Wetting and Evaporative Aggregation of Nanofluid ... - ACS Publications

Feb 23, 2014 - The wetting and evaporative aggregation of alumina nanofluids (Al2O3) are examined for CVD-synthesized graphene-coated (GC) surfaces ...
0 downloads 0 Views 5MB Size
Download PDF
Article pubs.acs.org/Langmuir

Wetting and Evaporative Aggregation of Nanofluid Droplets on CVDSynthesized Hydrophobic Graphene Surfaces Jae S. Park,† Kenneth D. Kihm,*,‡ Honggoo Kim,† Gyumin Lim,† Sosan Cheon,† and Joon S. Lee† †

School of Mechanical and Aerospace Engineering, Seoul National University, Gwanak-gu, Seoul 151-742, South Korea Department of Mechanical, Aerospace and Biomedical Engineering, The University of Tennessee, Knoxville, Tennessee 37996-2210, United States

‡

S Supporting Information *

ABSTRACT: The wetting and evaporative aggregation of alumina nanofluids (Al2O3) are examined for CVD-synthesized graphene-coated (GC) surfaces that are known as strongly hydrophobic (θcontact ≈ 90°). Our findings are compared to those associated with a hydrophilic cover glass (CG) substrate (θcontact ≈ 45°). The nanofluidic self-assemblies on the GC substrate are elaborately characterized in terms of the droplet wetting/crack formation, the particle migration time over the evaporative time (CR), the Derjaguin−Landau−Verwey− Overbeek forces (FDLVO), and the relative thermal conductivity (KR). The GC substrate forms relatively thicker and larger cracks and requires a longer evaporation time. Both the GC and CG substrates share approximately the same time constant CR, which suggests the formation of coffee-ring patterns for both substrates. The GC shows negative FDLVO, which implies a repulsive force between the nanoparticles and the substrate, and the CG shows a positive FDLVO of attraction. Furthermore, a more than 3 order of magnitude larger thermal conductivity of GC compared to that of CG drives significantly different particle/fluid motions near the drop edge areas between the two substrates. with different wetting characteristics,18,19 chemical or physical coating materials, and microfabricated surface morphologies.20,21 This article presents experimental examinations of the nanofluidic aggregation on the hydrophobic graphene-coated (GC) surface in comparison with the hydrophilic cover glass (CG) surface. Substantial elaboration is also presented to interpret the experimental findings for the physicochemical aspects, including the droplet profiles, evaporation rates, particle movement, particle volume fractions, particle interaction forces with the substrate, and relative thermal conductivities for the two distinctive substrate surfaces.

1. INTRODUCTION Graphene,1 the hexagonal lattices of carbon atoms in submicrometer thickness (0.335 nm), boasts extremely high electrical and thermal conductivities as well as optical transparency.2−5 Graphene is also very hydrophobic, and its hydrophobicity depends somewhat on the number of layers.5 These unique multiphysical properties make it possible to consider graphene layers as a future material to use for many potential purposes. A physical understanding of wetting and aggregation of nanofluids on the graphene surface will be necessary in considering a number of potential applications of graphene such as surface modification structuring, inkjet printing on the graphene surface, high-quality packaging using the hydrophobic graphene layers, a nanocrystalline surface coating with graphene layers, and others. The nanofluidic crystallization involves sophisticated multiphase phenomena of the evaporation of base fluid and the self-assembly of nanoparticles.6,7 Both theoretical8−11 and experimental efforts6,12,13 have widely developed the nanofluidic self-assembly research communities. The evaporative aggregation of nanoparticles is determined from the interactions between nanofluids and the substrate, which depend on the nanoparticle size, concentration, and composition.14,15 Recently, it was verified that microcavities could be formed under the self-assembly process of nanofluids16 and that the hydrophobicity of the substrate was an important factor in cavity formation.17,18 Diverse examinations have been previously conducted for various types of substrates © 2014 American Chemical Society

2. EXPERIMENTAL SECTION Dynamic Wetting and Nanoparticle Aggregation. Nanofluid samples were prepared by ultrasonically dispersing alumina (Al2O3) nanoparticles of 45 nm average diameter and 3.97 g/cm3 nominal density. We purchased Al2O3 powder manufactured by Nanophase Technologies Corporation and sold by Alfa Aesar Inc. under the product name NanoDur. Our tested nanofluids were prepared by the dispersion of these nanoparticles in distilled deionized pure water with no electrolytes, dispersants, or surfactants. A droplet of nanofluid (1.5 μL ± 0.15 μL) was dropped gently on a substrate using a micropipet to allow evaporation in an environment with controlled temperature (22 ± 0.5 °C) and humidity (35 ± 5%). Both dorsal and ventral Received: December 18, 2013 Revised: February 20, 2014 Published: February 23, 2014 8268

dx.doi.org/10.1021/la404854z | Langmuir 2014, 30, 8268−8275

Langmuir

Article

images were dynamically recorded to observe the movement of contact lines and cavity/crack formation (Figure 1a,b). In addition,

