface nanoroughness, a higher proportion of the adhesive could be involved in the viscoelastic dissipation during the debonding of the soft film from the substrate, which could be utilized in understanding the nature of soft adhesion.

Figure 2 .
Figure 2. The first derivative curve of the normalized heat flow of the PSA film for samples prepared from the center of the film (line) and the front of the film (dashed line) at a heating rate of 5°C/min.

Figure 3 .
Figure 3. Evolution of the mass of MBC30 latex with respect to its initial cast mass as a function of time during drying at various temperatures and (a) 15% RH and (b) 75% RH.

Figure 4 .
Figure 4. AFM phase images of the surface of the PSA films (10 μm × 10 μm) dried at various temperatures and humidities.

Figure 5 .
Figure 5. Surface nanoroughness of the PSA films dried at (a) 15% RH and (b) 75% RH versus drying temperature for the center (square) and front (circle) of the films and the average value (triangle).

Figure 6 .
Figure 6.The Marangoni number of the PSA films dried at various temperatures and (a) 15% RH and (b) 75% RH.

Figure 7 .
Figure 7. Cross-sectional FESEM images of the PSA films dried at (a and b) 15% RH and 25°C, (c and d) 15% RH and 40°C, (e and f) 75% RH and 25°C, and (g and h) 75% RH and 40°C (the designations in the parentheses stand for the center and front of the films, respectively).Images from the top to bottom belong to the cross-sections close to the top surface, at the center, and close to the bottom of the film, respectively.

Figure 8 .
Figure 8. Stress-strain curves of PSAs dried at various temperatures and (a, c, e) 15% RH and (b, d, f) 75% RH for samples prepared from the center (thick line) and front (thin line) of the films.

Figure 9 .
Figure 9. (a) Storage modulus and (b) loss modulus of the PSAs films dried at various temperatures.

Figure 10 .
Figure 10.Thermodynamic work of adhesion of the PSAs dried at various temperatures and (a) 15% RH and (b) 75% RH on PET.

Figure 11 .
Figure 11.Force-displacement curves obtained by atomic force spectroscopy on the PSAs dried at various temperatures and (a, c, and e) 15% RH and (b, d, and f) 75% RH for samples prepared from the center and front of the films.

Figure 12 .
Figure 12.Microscopic adhesion energy of the PSAs dried at various temperatures and (a) 15% RH and (b) 75% RH obtained by AFM.

Figure 13 .
Figure 13.The adhesion strength of the PSAs dried at various temperatures and (a, b) 15% RH and (c, d) 75% RH to the substrate.The PSAs were rolled using a (a, c) 2000 g and (b, d) 170 g roller.

Figure 14 .
Figure 14.(a) the viscoelastic dissipation ability and (b) the adhesion strength of the PSAs dried at various temperatures and 15% RH (square) and 75% RH (circle) and adhered to the PET substrate.The dashed and solid lines represent the fitted curves.

Table 1 .
The calculated values of � λ and Pe in the MBC30 latex drying process at various temperatures and humidities.