supakar-thesis-2016 (1)

EXPERIMENTAL STUDY OF LIQUID MARBLE FORMATION AND

DEFORMATION DYNAMICS

by

TINKU SUPAKAR, M.Tech. A Thesis

IN Chemical Engineering

Submitted to the Graduate Faculty of Texas Tech University in

Partial Fulfillment of the Requirements for

the Degree of MASTER OF SCIENCE

Approved Jeremy Marston

Chairperson of the committee

Gordon Christopher

Siva Vanapalli

Mark Sheridan Dean of Graduate School

August, 2016

Texas Tech University, Tinku Supakar , August 2016

ii

ACKNOWLEDGEMENTS First of all I would like express my profound respect to my thesis supervisor, Dr. Jeremy Marston, assistant professor, department of chemical engineering, TTU for his valuable guidance, constant encouragement and keen interest at every stage of this study, without which it would be extremely difficult to accomplish this thesis work.

I would also like to thank Dr. Siva Vanapalli, associate professor, department of chemical engineering, and Dr. Gordon Christopher, assistant professor, department of mechanical engineering, for their valuable suggestion to improve this thesis work.

I’m also thankful to department of chemical engineering, TTU, for providing me necessary facilities and financial support, to accomplish this thesis work. Special mentions to my loving husband, Dr. Amlan Ghosh, my elder sister, Mrs. Rinku Supakar, my parents, Mr. Bhim Sen Supakar and Mrs. Kabita Supakar and my in laws for all the love and affection they have showered on me. Their continuous support and encouragement has guided me through all my endeavors.

Last but not least, I really appreciate and am thankful to my lab colleagues and Lubbock friends who have helped me in all possible ways.

Texas Tech University, Tinku Supakar, August 2016

iii

TABLE OF CONTENTS Acknowledgements…………………………………………………………………….ii Abstract……………………………………………………………………………........v List of Tables…………………………………………………………………………...vi List of Figures……………………………………………………………………….....vii List of Symbols………………………………………………………………………....xi Overview……………………………………………………………………....................1

1.1 Liquid marble formation……………………………………………………...1 1.2 Liquid marble deformation…………………………………………………...4

2 Background and motivation………………………………………………….............6 2.1 Literature review……………………………………………………………...6 2.2 Research objectives…………………………………………………………..29 2.3 Organization of the thesis……………………………………………………30

3 Experimental Methodology………………………….....…………...……………….31 3.1 Liquid marble formation experiment set up…………………………………31 3.2 Liquid marble deformation experiment set up………………………………32 3.3 Hydrophobizing agent and power treatment ………………………………..33 3.4 Particle sizing and shape determination ……….……………………………34 3.5 Contact angle measurement…………….……………………………………34 3.6 Powder properties…………….……………………………………………...35 3.7 Video capture and analysis…………….…………………………………….36 3.8 Instrumentation…………….…………………………………………...........37

3.8.1 High speed camera…………….…………………………..37


iv

3.8.2 Objective lens…………….………………………………………38 3.8.3 Light Sources…………….……………………………………….38

4 Results and discussion………………………….....…………...…………………….40 4.1 Overview of impact dynamics for liquid marble formation………………...40 4.2 Maximum spread …………………………………………………………...44 4.3 Liquid marble formation ……………………………………………………47 4.4 Particle mobility and marble formation ……………………………………50 4.5 Transition to jammed interface……………………………………………..55 4.6 Most deformed shapes in marble formation ……………………………….56 4.7 Maximum spread of liquid marbles on impact……………………………..57 4.8 Comparison of spreading process of liquid marbles and pure droplets…….67 4.9 Liquid marble splashing behavior ………………………………………….71 4.10 Study of rupture in liquid marbles…………………………………………75

5 Closure …...………………………………………………………………………….79 5.1 Summary & Conclusions……………………………………………………79 5.2 Recommendations for Future Work…………………………………………81

References……………………………………………………………………………....82


v

ABSTRACT We present findings from an experimental study of the impact of liquid droplets onto powder surfaces, where the particulates are hydrophobic. We vary both the size of the drop and impact speed coupled with the size range of the powder in order to assess the critical conditions for the formation of liquid marbles, where the drop becomes completely encapsulated by the powder, and arrested shapes, where the drop cannot regain its spherical shape.

