Article pubs.acs.org/Langmuir
Shear Adhesion Strength of Aligned Electrospun Nanofibers Johnny F. Najem, Shing-Chung Wong,* and Guang Ji Department of Mechanical Engineering, University of Akron, Akron, Ohio 44325-3903, United States S Supporting Information *
ABSTRACT: Inspiration from nature such as insects’ foot hairs motivates scientists to fabricate nanoscale cylindrical solids that allow tens of millions of contact points per unit area with material substrates. In this paper, we present a simple yet robust method for fabricating directionally sensitive shear adhesive laminates. By using aligned electrospun nylon-6, we create dry adhesives, as a succession of our previous work on measuring adhesion energies between two single free-standing electrospun polymer fibers in cross-cylinder geometry, randomly oriented membranes and substrate, and peel forces between aligned fibers and substrate. The synthetic aligned cylindrical solids in this study are electrically insulating and show a maximal Mode II shear adhesion strength of 27 N/cm2 on a glass slide. This measured value, for the purpose of comparison, is 270% of that reported from gecko feet. The Mode II shear adhesion strength, based on a commonly known “deadweight” test, is 97-fold greater than the Mode I (normal) adhesion strength of the same. The data indicate a strong shear binding on and easy normal lifting off. Anisotropic adhesion (Mode II/Mode I) is pronounced. The size and surface boundary effects, crystallinity, and bending stiffness of fibers are used to understand these electrospun nanofibers, which vastly differ from otherwise known adhesive technologies. The anisotropic strength distribution is attributed to a decreasing fiber diameter and an optimized laminate thickness, which, in turn, influences the bending stiffness and solid-state “wettability” of points of contact between nanofibers and surface asperities.
1. INTRODUCTION This study is concerned with a simple proof-of-concept experiment of aligned electrospun nylon-6 with an average Tg of 47 °C, using a “dead-weight” test commonly used in industry. The fiber is in the solid state in contrast to elastomers that possess a Tg at subambient temperatures and are in the liquid state at room temperature. This research succeeds a series of preliminary studies evaluating the adhesive capability of electrospun polymer species.1−6 Electrospun polymer fibers are vastly different from what has been formed using elastomerbased pressure sensitive adhesives (PSA) reinforced by carbon and glass fibers7 and micrometer-scale wrinkled surfaces on thin films,8 in the sense that electrospinning and laid-flat nanoscale cylindrical solids are a unique technology in their own right. The electrospun nylon-6 is aligned by a rotating disc collector, and the setup makes use of a water beaker hanging onto a pulley system designed in our laboratory, as shown in Figure 1, also known as the “dead-weight” test in industry. Electrospinning occurs when the electrical forces at the surface of a polymer or adhesive solution overcome the surface tension and cause an electrically charged jet to be ejected. The electric force accelerates and stretches the adhesive jet, resulting in a decrease in the diameter and a concomitant increase in length. The cross-sectional area of the adhesive jet decreases by a factor as large as 1 million, and vice versa in length. The jet dries, solidifies, and is deposited onto a rotating disc collector. Electrospinning creates high-aspect ratio aligned fibers that, © 2014 American Chemical Society
with the right dimensions and laminate compliance, conform to surface asperities, to make numerous contact links per unit surface area, enhancing physical interactions such as van der Waals forces and capillary effects. In addition, because of the fiber diameters on the nanometer length scale, the solids can present mechanical interlocks as composite laminates. The morphology of the aligned nanoscale cylindrical solids is vastly different from previous work based on protruding carbon nanotubes,9 nanoforests,10 and lithographically molded protruding micropillars.11 Nature presents the most versatile dry adhesives that produce a clinging capability of 10 N/cm2.12,13 With growing interest in reusable clings and adhesives, new materials such as the types that employ elastomers, polyurethanes, acrylics, silicones, and other forms and molecular weights of PSA polymers are explored. Generally, these adhesives produce substantial Mode II shear adhesion strengths but are considerably difficult to remove from surfaces (Mode I). Commercial high-strength adhesives make use of low-Tg, liquid-state high-molecularweight polymers that flow and conform to surface asperities with minimal chemical interactions or cross-linkable species that permanently connect two surfaces together, at room temperature. Residues can be removed only by solvents, or by Received: April 1, 2014 Revised: August 6, 2014 Published: August 8, 2014 10410
dx.doi.org/10.1021/la502402y | Langmuir 2014, 30, 10410−10418
Langmuir
Article
Figure 1. Illustration of the “dead-weight” test setup. Aligned nylon-6 fibers are finger-pressed with a Teflon cloth sandwiched between the glass slide and the finger, and the Teflon cloth is completely removed prior to Mode II loading. The weight is added at the end of the wire. No additional external Mode I load is applied to the exposed nylon-6 fiber arrays laminated on glass during testing. Mode II shear and Mode I loads are subsequently performed at 0° and 90°, respectively, from the horizontal direction. A glass beaker filled with water is carried by the Mode II adhesion strength and the inherent strength of electrospun nylon-6 fibers. The exposed electrospun fabrics under loading are 2 mm in width and 3 mm in length on average. These aligned electrospun nylon-6 fibers demonstrate an optimal Mode II shear adhesion strength on the order of 27 N/cm2 until they fracture uniaxially.
