Designed to Fail: Flexible, Anisotropic Silver Nanorod Sheets for Low

Jun 23, 2016 - Table of Contents ... Much of the research into these devices has been focused on improving the robustness of the ..... dreaming to mim...
1 downloads 0 Views 3MB Size
Subscriber access provided by - Access paid by the | UCSB Libraries

Article

Designed to Fail – Flexible, Anisotropic Silver Nanorod Sheets for Low-Cost Wireless Activity Monitoring Layne Bradley, George K. Larsen, and Yiping Zhao J. Phys. Chem. C, Just Accepted Manuscript • DOI: 10.1021/acs.jpcc.6b04792 • Publication Date (Web): 23 Jun 2016 Downloaded from http://pubs.acs.org on June 29, 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.

The Journal of Physical Chemistry C 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 26

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 Journal of Physical Chemistry

Designed to Fail – Flexible, Anisotropic Silver Nanorod Sheets for Low-Cost Wireless Activity Monitoring

Layne Bradley1, George Larsen2, and Yiping Zhao1

1

Department of Physics and Astronomy, The University of Georgia, Athens, GA, 30601, U.S.A.

2

National Security Directorate, Savannah River National Laboratory, Aiken, SC 29808, U.S.A.

1 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

ABSTRACT: We describe the fabrication and properties of flexible, anisotropic silver nanorod sheets and investigate their potential to function as a sensor. Aligned and tilted silver nanorod (AgNR) arrays are incorporated into polydimethylsiloxane (PDMS), to form flexible conductive sheets. The electrical properties of these sheets are investigated and show large anisotropies, which are related to the alignment direction of the nanorods. Notably, the films show the greatest electrical resistance in the direction perpendicular to the nanorod alignment, and when strain is applied along this direction, the resistance increases monotonically with increasing loading/unloading cycles. In comparison, the resistance along the nanorod alignment direction remains constant over many strain cycles, and therefore, can serve as an internal reference or as a stable strain-gauge. These changes in resistivity are attributed to changes in the inter-nanorod connectivity and can be modeled using an effective medium approximation for anisotropic percolation. Stable piezoresistivity (in one orientation) and surface enhanced Raman scattering activity of the AgNR sheets make them attractive for flexible electronics applications such as electronic skin or as monitors for human-machine interactions. However, the ability to encode a surface’s dynamic history into material properties through resistance changes is a considerable simplification over other systems and can enable wireless activity monitoring where cost or demanding environments prevent more complicated devices from being implemented.

2 ACS Paragon Plus Environment

Page 2 of 26

Page 3 of 26

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 Journal of Physical Chemistry

1. Introduction Flexible electronic devices, created by combining soft materials and electronic circuits, promise to fundamentally change how living organisms interact with technology.1-2

While flexible

displays are expected to improve the ruggedness and usability of media devices,3 more unconventional applications have been envisioned. For example, flexible biomedical devices that can directly conform to the soft, dynamic, and intricate surfaces of the skin, heart, and brain have recently been demonstrated and can serve as the basis for implantable, highly sensitive electronics that can monitor and respond to biological signals.4-5 By fabricating such devices out of non-toxic, bioresorbable materials, researchers have created a new class of non-invasive, “transient” devices that are designed to operate for a certain period of time before being harmlessly dissolved within biological tissue.6-7

Flexible electronics are expected to have

impacts in other areas beyond the biomedical field; some examples include: edible, conformal silk-based sensors for food safety,8 robots with unique actuation mechanisms and more life-like movements,9 and soft energy conversion devices for harvesting non-traditional sources of mechanical energy.10 In all of these cases, it is critical that the device maintain its designed function throughout its lifetime, even for the intentionally transient devices. This means that the electrical properties of the device must not change after many cycles of bending, flexing, stretching, and similar deformations, or the device will fail to operate. A large number of approaches have been explored for fabricating such flexible electrical devices. In particular, different methods for embedding conductive nanostructures such as carbon nanotubes11, metal nanowires12-13, or metal nanotroughs14-15 into a bendable material have been widely pursued. Much of the research into these devices has been focused on improving the robustness of the electrical properties by engineering and optimizing the materials and geometries.16 This