By using different hydrophobization agents we find that a lower particle mobility may aid in promoting liquid marble formation at lower impact kinetic energies. From observations of the arrested shape formations, we propose that simple surface tensions may be inadequate to describe deformation dynamics in liquid marbles.

We also studied the dynamics of liquid marble deformation. For this we performed the impact experiments of liquid marble encapsulated with different sized particles sizes across a broad range of impact speed, V0 on substrates with different wettability. We hence determined the relation between the maximum spreading ratio, max and the impact parameters for the case of liquid marbles. We also compared the spreading process of liquid marbles upon impact on rigid surface to that pure water droplets experimentally and thereby studied the fingering pattern and rupture dynamics of liquid marbles in comparison to the case of pure water droplets.


vi

LIST OF TABLES

1.1 Summary of scaling for the normalized maximum spread diameter Dmax=D0 for different target surfaces.

3

2.1 Different models for the spreading factor βmax in literature as a function of the impact parameters

21

3.1 Summary of the properties of the particles used. Contact angles stated are based on the sessile drop technique. 36

4.1 Weber, Bond and Ohnesorge numbers for different drop diameters ( 1oD , 2 and 3 mm) as they impact the powder bed for speed V0 = 0.6–2.6 m/s.

23

4.2 Summary of marble formation for all drop and particle diameters.

48

4.3 Summary of splashing behavior for liquid marbles of all particle diameters and water droplet onto hydrophilic/hydrophobic surface.

74


vii

LIST OF FIGURES 2.1 A non-wettable powder particle contacting a liquid

surface. 8

2.2 Flow chart showing favoring condition for liquid marble formation.

12

2.3 Showing the gap between the freezing front and the particle coating due to nucleation inside a liquid marble. The arrow indicates the flow of water to the gap due to Marangoni effects.

16

2.4 Impact and spreading process of a pure liquid droplet onto a solid surface(a) before impact, (b) maximum spreading (c) maximum retract/rebound, and (d) equilibrium.

20

2.5 Showing a) Fingering b) Splashing pattern on drop impacts with rigid surfaces.

22

2.6 Splashing pattern of drop impact on two different surfaces a) Hydrophilic b) Hydrophobic.

23

2.7 Showing the rupture of a thin water film (thickness 25μm) produced by the impact of a 550 μm water droplet on a mirror-polished stainless steel surface at 40 m/s (a) Formation of holes, (b) growth of holes.

24

3.1 Schematic of the experimental setup used. 31 3.2 3.3

3.4

The experimental set up for liquid marbles impact study. The schematic view of the experimental set up for liquid marbles impact study. Microscope image (left) and binarized image (right) of particles ( 488pd m ) used for measurement of circularity.

32 33

35


viii

3.5 Images of water drops on (a) adhesive tape, (b) glass beads with 148pd m , and (c) glass beads with

488pd m used to derive contact angles.

35

3.6 Processed image for drop impact study. In the image the centroid of the drop is marked as green.

37

3.7 Processed image for a liquid marble (coated with 204 m particles) on impact with a hydrophilic plate. In the image the extrema points are marked as green.

38

3.8 Images of light source directed towards the surface of the impact.

40

4.1 Image sequences showing the formation of fully encapsulated spherical marbles during the impact of water droplets onto hydrophobic glass beads with 25pd m .

43

4.2 Image sequences showing the formation of deformed marbles with jammed interfaces during the impact of water droplets onto hydrophobic glass beads with

25pd m .

44

4.3 Normalized maximum spread diameter (Dmax/D0) versus the impact Weber number (We) for all droplet sizes and particles sizes used herein.

47

4.4 Plots of maximum spread versus impact speed highlighting the transitions from incomplete coverage (open squares) to complete coverage (filled circles) to jammed interfaces (filled diamonds).

51

4.5 Impact of 2.1 mm water droplets onto glass beads with mean diameter dp = 170 m coated with (a) Ultra ever dry and (b) Glaco MCZ.