Figure 2. (a) Schematic of the low-pressure, high-temperature chemical vapor deposition (CVD) system for graphene synthesis and (b) schematic illustration of the graphene growth on a copper foil: (1) 99.9999% pure copper foil, (2) recrystallization of the copper foil after annealing at 1000 °C, (3) initial stage of graphene growth forming graphene islands, and (4) complete growth of single-layered graphene on the copper foil. The CVD-synthesized graphene layer was transferred to a test substrate using the PMMA method (Figure 3a). The CVD graphene grown on the Cu foil (step 1) was spin-coated with PMMA (poly(methyl methacrylate)) solution and heat cured to 160 °C (step 2). The 46 mg/mL PMMA solution was prepared by dissolving PMMA powder (Sigma-Aldrich Inc.) in chlorobenzene. The extra graphene grown on the back side of the copper foil was removed by O2 plasma in a reactive ion (RI) etcher (step 3), and then the copper foil was dissolved by copper etchant containing ammonium peroxydisulfate H8N2O8S2 (step 4). After being rinsed thoroughly (step 5), graphene-PMMA was deposited on a cover glass and baked to 80 °C (step 6). The PMMA film was removed with acetone (Step 7), and finally the graphene deposited on the cover glass was cured in isopropyl alcohol to complete the transfer process (step 8). Figure 3b shows a qualitative inspection of the uniformity of the graphene layer as transferred to a cover glass by optically discriminating the visible light attenuation of approximately 2.3% per graphene layer.24 The Raman spectroscopy verifies the unique and quantitative characteristics of graphene by providing the intensity peak ratio of the 2D peak (∼2700 cm−1) to the G peak (∼1580 cm−1) to be 3.7 on average (Figure 3c). In addition, the negligible D peak (∼1350 cm−1) shows a high-quality graphene film with sufficiently large grain sizes.

Figure 1. Evaporative nanofluidic aggregation on a graphene substrate: (a) Schematic illustration of goniometric imaging (1), dorsal imaging with an upright microscope (2), and ventral imaging with an inverted microscope (3). (b) Dorsal (left) and ventral (right) views of a driedout nanofluid droplet of 1.5 μL initial volume with 1.25% volume Al2O3 nanoparticles of 45 nm nominal diameter. (c) Goniometric images of a similar nanofluid droplet at the beginning of evaporation (left) and after the completion of evaporation (right).

goniometric imaging (Figure 1c) allowed the monitoring of the evolution of the droplet geometry and contact angles by using ImageJ software that uses an angle tool after fitting the tangential and horizontal contact lines to the droplet profile. Tests were conducted for both a graphene-coated (GC) substrate and a pristine cover glass (CG) substrate for different nanoparticle volume concentrations ranging from 0.1 to 10%. Precleaned and factory-sealed cover glasses were used without additional cleaning to avoid any undesirable contamination by dust or stains. Fabrication of Graphene-Deposited Substrates Using CVD Synthesis. The graphene layer was synthesized by a low-pressure chemical vapor deposition (CVD) process, which was fully implemented at the Multiscale Optical Characterization Laboratory in the World Class University (WCU) Program at Seoul National University, Korea (Figure 2a). The CVD chamber pressure was maintained at approximately 10−3 Torr, and a 99.9999% purity copper foil of 25 μm thickness (Alfa Aesar Inc.) was used as a catalytic substrate.22,23 After 30 min of annealing the copper foil at 1000 °C for recrystallization, methane (CH4) gas was introduced to supply carbon atoms, allowing them to deposit on the copper foil with hydrogen (H2) gas as a cocatalyst (Figure 2b).

3. RESULTS AND DISCUSSION Experiments were conducted for the graphene-coated (GC) substrate in comparison with the cover glass (CG) substrate and for three different nanoparticle volume concentrations. Elaborations of the experimental findings were made to differentiate the distinctive wetting and dry-out characteristics of nanofluids on the GC substrate from those on the CG substrate. Formation of Aggregates and Crystallization. Figure 4 shows the dry-out patterns of nanofluid drops of 1.5 μL for a total of six tested cases: three different volume concentrations of Al2O3 nanoparticles (0.1, 1.25, and 10%) for both GC and CG substrates. The time evolution images for all of these conditions are shown in Figure S1 in the Supporting 8269

dx.doi.org/10.1021/la404854z | Langmuir 2014, 30, 8268−8275

Langmuir

Article

Figure 3. Preparation of a CVD graphene-coated substrate. (a) Schematic illustration of the PMMA-transfer process of CVD graphene film from a copper foil to a cover glass. (b) Optical microscope image of a CVD graphene layer coated on a cover glass. (c) Raman spectroscopy confirming a monolayer of CVD graphene and its pristine quality.

