Micro To Nanoscale Engineering of Surface Precipitates Using

12 mins ago - Nanoscale engineering has traditionally adopted the chemical route of synthesis or opto-chemical techniques like lithography requiring l...
0 downloads 11 Views 2MB Size
Subscriber access provided by Gothenburg University Library

Article

Micro To Nanoscale Engineering of Surface Precipitates Using Reconfigurable Contact Lines Prasenjit Kabi, Swetaprovo Chaudhuri, and Saptarshi Basu Langmuir, Just Accepted Manuscript • DOI: 10.1021/acs.langmuir.7b04368 • Publication Date (Web): 18 Jan 2018 Downloaded from http://pubs.acs.org on January 18, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Langmuir is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Micro To Nanoscale Engineering of Surface Precipitates Using Reconfigurable Contact Lines Prasenjit Kabia, Swetaprovo Chaudhuria,c*, Saptarshi Basua,b* a c

Interdisciplinary Centre for Energy Research,

b

Department of Mechanical Engineering,

Department of Aerospace Engineering, Indian Institute of Science, Bangalore, Karnataka

560012, India. Corresponding Author(s) *Emails: [email protected] [email protected]

KEYWORDS: interfacial re-engineering, directed self-assembly, non-spherical droplet evaporation, wall-less confinement, order-disorder, nanoscale. Nanoscale engineering has traditionally adopted the chemical route of synthesis or optochemical techniques like lithography requiring large process times, expensive equipment and an inert environment. Directed self-assembly using evaporation of nanocolloidal droplet can be a potential low-cost alternative across various industries ranging from semiconductors to biomedical systems. It is relatively simple to scale and re-orient the evaporation driven internal flow field in an evaporating droplet which can direct dispersed matter into functional agglomerates. The resulting functional precipitates not only display macroscopically discernible changes but also nanoscopic variations in the particulate assembly. Thus evaporating droplet forms an autonomous system for nanoscale engineering without the need for external resources. In this article develop an indigenous technique of interfacial re-engineering which is both simple and inexpensive to implement. Such re-engineering widens the horizon for surface patterning previously limited by the fixed nature of the droplet interface. It involves hand printing hydrophobic lines on a hydrophilic substrate to form a confinement of any selected geometry using a simple document stamp. Droplets cast into such confinements get modulated into a

ACS Paragon Plus Environment

1

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 30

variety of shapes. The droplet shapes control contact line behaviour, evaporation dynamics as well as complex internal flow pattern. Exploiting the dynamic interplay among these variables, we could control the deposit’s macro as well as nanoscale assembly not possible with simple circular droplets. We provide a detailed mechanism of the coupling at various length scales enabling a predictive capability in custom engineering particularly useful in nanoscale applications such as photonic crystals. Introduction There is an ever increasing demand in the manufacturing sector to find alternative pathways with lower economical as well as ecological overheads particularly in the areas of surface patterning, semiconductor devices, biomedical systems and additive manufacturing. By mimicking the behaviour of atoms and molecules at nano and mesoscopic scales, self-assembly allows fabrication of larger and more practical functional units1. Although material processing and waste generation are minimized as compared to top-down approaches, building practical, realworld applications while preserving the lab-scale precision of self-assembly poses a major challenge to application engineers. Bottom-up manufacturing processes profit from the nanoscale manipulation of matter as the high surface to volume ratio of atoms in nanostructures lead to enhancement of mechanical, electrical2, optical3, and electronic properties4-5. Directed selfassembly using templates leads to accelerated production throughput and better quality control of final products6. Central to such processes is the need to control or direct the interaction between the building blocks and the external environment7-11. Evaporative elimination of liquid from nano-colloidal drops leading to natural agglomeration of particles offers one attractive alternative in devising micro and nanoscale systems at ultra-low cost. In droplet based architecture, interfacial interactions at the three-phase contact line of a sessile drop coupled with the vapour mediated mass loss give rise to internal convective motion creating a simple and inexpensive template for self-assembly at hierarchical length scales. Though technologically simplistic to implement, the background processes involve complex multi-scale phenomena comprising of mass, momentum, energy transport, contact line dynamics, inter-particle as well as particle-substrate interactions12. The natural coupling between evaporation dynamics, contact line behaviour and internal fluid transport make it possible to effect micro-nano scale changes in assembled precipitate structure by simply perturbing the

