Influences of Substrate Adhesion and Particle Size on the Shape

Subscriber access provided by SUNY DOWNSTATE

Article

Influences of Substrate Adhesion and Particle Size on the Shape Memory Effect of Polystyrene Particles Lewis Michael Cox, Jason P. Killgore, Zhengwei Li, Rong Long, Aric W. Sanders, Jianliang Xiao, and Yifu Ding Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.6b00588 • Publication Date (Web): 29 Mar 2016 Downloaded from http://pubs.acs.org on April 2, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Influences of Substrate Adhesion and Particle Size on the Shape Memory Effect of Polystyrene Particles Lewis M. Cox,1,* Jason P. Killgore,2 Zhengwei Li,1 Rong Long,1 Aric W. Sanders,3 Jianliang Xiao,1 and Yifu Ding 1,4,*

1

Department of Mechanical Engineering University of Colorado at Boulder Boulder, CO 80309-0427, USA E-mail: [email protected]

2

Applied Chemicals and Materials Division

National Institute of Standards and Technology Boulder, CO 80305, USA

ACS Paragon Plus Environment

1

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

3

Page 2 of 29

Radio Frequency Technology Division

National Institute of Standards and Technology Boulder, CO 80305, USA

4

Material Science and Engineering Program University of Colorado at Boulder Boulder, CO 80309-0427, USA

KEYWORDS: Shape Memory, Polymer Particles, Particle Adhesion

Abstract

Formulations and applications of micro- and nano-scale polymer particles have proliferated rapidly in recent years, yet knowledge of their mechanical behavior has not grown accordingly. In this study we examine the ways that compressive strain, substrate surface energy, and particle size influence the shape memory cycle of polystyrene particles. Using nanoimprint lithography, differently sized particles are programmed into highly deformed, temporary shapes in contact with substrates of differing surface energies. Atomic force microscopy is used to obtain in situ measurements of particle shape recovery kinetics and scanning electron microscopy is employed to assess differences in the profiles of particles at the conclusion of the shape memory cycle.


2

Page 3 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Finally, finite element models are used to investigate the growing impact of surface energies at smaller length scales. Results reveal that the influence of substrate adhesion on particle recovery is size-dependent, and can become dominating at sub-micron length scales.

Introduction Tremendous research efforts have been spent on polymer particles in recent years, driven by the need for better drug carriers, biomarkers, building blocks for self-assembly, and sensors.1–10 This has brought unprecedented control over the structure and chemistry of the particles, highlighted by recent work on Janus and patchy particles.11 Particle shapes and the ability to change them when needed are important for a range of bio-applications.12–14 Hydrogel particles, which exhibit volume-phase transition (VPT) in water, are the most common type of stimuliresponsive particles.15–18 However, the VPT of a hydrogel particle causes simultaneous shape and volume change, as well as dramatic changes in the physical properties of the particle (e.g. modulus of a hydrogel particle can decrease from a few GPa to less than 1 MPa upon water uptake). In contrast, solventless, thermally-responsive particles, based on a liquid crystalline polymer, which do not exhibit a compromising of mechanical properties have also been reported.19 However, this approach requires specific liquid crystal chemistry, and achieves only a small degree of shape change between the ellipsoidal and spherical geometries. Recently, solventless, large shape changes in polymer colloidal particles were demonstrated based on the shape memory effect. Micrometer-sized, lightly crosslinked polystyrene (PS) particles were thermomechanically programmed using nanoimprint lithography (NIL). The particles assume highly deformed disc-like shapes with large compressive strains and can recover much of the original spherical shape upon heating or solvent exposure.11 This approach


3

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 29

is attractive because it is based on the entropic elasticity inherent to polymer networks, does not require (although it can possibly employ) use of solvents or complex chemistries, maintains a constant volume, and does not compromise the mechanical properties of the particles. To achieve controllable shape changes using this method, it is vital to understand the influences that both particle properties and conditions of programming and recovery have on the shapes of the particles. Particularly, the interfacial forces, which are negligible in the bulk shape memory effect, become increasingly dominant with a reduction in particle size. Currently, knowledge of the mechanical behaviors of micrometer- or sub-micrometer sized particles is scarce, even for particles that are structurally and chemically homogeneous. Recently, indentation based mechanical measurements on glassy polymer particles have revealed a strong particle-size dependency of the particle stiffness,20 and surprisingly high degree of plasticity under compression even at room temperature.21 Employing a soft colloidal probe atomic force microscopy technique, Erath et al. measured the stiffness of a micrometer-sized, rubbery polydimethylsiloxane (PDMS) particle and the resulting adhesion between it and surfaces with varying chemistries.22 However, the deformation of the PDMS particle was limited to a strain level of ≈ 5 % and only a single particle size was investigated. In this study, we systematically examined the constrained recovery behaviors of lightly crosslinked PS shape memory particles, as a function of programming strain, substrate surface energy, and particle size. Results reveal that the influence of substrate adhesion on the particle recovery is size-dependent, and can become dominant at sub-micron length scales.

