Electrowetting-on-dielectric behavior of aqueous droplets and gold nanofluid on an electrospun poly(vinyl chloride) microfiber layer

Morphology and wettability of dielectric layers are crucial properties that affect the electrowetting-on-dielectric (EWOD) performance of a working liquid. In this work, the use of a poly(vinyl chloride) (PVC) microfiber-modified PVC dielectric layer as an electrowetting-on-dielectric (EWOD) substrate is explored. Imaging with scanning electron and atomic force microscopy revealed intertwined microfibers on the original PVC film after PVC deposition via electrospinning. Hydrophobicity of the PVC layer was enhanced by the presence of microfibers, with the contact angle (CA) for a water droplet increasing from 84.9° to 123.9°. EWOD behavior of various liquids on the microfiber-modified PVC layer was investigated within a DC voltage range of 0 to 200 V DC. Gold nanofluid exhibited the largest change in CA of 57°, while smaller changes were observed for KOH (19.6°), KCl (21.2°), and water (21°). A mechanism explaining the observed reduction in contact angle for a gold nanofluid droplet is presented. Our results suggest the promising potential of PVC film-PVC microfiber as a hydrophobic dielectric material for EWOD applications involving diverse liquids, including gold nanofluids.


Introduction
The electrowetting-on-dielectric (EWOD) behavior of a working liquid is one of the governing factors in optimizing the performance of electrically-driven liquid-based devices [1][2][3].The phenomenon of EWOD refers to the change in contact angle (CA) or curvature of a liquid on a dielectric material in response to the presence of an electric field [4].This evolving, non-mechanical liquid manipulation technique has been attracting extensive attention due to its fast response time, low-power operation, simple instrumentation, and strong actuation effects as compared to other liquid actuation techniques [5].These advantages of EWOD have been recently exploited in various systems, including liquid lenses and prisms [6], liquid electronic displays [7], solar cells [8], and microfluidic devices [9].
One of the crucial components of an EWOD system is the dielectric layer.The governing principle of EWOD suggests that the change in a liquid's contact angle is directly affected by the electric permittivity of the dielectric material and inversely related to the thickness of the dielectric layer [4].Hydrophobicity and the micro/nanoarchitecture of the surface also affect the surface tension and movement of the fluid on the substrate [10].
Typically, inorganic dielectric materials such as SiO 2 [15,16], Ta 2 O 5 [17][18][19], ZrO 2 [20,21], ZnO [22][23][24], TiO 2 [25][26][27], and Si 3 N 4 [28][29][30] are utilized as dielectric layers due to their high electric permittivity (dielectric constant) and mechanical stability.They are deposited above a conducting layer using various deposition techniques such as anodization [18], atomic layer deposition [31], sputtering [32], electro-spraying [33], and electrochemical deposition [34].These inorganic materials are inherently hydrophilic [35] and necessitate the addition of hydrophobic layers on top of the dielectric layer.It should be emphasized that an EWOD substrate material should exhibit a reliable hydrophobicity to sustain droplets on the surface.The above deposition techniques, coupled with the need for an additional hydrophobic layer, limit the availability and costeffectiveness of inorganic materials as EWOD substrates [36].As an alternative, the use of dielectric polymers is being explored [36,37].A wide array of polymers possesses considerable dielectric properties for EWOD applications.In addition, the architecture of polymeric layers may be easily integrated and installed on the conducting layer using simple techniques such as drop casting, dip coating, and spin coating.Although their dielectric constants are usually lower than that of inorganic materials, recent work has demonstrated the successful fabrication of EWOD-based devices using polymers as a dielectric layer.Park et al fabricated an EWOD-driven multifunctional liquid lens using Parylene and Teflon as dielectric and hydrophobic layers, respectively [6]. Lee demonstrated a microfluidic device for bubble oscillation and transportation applications using polyimide and Teflon as a dielectric layer [38], while Kurawan and colleagues fabricated an EWOD-driven microfluidic device utilizing polydimethylsiloxane (PDMS) as a dielectric material for electrochemical detection [39].The devices mentioned above use a bilayer of dielectric polymers that acts as the main barrier preventing current flow [6,38,39].Such polymers, including PDMS, typically exhibit low hydrophobicity, with water contact angles in the range of 105°to 110° [40,41].Improving hydrophobicity requires the addition of layers of Teflon or fluorinated polymers, which have disadvantages of high cost, limited solvent compatibility, and adhesion issues [42,43].The production and disposal of fluorinated polymers is also subject to strict regulations due to environmental risks and potential health hazards [44], limiting availability of the material for users.
There remain other polymers that were reported to possess promising dielectric properties and hydrophobicity but have not yet been seriously considered for EWOD applications.One example is polyvinyl chloride (PVC) that was reported to have a dielectric constant (k) ranging from 3.29 to 10.3 [45].This is comparable with the dielectric constant of PDMS (k =2.7-3.69) and other commonly used dielectric polymer layers such as Teflon (k = 2.1) and Parylene (k = 2.2-3.15)[45,46].PVC has the advantages of being widely available, chemically nonreactive, and moisture-and heat-resistant which are important features of an EWOD substrate, particularly when working with an aqueous liquid at high voltage operations.However, there have been no reports on the utilization of PVC layers as the dielectric component of an EWOD system.
In this work, we investigated the potential of PVC as the dielectric and hydrophobic layer of an EWOD substrate.The PVC was deposited on ITO glass via a modified drop casting technique.The surface of the dielectric layer was further modified by adding PVC microfibers through electrospinning to enhance hydrophobicity.Electrospinning is a technique of generating fine nano-and microfibers from solution precursors and polymeric blends using a fairly high electric field [47].This technique allows the production of fibers with substantial control over the required morphology, density, and composition [48,49].Although electrospun polymer fibers and other fiber-like structures have been extensively used in tissue engineering, energy storage devices, sensors, membranes and filters and other applications [48,50], electrospun fibers have not been employed in electrowetting systems, based on our review of available literature.Our use of PVC addresses the need to find alternative dielectric materials and design strategies that enhance hydrophobicity while avoiding material and toxicity concerns associated with fluorinated coatings.
Additionally, EWOD behavior is examined for various aqueous droplets, such as KOH and KCl, and a nanofluid containing gold nanoparticles on PVC dielectric layers modified with microfibers.The aqueous droplets were chosen since they are typically used as working fluids in various microfluidic and actuator devices.To the best of our knowledge, this is the first work to report the EWOD behavior of aqueous droplets on PVC layers and PVC microfibers.Results from this work may offer new perspectives on substrate preparation and modification for EWOD devices, leading to improved performance in fluidic optics and other droplet applications.