The main drive in crack formation is the stress-generated tension in the evaporative nanocrystalline structure.25 Then, the cracks grow in the direction where the maximum amount of stress can be released so that the free energy of the system decreases. In the absence of surface functional charges of the nanoparticles,26 the crack spacing is proportional to the thickness of an aggregated layer. Likewise, the thickness of the aggregated layer tends to be larger with a larger contact angle of the droplet under evaporation. Therefore, the hydrophobic GC substrate with a larger contact angle generates thicker cracks with a larger spacing compared to that of the CG substrate, as shown by the ventral images in Figure 4. Note that the contact angles and other wetting characteristics of pure water with no aggregation on graphene surfaces are available from a published article by Rafiee et al.5 Although the wetting diameters for CG remain unchanged for most of the evaporation process, those for GC are subjected to a decrease as evaporation progresses. This is believed to be attributed to the relatively weak pinning strength on the hydrophobic GC substrate. The receding of the pinning is more pronounced for the lower nanoparticle concentrations (0.1 and 1.25%); however, the receding is negligibly small for the highest concentration of 10% (Figure S1). Goniometric imaging of nanofluidic drops allowed the recording of the time-evolving wetting diameters (Figure 5a), the contact angle variations (Figure 5b), and the maximum height of the drop (Figure 5c) during evaporative crystallization. The error bars represent the 95% confidence levels of three individual data sets. For the tested ranges of nanoparticle concentration from 0.1 to 10%, the physical observations for the evolution of nanofluidic drops upon contacting the substrates are that (1) the initial wetting diameter of the drop for the graphene-coated substrate (GC) is about 25% smaller on average than that for the cover glass substrate (CG), (2) the initial contact angle for GC is about 80% larger than that for CG, and (3) the initial drop height for GC is 45% higher than that for CG. All of these observations are consistent

Figure 4. Both the dorsal (upper two rows) and ventral (lower two rows) dry-out patterns of 1.5 μL Al2O3 nanofluidic droplets at various volume concentrations (0.1, 1.25, and 10%) on both the graphenecoated (GC) substrate and the cover glass (CG) substrate. The scale bar is 1 mm.

Information section. Although the dorsal images present the apparent shapes of the aggregations and the pinning patterns, the ventral views with natural fringes clearly identify the formation of the cracks that grow in the direction of the maximum release of the capillary-force-driven stresses during nanoparticle crystallization. 8270

dx.doi.org/10.1021/la404854z | Langmuir 2014, 30, 8268−8275

Langmuir

Article

Figure 6. Schematic comparison of evaporative flux distributions (J), evaporating surface areas (A), perimeters (P), and contact angles (θ) of identical nanofluidic droplets wetting different substrate surfaces: (a) the hydrophobic GC surface creates the uniform local flux along the surface with θ ≈ 90° and (b) the hydrophilic CG surface creates the intensified local flux near the contact line with θ ≈ 40°.

and electrical parameters that result in the dissimilar condition of particular interaction forces. This directly influences the aggregation of particles and deposition on a substrate. Also, the substrate can be modified with respect to its hydrophobicity by using surface-cleaning and -coating processes whereas some substrates are inherently hydrophobic. For instance, a cleaned glass exposed to dichlorodimethylsilane vapor converts the original hydrophilic surface to a hydrophobic surface.27 As a result, the droplet shows significant receding with evaporation to conform to a convex dry-out shape. In contrast, our nanofluidic droplet shows little or no receding with evaporation and results in a concave shape. The present work uses a graphene substrate that atomically exhibits hydrophobicity with a different surface potential and FDLVO, which will be described in the following sections. Particle Migration Time τpart versus Droplet Evaporation Time τevap. The aggregation of nanoparticles near the contact line is described with a dimensionless number CR that is given by30 τparticle CR = τevap (1)

Figure 5. Time-evolved goniometric data of the geometric parameters of evaporating nanofluidic droplets with different particle volume fractions on both the GC (solid symbols) and CG (blank symbols) substrates: (a) wetting droplet diameter vs the evaporation time, (b) contact angle vs evaporation time, and (c) droplet height vs evaporation time. Each of the symbols represents an average obtained from three individual measurements, and the error bars indicate the 95% confidence interval of the averages.