ACS Paragon Plus Environment

2

Page 3 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

droplet at millimetre scale. Recent studies demonstrate a diverse range of such perturbations like natural evaporation on homogenous as well heterogeneous surface wettability13-17, chemical composition of droplet or substrate18,19, confinement of droplet’s liquid-vapour interface20,21 as well as imposition of heat22-24 and vibration25,26, all of which tailor the evaporative process towards controlled surface patterning. By restricting the key parameters (evaporation-contact line dynamics-flow) to macroscopic length scales, sophistication and cost associated with nanoscale manipulation is deliberately simplified. Various kinds of precipitates ranging from “coffee-rings” to shape changing pellets

27,28

can be

possibly generated using droplet based template. Droplet evaporation forms the bedrock for all these applications ranging from ink-jet printing29,30, DNA microarrays31-33 and electronic circuit fabrication34 to name a few. As an example, colloidal photonic crystals have extensively depended upon methods such as ink-jet printing35, sedimentation36,37, emulsion drop method38,39 and dip-coating40,41 for fabrication where the deposition rate and contact line dynamics play a pivotal role in determining crystalline quality and reflectance signature. Recently evaporating droplets have demonstrated similar potential at fabricating photonic crystals on a smaller scale but with much lesser external intervention15, 29, 42,43. The shape of liquid drop on a flat inert surface is circumferentially symmetric44. As the internal flow pattern is guided by the droplet interface, its impact is rendered redundant in the azimuthal direction. Such redundancies severely limit design strategies. For example, while using drying of droplet to fabricate discrete electronic units on integrated circuits, shape control may lead to better ergonomic and repeatable design34. Methodologies involving re-shaping of the contact line allows introduction of concentration gradients crucial to many applications45-49.

ACS Paragon Plus Environment

3

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 30

Figure 1 How to impact multiscale surface patterns by using reconfigurable contact lines (a) Deploying droplet in printed perimeter (b) Confinement induced wettability contrast re-engineers the drop’s interface (c) The contact line and evaporation dynamics are coupled to the drop shape via the mode of evaporation (d) The modulated drop shape and the mode of evaporation control the internal flow pattern in such droplets (e) Evaporation induced deposit pattern. The deposit profile near the contact line of the cell and channel (black square) is shown. The deposit also displays nanometric variations in order-disorder across particle assemblies.

In the present work, we exploit the proof-of-concept proposed previously50 to re-engineer the natural interface of a sessile droplet into various exotic, non-spherical shapes as illustrated in Figure 1. The method is similar to micro-contact printing developed by Whitesides’ group51 although much simpler and cheaper to implement. It not only allows rapid prototyping of confinement structures on any substrate but also allows on demand reconfiguration. The confinement structures are essentially hydrophobic lines on plasma treated, highly wettable substrates52 as shown in Figure 1a. Nanofluid drops when dispensed within the confined area spread and morph into exotic non-spherical shapes as shown in Figure 1b. The height of these

ACS Paragon Plus Environment

4

Page 5 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

confinements are O (10-2) lower than the initial droplet height hence they are referred to as “wallless” confinements. The droplet spread is constrained by the higher energy barrier of the hydrophobic line. Induced asymmetry in drop shapes leads to asymmetry in contact line recession as well variations in the evaporation dynamics coupled to each other via mode of evaporation as shown in Figure 1c. The template driven drop shape induces a pressure driven recirculatory flow which is coupled to the evaporation induced particle transport as illustrated in Figure 1d. By redrawing the contact line, the millimetre form of deposit has been modified from its usual ring like appearance as shown in Figure 1e. Profilemetry scan of the deposit reveal its concentrated presence near the droplet edge. The confinement profile shown in Figure 1e justifies itself as practically “wall-less”. SEM of the deposit top reveals two distinct regions of nanoparticle packing which are affected by the rate of the capillary transport. Thus the deposit is modulated at multiple lengthscales in a controllable fashion by interfacially re-engineered droplets. Experimental section

Figure 2 (a) Steps for glass substrate preparation-plasma treatment followed by pattern imprint using custom made stamp. Printed boundaries are nearly 1 µm high. (b) Schematic illustration of imaging an evaporating droplet from top as well as side using a combination of microscope and DSLR camera. (c) Schematic illustration of micro-Particle Image Velocimetry (µ-PIV). Sequentially acquired images are pair-wise crosscorrelated to obtain flow vector map (right side).