Substrate preparation and PVC coating
The ITO glass substrates were cleaned by sonicating the sheets, followed by rinsing with acetone and deionized water.Substrates were then dried in a convection oven at 80 °C for 3 h.A manual dip-coating method was employed to introduce a PVC layer on the ITO glass substrate.20 ml of 15% PVC gel in THF/DMF (v/v, 4:1) was prepared in a beaker.A pre-cleaned glass slide (1 cm × 1 cm) was dipped in the PVC gel for 10 s and withdrawn rapidly by hand.Afterward, the ITO sheet with a PVC layer was soft baked on a hot plate at 100 °C for 1 h and then cured in a convection oven at 50 °C for another 3 h.

Surface modification with PVC microfibers
The PVC microfibers were added to the PVC-coated ITO sheets via electrospinning.In this method, 3 ml of 15% PVC gel in THF/DMF (v/v, 4:1) was placed in a syringe connected to a mechanical syringe pump that extruded PVC at a rate of approximately 2 ml/hour.The coated ITO glass sheets were mounted on a planar collector placed vertically, 18 cm from the tip of the syringe.PVC microfibers were deposited onto the ITO surface as the substrate was held stationary on the planar collector.Deposition time for the microfibers was 60 s.The operating voltage was set at 15 kV to produce consistent jets of PVC microfibers.Substrates with deposited microfibers were air dried for at least 1 h to remove any residual solvent.

Electrowetting-on-dielectric experiments on fabricated substrates
Standard electrowetting experiments were performed with the PVC-modified EWOD substrates and various aqueous droplets.A variable DC power supply (0-400 V) was used to apply external voltage across the droplet while contact angles were measured via the sessile method using an optical tensiometer (ThetaLite, Biolin Scientific).Figure 1 summarizes the fabrication methods and provides a schematic diagram of the EWOD setup.