with the fact that graphene-coated surfaces are strongly hydrophobic whereas the cover glass surface is hydrophilic.27−29 In contrast, the evaporation times for the GC substrates are about 20% longer on average compared to those for the CG substrates. Figure 6 schematically illustrates the physical distinction of the evaporation flux distributions along the drop surface for the two substrates. For the case of GC (θ ≈ 90°), the spherical drop provides a relatively smaller evaporation surface area and weaker pinning of the edge region, thus creating no thin edge region where the evaporation flux can be high. Thus, the evaporation mass flux tends to be uniform and moderate (Figure 6a). In contrast, the CG case (θ ≈ 45°) produces a relatively larger evaporation surface area and stronger pinning of the thin edge region with quite large evaporation flux distributions (Figure 6b). The contact angle of the droplet strongly depends on experimental conditions such as the size and surface potential of particles as well as the hydrophobicity of a substrate. So, two different types of particles will have different thermophysical

where the particle migration time τparticle relates the time for two adjacent nanoparticles to meet near the contact line and the droplet evaporation time τevap is defined as the time required for the droplet to reach the receding phase, not the entire evaporation time. For CR > 1, therefore, the particle migration is slow with large τparticle and/or the evaporation occurs quickly with smaller τevap, which weakens the aggregation of nanoparticles near the pinning area and creates a less-distinct coffee ring or assembly pattern on the droplet perimeter. For CR < 1, the particles migrate quickly to the contact line and/or the evaporation time is sufficiently long, which helps to form the assembly structure and creates thick and pronounced aggregated patterns. Although the original development of CR is available in detail elsewhere,30 here only the essential steps are repeated to make the present discussion self-standing. The particle migration time τparticle is the time required for two adjacent particles to assemble near the contact line, which is formulated as 8271

dx.doi.org/10.1021/la404854z | Langmuir 2014, 30, 8268−8275

Langmuir

Article

τparticle =

Lm 2 2Dp

measured as approximately 5 min for GC and 9 min for CG). The particle migration time τparticle decreases primarily with increasing nanoparticle concentration, which in turn decreases CR for both substrates. All of the experimental data points represent the average of three individual realizations using three identical droplets evaporated on different locations for both cases of the GC and CG substrates. The error bars based on the Kline−McClintock uncertainty35 and the detailed steps in estimating these uncertainties (also for the presented data shown in Figures 8 and 9) are additionally presented in the Supporting Information section.

(2)

where the mean distance between two adjacent particles is Lm = (Vd/n)1/3, with Vd being the volume of a single droplet and n being the number of nanoparticles contained in the droplet. The diffusion coefficient of particles Dp is given as its wellknown form of kBT/6πηr. The droplet evaporation time τevap is the time required for particles to form the dry-out pattern at the droplet perimeter, which is formulated as10,30 τevap =

θinitial − θreceding 16Dv (ns − n∞)

πρR i 2

(3)

where θinitial and θreceding denote the initial contact angle and the contact angle when the receding begins, respectively. For the GC substrate, θinitial = 85−90° and θreceding = 44−52°, and for the CG substrate, θinitial = 47−57° and θreceding = 6−25°. Dv is the molecular diffusion coefficient of the vapor evaporating into air, ρ is the density of the base fluid, and Ri is the initial radius of the droplet. The density differential of the saturated vapor at the liquid−vapor interface and the ambient vapor density, ns − n∞, is given by30 ⎛M P ⎞ ns − n∞ = ⎜⎜ w vs ⎟⎟(1 − RH) ⎝ R gT ⎠

(4)

where Mw is the molar mass of water, Pvs is the saturated vapor pressure of water, Rg is the ideal gas constant, T is the ambient temperature in Kelvin, and RH is the relative humidity. The extremely small CR values for both GC and CG substrates (Figure 7) support the massive aggregated layers observed near the droplet perimeter for both substrates (Figure 4). For the GC substrate, the relatively smaller initial drop radius Ri and the shorter time required to reach the receding phase (i.e., smaller τevap) provide slightly larger CR values compared to those of CG (the average receding time was

Figure 8. Derjaguin−Landau−Verwey−Overbeek forces (FDLVO) between the tested substrates (GC or CG) and Al2O3 nanoparticles for different nanoparticle concentrations. The repulsive or negative FDLVO between alumina particles and GC impedes the aggregation of nanoparticles at the pinned wetting line on the graphene-coated substrate, and the attractive or positive FDLVO for CG expedites the adhesion of particles onto the cover glass substrate. The numerical values of FDLVO and their uncertainty ranges are presented in the Supporting Information section.

