Ratchetlike Slip Angle Anisotropy on Printed Superhydrophobic

Jun 23, 2011 - Mark Barahman and Alan M. Lyons*. Department of Chemistry, College of Staten Island, City University of New York, 2800 Victory Boulevar...
0 downloads 0 Views 3MB Size
ARTICLE pubs.acs.org/Langmuir

Ratchetlike Slip Angle Anisotropy on Printed Superhydrophobic Surfaces Mark Barahman and Alan M. Lyons* Department of Chemistry, College of Staten Island, City University of New York, 2800 Victory Boulevard, Staten Island, New York 10314, United States

bS Supporting Information ABSTRACT: The fabrication and properties of superhydrophobic surfaces that exhibit ratchet-like anisotropic slip angle behavior is described. The surface is composed of arrays of poly(dimethylsiloxane) (PDMS) posts fabricated by a type of 3D printing. By controlling the dispense parameters, regular arrays of asymmetric posts were deposited such that the slope of the posts was varied from 0 to 50 relative to the surface normal. Advancing and receding contact angles as well as slip angles were measured as a function of the post slope and droplet volume. Ratchetlike slip angle anisotropy was observed on surfaces composed of sloped features. The maximum slip angle difference (for a 180° tilt angle variation) was 32° for 20 μL droplets on surfaces with posts fabricated with a slope of 50°. This slip angle anisotropy is attributed to an increase in the triple contact line (TCL) length as the droplet is tilted in a direction against the post slope whereas the TCL decreases continuously when the drop travels in a direction parallel to the post slope. The increasing length of the TCL creates an increased energy barrier that accounts for the higher slip angles in this direction.

I. INTRODUCTION The manipulation of individual droplets is of great interest in microfluidics research because it facilitates the ability to isolate and quantitatively study chemical reactions in small volumes. A wide variety of reactions in microdroplets have been recently reviewed.1 The impact of conducting the polymerase chain reaction (PCR) in droplets has received particular attention as a means of diagnosing disease from small sample quantities1,2 and conducting high-throughput genetic analyses.3 These microfluidic applications are conducted within microchannels so that the flow path can be controlled. Droplets are isolated from the walls of the microchannel as well as from each other by an oil phase in which the droplets are immiscible. In some cases, separation of the droplets from the oil and/or the partial solubility of droplet compounds in the oil limits the applicability of microchannel-based microfluidics. Superhydrophobic surfaces offer a valuable alternative to microchannel-based microfluidic devices because individual droplets can be translated across a surface without an oil phase to isolate them. The ability to manipulate fluids on superhydrophobic surfaces has led to some initial applications including the analysis of complex biological fluids4 and reserve microbatteries.5 The development of additional manipulation techniques would enhance the usefulness of these surfaces. In particular, a ratchetlike surface, where a droplet could be easily translated in one direction while limiting its ability to move backward along the same path would be useful for many applications including targeted liquid cooling.6 This ratchetlike behavior is an inherent element of pumped microfluidic r 2011 American Chemical Society

systems but has not been accessible on synthetic superhydrophobic surfaces. Some naturally occurring surfaces, such as the wing of the Morpho aega butterfly and rice leaves, exhibit ratchetlike droplet motion. In contrast, the slip angle of water droplets on a lotus leaf is uniform in all directions.7 Zheng et al.8 showed slip-angle anisotropy on the surface of the Morpho aega wing such that a water droplet could move easily away from the insect’s body (slip angle = 9°); however, when the wing was tilted in the opposite direction, the droplet was pinned on the surface and was thus prevented from rolling toward the insect. Nanostructures on the wing are hypothesized to switch between two different states, depending upon the position of the wing, and account for the anisotropic slip angle. When the wing is down, the nanostructures rise above the surface and thus reduce the solidliquid contact area, forming a discontinuous triple contact line (TCL) that results in less drag. When the wing is moved up, the nanostructures lie along the wing surface and thus present a relatively smooth surface along which the droplet makes nearly continuous contact resulting in pinning. Leaves of many other plants display superhydrophobic behavior.9 On rice leaves, a ratchetlike phenomenon similar to that observed on the Morpho aega has been reported7,10,11 and attributed to the arrangement of micropapillae on the surface. In some plants, it has been shown that the leaf directs water Received: April 3, 2011 Revised: June 19, 2011 Published: June 23, 2011 9902