Experimental Section


4

Page 5 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Materials. Two different cross-linked PS particles were used in this study, both of which were dispersed in aqueous solutions containing ≈ 1% particles by mass. The larger PS particles have a 4.25 µm diameter and a 2 % divinylbenzene (DVB) cross-linker density (PS05N/5689, Bangs Laboratory Inc.). The smaller PS particles have a diameter of 0.2 µm and a DVB cross-linker density of 3 % (S37392, Life Technologies). Both particles were used as received.

Thermomechanical Programming of PS Particles. PS particles were deposited onto flat silicon wafers by dipping the wafer into a particle suspension and then drawing it out at a rate of approximately 1 mm s-1. After deposition of particles, some of substrates were then modified with a low-surface-energy, self-assembled monolayer (SAM) of CF3(CF2)5(CH2)2SiCl3 (tridecafluoro-1,1,2,2-tetrahydrooctyltrichlorosilane,

Sigma-Aldrich)

to

modify

adhesion

between particles and substrate. Henceforth these modified surfaces will simply be referred to as fluorinated SAM (F-SAM) surfaces. PS Particles were thermomechanically programmed as schematically shown in Fig. 1, and discussed in greater details in our recent report

11

. The PS particles were compressed between

two parallel rigid plates: a Si substrate that contained the PS particles and another Si superstrate. The Si superstrate was first cleaned using piranha solution and then coated with the same F-SAM previously mentioned above to allow the superstrate release between the superstrate and the compressed PS particles after the programming process. All thermomechanical deformations were performed on a nanoimprinter (Eitre 3, Obducat) at 120 °C under a pressure of 1.5 MPa for 5 minutes. After this time period, the compressive force was maintained while the particles were cooled back to room temperature, at which point the pressure was relieved and the Si superstrate was removed from the compressed PS particles (Fig. 1). While most compressions were


5

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 29

performed using superstrates with a flat surface, in one demonstration NIL was performed using a superstrate with a line-and-space grating pattern was employed.

Figure 1. Schematic illustration of thermomechanically programming PS particles using NIL.

The PS particles remained in the temporary shape indefinitely at room temperature and recover their permanent shapes once heated to a temperature above their glass transition temperature (Tg). The geometric information of the PS particles at different stages of the shape memory cycle (i.e. permanent, temporary or programmed, and recovered) was obtained by atomic force microscopy (AFM, MFP 3-D Stand Alone AFM, Asylum) and scanning electron microscope (SEM, Zeiss, Supra 60) measurements. AFM measurements were performed on particles that had been recovered at a constant temperature for different lengths of time at a temperature of 120 °C. Additionally, in situ recovery of the programmed particles was examined using AFM measurements while continually heating the particles from 90 °C to 130 °C at a rate of 0.5 °C/min. For the SEM measurements, the PS particles were coated with a ≈ 5 nm thick Au layer to minimize the charging effect.


6

Page 7 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Finite Element Analysis of the “Programming” of the PS Particles. An axisymmetric finite element model was established using ABAQUS to obtain the total elastic energy stored in different sized particles under different compressive strains. For simulation, four-node bilinear axisymmetric quadrilateral (CAX4) elements are used to discretize the geometry of the polymer particle, and refined meshes were adopted to ensure the accuracy. The Young’s modulus and Poisson’s ratio used are E = 1.8 MPa and ν =0.49, respectively, for the polymer particle heated above Tg.23 The silicon superstrate is considered to be rigid, since it is much stiffer than the polymer particle, and thus its deformation is negligible. “Hard contact” is implemented to simulate the interface between the silicon superstrate and the polymer particle. Finally, displacement boundary conditions are assigned to the rigid plate to apply different levels of compression.