oriented fashion,5 and in this paper, we will examine the bending stiffness, orientation, and dimensional characteristics of these aligned electrospun cylindrical solids. “Wettability” is a concept known in adhesion science to describe how well the adhesives can maintain contact with surface asperities, resulting in intermolecular interactions. The degree of wetting (wettability) is determined by a force balance between adhesive and cohesive forces. The wettability of a surface can be measured using contact angle experiments. In this study, the solid fiber’s bending stiffness and dimensional characteristics, however, play a critical role in optimizing the van der Waals (vdW) interactions, hygroscopic and capillary effects, and solidstate wettability of nanofibers on substrates, and thus the Mode II shear adhesion strength between electrospun nylon-6 and surface asperities.
massive facestuck stretching and delaminations, and occasionally create a mess for cleanup. Fabrication of dry adhesives with anisotropic force distributions produces attractive alternatives for applications such as tapes, fasteners, treads of wall-climbing robots, clings, microelectronics, and biomedical applications. The “size” effect is prevalent in electrospun nanofibers.1−5,14−17 When the fiber dimensions migrate toward the nanometer length scale, the size effect becomes increasingly pronounced. On the other hand, on the basis of simple contact mechanics,18 high-aspect ratio structures will exhibit adhesion capability enhanced compared to that of a low-aspect ratio morphology,19−24 by generating additional contact points and a large surface area to volume ratio. As a result, electrospinning presents a versatile technique for fabricating nanoscale cylindrical solids with high aspect ratios that conform to surface asperities.4,15,25−29 The most adhesive nature of electrospun fibers is readily observed during electrospinning, and when fibers solidify on collecting electrodes. Little is known, however, about making use of such adhesive forces,1−6 upon fabrication of electrospun dry adhesives,30,31 which possess significant Mode II shear adhesion strength with easy Mode I normal detachment, for strong shear binding-on and low normal lifting-off effects. With the aid of microscopy and microanalyses, we investigate the effects of fiber diameter, fiber surface roughness, and thickness of laminate on their corresponding adhesion strengths, as a simple proof of concept of the capability of aligned electrospun polymer fibers. We first measured the adhesion work between two single free-standing electrospun polymer fibers,1−3 in cross-cylinder geometry and randomly
2. EXPERIMENTAL SECTION 2.1. Collection of Aligned Electrospun Fibers. Nylon-6 nanofibers from Sigma-Aldrich (CAS Registry Number 25038-54-4) are dissolved in formic acid (EMD Corp., CAS Registry Number 6418-6) and magnetically stirred overnight. A syringe pump is employed for maintaining a solution drop on the tip of a stainless steel needle. The needle is attached to a 5 mL capacity syringe filled with a polymer solution. Five different nylon-6 fibers with polymer concentrations of 12.5, 15, 17.5, 20, and 22.5% are electrospun. The needle is charged with a high voltage of 25 kV. The gap distance between the tip of the needle and the top of rotating disc collector is set at 150 mm. The collector has a diameter of 150 mm, and electrospun fibers are collected at an uptake velocity of 20 m/s. In this study, nylon-6 nanofibers are not dried after being electrospun and are kept at a temperature of 23 °C and a relative 10411
dx.doi.org/10.1021/la502402y | Langmuir 2014, 30, 10410−10418
Langmuir
Article
Figure 2. (a) Mode II shear adhesion strength as a function of laminate thickness, T, for aligned fiber laminates attached to a glass slide with a fingerpress preloading of ∼2 kg. (A) For T values of 40 μm, the Mode II adhesion strength is drastically reduced because of the decrease in bending flexibility and thus an increasingly constrained plasticity for Mode II deformation. (b) AFM scan of the glass slide that was tested for its adhesion to aligned electrospun nylon-6 fibers. The average roughness is ∼5 nm. humidity of ∼50% before being tested. For most dry or solid-state adhesives, the testing environment cannot completely rule out adsorption or moisture and absorption of hygroscopic materials. Some degrees of capillary, surface tension, plasticizing, and hygroscopic effects are unavoidable. These interactions are discussed in general as physical interactions, including vdW forces, in this paper as being among the aligned fiber laminates, fiber−fiber interactions, and glass surfaces, and the physical interactions are accounted for by the total values of shear adhesion strengths we empirically measure. 2.2. Fabrication of High-Aspect Ratio Nanofibers. After electrospun nylon-6 has been collected for 10 min, a section of the laminate is trimmed off using a sharp needle to prepare the surface for the collector for peeling. A small part of the laminate is then placed on a glass slide using the tapered tip of a needle. The glass slide is thereafter rotated in plane where laminates overlap one another.