dx.doi.org/10.1021/la201222a | Langmuir 2011, 27, 9902–9909

Langmuir droplets preferentially toward the roots of the plant, thus conserving water.12 Trichomes on the leaf surface are arranged with an orientation and pitch such that water droplets, either from rain or condensation, are preferentially directed toward one end of the leaf. Some forms of anisotropic fluid behavior have been reported on synthetic superhydrophobic surfaces. Long et al.13 fabricated patterned topographies comprising elongated microfeatures ranging from continuous parallel ridges to discontinuous ridges of varying lengths. Significant slip-angle anisotropy was observed between the directions parallel and perpendicular to the topography; however, ratchetlike behavior was not reported because the surfaces are symmetrical with respect to 180° rotation. The anisotropy with respect to tilt angles oriented 90° to each other (i.e., aligned parallel vs perpendicular to the topography) was attributed to higher energy barriers when the droplet spans the larger continuous gaps (and thus longer TCL) between parallel features during translation. Surfaces fabricated with asymmetric features have been reported,14 and it was found that droplet spreading occurred only in one direction. This preferential droplet spreading was found to increase in effectiveness with increasing feature slope. In all cases, the droplet spread in the direction that enabled greater liquidsolid contact to occur. Sawtoothlike ratchets have been fabricated by Zhang et al.,15 where the surfaces features were coated with nanoparticles to render them superhydrophobic. Anisotropic, ratchetlike droplet velocity was achieved on these surfaces by applying an external magnetic field that attracted water droplets in which superparamagnetic particles had been dispersed. The droplets were found to move more rapidly in a direction aligned against the slope of the sawtooth and more slowly when the droplet was moved in a direction aligned parallel to the sawtooth slope. The anisotropic behavior of droplets on synthetic surfaces has also been studied theoretically. Kusumaatmaja and Yeomans16 considered the behavior of a droplet placed between hypothetical surfaces patterned with asymmetric sawtooth ridges. They predicted that droplets would move more rapidly in the direction aligned along the slope of the sawtooth ridge and relatively more slowly when the surface was tilted in the opposite direction such that the droplet moved against the slope of the ridges. This type of droplet motion was predicted to occur only when the droplet was able to penetrate the top of the ridge partially and wet the slope of the sawtooth. Increased contact area between the fluid and the sloped surface was predicted to occur along the receding contact line, resulting in increased pinning and lower droplet velocities. In this article, we describe a process for fabricating superhydrophobic surfaces that exhibit ratchetlike behavior. The surface is composed of arrays of poly(dimethylsiloxane) (PDMS) posts fabricated by a type of 3D printing. By controlling the dispense parameters, asymmetric posts could be fabricated such that the slope of the posts was varied from 0 to 50° relative to the surface normal. Advancing and receding contact angles as well as slip angles were measured as a function of post asymmetry (i.e., slope) and droplet volume. The magnitude of this ratchetlike slip angle anisotropy was observed to be a function of the post slope. A mechanism to explain this phenomenon is presented.

II. EXPERIMENTAL SECTION A room-temperature vulcanizing (RTV) hydroxy-terminated dimethyl siloxane, manufactured by Dow Corning and sold as DAP clear

ARTICLE

Figure 1. Illustration of key parameters used in the dispensing robot system including (a) the height above the substrate where dispensing is initiated (Hd); (b) the height above the surface to which the dispenser is moved after the syringe is triggered (Hs) (note that the silicone post maintains contact with the dispensing tip); and (c) the dispensing tip translated at a specific angle. This movement causes the dispensed drop to form a sloped post as it disengaged from the tip.

Table 1. Properties of PDMS Surface Samples Prepared for CA and CAH and Slip Angle Measurementsa surface

slope of posts

(material-slope-array)

(deg)

(μm)

88

800 ( 10

DAP-21-20 DAP-29-5

21 ( 1 29 ( 1

20  10 55

870 ( 10 860 ( 10

DAP-28-20

28 ( 1

20  20

800 ( 10

DAP-35-5

35 ( 1

55

810 ( 10

DAP-35-20

35 ( 1

20  20

750 ( 50

DAP-44-20

44 ( 5

20  10

850 ( 60

DAP-50-10

50 ( 1

10  10

730 ( 10

DAP-00-8

a

feature height size of array

0

The post pitch is 500 μm in both the x and y directions.