Results and Discussion As we reported recently, PS micro-particles display reliable shape memory effect.11 A typical shape memory cycle consists of two primary stages: Programming a deformed (temporary) shape and recovering the original (permanent) shape. The two different sized PS particles studied here were lightly crosslinked with a DVB concentration of either 2 % or 3 %. With such a low concentration of crosslinker, the Tg of the PS particles will be only slightly higher than that of the bulk PS, which is ≈ 100 °C. Recent studies by Lu et al. showed that the Tg of PS particles containing a larger, 10 % DVB concentration was still less than 120 °C.24 For the smaller 0.2 µm particles free-surface chain mobility and substrate confinement effects may also have some effect on Tg, that is otherwise negligible in the larger particles.25 Nonetheless, both factors (low degree of crosslinking and the confinement effect) did not inhibit thermomechanical programming of


7

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 29

the PS particles. Specifically, both particle sizes displayed rubbery-like behaviors and were easily deformed during the programming process at 120 °C. After removal of the superstrates at room temperature both particles exhibited flattened programmed shapes. In the following, we describe the detailed analysis of the shape memory behaviors of both sizes of PS particles on both SiOx and F-SAM surfaces. Figure 2 shows the temperature-dependent particle heights from in situ AFM scans as the programmed particles were heated from 90 °C to 130 °C at a slow temperature ramp of 0.5 °C/min. Before annealing, the 4.25 µm particles were programmed into flattened discs with a compressive strain εp = abs((h- h0)/h0) ≈ 0.77, where h and h0 were the height of the programmed and undeformed PS particles (Fig. 2a), respectively. Detailed analysis of the shapes of the compressed PS particles confirmed that they behave similar to incompressible bulk elastomeric particles.26 Therefore, in this report, we only use the heights of the particles as a sole quantitative measure of particle shape at different stages of the shape memory cycle. No particle fracture or crazing was observed during the programming even for εp > 0.90, and PS particles were able to retain the programmed shape at temperatures below 100 °C. Such high ductility of the PS particles was attributed to both the nature of compressive deformation and the low degree of crosslinking of the particles. He et al. reported that even at room temperature, 5 µm PS particles crosslinked with 2 % and 3 % DVB were able to plastically deform with compressive strains up to εp = 0.8.27 No fractures were observed for their particles, albeit some crazes were observed near the circumferences of the compressed particles.


8

Page 9 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Figure 2. Heights h of the PS particles, normalized by the undeformed heights, h0, as a function of temperature when annealed at a heating rate of 0.5 °C/min. A) and B) are data for the 4.25 µm and 0.2 µm PS particles, respectively. Filled and empty symbols correspond to the recovery of the PS particles on SiOx and F-SAM substrates, respectively.

As shown in Fig. 2a, the degree of recovery h/h0 of the PS particles started to increase as temperature increased above 100 °C, driven by the entropic elastic energy stored in the deformed PS network. Between the SiOx and F-SAM substrates, the recovery of the larger PS particles was similar at the early stage. On both substrates, the temperature at the onset of recovery was between 102 °C and 104 °C, which coincides with the bulk Tg of high molecular weight PS.28 The later stage of the particle recovery was distinctively different between the two substrates. On the F-SAM substrates, the 4.25 µm PS particle had recovered h/h0 ≈ 0.7 at 115 °C whereas h/h0 ≈ 0.7 was not achieved on the SiOx substrate until 130 °C. Furthermore, recovery on the FSAM achieves a constant a value of h/h0 ≈ 0.82 at a temperature of 126 °C while the particle on the SiOx surface exhibited a gradual slowing of recovery, resulting in h/h0 ≈ 0.70 at the end of the experimental time window.


9

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 29

Figure 2b presents the recovery kinetics for the 0.2 µm PS particles on both substrates. Similar to the larger PS particle systems, the programmed particles were able to fix and retain their temporary shape (with εp ≈ 0.5 at temperatures below 100 oC. However, the onset recovery temperature of the PS particles on the F-SAM substrate was ~3 °C lower than that on the SiOx surface. Note that the programmed 0.2 µm particles exhibit h ≈ 100 nm, which is in the heightregime where a thickness-dependence of Tg for PS films has been observed

29–33

. Both the free

surface of the PS and the repulsive interface between the F-SAM layer and the PS particle should cause reduction in the Tg of PS. In comparison, on the SiOx surface, the free surface effect could be counter-balanced by the attractive interactions at the particle-substrate interface, which are known to increase Tg. As a result, the onset recovery temperature of the 0.2 µm particles on the SiOx surface appeared similar to that of the 4.25 µm particles (Fig. 2a). For the larger particles, the volume of near surface material expected to exhibit free-surface or confinement effects would be negligible resulting in no difference in onset temperature between substrates. The recoveries of the 0.2 µm particles on both substrates quickly plateaued after reaching 110 °C. In comparison, the 4.25 µm particles on SiOx continued to recover even above 120 °C. The 0.2 µm particle on the F-SAM exhibited h/h0 ≈ 0.75, in comparison to h/h0 ≈ 0.60 on the SiOx surface. Based on these in situ recovery experiments, the particles initiated recovery at ≈ 100 °C and reached steady state (or close to) at 120 °C, a period of time that spanned 40 min. Accordingly, for the following experiment, we chose to examine particle recovery after annealing at 120 °C for 1 hr. Figure 3 summarizes the experimental findings for degree of recovery in PS particles as a function of the compressive strain achieved after the thermomechanical programming process. During the programming process a 1.5 MPa pressure is applied to the superstrate and not to


10

Page 11 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

individual particles. Since the areal density of deposited particles varies after dip coating, a wide range of compressive strains is achieved in a single iteration. Each of the PS particles shown in Fig. 3 was characterized individually with the AFM.