DLVO Force (FDLVO). The DLVO (Derjaguin−Landau− Verwey−Overbeek) force is defined as the summation of the electrostatic (attractive or repulsive) force Felec and the van der Waals attraction force Fvdw between two components: FDLVO = Felec + Fvdw

(5)

The electrostatic force, Felec, is given as Felec = −

128πd pγsγpnkBTκ −1 2

18

⎛ Z ⎞ exp⎜ − −1 ⎟ ⎝ κ ⎠

(6)

where γs and γp are functions of the surface potential of the substrate (ψs) and of the particle (ψp):

Figure 7. Experimentally determined CR as a function of the nanoparticle volume fraction for the tested graphene (GC) and glass (CG) surfaces. The dimensionless number, CR, is defined as the ratio of the particle migration time (τparticle) to the droplet evaporation time (τevap). The exact values of CR and their uncertainty ranges are presented in the Supporting Information section.

⎛ eψ ⎞ γs = tanh⎜ s ⎟ ⎝ 4kBT ⎠

(6-a)

⎛ eψp ⎞ γp = tanh⎜ ⎟ ⎝ 4kBT ⎠

(6-b)

The surface potentials, ψs, are given as approximately −20 and +300 mV for CG and GC substrates, respectively.18,31,32 In the ambient environment, the surface potential of the graphene film is always positive as a result of hole doping.1 The surface potential of Al2O3 nanoparticles, ψp, is estimated from the zeta 8272

dx.doi.org/10.1021/la404854z | Langmuir 2014, 30, 8268−8275

Langmuir

Article

which ranges from 0.1 to 0.5 and is inversely proportional to the distance between the particle and the substrate. Figure 8 shows FDLVO as functions of nanoparticle concentrations calculated for both GC and CG substrates. The salient feature of FDLVO is its opposite signs for the two substrates. That is, the plus sign for CG means an attractive force between the Al2O3 nanoparticles and the cover glass, and the minus sign for GC indicates the repulsive force of the alumina particles against the graphene layer. At the low particle concentrations of 0.1 and 1.25%, FDLVO for CG is attractive in that the aggregation of nanoparticles is well pinned near the droplet periphery, as previously observed, and FDLVO gradually decreases with increasing concentration. In contrast, FDLVO for GC is repulsive in that the aggregated nanoparticles are not well pinned, which occurs earlier as evidenced by disfigured noncircular patterns for the 1.25 and 0.1% concentrations shown in Figure 4. With increasing nanoparticle concentration, FDLVO for both substrates approaches zero, and the distinction between the attraction and repulsion is diminished. This implies that both substrates can develop qualitatively similar dry-out patterns that remain circular with negligible receding of the edge area as shown for the case of 10% concentration in Figure 4. The surface potential of graphene increases with the number of graphene layers, and the repulsive force between the nanoparticles and the surface of multiple graphene layers will also increase. Also, the receding process for multiple layers can be more pronounced than for the case of monolayered graphene. Furthermore, it is known that the hydrophobicity and contact angle increase somewhat with the number of graphene layers.5 Therefore, the resulting particle assembly is anticipated to depend on the number of layers to an extent. Relative Thermal Conductivity (KR). The thermal conductivity differentials of the substrate and the nanofluid are expected to contribute by establishing Marangoni flows inside, ultimately affecting the dry-out patterns. The relative thermal conductivity is defined herein as

Figure 9. Relative thermal conductivity KR of the substrate (GC or CG) to that of the nanofluid. The Marangoni flows and particle migrations driven by the temperature gradients are schematically illustrated for the two substrates. GC drives the strong circulation of particles and demotes evaporation, whereas CG promotes the migration of particles to the pinned area and promotes evaporation. The numerical values of KR and their uncertainty ranges are presented in the Supporting Information section.

potential measurement data using the Nano-ZS instrument based on laser Doppler electrophoresis (Malvern Inc.). The surface potential of nanoparticles is estimated to be +53 mV. The number density of counterions n ranges from 5.35 × 1021 to 2.87 × 1022 ions/m−3 for 10 to 0.1% nanoparticle volume fractions. The distance between particles z is calculated from the particle concentration (i.e., z10% = 33.1 nm, z1.25% = 111.3 nm, z0.1% = 317.7 nm), and the Debye length κ −1 is given as approximately 1 μm in pure distilled deionized water (pH 7). Our samples are prepared by the dispersion of Al2O3 powder without any surfactant added, having very few electrostatic carriers. The Debye length in pure water at room temperature (25 °C) is given by κ−1(nm) = 0.334/(I(M))1/2,33 where I is the ionic strength expressed in molar concentration M (mol/ L). Because there is a 10−7 M concentration of both H+ and OH− ions in pure water (pH 7), the Debye length in this case is estimated to be 0.3/(10−7)1/2 nm ≈ 1 μm. The van der Waals attraction force, Fvdw, between the substrate and a nanoparticle of diameter dp is given by18 α 1 Fvdw = Ad p3 2 rtd 2 12 z (z + d p) (7)