aquarium silicone sealant 688, was used to fabricate PDMS superhydrophobic surfaces. The material is very viscous (>100 000 cP) as well as thixotropic because it contains ∼10 wt % amorphous silica. Such rheological properties enable the material to flow well under the shear induced during dispensing. When the shear is relieved after dispensing, the high apparent viscosity returns and so the dispensed shape is retained during curing. After being dispensed, the material was cured at room temperature for at least 24 h before contact angle measurements were made. The advancing and receding contact angles (θA and θR) and slip angles for a thin, flat film of cured DAP 688 are 115, 85, and 43 ( 3°, respectively. II.1. Robotic Dispenser. A robot (Janome-2203N) with (10 μm repeatability was used as an XYZ platform and trigger for a syringe controller (EFD Performus V). The silicone materials were carefully loaded into 10 cm3 polyethylene syringe barrels, avoiding air bubble entrapment, and capped with a 27 gauge (210 μm i.d.) tapered dispensing tip. The syringe was mounted on the robotic arm, and substrates (cleaned glass microscope slides) were placed on top of the robot’s vacuum stage. Arrays of silicone posts were formed by repeating the following dispensing cycle program: • position syringe tip at a predetermined dispensing height (Hd) above the substrate (Figure 1a); • trigger the syringe controller to dispense the silicone at a preset time and pressure; • lift the syringe normal to the surface: o for straight posts, the silicone will break away from the dispensing tip during this vertical translation; o for sloped posts, after a specified height (termed the slope height, Hs, in Figure 1b) is reached, the vertical motion is continued at a specified angle relative to the surface (Figure 1c); 9903

dx.doi.org/10.1021/la201222a |Langmuir 2011, 27, 9902–9909

Langmuir

ARTICLE

Figure 2. Optical microscope images of the DAP 688 PDMS posts described in Table 2. (a) DAP-00-8, (b) DAP-21-20, (c) DAP-28-20, (d) DAP-35-20, (e) DAP-44-20, and (f) DAP-50-10. the silicone will break away from the dispensing tip during this angled trajectory of the syringe; • translate the syringe to the next dispensing location and height (Hd). Dispensing with DAP silicone resulted in a flat-bottomed teardrop (or Hershey’s kiss-shaped) post shape. The diameter of the post decreases rapidly when moving from the base to the top along the right circular cone. In addition, the rheology of the material contributes to the formation of relatively small loops at the tops of some posts. II.2. Contact Angle and Slip Angle Measurements. A RameHart model 250 goniometer was used to measure contact angles and slip angles. On flat surfaces, the DROPimage Advanced software provided with the goniometer was used. A droplet of known volume of deionized (18 MΩ) water was carefully placed on top of the surface. On superhydrophobic surfaces, advancing and receding contact angles were measured by tilting the goniometer automated stage at a speed of 1°/s while recording a video of the drop. Still images of the drop were extracted from the video at a point immediately before the drop began to slip. These images were imported in Image-Pro software where the contact angles were measured using the manual angle measurement tool. The reproducibility of these measurements was (2°. For slip angle measurements, the automated stage on the goniometer was then tilted at a rate of 1°/s either to the right or to the left. For flat surfaces, the slip angle was defined as the angle when the trailing edge of the droplet began to move along the surface. On superhydrophobic surfaces, however, the slip angle was defined as the angle required to displace the droplet by two posts (1 mm) on both the advancing and receding edges of the drop. At least 10 measurements were made per direction for every droplet size and for each sample. The surface was dried after each measurement by blowing high-pressure, dry, filtered air over it for approximately 5 s.

III. RESULTS A series of surfaces were fabricated where the post slope was varied from 0° (straight posts) to 50°. The sample designation, post slope, size of the array, and post height are summarized in Table 2. Sample designations are constructed from hyphenation

of the material, slope, and array size (number of posts in the slip angle test direction) parameters. All posts were printed on 500 μm pitch square arrays in the plane of the glass slide. The base diameter of the posts was approximately 500 μm for all samples, tapering to a point of less than 25 μm in diameter. An analysis of microscope images showed that the height and slope variation for most samples was small ((1°). Sample DAP-44-20 showed the most variability with respect to the slope ((5°) and height (60 μm). Images of posts from selected surfaces described in Table 1 are shown in Figure 2. The size and separation of the posts are limited primarily by the polymer rheology and dispensing tip dimensions. The PDMS used is very viscous and thixotropic. Thus, a significant pressure drop develops as the polymer flows through the dispensing tip. A smaller-diameter tip would reduce the post diameter and enable us to print posts at finer pitches. However, the pressure drop across standard steel capillary tips was larger than the pressure that could be safely applied to the syringe (90 psi). Tapered plastic tips reduce the pressure drop but are available in limited sizes. The 210 μm tip used for these experiments is the smallest available. Smaller columnar features (e.g., 150 μm o.d.  1000 μm height) are possible when a low-viscosity polymer precursor is dispensed with steel capillary tips measuring 100 μm i.d.,17 and features as small as 75 μm diameter have been demonstrated. III.1. Surfaces Composed of Straight Posts. For surfaces fabricated with straight posts, droplets of water placed on these surfaces exhibited superhydrophobic properties (Figure 3). Static contact angle values ranged from 154 to 170° depending upon the droplet volume (additional information in SF3). Compared to most studies of superhydrophobic surfaces, the post pitch of these surfaces is comparable to the diameter of the droplets. Because of the relatively coarse pitch, the droplets partially penetrate into the surface and could be characterized as existing in a modified Cassie state. The measured advancing and receding CAs ranged from 157 to 161° and from 112 to 115°, respectively (Table 2). These CAs are considerably larger than the 115/85° θA/θR values measured on a flat DAP 688 surface. The slip angle 9904