3 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

issue is important for the applications described above, but an interesting possibility is raised by considering the reverse idea, the engineering of materials such that the electrical properties of a flexible device change in a predictable manner with use. By tracking these electrical properties, through resistance measurements (i.e., piezoresistive effect), for example, such a device could serve as a sensor that monitors the dynamic history of an active surface. While the concept of flexible strain-gauge sensors is not new,17 the proposed technique would offer a considerable simplification. Encoding a surface’s motional history into the material properties of the device through controlled failure would allow for the sensor to be operated remotely and be interrogated wirelessly through radio frequency, or later by electrical connection, without the added complications and costs of microprocessors and onboard memory. However, a notable concern with failure-based sensors is that of reliability. Therefore, it is of critical importance for such a device to have internal standards for cross-reference to ensure that the material’s properties change in a deliberate manner according to a model, as opposed to some other device failure. One way to achieve an internal reference is by taking advantage of anisotropic materials, where electrical properties vary depending orientation. In this case, if the electrical properties change for only one orientation, then the conductive properties of the stable orientation can serve as a reference for the device. An additional concern is that of cost since these types of sensors have an intentionally finite lifetime. Therefore, it is also of critical importance that such devices are fabricated using inexpensive materials and methods, or they become less attractive from an implementation standpoint. In this report, we describe the fabrication and properties of flexible, anisotropic silver nanorod sheets and investigate their potential to function as sensors. Aligned and tilted silver nanorod (AgNR) arrays are fabricated using oblique angle deposition,18-19 a simple and

4 ACS Paragon Plus Environment

Page 4 of 26

Page 5 of 26

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 Journal of Physical Chemistry

straightforward physical vapor deposition technique that is readily adaptable to different fabrication conditions and scales.20-21

The AgNRs are then incorporated into a ubiquitous

polymer, polydimethylsiloxane (PDMS), to form flexible, anisotropic conductive sheets. The electrical properties of these sheets are investigated and show large anisotropies, which are related to the alignment direction of the nanorods. Notably, the films show the greatest electrical resistance in the direction perpendicular to the nanorod alignment, and when strain is applied along this direction, the resistance increases monotonically with increasing loading/unloading cycles. In comparison, the resistance along the nanorod alignment direction remains constant over many strain cycles, and therefore, can serve as an internal reference or as a stable straingauge. These changes in resistivity are attributed to changes in the inter-nanorod connectivity and can be modeled using an effective medium approximation for anisotropic percolation. In addition to anisotropic conductivity, the flexible AgNR sheets exhibit piezoresistivity and surface enhanced Raman scattering (SERS), effects which could be combined to create multiplexed sensors. Due to their low-cost, ease of fabrication, and desirable properties, we believe these materials could serve as the foundation for wireless and disposable activity sensors.

2. Experimental Methods The AgNR were fabricated by a physical vapor deposition technique known as oblique angle deposition (OAD). For OAD, the substrate is rotated so that the angle between its surface normal and the incoming vapor is greater than 75°. At such oblique angles, the self-shadowing effect dominates the growth, resulting in the formation of tilted nanorod arrays22. Glass slides (Gold Seal, Cat. No. 3010) were used as the substrates for the initial AgNR formation by OAD. They were first cleaned with piranha solution, rinsed with deionized water, and dried by blowing

5 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

with nitrogen. The slides were then placed into a home built e-beam evaporation chamber and rotated to an oblique incident angle of θ = 85°. A calibrated quartz crystal microbalance (QCM) monitor positioned perpendicular to the vapor direction was used to monitor the thickness and rate of the evaporated Ag during the deposition. 1.5 µm of Ag (99.999% pure pellets from Kurt J. Lesker), as measured by the QCM, were deposited onto the substrates at a rate of 0.4 nm/sec. After being removed from the chamber, the substrates were placed within a customized mold. The top and bottom of the mold were two glass slides coated with Chlorotrimethylsilane (98+% Alfa Aesar) which prevented the PDMS from sticking to them. The sides of the mold consisted of 4 pieces of stainless steel blocks (1.42 mm thick) that could be arranged to fit around the AgNR-deposited glass slides. PDMS (Sylgard 184 Elastomer kit) was prepared by mixing 10 parts of the elastomer base to 1 part of the curing agent and then degassing for 30 minutes. The substrate, mold, and PDMS were placed in an oven at 85 °C for 5 minutes and then, as depicted in Figure 1a-b, the PDMS was poured on top of the AgNR substrate. The top part of the mold was slid into place, thus removing excess PDMS and providing a uniform thickness. The PDMS was cured at 85 °C for 45 minutes after which it was removed from the oven and allowed to cool to room temperature. The final thickness of the prepared PDMS sheets was measured with an electronic caliper and was found to be t = 0.64 ± 0.03 mm. The PDMS could then be peeled off of the glass slide (Figure 1c). Upon removal, the bulk of each nanorod was embedded within the PDMS, with a small amount of Ag residue remaining on the glass slide (Figure 1d). Square AgNR sheets (0.75 in. x 0.75 in.), both on glass prior to PDMS application and then after PDMS lift-off, were used to measure the anisotropic resistivity with a custom fourpoint probe set-up. The substrates were rotated counter-clockwise from 0° to 360° at intervals of every φ = 15°, where φ = 0° is along the tilting direction of the NR. Rectangular AgNR sheets (1

