Tunable Piezoresistivity from Magnetically Aligned Ni(Core)@Ag(Shell

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC

Applications of Polymer, Composite, and Coating Materials

Tunable Piezoresistivity from Magnetically Aligned Ni(core)@Ag(shell) Particles in an Elastomer Matrix Fang Peng, Keke Chen, Armen Yildirim, Xuhui Xia, Bryan D. Vogt, and Mukerrem Miko Cakmak ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.9b04287 • Publication Date (Web): 09 May 2019 Downloaded from http://pubs.acs.org on May 9, 2019

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 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 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.

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 36 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

ACS Applied Materials & Interfaces

Tunable Piezoresistivity from Magnetically Aligned Ni(core)@Ag(shell) Particles in an Elastomer Matrix Fang Peng†, Keke Chen†, Armen Yildirim#, Xuhui Xia†, Bryan D. Vogt†,*, Mukerrem Miko Cakmak#* †Department

#Departments

of Polymer Engineering, University of Akron, Akron OH 44325

of Materials Engineering and Mechanical Engineering, Purdue University, West Lafayette, IN 47907

KEYWORDS. “Z” alignment; PDMS; roll-to-roll; field induced surface topography; responsive composite

ABSTRACT: Core-shell (Ni@Ag) particles are aligned through the thickness of a poly(dimethylsiloxane) (PDMS) film using a magnetic field in a continuous roll-to-roll

ACS Paragon Plus Environment

1

ACS Applied Materials & Interfaces 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 2 of 36

process. The alignment of the particles dramatically decreases the percolation threshold for electrical conductivity through the thickness of the film by nearly an order of magnitude from 28 vol% without the field to 1 vol% with 52 mT magnetic field during curing. However, the magnetic forces lead to rough surface topography for intermediate Ni@Ag loadings, but confining the Ni@Ag/PDMS composite by a glass constraint provides a smooth surface. This difference in geometry changes the morphology of the vertically aligned “chains” of Ni@Ag particles where the “chains” are more aggregated when the film is unconstrained. As the Ni@Ag concentration is decreased below 3.6% for the constrained film, breaks in the aligned particles evident from X-ray tomography lead to pressure sensitive resistance across that film with a large decrease in resistance above a threshold pressure. The threshold pressure is demonstrated to be controllable from  15 to 290 kPa through the loading of aligned Ni@Ag in the PDMS, but this threshold pressure decreases on cyclic loading. These magnetically aligned composites represent a facile route to mechanically responsive films that could be used in a variety of applications where cyclic loading above and below the threshold pressure is not required, such as disposable pressure sensors for ensuring reliability of products through transportation and embedded structural health monitoring for identifying critical displacements.

INTRODUCTION


2

Page 3 of 36 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


Continuous roll-to-roll (R2R) processing has been actively explored for manufacture of a variety of emergent technologies, such as polymer photovoltaics,1 flexible electronics,2 smart packaging,3 and functional films,4, 5 due to its potential for high production rates and low manufacturing costs. A typical R2R process involves unwinding a substrate material from a roll that is subsequently modified by coating,6 deposition,7 printing,8 patterning,9 or other functionalization10 of the unwound substrates in a continuous process. For functional devices, this typically involves multiple processes and steps where components are built in layers on the substrate, which necessitate registry and alignment.11 Self-aligned imprint lithography provides one R2R compatible solution to overcome registry challenges.12 These lithographic methods are acceptable for high value products, but there opportunities for R2R production of functional films for commodity applications if the costs are sufficiently low. Self-assembly provide one potential avenue to low cost materials and has been demonstrated on flexible substrates.13 However structures when anisotropic, these tend to orient in the plane of the substrate, which commonly is counter to the desired orientation.14 Controlling orientation in self-assembled structures has been explored extensively in terms of directed self-assembly for nanofabrication in microelectronics.15, 16 One low cost route for desired self-assembled nanostructures is through field assisted alignment of block copolymers using electric,17 magnetic,18-20 or thermal21 fields. In addition to providing orientation to self-assembled materials, the application of external fields can