dx.doi.org/10.1021/la201222a |Langmuir 2011, 27, 9902–9909

Langmuir

ARTICLE

Figure 3. A 20 μL droplet of water on a DAP-00 surface. Partial wetting of the PDMS posts can be seen.

Table 2. Contact Angle Measurements (θA, θR, CAH) Measured with a 20 μL Droplet as a Function of the Tilt Direction for PDMS Surfaces with Straight and Sloped Posts contact angle surface DAP-00-8

CAH

contact angles

(with slope) (with slope) (against slope) adv = 157 ( 3

42 ( 2

rec = 115 ( 2 DAP-21-20 adv = 158 ( 2

41 ( 2 34 ( 2

DAP-35-20 adv = 153 ( 2

34 ( 2

rec = 119 ( 2

78 ( 2

adv = 150 ( 2 rec = 90 ( 2

60 ( 2

adv = 151 ( 2

61 ( 2

rec = 90 ( 2 36 ( 2

rec = 117 ( 2 rec = 118 ( 2

adv = 160 ( 2 rec = 82 ( 2

DAP-28-20 adv = 151 ( 2 rec = 117 ( 2

DAP-50-10 adv = 154 ( 2

49 ( 2

rec = 112 ( 2

rec = 117 ( 2

DAP-44-20 adv = 153 ( 2

adv = 161 ( 2

CAH (against slope)

Adv =161 ( 2

65 ( 2

rec = 96 ( 2 36 ( 2

adv = 128 ( 2

60 ( 2

rec = 68 ( 2

on this straight post surface (Table 3) is 21 ( 2°, which is half the angle measured on flat surfaces of the same material (43 ( 3°). The high CAs and the significantly reduced slip angles are consistent with superhydrophobicity. III.2. Surfaces Composed of Sloped Posts. Advancing and receding contact angles were measured on surfaces composed of sloped PDMS posts as shown in Table 2. CA values were measured in two directions: aligned with the slope (as defined in SF1) and against the slope (180° relative to the “with slope” direction). When the θadv/θrec angles were measured in the with slope direction, the values of both θA and θR were similar for all samples. For samples with post slopes ranging from 21 to 50°, values of θadv = 154 ( 3° and θrec = 118 ( 1° are comparable to those measured for straight posts (θadv = 157 ( 3° and θrec= 115 ( 2°). The contact angle hysteresis (CAH) in the “with slope” direction were similar for all samples, but consistently lower for samples with post slopes between 28 and 50°. The values of CAH (35 ( 1°) were 7° lower than for straight posts because of a combination of lower advancing CAs and higher receding CAs, indicating a somewhat smaller droplet surface interaction. When measured in the opposite direction, however, the CAH values were strongly impacted by the sample geometry. For the four samples prepared with slopes of between 28 and 50°, the