KR =

(9)

where subscripts S and L denote the substrate and liquid, respectively. When KR > 1.5, the Marangoni convective flows are driven along the drop surface from the pinning edge to the top,34 and for KR < 1.5, the Marangoni flow direction will be reversed. The thermal conductivities of the substrate, kS, are approximately 1.05 ± 0.05 and 370 + 650/−320 W/m·K for the CG and GC substrates, respectively,4 and the thermal conductivities of the nanofluids, kL, are 0.58 ± 0.03, 0.61 ± 0.03, and 0.66 ± 0.03 W/m·K for 0.1, 1.25, and 10% volume fractions, respectively.17 Figure 9 shows that KR for CG is on the order of unity (KR = 1.81, 1.72, and 1.59, respectively, for the three concentrations), and KR for GC is more than 2 orders of magnitude larger at 638, 607, and 561 because of the graphene’s extremely high thermal conductivity of 370 W/m·K in the cross-plane direction.4 Because KR for CG is the same order of magnitude as the critical value of 1.5, the Marangoni-driven flow for the glass substrate will be minimal, as schematically illustrated in Figure 9. The dashed contours show isotherms that were calculated using a simple heat conduction equation based on the first approximation, assuming a negligible effect of the internal convective flows on the overall thermal transport.

where A is the Hamaker constant. For the Al2O3−water−glass interaction, A ≈ 0.77 × 10−20 J, and for the Al2O3−water− graphite interaction, A ≈ 8.78 × 10−20 J. Furthermore, the Hamaker constant for the present case of Al2O3−water− graphene-coated glass can be determined from the following equation:5 Fvdw,graphene + glass = Fvdw,graphite(z = h) − Fvdw,graphite(z = h + d) + Fvdw,glass(z = h + d)

kS kL

(8)

Here, h is the distance between the nanoparticle and graphene layer, and d is the thickness of a single graphene layer (0.335 nm). αrtd is the retardation constant for the van der Waals force, 8273

dx.doi.org/10.1021/la404854z | Langmuir 2014, 30, 8268−8275

Langmuir As discussed earlier in both sections 3.1 and 3.2, the intense evaporation near the pinned edge for the CG substrate tends to lower the local temperature, and the Marangoni flow is driven from the top of the droplet to the contact line along the air/ liquid interface. The much larger KR for GC drives strong Marangoni flows toward the droplet top from the pinning edge along the interface. Furthermore, the graphene-coated (GC) substrate generates the highest temperature point near the droplet edge, and this also helps to drive the Marangoni flow direction away from the edge. To summarize, the evaporation-driven flow toward the pinned edge is more pronounced for the CG substrate, whereas the Marangoni-driven flow circulation is much more substantial for the GC substrate. Consequently, no circulating flow will be developed for CG and the nanoparticles are migrated readily to the contact line. This allows strong pinning in the early stage of evaporation and very little receding for the CG substrate. However, the onset of strong Marangoni circulation flows for the GC substrate tends to drive the nanoparticles away from the droplet edge area. This can lower the pinning strength of the nanoparticles and slow down the evaporative aggregation at the contact line. The Marangoni circulating flow for the case of the GC substrate can drive an earlier onset of the receding of the contact line than for the CG substrate.