52

4.6 Plot of maximum spread versus impact speed, with arrows indicating the onset of complete marble formation

54


ix

for different hydrophobization agents (red = Glaco MCZ, black = Ultra ever dry).

4.7 (a) Image sequence of a partially encapsulated droplet (D0 =1mm, dp =170 m ) with arrows indicating the location of an individual particle rolling across the interface. (b) Plot representing the motion of the individual particle in (a).

55

4.8 Sign of curvature relative to particle. 56 4.9 Image of an isolated particle dp =148 m at a drop

surface. 57

4.10 Arrested shapes formed during the impact of 2 mm water droplets with beds of hydrophobic 25 m particles.

59

4.11 Images taken from side showing impact of liquid marbles coated with particles of mean diameter dp = 170 m with (a) Hydrophilic plate (b) Hydrophobic plate.

61

4.12 Images taken from bottom showing impact of liquid marbles coated with particles of mean diameter dp = 170m with (a) Hydrophilic plate (b) Hydrophobic plate.

62

4.13 Impact and spreading process of a liquid marble coated with glassbeads (Glaco MCZ) with mean diameter dp = 170 m onto a hydrophilic surface.

62

4.14 Plots of normalised maximum spread versus impact weber number. The colors represent different particle diameters. Diamond symbols indicate impact on hydrophilic surface.

65

4.15 Normalized maximum spread diameter (Dmax/D0) versus the impact Weber number (We) for liquid marble impact on hydrophobic surface for all particles sizes used herein

65

4.16 Images taken from bottom showing impact onto hydrophobic glass plate of liquid marbles with particles

67


x

(Glaco MCZ coated) of mean diameter (a) dp = 204 m (b). dp = 66 m .

4.17 Images taken from bottom showing impact onto hydrophilic glass plate of liquid marbles with particles of mean diameter dp = 170 m coated with (a) Ultra ever dry (b) Glaco MCZ.

69

4.18 Plots of normalised maximum spread versus impact weber number for water droplets. Diamond symbols indicate impact on hydrophilic surface and circles indicates impact on hydrophobic surfaces.

70

4.19 Plots comparing the normalised maximum spread versus impact weber number of liquid marbles with all particles sizes used herein and water droplets. Diamond symbols indicate impact onto hydrophilic surface

71

4.20 Plots comparing the normalised maximum spread versus impact weber number of liquid marbles with all particles sizes used herein and water droplets. Circles indicates impact on hydrophobic surfaces

72

4.21 Comparison of impact and spreading process of a liquid marble coated with glass beads (Glaco MCZ and Ultra ever dry) with mean diameter dp = 170 m with a water droplet onto a hydrophilic surface.

73

4.22 Showing splashing behavior on impact onto hydrophilic and hydrophobic surfaces respectively. a) Hydrophilic surface- Prompt splash b) Hydrophobic surface- Receding break up c) Hydrophobic surface- drop rupture

76

4.23 Showing spreading and onset of splashing of a liquid marble coated with glass beads (Glaco MCZ) with mean diameter dp = 25 m onto a) hydrophilic surface b) hydrophobic surface.

76

4.24 Showing spreading and onset of splashing of a liquid marble coated with glass beads a) (Ultra ever dry) with

76


xi

mean diameter dp = 170 m b) (Glaco MCZ) with mean diameter dp = 25 m on impact with hydrophilic surface.

4.25 Comparing spreading and onset of splashing of a) water droplet of same diameter with b) a liquid marble coated with glass beads (Glaco MCZ) with mean diameter dp = 25 m on impact with a hydrophilic surface.

77

4.26 Comparing impact of liquid marble coated with glass beads (Glaco MCZ) with mean diameter a) dp = 25 m b) dp = 204 m onto a hydrophilic surface.

79

4.27 Comparing impact of liquid marble coated with glass beads (Glaco MCZ) with mean diameter a) dp = 25 m b) dp = 204 m onto a hydrophobic surface.

81


xii

LIST OF SYMBOLS g Acceleration due to gravity.