CAH values increased to 62 ( 2° primarily because of a significant decrease in θR. No significant decrease in θA was observed until the post slope increased to 50°. The sample with a post slope of 21° shows transitional behavior with CAH in the with slope direction being similar to the value measured on straight posts, whereas the CAH measured in the opposite direction shows the largest CAH increase because of a significant decrease in θR. Additionally, static contact angles were measured parallel and perpendicular to the slopes in order to probe for droplet elongation. Slight anisotropy was found (5°). This is shown in SF5. Values of the slip angle as a function of the post geometry and measurement direction are shown in Table 3 for 20 μL droplets. When the substrate was tilted in the direction aligned with the post slope, the slip angle was found to decrease slightly compared to straight posts (18 vs 21°). However, when the substrate was tilted in the opposite direction, the slip angle increased significantly and was found to be a function of the post slope. As the receding contact angle decreased, the slip angle in the direction opposite to the post slope increased significantly. For example, sample DAP-50-10 has the lowest θR (68 ( 2°) and the highest slip angle (50 ( 6°) and slip angle anisotropy (32 ( 4°). These observations are consistent with those of Long.13 Slip angle anisotropy, shown in Table 3, is observed over a wide range of drop volumes for all surfaces fabricated with sloped posts. Figure 4 shows the slip angle anisotropy as a function of the droplet volume for two such surfaces. (See the Supporting Information for the remaining three samples.) The absolute value of the slip angle generally decreases as the droplet volume increases, consistent with the increased force acting on the additional mass of the droplet. The measurements also became more reproducible as the droplet size increased. When droplets were induced to transition to the Wenzel state, no slip was observed even at a tilt angle of 90° because the hydrophilic nature of the glass substrate pins the droplets to the surface. The effect of post slope on the slip angle anisotropy is plotted in Figure 5. Increasing the post slope results in increased slip angle anisotropy. A relation originally developed by Furmidge18 and modified by Extrand19 was used to predict slip angles from CA measurements. Extrand modified the original relationship by adding the prefactor 2/π to account for the elongated shape of the droplet on a tilted surface compared to an ideal hemisphere. This can be written as sin R ¼

2w γ ðcos θR  cos θA Þ πmg LV

ð1Þ

where γ, m, w, and g are the liquidvapor surface tension, mass of the droplet, width of the droplet, and acceleration due to gravity, respectively. θA and θR are the apparent advancing and receding contact angles, respectively, and R is the critical angle of inclination or slip angle. Predicted values from the Furmidge relation (eq 1) are shown in Table 3 and plotted in Figure 5. As can be seen, there is good agreement between the measured slip angles and predicted values based on advancing and receding contact angles.

IV. DISCUSSION A droplet resting on a superhydrophobic surface will begin to move once the surface is tilted at a sufficiently steep angle such that the gravitational forces overcome the liquidsolid surface 9905

dx.doi.org/10.1021/la201222a |Langmuir 2011, 27, 9902–9909

Langmuir

ARTICLE

Table 3. Slip Angles Measured with a 20 μL Water Droplet as a Function of the Tilt Direction for PDMS Surfaces with Straight and Sloped Posts surface

slip angle (measured)

slip angle (predicted)

DAP-00-8

with = 22 ( 2

with = 20

DAP-21-20

against = 21 ( 1 with = 18 ( 1

against = 23 with = 18

against = 34 ( 4

against = 38

DAP-28-20 DAP-35-20

with = 18 ( 1

with = 17

against = 38 ( 1

against = 37

with = 18 ( 1

with = 17

against = 41 ( 1

against = 34

slip angle anisotropy (measured)

slip angle anisotropy (predicted)

1 ( 1

3

16 ( 3

20

20 ( 1

20

23 ( 1

17

DAP-44-20

with = 18 ( 4

with = 17

19 ( 4

21

DAP-50-10

against = 37 ( 5 with = 18 ( 2

against = 38 with = 18

32 ( 4

26

against = 50 ( 6

against = 44

Figure 4. Effect of tilt direction and droplet volume on the slip angle for two different PDMS surfaces: (a) 820 μL droplets on surface 28-20 and (b) 1030 μL droplets on surface DAP 21-10.

Figure 5. Effect of post slope and tilt direction on slip angle measured with (a) 20 and (b) 15 μL droplets. The legend shown in part a applies to both plots.

interactions that would otherwise pin the droplet to the surface. The angle at which the surface must be tilted for motion to occur (slip angle) is dependent upon the magnitude of the liquidsolid chemical interactions summed over the total length of the solidliquid interface or triple contact line (TCL). Thus, the

barrier to slip will depend upon both the chemical composition and the geometry of the solid surface at the receding TCL of the droplet.2025 By controlling the local surface topography of superoleophobic surfaces, a modified CassieBaxter relationship was developed26 that includes a local texture parameter that 9906

dx.doi.org/10.1021/la201222a |Langmuir 2011, 27, 9902–9909

Langmuir

ARTICLE

Figure 6. Photographs extracted from slip angle experiments of (a) straight posts, (b) sloped posts as the droplet moves along the slope of the posts (easy direction), and (c) sloped posts as the droplet moves away from the slope of the posts (hard direction).