ACS Paragon Plus Environment

5

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 30

Material preparation Polystyrene beads (Fluoro-Max Red Aqueous Fluorescent Particles, Thermofisher Scientific) are diluted in de-ionised (DI) water to 0.05 wt.% prior to being used for experiments. The prepared dispersion is sonicated for 5 minutes as a precaution against flocculation. Three different particle sizes are used-50 nm (R050), 200 nm (R200) and 860 nm (R900). The polymeric face of an ordinary document stamp is embossed with the design using heat treatment method. Commercial ink is used for printing purpose making it an ultra low cost method. Substrate preparation Printing the confinement: Figure 2a illustrates the confinement printing procedure. Ordinary microscope glass slides (Bluestar) measuring 75 x 35 x 1.35 mm are used for all experiments. They are cleaned in soap solution to remove any oil residue and then again washed in IPA (IsoPropyl Alcohol) and finally rinsed in DI water. The cleaned glass slide is then subjected to atmospheric plasma (10 kV and 45 MHz) using a hand held plasma generator (BD-20 AC, Electrotechnic Products) for a maximum time of 60 seconds to render the glass substrate superhydrophillic48. The custom made stamp shown in Figure. 2a, is then brought in contact with the glass slide to transfer the confinement pattern. The print is water soluble and hence needs to be blow dried for ~20 seconds. The confinement to be used for the experiment is locally treated with plasma again (~10 seconds) to ensure better spreadability. Printed patterns can be easily rinsed off using ethanol and water. Droplet deployment: A syringe pump (NE-1000, New Era Pump Systems) is used to dispense 2.4 µl of dispersion into the confinement. As soon as the dispensed liquid spreads and fills up the confinement perimeter, the glass slide is then transferred to a confined environment (to avoid any minor convective effects) for optical imaging. All experiments are performed in an airconditioned room where the temperature is maintained at 25˚C while the relative humidity is maintained between 42-47% using a room dehumidifier (Origin, Novita). Optical imaging: As shown in Figure 2b, the evaporating droplet is imaged from top as well as side throughout its entire lifetime. The side profile is illuminated using a combination of LED

ACS Paragon Plus Environment

6

Page 7 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

source and diffuser plate (for uniform lighting). Images are captured every 4 seconds using a Nikon 7200 fitted with 4.5x zoom lens assembly (Navitar). The top view is imaged every 1 second using a CCD camera (SC-30, Olympus) fitted onto a BX-51 (Olympus) microscope through a 5x objective. The residue obtained on the glass slide after the droplet dries out is preserved for characterization. Micro-Particle Image Velocimetry (µ-PIV): As shown in Figure 2c, micro-PIV measurements quantify as well as qualify the internal flow field of the droplet evaporating in various confinements. 860 nm particles are used for the experiments. The particle concentration is maintained at 0.08 wt% to ensure appropriate particle image density for accurate PIV calculation. As before the droplet is dispensed into a confinement and transferred to a confined environment where it is placed on 3-axis motorized stage fitted on a microscope (Flowmaster Mitas, Lavision). Illumination from the bottom side is done using a Nd-YAG laser (Nano-Piv, Litron Lasers) (λex= 532 nm) through an objective lens. The emission spectra (λem=612 nm) from the dye-coated particles are captured via the same objective lens using a dichroic mirror and collected by a CCD camera (Imager Intense, Lavision) also connected to the Flowmaster Mitas. All operations are synchronized by PTU (power terminal unit). Single frame images are acquired at an appropriate rate to ensure 3-4 pixel shift in consecutive images. Two objective lenses are used for this set of experiments. While using the 5x objective, the frame rate is varied between 0.5 -2 fps while the measurement plane is maintained at 1/4th of the instantaneous droplet height near the centreline of the droplet. For the 20x objective, the frame rate is fixed at 5 fps while the imaging plane is set very close to the substrate (~10 µm).Image processing and vector computation from the acquired frames is done using Davis 7.2 software from LaVision. Consecutive images are pair-wise cross-correlated to obtain instantaneous vector map of the internal flow field. Interrogation window (IW) is maintained at 64 x 64 pixels in the first pass and reduced to 32 x 32 pixels in the second pass with 50 % overlap between both windows. Fluorescence Studies: Two widely disparate particle sizes- 50 nm and 860 nm are used to study the dynamic growth of particle aggregation at the contact line. BX-FM (Olympus) is retro-fitted with a fluorescent module using Hg lamp for illumination. The incident light passes through a band-pass filter centred at 532 nm to separate the excitation bandwidth. The emission spectra