D0 Original diameter of the droplet before impact. Dmax Maximum Diameter of a drop during the spreading phase.

dp Powder particle diameter. Vi Impact velocity. Re Reynolds number. We Weber number. Bo Bond number. Oh Ohnesorge number. Viscosity of the fluid. Density of the fluid. κ Capillary length. Surface tension of the fluid. Ratio of the maximum diameter of a drop during the

spreading phase and original diameter of the droplet before impact.

Contact angle between the droplet and the powder. Packing fraction of the powder. Spreading coefficient. /SL Liquid spreading over solid coefficient.

S/L Solid spreading over liquid coefficient.

S Surface free energy of the solid state.

L Surface free energy of the liquid state.


xiii

SL Surface free energy between the solid and the liquid states. G Gibbs free energy change of an isolated system. m Micron.

GMCZ Glaco mirror coat zero. UED Ultra-ever dry


1

CHAPTER 1 OVERVIEW

1.1 Liquid marble formation Superhydrophobicity has been an exciting topic of research for many years and many

different techniques, typically incorporating both micro-texturing and vapor deposition, have been developed in order to render surfaces in such a state [1,2]. When water comes into contact with such a surface, it will exhibit a high contact angle and low degree of hysteresis. In particular, Water droplets resting on hydrophobic textured surfaces are typically in the Cassie–Baxter state [3] meaning that the bottom surface of the droplet is supported only by the tops of the surface asperities (pillars). This ultra-low contact area yields a very low-friction state whereby droplets can roll easily across a surface. A liquid marble, which is a liquid droplet encapsulated with solid particulate matter [4–7] essentially mimics this principle, whereby the marbles can roll and be transported in a very low-friction state [8–10]. In essence liquid marbles create a Cassie–Baxter state by having ‘‘pillars” (i.e. particles) embedded across the entire free-surface of the droplet. Given their mobility and robustness, liquid marbles have been studied and tested for a variety of potential applications, such as for gas sensing [11], synthesis of compounds/composites [13], blood typing and cell culture [12], and miniature chemical and biological reactors [10]. The mechanisms and conditions for liquid marble formation have been investigated previously [14–18] with the aim of quantifying the dependence on various parameters such as surface tension,viscosity, and droplet kinetic energy. Post-production, various phenomena such as evaporation [17], condensation [19] and freezing [20] have been observed. In all of these studies, the key parameters studied are shape and size of the individual particles encapsulating the liquid droplets. More generally, liquid marbles are a


2

specific class of particle laden interfaces, where solid particles lie at an interface [22] and induce interfacial mechanical properties beyond a simple surface tension [23]. In particular, particle-laden interfaces exhibit viscoelasticity and diminishing surface tension as they transition to a jammed state [25–28] and can form aggregates or ‘‘rafts” [24]. To date, only two studies [29,30] have reported jammed interfaces during liquid marble formations, which provides significant motivation for the present study; Given that liquid marbles are designed to be low friction transportable micro-reservoirs, their shape and mobility are key properties and understanding the transition towards a jammed interface is paramount. In the context of liquid marble formation, cratering [31,32,21,33–35] could be an important consideration as the compliance of the powder bed will influence the spreading dynamics and ultimately, the final coverage of powder on the surface of the drop. The spreading stage of a droplet during impact is typically quantified by the maximum spread, Dmax, (e.g. [36–42]), often scaled in terms of initial drop diameter, D0, and impact Weber number,

20 0D VWe .

For the inviscid case for impact on rigid solid surfaces, two primary scaling laws have been proposed: Firstly, where the kinetic energy at impact is completely transferred to surface energy at maximum deformation, whereby it can be derived that

1/2max 0D D We . Secondly, when considering that the impact itself induces an

acceleration term 20 0V D and assuming a roughly cylindrical shape at maximum

deformation, applying volume conservation leads to the scaling 1/4max 0D D We .

For powder surfaces, where the target surface can be deformed, the compliance of the powder bed must be considered. Katsuragi (2010, 2011) incorporated the density of the


3

bed to render the scaling, 1/40 bcD D We

, where Dc is the crater diameter (which is

equivalent to the maximum spread diameter) and b is the bulk density of the powder bed.