considers differentially small displacements of the TCL. The approach led to accurate predictions of θadv, θrec, CAH, and the slip angle on droplets in the Cassie state. Thus, the differential areal fraction of the solid substrate in contact with the TCL, as the TCL is displaced, is the most important factor in determining θadv and θrec as well as slip. For superhydrophobic surfaces composed of symmetrical features where the drop is in contact with only the top of the posts (pure Cassie state), the slip angle will be independent of the tilt direction. For these regular, symmetrical surface patterns, the change in TCL is largest when the receding edge of the droplet disconnects from a post and reattaches to the next post. For most superhydrophobic surfaces, the surface features are significantly smaller than the droplet area such that the receding TCL disconnects from only a few posts at a time and an average, directionally independent slip angle is observed. Only when the surfaces are composed of anisotropic features such as ridges13 can one observe a directional dependence on the slip angle. The magnitude of the slip angle is thus dependent upon the energy barrier formed by the largest incremental reduction in TCL.26 On surfaces prepared by 3D printing as described in this article, the PDMS posts are relatively large and the droplets partially wet these surface features. Because of these unique aspects, the liquidsolid energy contribution of individual surface features is greater compared to that of features prepared by micro- or nanofabrication. However, the behavior of water droplets on these surfaces is similar to that observed on microor nanotextured surfaces. For example, good agreement was

observed between predictions from the modified Furmidge relation and experimental observations of slip angles and slip angle anisotropy (Table 3 and Figure 5) as reported for microfabricated surfaces.26 One advantage of the relatively large size of the features that compose these surfaces is the ability to observe dropletsurface interactions with a light microscope under ambient conditions. Figure 6ac shows a water droplet moving across printed PDMS surfaces composed of either straight (Figure 6a) or sloped (Figure 6b,c) posts. Because of the coarse pitch (500 μm) and small feature diameters, the droplet mass cannot be supported exclusively on the tops of the posts. As a result, partial penetration of the droplet by the posts occurs. This partial wetting of the surface is consistent with a modified Cassie state.16,27 Some posts have thin curls at the tips. The formation of these tip embellishments is a product of the fabrication process and the rheology of the PDMS used. We found no significant differences in behavior across a surface or between surfaces containing features with fewer embellishments (such as DAP-21-20 and DAP-50-10) compared to surfaces containing more significant embellishments (such as DAP-00-8 and DAP-28-20). Because the droplet penetrates these fine scale features, their effect on droplet motion is expected to be small. As the surface with straight PDMS posts (one that has no slip angle anisotropy) is tilted, the trailing edge of the droplet moves up the trailing posts such that the TCL decreases in length in a continuous fashion (frames 13 in Figure 6a). Similarly, on a surface with sloped posts (Figure 6b), the trailing edge of the droplet moves up the trailing post and the TCL decreases 9907

dx.doi.org/10.1021/la201222a |Langmuir 2011, 27, 9902–9909

Langmuir

ARTICLE

Figure 7. Cartoons illustrating the wetted portion (blue) of an individual trailing edge post (yellow) during a tilt experiment. As the droplet moves in a direction parallel to the slope of the posts, the TCL shrinks continuously.

Figure 8. (a) Perspective view. (b) Backside view. As the droplet moves in the direction opposite to the slope of the posts, the TCL grows until a maximum value is achieved. The droplet remains pinned at this position until the gravitational force becomes sufficiently large and the droplet is released.

continuously as the droplet moves across the surface in a direction parallel to the post slope. The slip angles associated with these two geometries are within (1° of each other. However, when the surface is tilted in the opposite direction (Figure 6c), the droplet behavior is significantly different. As for all surfaces, the trailing edge of the droplet starts with the TCL around the middle of the trailing post, parallel to the substrate. As the tilt angle is increased, the droplet moves such that the trailing edge of liquid moves up the post. Thus, TCL increases in length during the initial stages of droplet motion during tilting. The increased TCL length presents a larger barrier to further droplet movement, so the tilt angle must be increased, relative to motion in the opposite direction, before the gravitational forces are sufficient to overcome the larger solidliquid interactions. Before release (Figure 6c(3)), the TCL assumes an elliptical shape where the major axis is parallel to the slope of the trailing post. This elliptical TCL is greater in length than the TCL at the start of the experiment. The droplet remains pinned in this position until the tilt angle increases sufficiently such that the gravitational force overcomes the forces at the TCL. Once this barrier is overcome, the TCL ellipse begins to shrink and the drop is rapidly released from the trailing post. A visual examination of these still images and videos of the tilt experiments emphasize the importance of incremental TCL