■

REFERENCES

(1) Novoselov, K. S.; Geim, A. K.; Morozov, S. V.; Jiang, D.; Zhang, Y.; Dubonos, S. V.; Grigorieva, I. V.; Firsov, A. A. Electric Field Effect in Atomically Thin Carbon Films. Science 2004, 306, 666−669. (2) Balandin, A. A.; Ghosh, S.; Bao, W.; Calizo, I.; Teweldebrhan, D.; Miao, F.; Lau, C. N. Superior Thermal Conductivity of Single-Layer Graphene. Nano Lett. 2008, 8, 902−907. (3) Berciaud, S.; Han, M. Y.; Mak, K. F.; Brus, L. E.; Kim, P.; Heinz, T. F. Electron and Optical Phonon Temperatures in Electrically Biased Graphene. Phys. Rev. Lett. 2010, 104, 227401. (4) Cai, W.; Moore, A. L.; Zhu, Y.; Li, X.; Chen, S.; Shi, L.; Ruoff, R. S. Thermal Transport in Suspended and Supported Monolayer Graphene Grown by Chemical Vapor Deposition. Nano Lett. 2010, 10, 1645−1651. (5) Rafiee, J.; Mi, X.; Gullapalli, H.; Thomas, A. V.; Yavari, F.; Shi, Y.; Ajayan, P. M.; Koratkar, N. A. Wetting Transparency of Graphene. Nat. Mater. 2012, 11, 217−222. (6) Prevo, B. G.; Velev, O. D. Controlled, Rapid Deposition of Structured Coatings from Micro- and Nanoparticle Suspensions. Langmuir 2004, 20, 2099−2107. (7) Park, J.; Moon, J. Control of Colloidal Particle Deposit Patterns within Picoliter Droplets Ejected by Ink-Jet Printing. Langmuir 2006, 22, 3506−3513. (8) Deegan, R. D.; Bakajin, O.; Dupont, T. F.; Huber, G.; Nagel, S. R.; Witten, T. A. Capillary Flow As the Cause of Ring Stains from Dried Liquid Drops. Nature 1997, 389, 827−829. (9) Deegan, R. D.; Bakajin, O.; Dupont, T. F.; Huber, G.; Nagel, S. R.; Witten, T. A. Contact Line Deposits in an Evaporating Drop. Phys. Rev. E 2000, 62, 756−765. (10) Popov, Y. O. Evaporative deposition patterns: spatial dimensions of the deposit. Phys. Rev. E 2005, 71, 036313. (11) Hu, H.; Larson, R. G. Analysis of the effects of marangoni stresses on the microflow in an evaporating sessile droplet. Langmuir 2005, 21, 3972−3980. (12) Nikolov, A. D.; Wasan, D. T. Mechanisms of the assembly of nano- and microparticle two-dimensional structures in a wedge film. Ind. Eng. Chem. Res. 2008, 48 (), 2320−2326. (13) Jung, J. Y.; Kim, Y. W.; Yoo, J. Y.; Koo, J.; Kang, Y. T. Forces acting on a single particle in an evaporating sessile droplet on a hydrophilic surface. Anal. Chem. 2010, 82, 784−788. (14) Nguyen, T. A. H.; Nguyen, A. V.; Hampton, M. A.; Xu, Z. P.; Huang, L.; Rudolph, V. Theoretical and experimental analysis of droplet evaporation on solid surfaces. Chem. Eng. Sci. 2012, 69 (), 522−529. (15) Vafaei, S.; Wen, D.; Borca-Tasciuc, T. Nanofluid surface wettability through asymptotic contact angle. Langmuir 2011, 27, 2211−2218. (16) Kim, I.; Kihm, K. D. Unveiling hidden complex cavities formed during nanocrystalline self-assembly. Langmuir 2009, 25, 1881−1884. (17) Kim, I.; Kihm, K. D. Hidden cavity formations by nanocrystalline self-assembly on various substrates with different hydrophobicities. Langmuir 2012, 28, 9195−9200. (18) Bhardwaj, R.; Fang, X.; Somasundaran, P.; Attinger, D. Selfassembly of colloidal particles from evaporating droplets: role of DLVO interactions and proposition of a phase diagram. Langmuir 2010, 26, 7833−7842. (19) Adachi, E.; Dimitrov, A. S.; Nagayama, K. Stripe patterns formed on a glass surface during droplet evaporation. Langmuir 1995, 11 (), 1057−1060.

ASSOCIATED CONTENT

S Supporting Information *

Time-dependent ventral microscopy images of 1.5 μL Al2O3 nanofluidic droplets under evaporation. Numerically tabulated values of CR, FDLVO, and KR corresponding to Figures 7−9, respectively. This material is available free of charge via the Internet at http://pubs.acs.org.

■

ACKNOWLEDGMENTS

This research is partially supported by the World Class University (WCU) Program (R31-2008-000-10083-0) and partially by the Nano-Material Technology Development Program (R2011-003-2009), both through the National Research Foundation (NRF) funded by the Korean Ministry of Education, Science and Technology.

4. CONCLUSIONS Nanofluidic wetting and evaporative aggregation on a hydrophobic graphene-coated (GC) substrate are comparatively discussed along with those of a hydrophilic cover glass (CG) substrate. A few fundamental findings from the experimental investigations are the following: (1) The ratio of the particle migration time to the evaporation time (CR) levels are extremely low for both GC and CG substrates, which makes both of them available for massive coffee-ring patterns. (2) The repulsive DLVO force of the GC substrate interrupts the pinning of nanoparticles, and the attractive DLVO force of the CG substrate enhances the pinning. (3) Extremely high thermal conductivity of the GC substrate with respect to that of the nanofluid (KR) induces Marangoni flow from the droplet edge to the top that weakens the pinning strength of nanoparticles on the graphene-coated substrate. (4) Points 2 and 3 support the receding of the contact line for GC substrates, resulting in noncircular dry-out patterns, which is more pronounced for low particle concentrations. With increasing particle concentration, the GC substrate recovers to a circular aggregation pattern with minimum receding of the edge area.