Marston et al. (2010) found that the maximum spread max 0D D scaled as We1/5 for the impact of low-surface tension or low-viscosity liquid drops onto dry, wettable powders for

10 1000We , which decreased to 1/10 1/5We for pre-wetted powders, but for hydrophobic powders, the scaling We2/5 was observed for 30 100We . Nefzaoui & Skurtys (2010) also observed a 1/5 scaling law exponent. A summary of these observed scaling laws is presented in Table 1. Thus from previous theoretical arguments and empirical observations, the normalized maximum spread during impact onto hydrophobic powder beds should be described by the Weber number raised to some exponent = 0.2–0.5. However, there is no widely accepted scaling law for the impact of liquid drops onto loose, hydrophobic powder surfaces. This is an important aspect as it is pertinent to the formation of liquid marbles. As such, one focal area in this work was to conduct an experimental campaign to provide empirical evidence for a scaling law. To achieve this, we have complemented, and extended, previous works by performing liquid marble formation experiments across a broad range of parameters, including the contact angle, , between the droplet and the particles, particle diameter, dp, initial droplet diameter, D0, and impact speed, V0. We have also examined grain mobility on the liquid droplet surface, which in turn affects the encapsulation of the liquid droplet with the particles. In doing so, we have expounded upon the threshold conditions for the formation of fully covered (spherical) and deformed (non-spherical) liquid marbles, originally reported in [29, 30].


4

Table 1.1. Summary of scaling for the normalized maximum spread diameter Dmax=D0 for different target surfaces. Target Surface Observed Scaling Reference

Dry solid (Inviscid) We1/4,We1/2 [36-38,42] Dry solid (Viscous) Re1/5Ꞙ(WeRe-2/5 ) [39,40] Superheated solid We2/5 [41]

Dry wettable powder We1/5 [21,43] Pre wetted powder We1/10-1/5 [44]

Dry hydrophobic powder We2/5 [29] 1.2 Liquid marble deformation

Many studies have been conducted on the impact of pure liquid droplets on the rigid surfaces because of its relevance both in nature and in industry. An important study of a drop impact dynamics on a solid substrate is the, maximum spreading , Dmax because this quantifies how much of a surface can be covered with a single droplet . Numerous relations between the maximum spreading and the impact parameters have been established. Most of the existing scaling laws are of the form, max max 0/D D and are based on two dimensionless parameters, weber number (We) and reynolds number (Re) based on the energy balance, in terms of kinetic energy and surface energy prior to impact and surface energy along with viscous dissipation at its maximum spreading. Recently more detailed models have been included which involves corrections due to the initial surface energy, dynamic contact angle, or viscous dissipation [53][69-74]. However among the existing models, the scaling law by Clanet et al. [37] provides good quantitative agreement with experimental data for impact on different substrates. This scaling law has commonly been used to fit experimental data for maximum spread of droplets on impact with rigid surfaces. Clanet et al. (2004) derived the scaling law, 1/4

max We considering that the drop during


5

its impact experienced an effective acceleration, 2 /o oU D as its velocity reduces to zero after hitting a solid .

However like pure liquid droplets very few studies concern the impact of particle laden and coated droplets. Planchette et al. (2011) investigate the properties of liquid marbles when impacting onto a solid substrate. They analyzed the threshold conditions for different behaviors during impact like non bouncing, bouncing and rupture. Thus in the second part of this work we have tried to determine the relation between the maximum spreading ratio, max and the impact parameters for the case of liquid marbles. For this we have performed the liquid marble impact experiments across a broad range of impact speed, V0 on substrates with different wettability. We have also compared the temporal evolution of liquid marbles upon impact on rigid surface to that of pure water droplets. The splashing threshold for pure liquid droplets is characterized by the so-called splashing parameter proposed by Stow & Hadfield (1981) and Mundo et al. (1995) defined as ReK We . For impacts on rigid surfaces 3000K typically indicates the onset of splashing. Here we have tried to investigate the threshold splashing parameter for the case of liquid marbles. Further, we have studied the difference in splashing pattern at two different surfaces of different wetting characteristics, one being hydrophilic and other hydrophobic. We also studied the fingering pattern and rupture dynamics of liquid marbles in comparison to the case of pure water droplets. The wettability of the substrate has a significant effect on rupture dynamics for pure water droplets [42]; we therefore investigated in this work how film rupture for liquid marbles depends on the wettability of the substrate.