Characterization techniques
Surface topography of the fabricated substrates was examined using atomic force microscopy (AFM, XE-70 Park Systems) while two-dimensional morphology was examined using scanning electron microscopy (SEM, Phenom XL desktop Thermoscientific).Determination of functional groups present in the sample was done using Fourier transform infrared spectroscopy (FTIR, Shimadzu Prestige-21).Hydrodynamic sizes of gold nanoparticles were measured using a dynamic light scattering (DLS, Malvern Zetasizer) instrument.Micrographs of the samples were captured using an optical microscope (AmScope).Contact angle measurements were verified by image drop analysis using ImageJ software [51].

EWOD substrate surface features and contact angle measurements
Surface structure is critical in determining the hydrophilicity or hydrophobicity of a material.Our ITO glass substrates were coated with PVC to prevent the transport of charges from the droplet to the conducting ITO layer.Such charge transport should be avoided to prevent the electrolysis of water molecules.As seen in figure 2(a), the dip coating step resulted in a PVC layer with a thickness of around 9.9 μm.After the addition of PVC microfibers, a total substrate thickness of about 15.7 μm was observed in optical micrographs of the samples (figure 2(b)).The topography of the fabricated substrate was investigated using AFM measurements in non-contact mode.After depositing PVC microfibers on the PVC layer, the presence of randomly oriented PVC microfibers with diameters of roughly 1 μm was observed (figure 2(c)).Intertwined microfiber structures are also present in the SEM image (figure 2(d)).The microfibers were uniformly distributed across the surface of the glass but were randomly intertwined due to the motion of the polymer jet.Since our sample was relatively small, the surface was completely covered by microfibers.
FTIR spectra of the PVC powder precursor, dip coated PVC film, and PVC microfibers made through electrospinning were obtained to confirm the chemical identity of PVC in its different forms.Results presented in figure 3 revealed the consistency of the vibrational bands of anticipated functional groups in all samples, Water CA for ITO glass and PVC films with and without a microfiber layer was noted before electrowetting experiments, as illustrated in figure 4. As expected, ITO glass in figure 4(a) demonstrates poor hydrophobicity due to its hydrophilic nature as an oxide-rich material.When a PVC layer was added, an increase in the contact angle from 41.5°to 84.9°was observed in figure 4(b), affected by the hydrophobic characteristics of the PVC film.Water CA was enhanced even further, up to 123.9°, due to the presence of PVC microfibers (figure 4(c)).This significant increase in CA is attributed to a wetting transition from a Wenzel to Cassie-Baxter state [51,52].In a Wenzel state, the aqueous droplet is in contact with a relatively smooth dielectric surface, as in figure 4(d), yielding a smaller water CA value.Meanwhile, in figure 4(e), the hydrophobic and irregular ridges of PVC microfibers trap air and prevent the water droplet from penetrating the flat PVC layer, leading to a Cassie-Baxter wetting state.In this state, air pockets present within the intertwining microfibers create micro-interfaces of water-air beneath the aqueous droplet that increase the observed contact angle [52][53][54].