6 ACS Paragon Plus Environment

Page 6 of 26

Page 7 of 26

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 Journal of Physical Chemistry

cm wide and 5 cm long) were used to measure the resistance while a strain was applied. A home-built stretching apparatus was used to perform the stretching and compression cycles (see Figures S1 and S2, Supporting Information). The AgNR PDMS sheets were held at either end by clamps consisting of a non-conductive acrylic base that was secured to the stretching apparatus, but electrically insolated the system from it, and an aluminum top that could be tightened around the sheets by screws. The clamps holding the PDMS sheet were originally positioned 2 cm apart, and the sheet was then stretched and compressed ± 5 mm (resulting in ± 25% strain) at a rate of 0.25 mm/sec. With the conductive portion of the AgNR sheet in contact with the aluminum top of the clamps, alligator clips were connected directly to that portion of the clamps so that resistance measurements (20 readings every 0.7 seconds) could be taken during the cycles. A Keithley 2000 multimeter was used for all resistance and resistivity measurements. A C# computer program was written to simultaneously control the stretching apparatus through an Arduino Uno board and communicate with the multimeter to record the measurements.

3. Experimental Results and Discussion Oblique angle deposition (OAD) is used in order to create the silver nanorods for the AgNR flexible sheets, as described in the preceding section. Figure 2a shows representative scanning electron microscopy (SEM) images of the AgNRs obtained from OAD. As can be seen in the images, the films are composed of randomly distributed nanorods having similar yet irregular morphologies. In spite of their irregularities, the nanorods are uniformly aligned and tilted toward the direction of the vapor incidence. Thus, anisotropy in these nanostructured thin films arises from the arrangement of a collection of similarly shaped nanorods. The surface coverage of the nanorods can be described by density ρ = 19 ± 3 rods/µ2, and the geometry of the

7 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

individual nanorods can be described by statistically averaged parameters: length L = 1.0 ± 0.2 nm, diameter d = 0.10 ± 0.02 nm, and tilting angle θ = 13 ± 2°. After deposition onto glass substrates, the AgNR film is embedded in PDMS, which is then peeled away after the polymer has cured (Figure 1) to create flexible AgNR sheets. Figure 2b shows representative SEM images of the AgNRs embedded in PDMS. The polymer coats the nanorods and intercalates between them while preserving the structure of the individual nanorods and the arrangement of the array. The electrical properties of the AgNR films and flexible sheets are investigated by 4point probe resistivity measurements, and the results are shown in Figure 3. The resistivity measurements seen in the plots have a distinctive anisotropic shape that agrees well with previous reports.23 In particular, the resistivity is the lowest when measured along the nanorod tilting direction, where the azimuthal angle φ ≈ 0°, and increases monotonically until φ ≈ 90°, which is perpendicular to the nanorod tilting direction. The resistance proportionally increases for all angles after the AgNRs are embedded in PDMS, while the anisotropic property is preserved. Several different factors could contribute to the increase in resistance seen after the AgNRs are embedded in PDMS: conductive junctions between the nanorods that are close to glass substrate surface may not embed in and peel off with the PDMS sheet; the mechanical action of the peeling step might fracture nanorods and inter-nanorod junctions; and/or the contact resistance between the measurement probes and the AgNRs could be changed by the polymer matrix. The flexible AgNR sheets do exhibit piezoresistivity, as shown in Figures 4a and 4b, which respectively show the measured resistances (using 2-point measurements) of the AgNR sheets for different values of strain when the sheets are stretched and compressed over many

8 ACS Paragon Plus Environment

Page 8 of 26

Page 9 of 26

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 Journal of Physical Chemistry

cycles in the directions parallel with (φ = 0°) and perpendicular to (φ = 90°) the nanorod tilting direction. For both orientations, the resistance monotonically increases with the magnitude of strain, though the resistance shows greater increases for positive strain values. The increase in resistance for negative strains is an interesting result, as one would expect the resistance to decrease for metallic conductors since the compression increases the cross-sectional area and decreases the length of the conduction path.