ACS Paragon Plus Environment

7

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 8 of 30

from the droplet is collected via 612 nm high pass filter and sent to the CCD camera (SC-30) through the dichroic mirror. All images are captured using the 50x objective at a rate of 1 fps. Field Electron Scanning Electron Microscope (FE-SEM): Samples preserved from evaporation experiments are sputtered with 10 nm gold particles to ensure sample conductivity to electron beam. The coated sample is then examined using Ultra 55 Mono-CL (Zeiss). The electron beam is maintained at 3-5 kV while InLens detector is used to capture images. Working distance between sample and detector is maintained at 10 mm. Optical Profilemetry: The preserved samples are also scanned using a Taylor Hobson profilometer. A 50 x objective lens is used to acquire sample images at a distance of 3.78 mm from the sample surface. The acquired images are processed using CCI Taylor software. The height profile of the deposit is extracted at multiple locations along the hydrophobic perimeter. The inner region is almost devoid of any deposit and no discernible signal could be obtained from it. Results and discussion Drop shape in wall-less confinement

Figure 3(a) Schematic to illustrate the top and sideview acquired during the evaporation process. The contact radius in cell is rα and in channel is given as rβ. The ratio γ designates the template design. (b) Top and sideview of the droplet (loaded with 50 nm particles) after deployment in each template. Both scale bars equal 1 mm.

ACS Paragon Plus Environment

8

Page 9 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Printing of hydrophobic lines on hydrophilic substrates as detailed in “Methods” section creates wells for droplet containment. The printed wells or templates have smooth corners to reduce hydrodynamic resistance towards the spread of deployed droplet53 allowing it to occupy the entire perimeter. The maximum drop height is always less than capillary length of water (2.7 mm). Figure 3a illustrates the top and sideview of the droplet within the confinement. The contact radius of either of the bulbous ends referred to hereafter as the cell region is rα. The halfwidth of the straight portion denoted as the channel is rβ. The ratio γ= rβ/ rα is the aspect ratio and shall be used hereafter to distinguish any given template. Figure 3b shows the optical view of the droplet’s top and side profile after being deployed in different templates. Although Figure 3b presents droplets seeded with 50 nm particles, no discernible change is observed for different particle diameters (200 nm and 860nm). For appropriately small volumes, the droplet assumes a stable configuration54. For γ=1, the side profile is similar to a circular sessile droplet while the topview shows an elongated form. As γ is reduced from 1 to 0.45, the elongated droplet appears to be constricted at the centre while from the side one observes two cells connected by a thin channel of liquid. Since droplet deployment is always in the cell region, it follows intuitively that as γ is reduced, the volume of liquid spreading to the channel region reduces resulting in the thinner strip. Thus a single droplet volume is sub-partitioned into smaller yet inter-connected liquid segments by simply deploying in different confinements. L=10.2mm

initial conditions

γ

rα(µm)

rβ(µm)

hα(µm)

hβ(µm)

θα (degrees)

θβ(degrees)

0.45

895

402.75

338.17

102.44

43.18

29.45

0.56

895

501.2

274.65

146.46

35.63

33.29

0.68

895

608.6

240.27

170.2

31.4

32.705

0.78

895

698.1

213.22

192.17

28.019

31.79

0.9

895

805.5

139.29

162.18

18.515

19.435

1

895

895

133.74

164.96

17.79

21.85

Table 1: The initial values of r (contact radius), h (height) and θ (contact angle). Subscript α and β represent cell and channel respectively.