Electrowetting-on-dielectric behavior on a PVC layer
In electrowetting experiments, a DC voltage was applied across the ITO glass and droplet through a platinum electrode and the change in contact angle at increasing voltage levels was recorded.Figure 5(a) shows the actuation of a representative water droplet on ITO glass with a PVC film for an applied voltage range of 0 to 200 V.In the absence of an electric field, CA of the water droplet was found to be 84.9°.A decrease in CA of 5°w as observed after applying 60 V, with a final CA of 63.3°at the maximum voltage of 200 V. Enhanced hydrophobicity of the EWOD substrate with a PVC layer and added PVC microfibers is evident in figure 5(b).Initial CA for this configuration was 123.9°, decreasing by 7°at 60 V of applied voltage.At 200 V, the CA dropped to 102°.It is worth mentioning that the observed change in water contact angle (ΔCA) on the PVC layer with microfibers is comparable to or better than EWOD systems with different dielectric materials such as ZnO nanorods (∆ = CA 13°) [22], low-density polyethylene plastic (∆ = CA <10°) [55], porous polytetrafluoroethylene (∆ = CA <10°) [56], Teflon-coated SiO 2 (∆ = CA <10°) [33], and SU8 2005/PDMS (∆ ) = <  CA 10 [57] within the range of 0 to 110 V.The droplet was stable on the surface after being actuated at any voltage.No reversibility of the droplet's contact angle was observed when the applied voltage was removed, reduced, or reversed.The Young-Lippman equation describes the contact angle as a function of applied voltage: where q o is the initial contact angle with no applied voltage, V is the applied voltage, g lg is the liquid-gas surface tension, ε 0 is the dielectric permittivity in vacuum, ε is the relative dielectric permittivity of the dielectric material, and d is the thickness of the dielectric layer.In figure 5(b), experimental CA values for the PVC film follow the Young-Lippman equation only up to a certain applied voltage.Deviation from theory is observed at voltages larger than 100 V, indicating CA saturation beyond this applied potential.Saturation may be attributed to the breakdown of the dielectric layer, charge pinning on the droplet edges, or micro-expulsion of water droplets at higher voltages [58].When using the measured thickness (d = 15.7 μm) of the PVC film with PVC microfibers in the Young-Lippman equation, the theoretical curve in figure 5(d) fails to capture the trend of experimental CA values.We note that the Young-Lippman equation assumes a smooth surface and the effect of a rough surface topology is not accounted for [59].Similar discrepancies have been reported by previous studies investigating the wetting behavior of fluids on nanostructured surfaces [22,60].Much better agreement between experiments and the Young-Lippman equation is achieved with a thickness of d = 10.5 μm.Considering an effective thickness leads to the Young-Lippman equation remaining valid for a dielectric layer comprising a PVC film and PVC microfibers until CA saturation occurs at 110 V, comparable with the performance of an unmodified PVC film.This effective thickness is attributed to liquid penetrating gaps in the intertwined PVC microfibers as an electric field is applied to the droplet.Such behavior is expected for rough or irregular surfaces, as proposed by the Cassie-Baxter to Wenzel state transition model [22,61,62].The wetting transition from Cassie state to Wenzel state using an applied voltage has been observed in previous studies [22,[61][62][63].Applying significant levels of voltage beyond the breakdown threshold induced leakage currents through air pockets [22].Subsequent disturbances in the equilibrium energy and interfacial surface tensions occur which result in depinning of the droplet and deviation from the Cassie-Baxter model [62,63].This wetting transition typically stays even after removing the applied voltage.With our experiments, we observed a consistent transition from Cassie-Baxter state to Wenzel state the voltage was applied for an extended period.The droplet remained in a Wenzel state even when the applied voltage was switched off, making the wetting transition irreversible.