This result points to a different conduction

mechanism for the AgNR sheets as compared to bulk metallic sheets. The piezoresistivities of both orientations also display hysteresis, where the resistance for a given strain value depends on whether strain is being loaded or unloaded. Since the AgNR are not rigidly embedded in the PDMS they can shift slightly as the strain is applied and then subsequently released. The dynamics of these small movements alters the node connectivity between the nanorods which could lead to this observed hysteresis. However, the behavior of the different orientations, φ = 0° and φ = 90°, do differ in an important aspect, and that is in their cyclic stability. When stretched and compressed along θ = 0°, the resistance values are stable at 0%, 25%, and -25% strains over 100 cycles (Figure 4c). On the other hand, the resistance values monotonically increase with each successive cycle when the AgNR sheet is stretched and compressed in the φ = 90° direction (Figure 4d). Thus, due to their anisotropy, these flexible AgNR sheets display the unique phenomenon of simultaneously exhibiting stable electrical resistance along the AgNR tilting direction and slowly varying electrical resistance in the perpendicular direction. The stable resistance measurements along φ = 0° over many cycles would make these AgNR sheets a candidate to provide a reliable self-calibration response in a strain sensor, while the varying resistance measured along φ = 90° provides a method for monitoring the sensor’s usage.

9 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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 26

In order to better visualize the trends in the hysteresis data, we calculate the hysteresis area, which is defined here as the area of the unloading curve subtracted from the area of the loading curve obtained from the resistance versus strain data. The hysteresis area is relatively stable for strains along φ = 0° after 10 or more cycles (Figure 4e), with the slopes of the graphs approaching zero. If the hysteresis is being caused by small shifts of the AgNR, this implies that they become repeatable after these first few cycles. Combined with the near-zero slope of the resistance measurements in Figure 4c, the stable hysteresis area implies that the AgNR sheet provides consistent piezoresistive response along the φ = 0° direction after many cycles. However, the magnitude of the hysteresis area increases with cycle number for strains along φ = 90° direction (Figure 4f) which could provide further information for usage monitoring. It should be mentioned that the apparent oscillatory behavior of the hysteresis areas is an artifact due to the timing of the resistance measurements. The measurements from the multimeter were recorded at a rate slightly out of phase with the stretching and compressing cycle rate; thus, while resistance was consistently measured at strains of -25%, 0%, and 25%, there was some variation at which strains it was measured between these points from cycle to cycle. This variation created the oscillatory artifact in the calculations of the hysteresis area. It is important to note that the hysteresis area grows increasingly negative for φ = 90° stretching cycles, while for φ = 0° the hysteresis areas remain positive. In other words, the resistance temporarily decreases after loading strains along φ = 0°, but permanently increases after loading strains along φ = 90°. This observation highlights the differences in the conductivity mechanisms for the two different orientations of the AgNR sheets, which will be discussed in greater detail in Section 4 below.

10 ACS Paragon Plus Environment

Page 11 of 26

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 Journal of Physical Chemistry

A demonstration of an application of the flexible AgNR sheets can be seen in Figure 5, which shows the resistance versus time for an AgNR sheet conformably attached to a person’s flexing finger. The bending motion of the finger stretches the AgNR sheet causing the resistance to change due to the piezoresistive effect.

The rate of change and the peak value of the

resistance are directly related to how fast and how far the finger is bent, respectively. Since the loading/unloading and resistance measurements are conducted along φ = 0° in this example, the observed resistance versus stretching relationship is stable over many cycles. On the other hand, for the AgNR sheets, the stretching induced deformation is not limited to just one direction, and loading/unloading cycles will cause a permanent change to the resistivity of the φ = 90° direction.

Therefore, the activity history becomes permanently encoded into the material

properties of the device and can be determined by measuring the resistance of the φ = 90° direction. Another advantage of using AgNRs as the anisotropic basis for activity sensors is that AgNRs have been previously shown to be highly sensitive SERS-active substrates.22 Figure 6 shows the SERS spectra of the flexible AgNR sheets with and without the Raman reporter molecule, trans-1,2-bis(4-pyridyl)-ethylene (BPE), for different applied strains along φ = 0° and φ = 90°. No strong peaks are observed in any of the spectra for the blank substrates, but the characteristic BPE pattern emerges after it is applied to the flexible AgNR sheet. For φ = 0°, the applied strain initially decreases the SERS peak intensity, but as the strain continues to increase, the SERS peak intensities increase eventually greatly exceeding the 0% strain value. In contrast, the applied strain along φ = 90° leads to an overall decrease in SERS peak intensity. A possible mechanism for these different behaviors is related to the elastic changes to inter-nanorod spacing and geometry. SERS intensity is enhanced by the large electric fields between closely coupled