6

CHAPTER 2 BACKGROUND AND MOTIVATION

2.1 Literature review Formation of Liquid Marbles

Liquid marbles can be produced by rolling of a liquid droplet over a bed of hydrophobic glass beads. However it is a crude method to produce full encapsulations presented in some previous works [58][18]. A more efficient way to produce liquid marbles is by a normal impact of a liquid drop on a bed of hydrophobic glass beads [14]. When a drop of aqueous liquid is placed on a bed of hydrophobic powder, the powder particles do not spontaneously spread over the drop surface. Instead, the liquid drop must be brought in contact with powder beads so that it can pick up powder particles to obtain a full powder coverage and form a liquid marble. From a thermodynamic perspective, whether liquids will spread on solid surfaces or not depends on the balance of interfacial energy coefficient,

/S L . The spreading coefficient, /S L is defined as / L SLS L s where, S , L is

surface free energy of the solid state and liquid state respectively and SL is the surface free

energy between the solid and the liquid states. The value of /S L can be further calculated considering the work of adhesion due to the polar and non-polar intermolecular interactions [58] as given in the following expression [58]:

/ 4 2d d P PS L S LS L sd d P P

S L S L

(1)

Thus we can write the thermodynamic equation for the final state when a liquid droplet spreads over a solid surface as:


7

/ L SLS L sG A A (2)

Where G is the gibbs free energy and A is the total surface area of the system. If

/S L >0 i.e. ( initial finalG >0) thermodynamically the liquid has a stronger tendency to increase its contact with the solid than to reduce its contact area with the solid. The liquid will therefore spread on the solid surface spontaneously. The spreading of a liquid over a solid surface bears thermodynamic similarity to the spreading of a liquid over another, but immiscible, liquid. The spreading coefficient, /S L however does not predict the behavior of the powder on the surface of the droplet. It can only predict how much of the droplet is coated by the powder. When a solid powder aggregate covers a liquid surface, the behaviour of the molecules in the powder particles is different from the behaviour of molecules of the liquid as they spread over a solid surface. For a hydrophobic powder, solid powder particles do not increase their surface area as they spread. Instead, powder aggregates merely disintegrate as they move towards the free liquid surface. The only change in this process is that a fraction of the solid powder surface becomes a solid/liquid interface. This critical difference between these two spreading processes need to be addressed in sufficient detail. The model of solid powder spreading over a liquid surface using spreading coefficient, /S L assumes the two spreading processes of liquid over solid and solid particles over liquid droplet are similar and can be described by similar physics. However Nguyen et al. [58] have shown that this spreading coefficient, /S L is an incorrect parameter for predicting powder spreading over liquid surfaces.

The ratio of the solid/liquid interface area and the liquid surface area it replaces is defined according to the following equation (see figure 2.1):


8

21 /h r (3)

Figure 2.1. A hydrophobic powder particle contacting a liquid surface, ref [58].

According to Nguyen et al. [58] the ratio, is almost always between 1 and 2, where β=1 represents the limit of non‐wetting and β=2 represents the Cassie‐Baxter effect. The total initial surface free energy of solid particles-liquid drop system is ( S S L LA A ), and the total surface free energy after the solid particles spread over fraction of the liquid droplet is (γL[1− Φ]AL + (1+β)Φ ALγLS + [AS−(1+β)ΦAL] γS). The free energy change of the powder spreading process over the liquid droplet will then be,

( )L A LS L S LSG A W A (4)

Where, ( )A LSW is the work of adhesion between liquid and solid defined in terms of thermodynamic spreading coefficient for a liquid spreading over a solid surface as

/ ( ) (L)L S A LS CW W . However equation (4) considering spreading coefficient of solid over liquid (λS/L) does not capture the actual physical process of powder spreading over a liquid phase and therefore we cannot describe the spreading process of solid particles over liquid droplet using the spreading coefficient, /S L . We rather need to investigate other factors to determine for whether or not an impacting or rolling droplet will lead to the successful formation of a liquid marble.