Subsequently, laminates with measurable thicknesses are obtained. These laminates are then installed in a mounting grip. 2.3. Atomic Force Microscopy (AFM). Contact-mode AFM (Veeco Instruments, Inc.) is performed to investigate the elastic modulus of nanofibers. A tip radius of 12 nm and a cantilever spring constant of 0.125 N/m are employed. The diameter and surface roughness of fibers are determined using a tapping-mode operation. A tapping probe (MikroMasch Inc.) with a tip radius of 10 nm and a cantilever spring constant of 40 N/m is employed. Scanning is conducted at a frequency of 0.5 Hz, and data are analyzed by employing SPI3800N probe station software (Seiko, Inc.). The surface roughness of the nanofiber is obtained by tracing a line along the fiber axis. An average roughness of the fiber surface can then be determined. A separate scan of the surface roughness of the glass slide is performed, and it is determined the surface on average possesses a roughness of 5 nm. The glass slides are used as received after being gently cleansed to 10412
dx.doi.org/10.1021/la502402y | Langmuir 2014, 30, 10410−10418
Langmuir
Article
Figure 3. (A−E) Scanning electron microscopy (SEM) images of electrospun nylon-6 fibers with different nanofibers diameters of 300, 135, 115, 85, and 50 nm, respectively. (F) Nanofiber diameter as a function of polymer concentration in solvents. Decreasing the polymer concentration will lead to a decrease in nanofiber diameter and results in an increase in the packing density and aspect ratio of fibers.
μm. This greatly reduced Mode II adhesion in shear is attributed to the loss of bending flexibility and thus fiber conformity as an airtight laminate to surface asperities. In the adhesion science literature, this is often known as wettability,32 but, in this study of nanoscale cylindrical solids, it is more sensibly termed solid-state “wettability”. Because it is the first time aligned electrospun laminates have been introduced into shear adhesion work,1−5 the change in adhesion strength versus laminate thickness, T, is described in a more generic fashion. Mode II shear adhesion testing is performed at an angle, θ = 0°, with the glass slide (see Figure 1). The highest shear adhesion strength, 27 N/cm2, which is in fact the nanofiber fracture strength in the optimized case, because the adhesion strength is stronger than the uniaxial tensile strength of nanofibers, is reached for laminates having fibers with a diameter d of 50 nm and a T of 40 μm. This value is subsequently revalidated and reproduced by Fu and co-workers with a different batch of electrospun nylon-6.33 A glass beaker is filled with water and carried by a small piece of laminate with dimensions of 3 mm × 2 mm. This shear adhesion strength is 415% greater than that reported for polycarbonate-based dry adhesives.34 Generally, most solid-state cylindrical solid fibers show rather weak adhesion strength on glass slide as shown in Figure 2a. It is plausible that the empirical observation of the transition of a decrease in shear adhesion strength as T increases from 40 to 80 μm is caused by an increase in the stored elastic energy, Gc, that can initiate a crack that delaminates the adhesive from the adherend, but this could not satisfactorily explain the transition of “low−high−low” shear adhesion strength as observed for the nylon fiber with a d of 50 nm, which is our optimized fiber morphology. In mechanics,35−41 “plane strain” conditions are generally used to describe constrained plastic flow in twodimensional planes, thus limiting the ability of solid-state specimens to attain a higher geometry-dependent critical stress intensity factor, Kc, elastic energy release rate, or crack driving force, Gc, impact strength, deviatoric stress state, or distortional deformation. Plane strain deformation, or small-scale yielding,