3


Page 4 of 36

be used to generate structures in polymer blends22 and polymer nanocomposites.23 For example, the magnetic alignment of carbon nanotube containing polymer composites significantly enhance their thermal and electrical properties.24 The alignment of iron particles in an elastomeric matrix provides tunability of the elastic properties for the composite.25 External fields, such as electric and magnetic, provide opportunities to manipulate properties without changing the composition.21 Functionalized magnetic particles at oil/water interfaces can be directed into 2-D assemblies in the presence external magnetic field.26 By polymerized the oil phase, the directed assembled particles can be locked in place through “fossilized liquid assembly” to directly demonstrate how these particles respond to the magnetic field.26 The fabrication of these types of field directed materials can be scalable with recent advances with roll-to-roll (R2R) tools that are equipped with electric and magnetic field stations to provide continuous production of aligned materials that has been recently demonstrated in our laboratories.27 For R2R processes, the materials are typically cast onto substrates; the casting process tends to orient anisotropic particles parallel to the substrate, but subsequent application of fields can provide “Z” (out of plane) orientation.28 For example, dielectric particles, such as BaTiO3, polarized in an applied electric field tend to align themselves into chain-structures through their dipole-dipole interactions and provide a route to “Z” direction alignment of BaTiO3 particles29 in a continuous R2R process. Similarly for highly anisotropic particles such as graphite30 or clay,31 an external


4

Page 5 of 36 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


electric field can preferentially align the long axis through the thickness of the film to yield “Z” orientation. An external magnetic field can also provide “Z” direction alignment of magnetically responsive particles, such as Ni.32 The magnetic alignment of electrically conductive particles can lead to conductivity through the thickness of the film at concentrations well below the random percolation threshold due to the organization of the particles by the field.32 For glassy polymers, this alignment leads to a permanently conductive material with conductivity only in thickness direction with no in-plane conductivity as the conductive nanocolumns are separated by an insulating polymer matrix preventing electron transport in the plane. However, there are many cases where elastomers are desired to enable greater deformation of the functional material, such as desired for stretchable electronics33 and electronic skin.34 When considering filled elastomers, the ability of the network to deform, which can lead to re-arrangement of the filler material, can lead to variable properties35 and the mechanical mismatch between the filler and matrix can limit the durability.34 There have been significant efforts to minimize the variability in the electronic properties for stretchable electronics devices, which has included conductive self-organized pathways36 and oriented nanostructures37 that can deform without loss of conductivity. However, the variability in properties from deformation can be beneficial for some applications. For example, elastomeric polymers filled with anisotropic conductive fillers can exhibit piezoresistivity based on the deformation of the elastic matrix to induce or break percolation


5


Page 6 of 36

for conductive fillers.38-40 One of the most common matrices is polydimethylsiloxane (PDMS), which has been loaded with gold nanowires38, silver nanowires39 or carbon nanotubes40 to generate piezoresistive elastomers. These materials have been demonstrated as structural health monitors41 and wearable strain sensors42. However, one of the drawbacks of these approaches is the high cost and limited commercial availability of the metal nanowires, while the purity, length distribution, and chirality of the carbon nanotubes43 are important considerations for obtaining repeatable performance. These limitations are particularly problematic for cost sensitive applications of piezoresistive composites. Previously, magnetic field aligned PDMS/Ni particles composites with low particle concentration were demonstrated to exhibit piezoresistive behavior,44 but these materials suffered from high resistivity (105 ohm·m) in the “conductive” state and required pressures on the order of GPa in order for the switch between insulating and conducting properties, which limits their utility for many applications.

In this work, we present a scalable R2R process to fabricate elastomeric films with tunable piezo-resistivity through the use of magnetic fields to align nickel (core)-silver (shell) particles (Ni@Ag) in a PDMS matrix. The Ni@Ag particles are commercial, low cost filler materials that were developed to decrease the cost of Ag fillers for conductive adhesives for photovoltaic applications. The application of the external magnetic field leads to alignment of the particles into long chain structures that are orientated with their


6

Page 7 of 36 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


long axis along “Z” direction of the film. These chains of Ni@Ag particles are in loose contact with the connectivity increasing as the loading of Ni@Ag in the PDMS increases. This structure results in electrical conductivity through the thickness of the film when a threshold pressure is applied. The threshold pressure for electrical conductivity is strongly dependent on the Ni@Ag particle content, which modulates the threshold from 0 kPa to 70 kPa as the Ni@Ag loading decreases from 5.3 vol% to 0.25 vol%. The ability to use commercially available, low cost materials to generate tunable pressure sensors through a simple scalable R2R fabrication approach could be enabling for low cost sensor and electrical switch applications in stretchable, flexible, and wearable electronics.