11 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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 26

plasmonic nanostructures, and the largest fields are located near the tips and sharp edges. As the AgNR sheet is stretched in either direction, the in-plane distance between the nanorods begins to increase, but at the same time, the thickness of the elastic sheet begins to contract pulling the nanorods closer together in the out-of-plane direction. As the sheet is stretched along φ = 0°, the nanorods slide along each other, bringing the nanorod tips closer together in both the in-plane and out-of-plane direction, which should enhance the SERS response after the initial separation. Measurements of the optical transmittance (see Figure S3a) under different polarization conditions when the AgNR sheet was stretched this way show very little change thus support the argument of increased alignment of the nanorods. On the other hand, stretching the AgNR sheet along φ = 90° increases the separation between the nanorods, and any benefit due to out-of-plane contraction is not enough to overcome the increase in lateral nanorod separation. This increased separation also causes the substrate to have a polarization effect as can be observed in the optical transmittance measurements (see Figure S3b). Overall, the SERS activity of the flexible AgNR sheets is a benefit for sensing applications, as the strain monitoring could be combined with chemical or biochemical detection to create a multiplexed sensor platform. The spectral intensity of a Raman reporter could also function as wireless signal for strain monitoring. Finally, the relative SERS intensity is an additional measurable parameter that contains information describing the bending direction of the AgNR sheet. By combining the SERS data with the two direction resistance data, the 2dimensional bending direction can be uniquely determined.

4. Anisotropic Effective Medium Theory

12 ACS Paragon Plus Environment

Page 13 of 26

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 Journal of Physical Chemistry

An anisotropic effective medium approximation (EMA) can be employed to better understand the electrical behavior of the flexible AgNR sheets before, during, and after cycling. The EMA for anisotropic bond-percolation of a two-dimensional square lattice (AB2-EMA) is analytically solvable,24 and has been previously shown to be a reasonable approximation for conductivity of AgNRs.23 In this case, the metal nanorod array can be thought of as a two-dimensional square lattice of resistors, and the two directions that define the individual nodes correspond with φ = 0° and φ = 90°. By assuming the resistivities of the individual conductors that make up the lattice are independent of direction, then the anisotropy and effective resistances of the array solely results from the difference in node connectivity in the φ = 0° and φ = 90° directions. Relating this model back to the AgNR sheet, these assumptions imply that individual nanorods generally have similar resistivities that can be described by an average value, and that the electrical anisotropy is due to a different number of inter-nanorod connections along the φ = 0° and φ = 90° directions. The different probabilities of nanorod being connected in the φ = 0° and φ = 90° directions gives rise to the different “effective” resistivities measured experimentally. The probability distribution of the conductive pathways in AB2-EMA model for the AgNR sheets is given by pφ(σ) = pφ δ(σ – G) + (1 – pφ) δ(σ),

φ = 0°, 90°,

(1)

where δ(u) is the Dirac delta function. This describes a system where the individual conductances for direction φ have value G with probability, pφ, and have conductances equal to zero with probability (1 – pφ). The solution to the AB2-EMA with the probability distribution given by Eq. (1) is σ0° = G [p0° – (1 – x)] / x,

(2)

σ90° = G (p90° – x) / (1 – x),

(3)

13 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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 26

where x is the solution to tan2[(1/2) π x] = [x (p90° – x)] / {(1 – x )[p0° – (1 – x)]}.

(4)

The effective conductivities of the AgNR sheets for φ = 0° and φ = 90° are given by σ0° and σ90°, respectively, and a non-vanishing solution is given by Eqs. (2) – (4) only when the percolation threshold is reached, p0° + p90° > 1. According to the AB2-EMA model above, piezoresistance should only arise from changes in the probability distribution of conductive pathways.

Therefore, we define two more

parameters p+25% / pφ and p-25% / pφ, which are the probabilities of there being a conductive pathway under 25% and -25% strains, respectively, normalized by the probability of there being a conductive pathway for 0% strain for strains applied along the φ direction. Using previously determined conductive parameters for the individual AgNRs and assuming p0° = 1 initially,23 the AB2-EMA has been used to model the resistance behavior of the flexible AgNR sheets at 0, 25, 50, 75, and 100 cycles for 0%, 25%, and -25% strains. The theoretical results have been compared to the experimental data for strains applied along φ = 0° and φ = 90° (Figures 7a and 7b, respectively), and they agree quite well. The probability distributions used in the modeling are presented in Figures 7c and 7d. For the 0% strains, p0° and p90° behave as one would expect; p0° changes relatively little after 100 cycles moving from 1 to 0.98. On the other hand, p90° is initially quite low at 0.406, and moves even lower after 100 cycles to 0.265. The differences in conductive pathway loss is what makes the resistivity and piezoresistivity stable for φ = 0° and unstable for φ = 90°. Furthermore, since p0° ≈ 1, percolation along the highly conductive φ = 0° path is essentially decoupled from the percolation properties along φ = 90°. It is also important to note that the decrease in p90° is strongly linear, and it is this steady decline which makes the AgNR flexible sheets attractive for failure based sensing. The probabilities ratios, p+25% / p0° and