■

■

Article

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. Notes

The authors declare no competing financial interest. 8274

dx.doi.org/10.1021/la404854z | Langmuir 2014, 30, 8268−8275

Langmuir

Article

(20) Zhai, L.; Cebeci, F. Ç .; Cohen, R. E.; Rubner, M. F. Stable superhydrophobic coatings from polyelectrolyte multilayers. Nano Lett. 2004, 4 (), 1349−1353. (21) Shirtcliffe, N. J.; McHale, G.; Newton, M. I.; Perry, C. C. Wetting and wetting transitions on copper-based super-hydrophobic surfaces. Langmuir 2005, 21, 937−943. (22) Li, X.; Cai, W.; An, J.; Kim, S.; Nah, J.; Yang, D.; Piner, R.; Velamakanni, A.; Jung, I.; Tutuc, E.; Banerjee, S. K.; Colombo, L.; Ruoff, R. S. Large-area synthesis of high-quality and uniform graphene films on copper foils. Science 2009, 324, 1312−1314. (23) Bae, S.; Kim, H.; Lee, Y.; Xu, X.; Park, J. S.; Zheng, Y.; Balakrishnan, J.; Lei, T.; Ri Kim, H.; Song, Y. I.; Kim, Y. J.; Kim, K. S.; Ö zyilmaz, B.; Ahn, J. H.; Hong, B. H.; Iijima, S. Roll-to-roll production of 30-in. graphene films for transparent electrodes. Nat. Nanotechnol. 2010, 5, 574−578. (24) Nair, R. R.; Blake, P.; Grigorenko, A. N.; Novoselov, K. S.; Booth, T. J.; Stauber, T.; Peres, N. M. R.; Geim, A. K. Fine structure constant defines visual transparency of graphene. Science 2008, 320, 1308. (25) Xu, P.; Mujumdar, A. S.; Yu, B. Drying-induced cracks in thin film fabricated from colloidal dispersions. Drying Technol. 2009, 27, 636−652. (26) Carle, F.; Brutin, D. How surface functional groups influence fracturation in nanofluid droplet dry-outs. Langmuir 2013, 29, 9962− 9966. (27) Kuncicky, D. M.; Velev, O. D. Surface-guided templating of particle assemblies inside drying sessile droplets. Langmuir 2008, 24, 1371−1380. (28) Shin, D. H.; Lee, S. H.; Jung, J. Y.; Yoo, J. Y. Evaporating characteristics of sessile droplet on hydrophobic and hydrophilic surfaces. Microelectron. Eng. 2009, 86, 1350−1353. (29) Shao, F. F.; Neild, A.; Ng, T. W. Hydrophobicity effect in the self assembly of particles in an evaporating droplet. J. Appl. Phys. 2010, 108, 034512. (30) Shen, X.; Ho, C. M.; Wong, T. S. Minimal size of coffee ring structure. J. Phys. Chem. B 2010, 114, 5269−5274. (31) Datta, S. S.; Strachan, D. R.; Mele, E. J.; Johnson, A. T. C. Surface potentials and layer charge distributions in few-layer graphene films. Nano Lett. 2009, 9, 7−11. (32) Hao, G. L.; Qi, X.; Li, J.; Yang, L. W.; Yin, J. J.; Lu, F.; Zhong, J. X. Surface potentials of few-layer graphene films in high vacuum and ambient conditions. Solid State Commun. 2011, 151, 818−821. (33) Israelachvili, J. Intermolecular and Surface Forces; Academic Press: Orlando, FL, 1985. (34) Ristenpart, W. D.; Kim, P. G.; Domingues, C.; Wan, J.; Stone, H. A. Influence of substrate conductivity on circulation reversal in evaporating drops. Phys. Rev. Lett. 2007, 99, 234502. (35) Kline, S. J.; McClintock, F. A. Describing uncertainties in singlesample experiments. Mech Eng. 1953, 75, 3−8. (36) Mintsa, H. A.; Roy, G.; Nguyen, C. T.; Doucet, D. New temperature dependent thermal conductivity data for water-based nanofluids. Int. J. Therm. Sci. 2009, 48, 363−371.

8275

dx.doi.org/10.1021/la404854z | Langmuir 2014, 30, 8268−8275