length changes as the surface is tilted. (See the Supporting Information for videos and SF3 for additional optical photographs of a droplet moving away from the camera and against the direction of the post slope.) A schematic illustrating the change in the TCL as a function of the tilt direction is shown in Figures 7 and 8. As the droplet moves in a direction parallel to the slope of the posts (Figure 7), the TCL shrinks continuously. Thus, once the tilt angle is increased to overcome the initial TCL energy barrier, the droplet will move with no additional barrier to surmount. When the droplet is tilted such that it moves in a direction opposite to the post slope (Figure 8), the TCL grows until a maximum value is achieved. The tilt angle must be increased sufficiently to overcome the stabilization energy resulting from this increased TCL. The increased energy barrier resulting from the longer TCL is responsible for the slip-angle anisotropy. Increasing the slope of the features appears to increase the length of the TCL and thus the tilt angle required to overcome the liquidsolid forces pinning the droplet. This is reflected by the lower θR values in Table 2. The increased slope would not significantly affect the TCL when the droplet moves in a direction parallel to the post slope. The behavior of these printed surfaces appears to validate the predictions made by Kusumaatmaja.16 As with their model asymmetric (i.e., sawtooth) features, anisotropic slip is observed 9908

dx.doi.org/10.1021/la201222a |Langmuir 2011, 27, 9902–9909

Langmuir when the droplet is able to penetrate the features partially and make contact with the sloped, geometrically asymmetric surface features. In addition, a slight elongation of the droplet was observed in the direction parallel to the post slope, as previously demonstrated by Kusumaatmaja et al.28 and Chung et al.29 on surfaces patterned with parallel grooves.

V. CONCLUSIONS In this work, we demonstrate slip-angle anisotropy on synthetic superhydrophobic surfaces fabricated using a novel 3D printing technique. Surfaces composed of these relatively large, 500-μm-pitch PDMS features displayed superhydrophobic properties that are in excellent agreement with relations used for surfaces made with features 2 to 3 orders of magnitude smaller. Optical microscopy was used to study the liquidsolid interactions during tilting. When the surface was titled parallel to the feature slope, the droplet began to slip once the liquidsolid forces were exceeded by gravity and the TCL decreased continuously as the drop was displaced. When the surface was tilted in a direction opposite to the post slope, the TCL was observed to increase as the droplet began to slip. The larger TCL increases the magnitude of the liquidsolid interactions, so a greater tilt angle is required to overcome this barrier before the droplet can slip. Three-dimensional printing of PDMS features enables the fabrication of surfaces with isotropic or anisotropic superhydrophobic behavior in any arbitrary location on the substrate through simple programming of a robotic printer. The incorporation of regions with this ratchetlike slip-angle anisotropy significantly increases the utility of superhydrophobic surfaces for microfluidic applications because unidirectional droplet motion can be achieved without pumping a carrier fluid. Although the features used in this study are relatively large, the design concepts can be applied to smaller features that could support droplets with a smaller diameter. Future work is underway to develop polymer printing technologies that can fabricate features that are more than 10 times smaller using a tip-transfer process while maintaining low fabrication costs and design flexibility. ’ ASSOCIATED CONTENT

bS

Supporting Information. Post slope and height measurement procedure. Static contact angles on superhydrophobic surfaces with straight posts. Effect of droplet volume, post slope, and tilt direction on the slip angle. Optical photographs of a droplet moving away from the camera and against the direction of the post slope. Effect of post slope orientation on the static contact angle and droplet shape. Video 1: 20 μL droplet moving in the direction aligned with the post slope (easy direction) on DAP 21-20. Video 2: 20 μL droplet moving in the direction against the post slope (hard direction) on DAP 21-20. Video 3: 20 μL droplet moving to the left on a surface with straight posts (DAP 00). Video 4: 20 μL droplet moving to the right on a surface with straight posts (DAP 00). This material is available free of charge via the Internet at http:// pubs.acs.org.