14 ACS Paragon Plus Environment

Page 15 of 26

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 Journal of Physical Chemistry

p-25% / p0°, are more stable than those for φ = 90°, and both slightly increase in a linear fashion over 100 cycles. The increase in resistance for negative strains, which is atypical for metallic conductors, is captured by p-25% / p0° < 1. This suggests that the out-of-plane movement from buckling under compression temporarily disrupts charge percolation pathways along φ = 0°. For φ = 90°, the strain-induced changes to percolation pathways are different for tensile and compressive strains. For positive strains, the modeling suggests that p+25% / p90° = 1, even after 100 cycles, meaning that positive strains do no disrupt the percolation pathways along this direction. This indicates that it is the zigzag motion of electrons, moving mostly along the highly conductive φ = 0° direction that allows charge to percolate along φ = 90°. On the other hand, p25%

/ p90° is initially equal to 1, but continuously and rather significantly decreases over 100

cycles.

This result suggests that the buckling motion induced by the compressive strains

produces the irreversible changes in the conduction pathways for φ = 90°. The theoretical finding that buckling is the source of conduction degradation along φ = 90° also agrees well with the increasingly negative hysteresis areas seen for buckling in Figure 4f.

5. Conclusion Flexible AgNR sheets that exhibit anisotropic electrical behavior in regards to both their resistivity and their durability have been investigated as a sensing platform, and the underlying mechanisms have been analyzed using anisotropic effective medium theory.

The stable

piezoresistivity (in one orientation) and the SERS activity of the AgNR sheets make them attractive for flexible electronics applications such as electronic skin or as monitors for humanmachine interactions. However, the ability to encode a surface’s dynamic history into material properties is a considerable simplification over other systems and can enable wireless activity

15 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

monitoring where cost or demanding environments prevent more complicated devices from being implemented. Healthcare is one field where such a device could play an important role, and some examples of use could be as disposable smart bandages or as wearable tremor monitors for Parkinson’s patients.25-26 The AgNR sheets could also function as wireless wear monitors for infrastructure and components that have a finite lifetime, or they could monitor a container’s handling during shipping. Finally, this proof-of-concept report investigated AgNRs embedded in a flexible polymer, PDMS, which is well-suited for monitoring human-centric activities. However, the OAD-based process is compatible with a wide range of materials with different moduli, and the embedding medium could be tuned to better match specific applications. AgNRs combined with thin membranes could monitor infinitesimal strains,27 while more rigid embedding media could be used to monitor the activity of inelastic structures. Overall, the anisotropic electrical properties of these easily fabricated AgNRs offer a unique platform for sensing applications, and should be explored further.

Supporting Information Photos of the home-built stretching apparatus, photos of the response of the AgNR sheets under different strain conditions, and results from optical transmission measurements. This information is available free of charge via the Internet at http://pubs.acs.org

Corresponding Author * Email: [email protected]. Telephone: +1-706-542-7792.

Acknowledgment: This work was supported in part by the National Science Foundation (CBET1064228).

16 ACS Paragon Plus Environment

Page 16 of 26

Page 17 of 26

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 Journal of Physical Chemistry

REFERENCES: 1.

2. 3.

4.

5. 6.

7.

8.

9. 10.

11.

12. 13.

14.