Electrowetting-on-dielectric of aqueous and nanofluid droplets
The PVC film-PVC microfiber substrate was employed in additional experiments to examine the EWOD behavior of droplets of 1 M KOH, 1 M KCl, and gold nanofluid.Gold nanoparticles within the nanofluid exhibit a quasi-spherical morphology with a diameter of around 30 to 50 nm, as verified in the SEM image of figure 6(a).The mean hydrodynamic diameter of the nanoparticles from dynamic light scattering measurements was recorded to be 33.6 ± 1.9 nm, based on the histogram in figure 6(b), with a polydispersity index of 0.4, signifying good stability and dispersity in water.The difference in nanoparticle size measurement between SEM and DLS has been associated with different factors such as nanoparticle aggregation during the imaging process, concentration, and nanoparticle-liquid interaction [64,65].Larger particle sizes obtained in the SEM images are attributed to the aggregation of nanoparticles during the sample preparation for the imaging process.Since the nanoparticles were sonicated and suspended in the water during DLS measurement, aggregation may have been minimized and the hydrodynamic diameter of individual nanoparticles was obtained.We note that nanoparticles with diameters of about 30 nm are seen in the SEM image, coinciding with the hydrodynamic size obtained from DLS.Initial CAs for aqueous KOH, KCl, and nanofluid droplets without an applied voltage were 124.2°, 124.6°, and 121.8°, respectively.Within the voltage range of 0 to 200 V, a recorded change in the CA of 19.6°and 21.2°w as recorded for the KOH and KCl droplets, respectively.The modification in CA is very close to the actuation observed with a water droplet, as summarized in figure 7(a), suggesting that the presence of ions does not significantly affect EWOD behavior, even in the case of KOH and KCl solutions where ionic conductivity is expected to be much greater than in water.Experimental data also match the trend predicted by the Young-Lippman equation when an effective thickness of d =10.5 μm is applied.
The presence of gold nanoparticles within the droplet resulted in a more pronounced difference in EWOD behavior over the same voltage range.Initial CA of nanofluid on the PVC film-PVC microfiber substrate was measured to be around 121.8°which decreased to 64.8°after applying 200 V, corresponding to a total change of around 57°.Our experimental CA values deviate significantly from the expected behavior if parameters valid for other aqueous liquids are used in the Young-Lippman equation, as seen in figure 7(b).The observed CA behavior for nanofluids in this EWOD system is comparable with the exhibited reduction of CA values for Bi 2 Te 3 [66], core/shell CdSe/CdS and CdTe/CdS quantum dots [67], and gold nanofluids [68] at similar applied voltages.the succeeding discussion, we present a possible mechanism for how droplet actuation is influenced by suspended gold nanoparticles altering the liquid-gas surface tension of the droplet.
Despite numerous studies on EWOD behavior of nanofluids containing various nanoparticles such as quantum dots and gold nanoparticles being available in the literature, there is still no definitive mechanism for the enhancement of electrically-driven CA change in nanofluid droplets.Figure 8 is a schematic illustration of how the suspended Au nanoparticles may affect the liquid vapor surface tension and electrowetting behavior of the droplet.Figure 8(a) illustrates the suspended Au nanoparticles in the droplet.In the work of Patacsil et al [69] and Kumar et al [70], the nanoparticles (figure 8(a)) were considered as polar dielectrics that were suspended in the liquid medium.When a potential difference is applied, nanoparticles are polarized, creating a series of capacitors within the droplet as shown schematically in figure 8(b).With this assumed mechanism, the resulting capacitance of the nanofluid-based EWOD system is approximately equal to the capacitance of the dielectric layer; the presence of nanoparticles has a negligible effect on the total capacitance of the droplet.
Another possible driving factor that led to an enhanced EWOD behavior in nanofluids is the decrease in liquid-gas surface tension due to the suspended nanoparticles [71,72].Tohgha and colleagues reported an apparent decrease in the liquid-gas surface tension of quantum dot-based nanofluids [73].Figure 8(c) illustrates how surface tension in a pure water droplet is the result of attraction between molecules.In water, the weak negative charges of oxygen atoms tends to attract the weak positively charged hydrogen atoms.The uneven charge distribution and polarization cause water molecules to attract each other.At the interface, interfacial water molecules experience an inward pull due to the attractive forces of the molecules.This cohesive force is manifested as the surface tension at the water-gas interface.In the case of a nanofluid, the presence of suspended nanoparticles affects the attraction among water molecules.The commercial nanofluid in our experiments is hydrodynamically stable, implying a strong repulsive force between suspended nanoparticles and attraction with surrounding water molecules.This repulsive interaction among the nanoparticles reduces the intermolecular attraction of water molecules, thereby decreasing the water-water attraction and liquid-gas surface tension [73], as illustrated in figure 8(d).It is well-known that the inherent high surface area to volume ratio of nanoparticles actively affects the behavior of the fluid in which they are suspended [69][70][71][72][73].
In figure 9, we plot q q cos cos 0 as a function of V 2 to estimate the effective g lg for the droplets employed in our PVC film-PVC microfiber EWOD system.Corresponding slopes for each fluid are extracted based on data points prior to the onset of CA saturation.From equation (1), these slopes give the capacitance divided by g .lg Since our EWOD substrates have identical dielectric compositions and thicknesses, the capacitance is considered uniform for all samples.As expected, behavior of aqueous droplets of KOH and KCl is found to have a slope similar to that of water.Effective liquid-gas surface tensions are calculated to be 0.070 N m −1 , 0.064 N m −1 , and 0.066 N/m for water, KOH and KCl, respectively.We note that the g lg value obtained for water is consistent with the actual g lg of water at room temperature [69][70][71][72].The g lg for the nanofluid was determined to be markedly different from that of the aqueous solutions, with a calculated value of 0.025 N m −1 .This surface tension falls within the range of g lg reported for nanofluids containing quantum dots [73] and gold nanoparticles [69].Using this effective g lg in the Young-Lippman equation, better agreement is achieved between the theoretical curve and experimental data for the effect of applied voltage on CA (figure 7(b)), specifically at low voltages.Accounting for a decrease in droplet surface tension due to electrostatic repulsion caused by suspended nanoparticles clearly leads to a more accurate mechanism for EWOD behavior of the nanofluid.Our current model still does not completely describe the observed droplet actuation, given the remaining difference between the theoretical curve and experimental results even after considering reduced surface tension, particularly at higher voltages.We suspect that this discrepancy could be attributed to effects outside the scope of the Young-Lippman model, such as nanoparticle layering or deposition on the solid-liquid interface.Due to the surface charge of suspended nanoparticles, migration and deposition on the substrate surface through electrophoresis could also contribute to the change in wettability.A more detailed investigation of nanoparticle-related phenomena is recommended for future work.