remove dust particles on glass surfaces using a microfiber lens cleaning cloth. 2.4. Adhesion Testing Using a Beaker and Pulley System, Also Known as the “Dead-Weight” Test. The nanofiber laminate is finger-pressed with a Teflon fabric to help ensure compliant contact on a glass slide, and then the loading weight is added to the end of the wire. The normal preload applied is estimated to be relatively high to ensure good contact between nanoscale contact links and the glass slide. The preload also serves to ensure all air pockets trapped between laminates and the glass substrate are being rubbed out completely, enhancing the physical contacts between aligned nanoscale cylindrical solids and surfaces. Cold welding was not observed in our previous study5 of randomly oriented electrospun porous membrane, but this cannot be completely ruled out. Nevertheless, any physical interactions between the nanoscale polymer and SiO2 are considered as part of the contributing factors to the shear adhesion force under relatively dry conditions. The Teflon applicator layer is subsequently completely removed (see the movie provided as Supporting Information and Figure 1) prior to Mode II shear loading. Note that this type of shear loading geometry is also known as Mode II in the mechanics literature. The loading weight consists of a beaker that is incrementally filled with water until the laminates fracture. Therefore, the shear adhesion strength between the laminate and the glass slide is stronger than the fracture strength of the nanofibers themselves. The contact area between the laminate and a glass slide has a width of 3 mm and a length of 2 mm. The only two parameters that vary are laminate thickness and nanofiber diameter. No additional external Mode I load is applied to the laminate while it is being tested by the loading weight when it is applied in the Mode II direction.
3. RESULTS AND DISCUSSION 3.1. Effects of Laminate Thickness (T) on Shear Adhesion. As shown in Figure 2a, the nanofiber laminate displays low Mode II shear adhesion strengths for thicknesses of 40 μm, the shear adhesion strength of the laminate is drastically reduced with an increase in T and reaches considerably lower values at T values of ∼80 10413
dx.doi.org/10.1021/la502402y | Langmuir 2014, 30, 10410−10418
Langmuir
Article
corroborate the hypothesis that aligned electrospun nylon-6 fibers can operate as dry reusable adhesives when the dimensions of aligned fibers and laminate thickness are optimized. Microscopically, we cannot observe significant moisture, capillary, or hygroscopic effects of nylon-6, as shown in Figure 3, but they cannot be ruled out in our testing environment. More detailed physical analyses are provided in the discussion that follows. The surface boundary effect of fibers contributes to Mode II adhesion by making additional contact links per unit area and enhancing fiber conformability by reducing bending stiffness. For side wall contacts of fibers, the attractive force per unit length between the nanofiber and substrate is
is well established to occur (i) in thick specimens, (ii) around crack tip triaxiality, (iii) at low temperatures, and/or (iv) under high-speed straining conditions. The measured plane strain values are sometimes considered as “material” properties. Most solid films undergo plane stress deformation except for film ligaments approaching near-zero crack extension, that is, at the initiation of the incipient crack.41 The general definition of “constrained plastic flow” and thereby small-scale yielding (plane strain) with limited material ligament is adopted in this paper. Figure 2 presents a plausible alternative explanation of the “low−high−low” transition in shear adhesion strength based on the “plane strain−plane stress−plane strain” equivalent transition, as previously described in the Mode I essential fracture work of thin polymer films by Mai, Cotterell, Hashemi, Karger-Kocsis, and co-workers.35−41 As the material thickness increases, the bending stiffness and system stored elastic energy increases drastically, the fiber laminate of which appears to be under small-scale yielding (plane strain), and the surface conformity or solid-state “wettability” decreases sharply. As the thickness decreases to 40 μm, the aligned laminate appears to exhibit more extensive plastic flow and large-scale yielding, and thus, plane stress conditions ensue. Such a state of stress, viz., plane stress with larger-scale yielding, gives rise to an apparent increase in face separation work per unit crack extension, or an increased thickness-dependent Gc. In the case of Mode II shear adhesion, enhanced conformability in airtight laminate to surface asperities is observed. Nevertheless, when the laminate dimension is further reduced, that is, less than 40 micrometer, the lack of available material to serve as adhesives and adherends causes a drastic drop in measurable adhesion. Such a phenomenon is common when a ligament of polymer films is reduced to close to zero ligament length prior to crack propagation under the work of fracture measurements,35,41 and it is generally known as a return to “near” plane strain conditions41 due to