’ AUTHOR INFORMATION

ARTICLE

’ ACKNOWLEDGMENT This research was supported by the New York State Foundation for Science, Technology and Innovation (NYSTAR) Faculty Development Program. Additional support from the NYSTARfunded Center for Engineered Polymeric Materials (AML) and the NSF-funded STEAM program (MB) is gratefully acknowledged. We thank Dr. QianFeng Xu for helpful discussions. ’ REFERENCES (1) Theberge, A. B.; Courtois, F.; Schaerli, Y.; Fischlechner, M.; Abell, C.; Hollfelder, F.; Huck, W. T. S. Angew. Chem., Int. Ed. 2010, 49, 5846–5868. (2) Markey, A. L.; Mohr, S.; J.R. Day, P. J. R. Methods 2010, 50, 277–281. (3) Gonzalez, A.; Ciobanu, D.; Sayers, M.; Sirr, N.; Dalton, T.; Davies, M. Biomed. Microdev. 2007, 9, 729–736. (4) Gentile, F.; Das, G.; Coluccio, M. L.; Mecarini, F.; Accardo, A.; Tirinato, L.; Tallerico, R.; Cojoc, G.; Liberale, C.; Candeloro, P.; Decuzzi, P.; DeAngelis, F.; DiFabrizo, E. Microelectron. Eng. 2010, 87, 798–801. (5) Lifton, V. A.; Taylor, J. A.; Vyas, B.; Kolodner, P.; Cirelli, R.; Basavanhally, N.; Papazian, A.; Frahm, R.; Simon, S.; Krupenkin, T. Appl. Phys. Lett. 2008, 93, 043112. (6) Basavanhally, N. R., Hodes, M. S., Kolodner, P. R., Kornblit, A., Krupenkine, T. N., Lee, W., Lyons, A. M., Salamon, T. R., Vyas, B. Thermal Energy Transfer Device. U.S. Patent 7,832,462, Nov 16, 2010. (7) Sun, T.; Feng, L.; Gao, X.; Jiang, L. Acc. Chem. Res. 2005, 38, 8. (8) Zheng, Y.; Gao, X.; Jiang, L. Soft Matter 2007, 3, 178182. (9) Neinhaus, C; Barthlott, W. Ann. Bot. 1997, 79, 667–677. (10) Guo, Z.; Liu, W. Plant Sci. 2007, 172, 11031112. (11) Feng, L.; Li, S.; Li, Y.; Li, H.; Zhang, L.; Zhai, J.; Song, Y.; Liu, B.; Jiang, L.; Zhu, D. Adv. Mater. 2002, 14, 1857–1860. (12) Shirtcliffe, N. J.; McHale, G.; Newton, M. I. Langmuir 2009, 25, 14121–14128. (13) Long, C. J.; Schumacher, J. F.; Brennan, A. B. Langmuir 2009, 25 (), 12982–12989. (14) Chu, K. H.; Xiao, R; Wang, E. N. Nat. Mater. 2010, 9, 413–417. (15) Zhang, J.; Cheng, Z.; Zheng, Y.; Jiang, L. Appl. Phys. Lett. 2009, 94, 144104. (16) Kusumaatmaja, H.; Yeomans, J. M. Soft Matter 2009, 5, 2704–2707. (17) Lyons,A. M.; Mullins, J.; Barahman, M.; Erlich, I.; Salamon, T. 20th Annual International Solid Freeform Fabrication Symposium; Austin, TX, USA, August 35, 2009; pp 488497. (18) Furmidge, C. G. L. J. Colloid Sci. 1962, 17, 309–324. (19) Extrand, C. W.; Gent, A. N. J. Colloid Interface Sci. 1990, 138, 431–442. (20) Gao, L.; McCarthy, T. J. Langmuir 2006, 22, 6234–6237. (21) Gao, L.; McCarthy, T. J. Langmuir 2007, 23, 3762–3765. (22) Gao, L.; McCarthy, T. J. Langmuir 2009, 25, 14105–14115. (23) Yang, J. T.; Yang, Z. H.; Chen, C. Y.; Yao, D. J. Langmuir 2008, 24, 9899–9897. (24) McHale, G.; Shirtcliffe, N. J.; Newton, M. I. Langmuir 2004, 20, 10146–10149. (25) McHale, G. Langmuir 2007, 23, 8200–8205. (26) Choi, W.; Tuteja, A.; Mabry, J. M.; Cohen, R. E.; McKinley, G. H. J. Colloid Interface Sci. 2009, 339, 208–216. (27) Lafuma, A.; Quere, D. Nat. Mater. 2003, 2, 457–460. (28) Kusumaatmaja, H; Vrancken, R. J.; Batiaansen, C. W. M.; Yeomans, J. M. Langmuir 2008, 24, 7299–7308. (29) Chung, J. Y.; Youngblood, J. P.; Stafford, C. M. Soft Matter 2007, 3, 1163–1169.

Corresponding Author

*E-mail: [email protected]. 9909

dx.doi.org/10.1021/la201222a |Langmuir 2011, 27, 9902–9909