EXPERIMENTAL SECTION Materials. Sylgard 184 Silicone Elastomer was purchased from Dow Corning Corporation. Silver coated nickel (Ni@Ag) particles containing 15 wt% Ag were purchased from Novamet Specialty Products Corp., USA. The average particle size was 49 ± 12 μm (see Figure S1). The substrate for the R2R line was thermally stabilized poly(ethylene terephthalate) (PET, 125 μm thick, Terphane, Inc.). Particle alignment and film fabrication. Sylgard 184 Silicone Elastomer base and curing agent was mixed at a ratio of 10:1 by a planetary vacuum mixer (Thinky ARV-310) at 1500 rpm for 5 min. The PDMS was then pre-cured at 50°C for 22 min and then mixed with the Ni@Ag particles at the desired volume fraction (0.25 vol% to 5.3 vol%) using the planetary


7


Page 8 of 36

vacuum mixer at 1500 rpm under vacuum for 5 min. This composite mixture was then cast within 1 min to avoid significant variation in the cure state of the PDMS during the alignment. Two different processing procedures were used to generate samples with flat and textured topography as shown in Figure 1. For fabricating flat samples, the PDMS / Ni@Ag mixture was cast into a 1 mm deep glass mold and covered with a Mylar sheet on the roll-to-roll processing line27 as shown in Figure 1A. A pair of electromagnets applied a uniform magnetic field to the PDMS/Ni@Ag mixture along its thickness direction to produce “Z” alignment of the particles as shown in Figure S2. Within the magnetic field (typically 52 mT unless otherwise noted), the PDMS/Ni@Ag mixture was heat cured by hot air (70 °C) for 60 min. To demonstrate that the alignment did not require an expensive electromagnet, an analogous alignment process was investigated where the same Mylar capped PDMS/Ni@Ag mixture was aligned between a pair of 3”1” 1/4" neodymium magnets (Amazon) during heat-curing in an oven at 90 °C for 60 min.


8

Page 9 of 36 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


Figure 1. Schematic illustration of the two processes used to generate magnetically aligned Ni@Ag in PDMS: (A) molded materials cast within a glass mold and covered with Mylar to generate flat samples and (B) unconstrained roll-to-roll alignment of PDMS/ Ni@Ag where the material is produced in a continuous manner.

This processing results in a well-defined morphology for the composite, but this processing route with a mold does not lead to a readily manufacturable material. Aligned films of Ni@Ag in PDMS were also fabricated in a continuous process on the R2R machine using the electromagnet (Figure S2).27 The PDMS/Ni@Ag mixture (prepared as described previously) was cast using a doctor blade with 1 mm gap onto the moving PET substrate (4 cm/min). The mixture was simultaneously aligned in the electromagnets with a magnetic field of 52 mT and heat-cured with hot air (140 °C) as shown pictorially in Figure 1B. The cured composite was removed by simply peeling from the PET substrate to generate free standing films approximately 1 mm thick.

Characterization. The structure of the assembled particles within the composite films was assessed using an X-ray MicroCT scanner (Bruker Skyscan1172) operating at 50 kV and 200 μA. The large difference in electron density between the PDMS and the Ni@Ag particles enabled elucidation of the 3D distribution of the Ni@Ag particles within the film. Transmission X-ray images were recorded at 0.3° rotational steps over 360° of rotation.


9


Page 10 of 36

NRecon software was used to reconstruct the cross-section images, which were then imported into the Skyscan CT Analyzer (V1.1) software to construct the full 3D morphology.

The surface morphologies and cross-sections of the films were characterized by using an optical microscope (Olympus CX31) and/or field-emission scanning electron microscopy (FESEM, JEOL-7401). The cross-sections were prepared by cryofracture through immersed of the sample in liquid nitrogen and the sample was then broken by deflection. The cryofracture surfaces of the samples were then assessed using SEM microscopy. Before the SEM imaging, the top surface and cross-sections of the samples were sputter coated with silver to prevent surface charging during the electron microscopy measurements. To better assess the alignment of particles through the full thickness, optical images were obtained of both the top and bottom of the PDMS/Ni@Ag films in transmission mode.