constrained plasticity and the material confinement effect. The transitions from inadequate material contact to highly conformable surface contact and, subsequently, to laminate with an overly high bending stiffness and stored elastic energy to conform to surface asperities are described in the schematics as shown in Figure 2. More discussion about equivalent constrained plastic flow and its transitional phenomena in work of fracture can be found elsewhere.35,41−43 3.2. Effects of Fiber Packing Density, Nanofiber Diameter (d), and Fiber Surface Roughness on Adhesion. As shown in Figure 3, the packing density and aspect ratio of fibers are both noticeably enhanced with a decrease in d. This enhancement is suggested to augment physical interactions, including the well-known vdW between the fibers and substrate, leading to an upsurge in Mode II adhesion strength. As shown in Figure 3F, d extensively decreases with decreasing concentrations of polymer solutions (R), as described in section 2.1. The fiber packing density is characterized via SEM (FEI Quanta 200). Theoretical studies have shown that significant adhesion could be obtained via a reduction in size. Furthermore, the side contact of fibers with a substrate over a large contact area causes a significant increase in adhesion. Most biomimicking adhesive literature10,11,44−51 focuses on nanoprotrusions as the mechanism of choice for dry adhesion. Nevertheless, recent findings on the adhesion of geckos52 show that cylindrical solids laid flat on surfaces are found to be operative in gecko locomotion. The findings
Fv = (A d) /(16D2.5)
(1)
where A is the Hamaker constant and D the gap distance between the surface of the nanofiber and the substrate.53,54 A cutoff gap distance D = D0, which represents the effective separation between the nanofiber and substrate and at which the maximal Fv (FvM) is estimated. The total FvM is FvMK = (NLA d )/(16D0 2.5)
(2)
where N represents the total number of fibers along the contact width between the nanofiber arrays and the substrate (W) and N = W/d. Replacing the value of N in the previous equation, we obtain FvMK = (LWA)/(16D0 2.5 d )
(3)
For constant values of W and the contact length between the nanofibers and the substrate (L), FvMK radically increases with a decrease in d. These results suggest a substantial increase in the Mode II shear adhesion strength with a decrease in d. As shown in Figure 4, the surface roughness of fibers considerably decreases with a decrease in d. The latter decrease
Figure 4. Surface roughness effect of a nanofiber as a function of nanofiber diameter, d. With a decrease in d, the Mode II shear adhesion strength between the laminate and substrate increases dramatically.
leads to a significant enhancement in Mode II shear adhesion strength between the electrospun laminate and a glass substrate due to considerable proliferation of the effective nanoscale connectors per unit area. 3.3. Mode I Adhesion Force. Electrospun laminates are easily peeled off a glass slide for θ = 90°. For a contact area of 3 mm × 2 mm on a glass slide, the Mode I adhesion force of the electrospun laminate is roughly 0.015 N, regardless of d and T. 10414
dx.doi.org/10.1021/la502402y | Langmuir 2014, 30, 10410−10418
Langmuir
Article
Meanwhile, the maximal Mode II shear adhesion force for nanofibers is 1.6 N (fibers with a d of 50 nm), which leads to a V of 100. Thus, these nylon-6 fiber laminates are 10 times easier to detach from a glass slide in the Mode I direction as dry adhesives. This significant V value is suggested to be due to nanoscale cylindrical solids in this work. Thus, a significant decrease in the number of nanoscale contact links per unit area between the nanoscale cylindrical solids and the substrate arises during vertical detachment, which leads to an easy Mode I lifting off. 3.4. Effect of Nanofiber Bending Stiffness on Adhesion. The bending stiffness of a nanofiber is b = EI, where I is the moment of inertia of the cross section of a nanofiber.34 The indentation of an AFM cantilever tip into the nanofiber is directly proportional to the lateral deflection of the cantilever (ΔX).14,54 According to the Hertz model, ΔX is defined as h = {[ 2(1 − v 2)f ]/(4a 0.5E)}2/3
I = (πd 4)/64
the relative moment of inertia is Ir = I /I0 = (d /d0)4
(7)
where d0 is 300 nm and I0 is I for d0. The relative bending stiffness is
br = ErIr
(8)
And thus: br = Er (d /d0)4
(9)
As shown in Figure 6B, 1/br significantly increases with a decrease in d for fibers with d values of