Bauer, S.; Bauer‐Gogonea, S.; Graz, I.; Kaltenbrunner, M.; Keplinger, C.; Schwödiauer, R., 25th Anniversary Article: A Soft Future: From Robots and Sensor Skin to Energy Harvesters. Adv. Mat. 2014, 26, 149-162. Sun, Y.; Rogers, J. A., Inorganic Semiconductors for Flexible Electronics. Adv. Mat. 2007, 19, 1897-1916. Schindler, A.; Brill, J.; Fruehauf, N.; Novak, J. P.; Yaniv, Z., Solution-Deposited Carbon Nanotube Layers for Flexible Display Applications. Phys. E: Low-Dim. Sys. and Nanostructures 2007, 37, 119-123. Hammock, M. L.; Chortos, A.; Tee, B. C. K.; Tok, J. B. H.; Bao, Z., 25th Anniversary Article: The Evolution of Electronic Skin (E‐Skin): A Brief History, Design Considerations, and Recent Progress. Adv. Mat. 2013, 25, 5997-6038. Trung, T. Q.; Lee, N. E., Flexible and Stretchable Physical Sensor Integrated Platforms for Wearable Human‐Activity Monitoringand Personal Healthcare. Adv. Mat. 2016. Kim, D.-H.; Viventi, J.; Amsden, J. J.; Xiao, J.; Vigeland, L.; Kim, Y.-S.; Blanco, J. A.; Panilaitis, B.; Frechette, E. S.; Contreras, D., Dissolvable Films of Silk Fibroin for Ultrathin Conformal Bio-Integrated Electronics. Nat. Mat. 2010, 9, 511-517. Hwang, S. W.; Song, J. K.; Huang, X.; Cheng, H.; Kang, S. K.; Kim, B. H.; Kim, J. H.; Yu, S.; Huang, Y.; Rogers, J. A., High‐Performance Biodegradable/Transient Electronics on Biodegradable Polymers. Adv. Mat. 2014, 26, 3905-3911. Tao, H.; Brenckle, M. A.; Yang, M.; Zhang, J.; Liu, M.; Siebert, S. M.; Averitt, R. D.; Mannoor, M. S.; McAlpine, M. C.; Rogers, J. A., Silk‐Based Conformal, Adhesive, Edible Food Sensors. Adv. Mat. 2012, 24, 1067-1072. Lu, N.; Kim, D.-H., Flexible and Stretchable Electronics Paving the Way for Soft Robotics. Soft Robotics 2014, 1, 53-62. Qi, Y.; Jafferis, N. T.; Lyons Jr, K.; Lee, C. M.; Ahmad, H.; McAlpine, M. C., Piezoelectric Ribbons Printed onto Rubber for Flexible Energy Conversion. Nano Lett. 2010, 10, 524-528. Shim, B. S.; Zhu, J. A.; Jan, E.; Critchley, K.; Kotov, N. A., Transparent Conductors from Layer-by-Layer Assembled Swnt Films: Importance of Mechanical Properties and a New Figure of Merit. ACS Nano 2010, 4, 3725-3734. Durham, J. W.; Zhu, Y., Fabrication of Functional Nanowire Devices on Unconventional Substrates Using Strain-Release Assembly. ACS Appl. Mater. Inter. 2013, 5, 256-261. Kim, J.; Lee, M. S.; Jeon, S.; Kim, M.; Kim, S.; Kim, K.; Bien, F.; Hong, S. Y.; Park, J. U., Highly Transparent and Stretchable Field-Effect Transistor Sensors Using GrapheneNanowire Hybrid Nanostructures. Adv. Mat. 2015, 27, 3292-3297. An, B. W.; Hyun, B. G.; Kim, S. Y.; Kim, M.; Lee, M. S.; Lee, K.; Koo, J. B.; Chu, H. Y.; Bae, B. S.; Park, J. U., Stretchable and Transparent Electrodes Using Hybrid Structures of Graphene-Metal Nanotrough Networks with High Performances and Ultimate Uniformity. Nano Lett. 2014, 14, 6322-6328. 17 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

15.

16.

17. 18.

19. 20. 21.

22.

23. 24. 25. 26.

27.