The simultaneous measurements of mechanical and electrical properties of the PDMS/Ni@ Ag were carried out using a customized tensile machine reported elsewhere32, 44

with a custom designed compression clamp shown in Figure S3. This compression

clamp mounted on the tensile machine was designed to compress the sample between the two electrodes during travel of the two crossheads in opposite directions. The bottom


10

Page 11 of 36 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


electrode was installed on a compression spring (measured spring constant= 11.389 kN/m), which allows a linear increase of the applied force during compression. During testing, the through-thickness resistance was measured using a Keithley 6487 Picoameter integrated into the data acquisition system. These measurements were performed in compression at 1 mm/min using 10mm  10mm  1mm PDMS/Ni@Ag films placed between two copper sheets. DupontTM PE410 ink-jet silver ink was applied between the copper sheets and cured for 20 min at 80°C to improve the electrical contact of the PDMS/Ni@Ag films with the electrodes.

RESULTS AND DISCUSSION The application of the magnetic field during the curing process of the PDMS dramatically impacts the morphology of the composite as shown in Figure 2. The microcomputed X-ray tomography images show that the Ni@Ag particles sediment during the curing of the PDMS (Figure 2A) without any applied field. This leads to a surface that is devoid of any particles (Figure 2B). This morphology leads to low electrical surface resistivity (0.262 in ohm/sq) on the bottom surface due to the formation of percolated network of Ni@Ag that have settled during curing. However, the resistivity through the thickness of the sample (> 1010 ohm·m) is very large due to the insulating characteristics of PDMS; a similar resistivity is found when considering the top surface of the film. With the application of a low strength magnetic field (52 mT) using an electromagnet (Figure


11


Page 12 of 36

S2), the morphology of the composite is drastically altered as shown in Figure 2C. In this case, the Ni@Ag are aligned into chains through the thickness of the film (Figure 2D). This alignment leads to high surface resistivity on both side due to the isolation of the particles in the plane of the film, while the through-thickness resistivity is low due to the percolation of particles in the chains. Increasing the magnetic field strength to 225 mT leads to a limited change in the morphology or electrical properties of the composite (Figure 2E). There is a narrowing of the clustered chains of Ni@Ag particles at the higher field strength (Figure 2F).


12

Page 13 of 36 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


Figure 2. X-ray tomography images of composite films containing 3.6 vol % Ni@Ag particles in PDMS and cured at 70 °C for 1 h using (A) no magnetic field with dashed line being a 0.3 mm cross-section slice shown in (B), (C) 52 mT field with the electromagnet with (D) 0.3 mm thick cross-section of this aligned sample, (E) 225 mT field with the electromagnet with (F) 0.3 mm thick cross-section of this aligned sample. This change in morphology can be explained by the balance between the surface energy of the columnar domains of particles and the demagnetization energy.45, 46 From these thermodynamic arguments, the size of these domains should scale as the field strength to the -2/3 power.45 This scaling is consistent with the decrease in the size of the columns of Ni@Ag at increased magnetic field strength in comparison of Figure 2C and 2E). The Ni@Ag particles can also be well aligned using a permanent magnet as well as shown in Figure S4 as the magnetic field strength required for assembly (phase separation) of the particles is relatively small for the concentrations examined.45 The morphology of these structures should be controlled by the combination of magnetic field strength, concentration of the particles, their magnetic moment, and the thickness,45,

46

which

provide significant processing space to control the morphology and associated properties of these assembled composites. In order to better understand the electrical conductivity through the thickness of the aligned films, samples were prepared using different filler loadings but the same magnetic field strength (52 mT) and gap thickness (1 mm). As shown in Figure 3A, all samples exhibit


13


Page 14 of 36

a resistivity lower than reported volume resistivity (2.91012 ohm·m) of Sylgard 184 PDMS,47 even at sub-threshold pressures. This is due to the aligned Ni@Ag particle columns that nearly percolate through the thickness direction. The measured resistivity of the composite film containing 0.25 vol% particles ranges from 1010 – 1012 ohm·m during the compression. At this low loading, the composite is not piezoresistive within the range of applied pressure examined (109 ohm·m.