Summary
We report the successful fabrication of a PVC film-PVC microfiber dielectric substrate for EWOD applications, with enhanced hydrophobicity and contact angle actuation.Static CA for a water droplet on a PVC film dielectric substrate increased from 84.9°to 123.9°with the addition of PVC microfibers deposited via electrospinning.Total change in CA within a DC voltage range of 0 to 200 V for the PVC film substrate was extended from 17°to 21°when PVC microfibers were added.EWOD behavior of 1 M KOH and 1 M KCl solutions on the PVC film-PVC microfiber substrate was nearly identical to that of water.A wider range of droplet actuation, from an initial CA of 121.8°to a final value of 64.8°after applying 200 V, was observed with a nanofluid containing gold nanoparticles.The more pronounced wetting of the nanofluid is attributed to the reduction of liquid-gas surface tension, from 0.070 N m −1 (water) to 0.025 N m −1 , by suspended nanoparticles.
Our results demonstrate the feasibility of PVC film-PVC microfiber as the dielectric material in EWOD architectures.The novel substrate reported here is compatible with aqueous and nanofluid droplets and may easily be configured for various applications, with low voltages needed to achieve significant contact angle actuation.We see potential utility of the composite PVC substrate in fluid sensing, microchannel transport, and fluidic optics for beam steering.The results obtained from the EWOD behavior of nanofluids have also contributed to our knowledge of this complex electrically-driven phenomenon in nanofluids.

Figure 1 .
Figure 1.Summary of the fabrication and experimental workflow implemented in the study, illustrating the drop casting and electrospinning steps, and EWOD experiment.

Figure 2 .
Figure 2. Surface features of PVC-modified ITO glass: optical micrographs of substrate cross-sections after (a) dip coating with PVC and (b) electrospinning of PVC microfibers.The presence of microfiber structures is confirmed by (c) AFM and (d) SEM imaging.

Figure 3 .
Figure 3. FT-IR spectra of the PVC powder, film, and microfibers.

Figure 4 .
Figure 4. Photographs of water droplets and corresponding contact angles on (a) ITO glass, (b) PVC film, and (c) PVC film with PVC microfibers, with (d) the originally smooth PVC layer becoming (e) a rough surface due to the presence of microfibers which results in an increase in hydrophobicity.

Figure 5 .
Figure 5. Water droplet actuation on (a) a PVC film and (b) a PVC film-PVC microfiber layer at different voltages.CA versus voltage plot for a water droplet on (c) a PVC film and (d) a PVC film-PVC microfiber layer.

Figure 6 .
Figure 6.Nanofluid properties: (a) SEM image of gold nanoparticles from the nanofluid, with a magnified view showing quasispherical shapes (inset), and (b) size distribution of nanoparticles suspended in water obtained from dynamic light scattering.

Figure 7 .
Figure 7. Contact angle versus applied voltage for: (a) various aqueous droplets, with solid lines representing the theoretical Young-Lippman curve for d =15.7 μm and dashed lines representing the curve for an effective thickness d =10.5 μm, and (b) a nanofluid with suspended gold nanoparticles, with solid lines indicating theoretical behavior using d = 10.5 μm and g lg for water while the dashed line is the Young-Lippman curve using an effective g lg .

Figure 8 .
Figure 8. Schematic diagrams of nanofluid droplets with (a) suspended nanoparticles (red circles) at applied = V 0 becoming (b) polarized at ¹ V 0. The droplet surface tension is due to the (c) intermolecular attraction in the water and (d) is reduced by nanoparticles disturbing the water-water attraction.