ACS Paragon Plus Environment

9

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 30

Initial features of the droplet post deployment are tabulated in Table 1. Since the liquid occupies the entire perimeter of the template, the values of rα and rβ are self-evident. The initial difference of curvature between cell and channel (contact angle and apex height) is listed in columns (4-7). Consistent with the observations in Figure 3, the initial disparity in curvature between cell and channel gradually diminishes as γ is increased. Contact line dynamics

Figure 4 Contact line dynamics of an evaporating liquid drop deployed in various templates. The drying state is specified by t/T where t is the acquisition instant of the image and T is the total droplet lifetime. The templates discussed are (a) γ=1 where contact line depins in cell before doing so in channel (b) γ=0.68 where the the slip in cell and channel is nearly simultaneous (c) γ=0.45 where the contact line slip occurs in channel before cell. Blue dashed line represents contact line within the template perimeter signifying the liquid coverage in the region. Scale bars equal 0.5 mm. All droplets shown here are loaded with 50 nm particles.

ACS Paragon Plus Environment

10

Page 11 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

A total of six different templates (γ=1 to 0.45) and three different particle diameters are reported here. For brevity we shall restrict our present discussion to only three cases (γ=1, 0.68 and 0.45) involving droplets loaded with 50 nm particles. Figure 4a shows the contact line (CL) recession in case of γ=1. Events are reported as a fraction of the droplet lifetime T to account for experimental errors introduced by small fluctuations in ambient conditions and droplet deployment. The CL first de-pins in the cell region at t/T~0.87 while still being pinned in the channel region. At t/T~0.9, the CL has completely receded out of the cell region but no discernible retraction is observed in the channel. At t/T~0.95, CL de-pins in channel. Subsequently, CL from both cell and channel rapidly converge. However, as the droplet CL is now completely free from the template perimeter, it is no longer under the effect of dual wettability and relaxes from an elongated to a more spherical cap shape at t/T=0.9855. In case of γ=0.68, the lag between the CL shift in cell and channel is now reduced (∆t/T~0.03 vs. ∆t/T~0.08 for the previous case of γ=1). At t/T~ 0.95, the CLs in both cell and channel recede rapidly (Figure 4b) but unlike γ=1, the droplet is still highly elongated. To achieve equilibrium, it must break into smaller droplets (to minimize surface area) thereby creating two centres of CL convergence as opposed to a single one in case of γ=1. Finally in case of γ=0.45, CL in cell remains pinned while it depins in channel (Figure 4c). Eventually, the thinning liquid mass in the channel necks (t/T~0.9) leading to breakup into two separate droplets. The dual centres of CL convergence are now located within the cell regions. Thus, with decreasing value of γ, the CL either stays as a single entity (non-necking condition) or splits into two (necking condition). Experiments repeated for all the templates using particle diameters 200 nm and 860 nm show no variance in this phenomenon.

ACS Paragon Plus Environment

11

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 30

Figure 5 (a) For different particles (diameter dp =50 nm, 200 nm and 860 nm) onset of contact line slip (τα=tslip/T) in different templates (left axis) varies monotonically with its initial volume (v0,α) (right axis) in cell region. The CL slip appears independent of particle size. (Inset) shows the first observable instant of contact line slip in cell for a droplet having 860 nm particles (b) Similar plot for contact line slip (τβ) in channel region (c) Ratio of contact line slip in cell and channel (ζ=τβ/ τα) plotted for each template design depicts the necking to non-necking transition during evaporation (inset) the legend used for the plots of Figure 4. (d) Initial volume segregation in cell and channel shown by plot of ᴧ0= v0,β/ v0,α.

Onset of contact line slip (τ) is plotted for all template designs and different particle sizes where the symbols α denote the cell and β denotes channel. Figure 5 (a & b) show the instance of CL shift in the same template for different particle sizes. Quantitatively CL dynamics appear independent of size of dispersed particle. Experimentally from Figure 5a we obtain  = 0.79 .

[1]

while from Figure 5b it is observed to be  = 0.97 .

[2]

Figure 5c presents a regime map of the CL retraction by plotting ζ=τβ/ τα with respect to γ. Although ζ=1 is denoted as the cross-over point between the occurrence of necking and nonnecking regime, the template design effects a smooth transition between the two. Figure 5a and 5b also demonstrate that by increasing the initial volume in either cell or channel, the onset of CL shift maybe delayed. The volume partition of a single droplet as achieved by deploying it in different templates is shown in Figure 5d. In our previous work, we had simply demonstrated the binary nature of necking for a single particle diameter50. Experiments with different templates reveal the continuous nature of the phenomenon. The decoupled nature of the CL dynamics and particle size imply the inherent scalability of the proposed method.