14

Page 15 of 36 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


The relationship between the resistivity () of a polymer/conductive particle composite and applied pressure has been reported to follow:50 ρ(𝑃)

ρ(0) = 𝑒

―2𝜆Δ𝑑

= 𝑒 ―2𝑃𝜆𝑑/𝐸

(1)

where (P) is the resistivity at applied pressure P, 𝜆 is the tunneling parameter when assuming identical tunneling junctions for all particles, d is the interparticle distance, and E is the modulus of the composite. PDMS filled with aligned conductive particles at loadings of ~ 15 % followed this scaling.50 For the composites examined here, this resistivity-pressure relationship does not hold with a threshold stress required to observe an appreciable decrease in resistance.


15


Page 16 of 36

Figure 3. (A) Piezo-resistivity of aligned (flat) Ni@Ag composites for different particle loadings in the composite. Measurements were performed on loading (increasing pressure). (B) Threshold critical pressure associated with conductivity percolation of the aligned Ni@Ag composites. To understand the origins for the critical pressure in these Ni@Ag composites, the morphology of Ni@Ag particles was investigated using optical microscopy as shown in Figure 4. At all concentrations, both the top and bottom surfaces show exposed particles


16

Page 17 of 36 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


(black regions with clear edges). However, at low concentration (0.25 - 1 vol%), significantly more particles are exposed at the bottom than at the top surface. As shown in Figures 4A-4B, the exposed versus buried (see red circles) particles can be identified as the surface is in focus and the particles that are buried in the matrix are out of focus in the micrographs. The fraction of out-of-focus particles is significantly larger for the top of the film than the bottom, which suggest the formation of aligned particle columns that are too short to percolate through the thickness of the sample. As shown in Figures 4C4D, these buried particles become less common at the top surface at higher concentrations (2-3.6 vol% Ni@Ag), but the particles appear to aggregate into clusters that percolate through the film. This aggregation suggests that the particle-particle interactions become important at these higher concentrations to generate structures besides simple chaining of particles typically observed with magnetic fields.23, 25 However, examination of these optical images for the top and bottom of the aligned composites (Figure 4) suggests that the structure becomes more uniform at high Ni@Ag particle concentrations as the images are more similar, although some regions of the aggregated structure are out of focus at the top of the film, but not at the bottom. Nonetheless, this result is consistent with better formed columns of Ni@Ag particles that propagate through the film thickness at higher loadings, which could start to explain the differences in piezoresistive behavior for the composites.


17


(A)

Page 18 of 36

500μm

Top

500μm

Bottom

500μm

Top

500μm

Bottom

(C)

500μm

Top

500μm

Bottom

(D)

500μm

Top

500μm

Bottom

(B)

Figure 4. Optical micrographs of the top and bottom of the flat-PDMS/Ni@Ag films with different Ni@Ag loading: (A) 0.25 vol%, (B) 1 vol%, (C) 2 vol% and (D) 3.6 vol%. Several Ni@Ag particle columns that do not percolate through the thickness of the film are shown with the red circles. These particles are out of focus as they are embedded within the PDMS and not at the surface.


18

Page 19 of 36 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


To better understand these morphological differences, the internal morphology was assessed with X-ray microCT for the magnetically aligned PDMS/Ni@Ag films. Figure 5A shows the distribution of Ni@Ag particles inside the PDMS for the film containing 1 vol% Ni@Ag. The magnetic field leads to assembly of the Ni@Ag particles into small columns. In Figure 5A, most particle columns are composed of predominately a chain of single particles although some particle aggregation is observed at the bottom of the particle columns (Figure 5B). Increasing the concentration to 3.6 vol% of Ni@Ag (Figure 5C) leads to a change in the morphology of the particle columns, where a majority of the columns are composed of aggregates of particles and some columns are connected to neighboring columns at the bottom of the film through larger aggregates (Figure 5D). The different morphology observed for the top and the bottom of PDMS/Ni@Ag films from examination of the tomography cross sections is likely from gravity driven sedimentation of Ni@Ag particles (see Figure 2B) in the liquid PDMS precursor prior to curing of the PDMS. Some sedimentation occurs prior to application of the magnetic field, which leads to a higher concentration of Ni@Ag particles near the bottom than near the top surface. In the magnetic field, the Ni cores of Ni@Ag particles are magnetized and gain a magnetic dipole. The dipole-dipole interaction between neighboring particles leads to formation of chain-like structures. The physics associated with the chaining of these particles in an external field have been reported.45, 46 At higher concentration, each chain of particles is