Figure 3. Recovered PS particle heights, normalized by the heights of the undeformed particles, as a function of compressive strain exerted after thermomechanical programming, for a) 4.25 µm and b) 0.2 µm particle systems. Different symbols represent substrate surface chemistry (native oxide or F-SAM surface), as well as different annealing time (1 and 16 hr) used. The dotted line represents the expected trend if no particle recovery occurred after annealing.

On the SiOx substrates, the 4.25 µm particles with small compressive strains (εp < 0.4) recovered ~ h/h0 = 0.90 of the original particle height after annealing (Fig. 3a). For particles with compressive strains εp > 0.4, a slight decrease of particle recovery with increase of compressive strain is observed (m = -0.17, the slope of a linear fit, N = 17). While the difference is slight, at larger compression ratios (εp > 0.6) the recovery of 4.25 µm on SiOx ceases to exhibit a clear


11

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 29

dependency on compressive strain. Specifically, the PS particle with a compressive strain of εp = 0.9 recovered h/h0 =0.74 of the original particle height. In comparison, the recovered particle height on the F-SAM substrate (h/h0 = 0.86 ± 0.02, average ± standard deviation, N = 13) appeared independent of the compressive strain of the programmed particles. Increasing the annealing time from 1 hr to 16 hr indeed caused more recovery for the particles on the SiOx surface, consistent with observations in Fig. 2a. However, the recovered heights were still below that achieved on the F-SAM substrate. Additional annealing time for particles on the F-SAM substrate resulted in no change in recovery and therefore such data are not shown. Figure 3b presents the results obtained for the 0.2 µm particles. On the F-SAM substrates, the recovered particle heights (h/h0 = 0.81 ± 0.02, N = 16) appeared independent of the compressive strain of the programmed particles, which is similar to that observed for the 4.25 µm particles, although the total recovery is slightly less. In stark contrast, the recovered particle heights on the SiOx surface dramatically decreased with increase of compressive strain, with m = -1.36 (the slope of a linear fit, N = 19), which is 8 times greater compression-dependence than that observed for the 4.25 µm particles on the SiOx substrate. Figure 4 presents side-view images of the recovered 4.25 µm and 0.2 µm PS particles on both SiOx and F-SAM substrates. In the 4.25 µm particle systems, a very small, yet highly deformed necking shape appears at the point where the particle edge contacts the substrate (marked by the arrows in Fig. 4). Using the mid-points of the neck in the SEM images, the local contact angles between the PS particle and the SiOx substrate were approximately 29° ± 6° (N = 5) for the 4.25 µm particles. The residual contact, represented by the ratio between the contact radius a and the particle diameter (2R), a/2R ≈ 1.10 for the 4.25 µm particles. In comparison, no neck formation was observed for the 4.25 µm particles on the F-SAM surfaces. The contact angle was measured


12

Page 13 of 29

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

to be 130° ± 7° (N = 5), which is significantly larger than that on the SiOx surface. The residual contact of the recovered particle was a/2R ≈ 0.54, much smaller than that on the SiOx surface. However, since the neck accounts for only a very small portion of the particles’ volume, the recovered particle heights on both substrates do not differ significantly. For the 0.2 µm particles, no clear neck was observed on either SiOx or F-SAM surfaces. Quantitatively, the contact angles were measured to be 58 ± 3° (N=5) on the SiOx surface and 87 ± 1° (N = 5) on the F-SAM surface, with corresponding residual contact of a/2R ≈ 1.54 and a/2R ≈ 1.13, respectively.


13

Langmuir

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 29

Figure 4. Side-view SEM images of recovered 4.25 and 0.2 µm particles, taken at 90° and 80° angles, respectively. The inset shows the magnified view of the necking area including the profile of the 0.2 µm particles.

The above experimental findings indicate that the size of the particles and the contact adhesion between the substrate and the PS particles significantly affect the recovered particle heights and shapes.

Understanding this behavior requires analysis of the interplay between adhesion and

elastic energy. Adhesion energy will be predicted using the Johnson-Kendall-Roberts (JKR) theory. The JKR theory was derived for infinitesimal deformations (δ

Influences of Substrate Adhesion and Particle Size on the Shape

Recommend Documents