ACS Paragon Plus Environment

12

Page 13 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

Modes of evaporation

Figure 6 (a) Modes of evaporation shown via contact angle (CA-θ/θ0) and contact radius (CR-r/r0) regression over the droplet lifetime (t/T) where θ0 is the initial contact angle, r0 is the initial contact radius, T is the evaporation period of the entire droplet while symbols α and β denote cell and channel respectively. CR regression is shown for two different particle dispersion-50 nm and 860 nm. (b) Total volume regression (ṽt=vt/vt,0) for 50 nm and 860 nm particle dispersions where vt,0 is the total initial volume (c) Normalized height ĥ =

 ,, ,

 ,, ,

variation over total lifetime shown for 50 nm particle dispersion.

For an axisymmetric sessile droplet with circular contact line, the entire evaporation period could be divided into three phases56. In constant contact radius (ccr) mode, the contact angle reduces gradually to accommodate solvent loss while the contact radius stays pinned. In constant contact

ACS Paragon Plus Environment

13

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 30

angle mode (cca), the contact line slips while the contact angle remains constant. Finally, in mixed mode, both the contact radius and angle shrink. The sequence through which the various modes of evaporation progress is generally dependent on interfacial interactions at the threephase contact line. Generally, for glass substrates, initial stage of evaporation is ccr followed by cca or the mixed mode. The modes of droplet evaporation are shown in Figure 6a. The contact radius (r/r0) in both cell and channel for all values of γ remains pinned for nearly 90 % of the evaporation period. The onset of slip sets in at different times for different configurations. Once initiated, the contact line slips rapidly. The contact angle (θ/θ0) regression is similar for both cell and channel of the droplet in different confinements which is shown in Figure 6a. The volume of the cell region is computed by assuming it to be a spherical cap which is facilitated by using a circle instead of a rectangular perimeter50 for the cell. The cell volume is  !

 = "#

%$!

2 − 3)*+, + )*+,  

[3]

while the channel is considered to be a cylindrical cap whose volume is given by  /

.  = "#

%$



/ 0 ° .

2, − sin 2, 4

[4]

Here rα, rβ, θα and θβ are illustrated in the inset of Figure 6a while l is the length of the channel section. The rate of solvent loss for a spherical sessile droplet53 may be expressed as 67 =

89 8:

=−

;  ?@ A

B,

[5]

where J is the solvent flux from the droplet, A is the surface area of the drop, M is the molar mass of water (18g/mol), D is the diffusion constant (2.54x10-5 m2/s), r is the contact radius, RH is the relative humidity, ρ is the solvent density (0.997 g/ml), Cs is the saturated vapour concentration at ambient conditions (1.278 mol/m3) and f(θ) accounts for the droplet shape57. Using equation 5, the rate of solvent loss (JA) is calculated separately for cell and channel. The variation in total solvent loss (JA) for different values of γ is within 10% which accounts for the globally similar trend of volume regression as shown in Figure 6b. However, the individual JA

ACS Paragon Plus Environment

14

Page 15 of 30 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Langmuir

values for cell and channel are not similar. The mean droplet height within template is observed to decay gradually as seen from Figure 6c. By computing the initial value of total 67: and considering that the contact radius is constant throughout the droplet lifetime, the evaporation time maybe estimated as CD,D": =

F

EH G 89

IJGKH

=

LG

[6]

IJGKH

By evaluating equation 6 we get

:M,MNO :M,M>G

~ 0.88 for γ=0.45 and ~0.66 for γ=1. Equation 5 holds true

for spherical cap geometry. For smaller values of γ, higher fraction of drop volume consists of a spherical cap allowing better estimation of droplet lifetime using equation 6. However, as γ increases, a larger portion of the droplet volume assumes the cylindrical cap shape in the channel thus deviating from the timescales estimated using equation 6. The assumption of the contact radius being pinned for the entire lifetime instead of only 90% as shown in Figure 6a may also result in minor errors. Evaporation and geometry controlled flow field

ACS Paragon Plus Environment

15

Langmuir 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 30

Figure 7 Discussion on internal flow field (a) Schematic representation of double curvatures in droplet of γ