19


Page 20 of 36

closer to neighboring particle chains on formation, therefore are more likely to be attracted by neighboring particle chains to form particle aggregates in the column. Another important morphology difference between these two films are the gaps between adjacent Ni@Ag particles within the vertically aligned chains of particles. As shown in Figure 5B, with 1 vol% Ni@Ag, gaps between adjacent particles can be clearly observed in each aligned particle column from the X-ray microCT reconstruction. Meanwhile, Figure 5D suggests that with 3.6 vol% Ni@Ag, particles are more densely packed in each particle column. Much smaller gaps are observed at this higher Ni@Ag and not for every column. The resolution of the X-ray microCT is approximately 2.5 µm, so substantive gaps (in terms of electron tunneling) likely exist in these columns as well. These gaps in particle columns are discontinuities in the electrically conductive network, which would result in increased through-thickness resistivity due to requirement of electron tunneling through the insulating PDMS. During compression of the film, these gaps can be closed to generate new continuous conductive paths through the thickness as a consequence of the elastomeric nature of PDMS matrix and “rigid-body”-like Ni@Ag fillers. Below the critical pressure, gaps still exist in each column that lead to high electric resistivity. At the critical pressure, some gaps are closed due to dimension change in thickness direction. Therefore, new conductive paths are formed and the resistivity decreases drastically. The pressure range over which the rapid decrease in resistivity occurs is likely related to the distribution of gap sizes. This mechanism based on closing gaps between particles is


20

Page 21 of 36 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


different than that described by eqn(1) as in this case the decrease of resistivity relies on the closing of large interparticle gap that are randomly distributed along the chain of aligned particles instead of the uniform decrease of interparticle distance. As there are more and wider interparticle gaps with only 1 vol% Ni@Ag, larger compression strain and hence higher compression pressure is required to close these gaps. This mechanism explains the presence of a critical pressure for high conductivity and its concentration dependence based on the microCT images in Figure 5.

Figure 5. MicroCT images of flat-PDMS/Ni@Ag film containing 1 vol% of Ni@Ag for (A) full 3D representation of the Ni@Ag distribution with surface projection shown in the inset and (B) a 0.3 mm-thick section as noted by the dashed box in (A). The chains of


21


Page 22 of 36

Ni@Ag tend to bundle when the concentration is increased to 3.6 vol% as shown for (C) the full 3D representation and (D) cross-section that better illustrates the bundling as noted by the dashed box in (C). Figure 6 shows the reversibility of the piezoresistivity of Ni@Ag /PDMS composite over cyclic loading and unloading measurements. During the first cycle, the resistivity of the sample decreases during compression then increases back to high resistivity when the applied pressure is released. However, there is a large hysteresis in the resistivity loop with the critical pressure during unloading decreasing to nearly half that of the critical pressure on loading. During the 2nd loading cycle, the resistivity beings to drop precipitously at 161 kPa, which is much lower the original pressure required on loading to achieve the conductive state. However, this pressure is greater than the pressure during the initial unloading. Unloading during the 2nd cycle leads to a lower threshold pressure below which high resistivity is measured. This hysteresis in the pressure dependencies of the electrical resistivity suggests that there is non-recoverable changes in the connectivity between particles upon compression. Although nanosized Ag can be welded through ohmic heating,51 the current (too low) and size of the Ni@Ag particles (too large) are not appropriate to weakly weld the particles. However, the Ni@Ag particles (Figure S1) exhibit rough surfaces that may promote their adhesion on contact. In addition, the mechanical disparity at the interface between the Ni@Ag and PDMS is large, which can promote failure at the interface under mechanical load.34 Changes at this interface could provide a