Page 18 of 26

An, B. W.; Gwak, E. J.; Kim, K.; Kim, Y. C.; Jang, J.; Kim, J. Y.; Park, J. U., Stretchable, Transparent Electrodes as Wearable Heaters Using Nanotrough Networks of Metallic Glasses with Superior Mechanical Properties and Thermal Stability. Nano Lett. 2016, 16, 471-478. Gonzalez, M.; Axisa, F.; Bulcke, M. V.; Brosteaux, D.; Vandevelde, B.; Vanfleteren, J., Design of Metal Interconnects for Stretchable Electronic Circuits. Microelect. Rel. 2008, 48, 825-832. Siddall, G.; Smith, G., A Thin Film Resistor for Measuring Strain. Vacuum 1959, 9, 144146. Hawkeye, M. M.; Brett, M. J., Glancing Angle Deposition: Fabrication, Properties, and Applications of Micro-and Nanostructured Thin Films. J. of Vac. Sci. & Tech. A 2007, 25, 1317-1335. Hawkeye, M. M.; Taschuk, M. T.; Brett, M. J., Glancing Angle Deposition of Thin Films: Engineering the Nanoscale; John Wiley & Sons, 2014. Robbie, K.; Sit, J.; Brett, M., Advanced Techniques for Glancing Angle Deposition. J. of Vac. Sci. & Tech. B 1998, 16, 1115-1122. Larsen, G. K.; He, Y.; Ingram, W.; LaPaquette, E. T.; Wang, J.; Zhao, Y., The Fabrication of Three-Dimensional Plasmonic Chiral Structures by Dynamic Shadowing Growth. Nanoscale 2014, 6, 9467-9476. Chaney, S. B.; Shanmukh, S.; Dluhy, R. A.; Zhao, Y.-P., Aligned Silver Nanorod Arrays Produce High Sensitivity Surface-Enhanced Raman Spectroscopy Substrates. Appl. Phys. Lett. 2005, 87, 031908. Song, C.; Larsen, G. K.; Zhao, Y., Anisotropic Resistivity of Tilted Silver Nanorod Arrays: Experiments and Modeling. Appl. Phys. Lett. 2013, 102, 233101. Bernasconi, J., Conduction in Anisotropic Disordered Systems: Effective-Medium Theory. Physical Review B 1974, 9, 4575. McLister, A.; McHugh, J.; Cundell, J.; Davis, J., New Developments in Smart Bandage Technologies for Wound Diagnostics. Adv. Mat. 2016. Rigas, G.; Tzallas, A. T.; Tsipouras, M. G.; Bougia, P.; Tripoliti, E. E.; Baga, D.; Fotiadis, D. I.; Tsouli, S. G.; Konitsiotis, S., Assessment of Tremor Activity in the Parkinson’s Disease Using a Set of Wearable Sensors. Info. Tech. in Biomed., IEEE Transactions on 2012, 16, 478-487. Larsen, G. K.; Zhao, Y., Buckle-Driven Delamination of Hydrophobic Micro-, Nano-, and Heterostructured Membranes without a Sacrificial Layer. Nanoscale 2013, 5, 1085310857.

18 ACS Paragon Plus Environment

Page 19 of 26

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 Journal of Physical Chemistry

FIGURES:

Figure 1. (a) Aligned Ag nanorods are deposited on glass using OAD. (b) PDMS is applied to the substrate and cured. (c) - (d) Following curing, the PDMS is peeled from the glass with the aligned Ag nanorods embedded.

19 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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. Top-view SEM images of Ag nanorod substrates (a) before and (b) after PDMS application. Insets show the cross-section view of the corresponding structure. All scale bars are 500 nm.

20 ACS Paragon Plus Environment

Page 20 of 26

Page 21 of 26

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 Journal of Physical Chemistry

Figure 3. Resistivity of the AgNR film on glass and embedded in a PDMS sheet as a function of azimuthal angle, φ, with respect to the AgNR tilting direction.

21 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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 4.

Page 22 of 26

The piezoresistive effect of the flexible AgNR sheets for various tensile and

compressive loading cycles are shown when the resistance is measured along the (a) φ = 0° and the (b) φ = 90° directions of the flexible AgNR sheets. The resistance versus cycle number for 25%, 0%, and 25% strains when measured along (c) φ = 0° show a consistent piezoresistive response suitable for sensing applications, whereas the predictively changing values when measured along (d) φ = 90° would allow usage monitoring. The calculated hysteresis areas for buckling and stretching cycles applied along (e) φ = 0° have a slope which approaches zero further making readings more consistent while and the amplitude of them continues to increase (f) φ = 90° providing additional usage or wear information.

22 ACS Paragon Plus Environment

Page 23 of 26

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 Journal of Physical Chemistry

Figure 5. A manual demonstration of application as a gauge sensor. The resistance change corresponds to the stretching of the substrate from the bending motion of the finger with the peak value being related to the amount the finger is bent.

23 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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 24 of 26

Figure 6. SERS spectra for the AgNR sheets before and after application of BPE (10-3 M) for various strains applied along (a) φ = 0° and (b) φ = 90°. (c) Relative intensity of the 1207 cm-1 peak versus strain for φ = 0° and 90°.

24 ACS Paragon Plus Environment

Page 25 of 26

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 Journal of Physical Chemistry

Figure 7. Comparison between the experimental results and the AB2-EMA model for (a) φ = 0° and (b) φ = 90°. Probabilities used in the AB2-EMA model for (c) φ = 0° and (d) φ = 90°.

25 ACS Paragon Plus Environment

The Journal of Physical Chemistry

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

TOC/Abstract Graphic:

26 ACS Paragon Plus Environment

Page 26 of 26