22

Page 23 of 36 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


pathway to alter the connections between Ni@Ag particles. SEM cross-section images (Figure S5) illustrate that the PDMS is fully in contact with the Ni@Ag prior to the compression cycles, but the Ni@Ag appears to delaminate from PDMS matrix during compression. The change in the adhesion between the PDMS and particles could provide a lower energy path to close the inter-particle gap, so the lower threshold pressure for percolation could be due to slip of the particles which does not require as significant compression of the PDMS matrix. For the third loading-unloading cycle, the threshold pressure on loading is similar to the 2nd cycle, but the width of the hysteresis loop is increased with the high resistivity achieved plateau at the lower pressure for the 3rd cycle. Additional cycling (4th and 5th) leads to similar piezoresistive to the unloading of the 3rd cycle, although the threshold pressure continues to decrease slightly on each subsequent cycle. As the adhesion between the filler (Ni@Ag) and matrix is poor, there may be questions about the fatigue resistance of these materials. To understand the influence of these changes on the mechanical response of the PDMS/Ni@Ag materials, a cyclic fatigue test was performed as shown in Figure S6. There is a decrease in the effective compressive modulus on the first cycle, but there is no statistical change in the mechanical response between the 2nd and 50th compression cycle. Similar decreases of threshold critical pressure after the first compression cycles are also observed for samples with different loadings of Ni@Ag samples as shown in Figure S7. For device applications, this prior history dependence of the piezoresistive behavior is


23


Page 24 of 36

generally undesired, but there are instances where this history dependence could be beneficial. This dependence on prior loading history could be a simple anti-tampering component52 or determine if a device/product was subjected to undesired pressures at some time. However as the piezoresistive behavior appears to reach a quasiequilibrium state after 3 loading-unloading cycles, an initial ‘burn-in’ period of 3+ cycles could be used to condition the material as well as increase the pressure sensitivity of the material.

Figure 6. Resistivity (in ohm·m) of magnetically aligned flat PDMS/Ni@Ag film containing 1 vol% Ni@Ag particles during loading (compression) (—) and unloading (-----) cycles.


24

Page 25 of 36 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


In the cases presented thus far, the composite is covered during the magnetic field application and curing to generate flat samples. However without this cover, the Ni@Ag particles can generate surface features during the magnetic field assisted alignment as shown in Figure 7. With 1 vol% Ni@Ag, the film surface is nearly flat, like the previously reported materials. However at higher concentration (3.6 vol%-10.2 vol%) of Ni@Ag, large spiky features appear on the surface of the films. Further increasing the concentration of Ni@Ag to 18 vol % leads to a significant reduction in these surface features with only tiny spikes at the surface. This evolution in the surface morphology can be explained by the same underlying physics that defines the size of the features through a competition between surface energy and demagnetization.45, 46 In this case, the features on the surface result from the ability to narrow the chain aggregate dimensions to avoid the energy penalty of demagnetizing field, which is greater than the gravitational force on the Ni@Ag particles. The feature height is limited by the contributions of surface energy of the PDMS and gravity. As the loading of Ni@Ag increases, the transparency of the composite decreases. With 1 vol % Ni@Ag, there is only a slight decrease in the light transmission. At 5.3 vol% Ni@Ag, the grid lines can still be observed behind the sample, but at 10.2 vol% the sample becomes fully opaque.


25


Page 26 of 36

Figure 7. Photographs of aligned PDMS/Ni@Ag films with different Ni@Ag volume fractions produced by continuous roll-to-roll fabrication.

Figure 8A illustrates the spiky features in the surface of the 5.3 vol% Ni@Ag composite from a SEM cross-section. In this case, the spikes are approximately 0.5 mm wide and 1 mm high. These spikes are comprised predominately of Ni@Ag particles as determined by microCT (Figure 8B, lower left magnified region). Without the constraint at the surface of PDMS during alignment, the Ni@Ag particles can overcome gravity and the surface tension of PDMS to form spiky aggregates that protrude from the surface of PDMS due to the force provided by the applied magnetic field. Such surface features are similar to those used for reusable dry contact electrocardiography (ECG)/electroencephalograms (EEG) sensors.53 These electrodes are fabricated with conductive materials, which are connected to an amplification circuit and in contact with skin during use, but dry electrodes often suffer from poor signal quality due to the contact impedance between


26

Page 27 of 36 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


electrodes and the Stratum Corneum. To improve the signal quality, surface features manufactured by molding,54 micromachining55 or 3D-printing56 act to reduce motion artifact and the contact through increasing the local pressure. The R2R manufacturing of aligned PDMS/Ni@Ag provides a potential scalable, low-cost alternative method for the fabrication of flexible dry-contact electrodes with low resistivity (

Tunable Piezoresistivity from Magnetically Aligned Ni(Core)@Ag(Shell

Recommend Documents