9

McEleney et al. [59], in another research study reported that the temperature gradient could be one of the major factors affecting the movement of the powder over the liquid surface. The assumption was that the top part of the drop is exposed to the light for experimental investigation while imaging, so the temperature at the top of the drop will be higher than that at the bottom. This temperature difference could drive the motion of the particles over the surface of the drop [59] due to marangoni effect. However his reasoning was not supported in any other publications. More extensive research is required on this mechanism. The formation of liquid marbles has been examined extensively by Hapgood and co‐workers [14, 15, 16, 17, and 18]. They have identified for a given fluid, the impact

kinetic energy, 21. 2 iK E mu , where m is the mass of the liquid drop and ui is the impact

speed, has a profound influence on the marble formation rather than the difference between works of cohesion and adhesion due to the spreading of powder all around the surface of the liquid drop denoted by the solid over liquid spreading coefficient, /S L as stated by Nguyen et al. [58] . Hapgood and co‐workers [14, 15, 16, 17, and 18] have developed a rule‐of‐thumb flow‐chart for determining whether or not an impact will lead to the successful formation of a liquid marble based on empirical data, which is reproduced here in figure[3]. These criteria give a starting point for selection of materials for investigating liquid marble formations. We tried to quantify all the factors which control the particle encapsulation over the liquid droplet which are as follows. Effect of droplet and particle size

The criterion for complete marble formation has already been reported in literature [18]. Eshtiaghi et al. [14] found that for complete particle coverage of an impinging liquid


10

droplet onto powder bed, 0 25 pD d where D0 is the droplet diameter and pd is particle diameter. Moreover the smaller sized beads covers the liquid droplets easily leading to complete encapsulation compared to the bigger beads for the same liquid droplet size. This is because as the particle sizes of the powder decrease, the inter‐particle attraction forces between those particles become stronger, and therefore the particles tend to move over the liquid surface with ease compared to bigger particles. However the particle size of the powder bed is not the only the deciding criterion. Other factors like the impact kinetic energy, viscosity and surface tension of the liquid droplet plays the key role in particle encapsulation over the liquid droplet as discussed below [14, 15, 16, 17, and 18]. Effect of kinetic energy

Eshtiaghi et.al [14] suggested that for liquid marble formation surface energy is not the key driving force rather it is impact kinetic energy of the falling liquid droplet. As the kinetic energy is increased, the percentage coverage of the liquid droplet by powder increases. He performed the experiments to examine the efficacy of the formation of marbles due to the impact alone, without manual manipulation or mechanical agitation and deduced the following equation for the percentage coverage of the liquid drop:

(%) 1 bECoverage A e (5) Where, A is the maximum extent of liquid marble coverage, b represents the ease of liquid marble formation (coverage/unit energy) and E, is the kinetic energy of impact. High values of b mean that only a small increment in kinetic energy is required to produce a considerable increase in liquid marble coverage. Effect of fluid viscosity on particle coverage


11

The increase in viscosity of the solution was found to significantly impair the deformation and recoil of the drop upon impact. This consequently gives lower coverage because so much of the kinetic energy is dissipated by viscous forces [14][43]. Thus viscous liquid droplet exhibit lower coverage compared to pure water droplets for the same sized particles. Effect of surface tension on particle coverage Upon impact the drops was deformed and flatten. The surface tension causes the drop to recoil back towards a spherical shape. However surface tension varies over a narrow range as compared to several orders of magnitude variations in fluid viscosity. This suggests that surface energy effects are not the main factor in determining the liquid marble powder coverage. In general fluids with higher viscosity and lower surface tension require higher kinetic energy input to produce the same liquid marble coverage[14][43].

Figure 2.2. Flow chart showing favoring condition for liquid marble formation [18].


12

Properties of Liquid Marbles Various research study has been done [5][7][19][20][24] to analyze the properties of liquid marbles pertaining to different applications which are described as follows:

1. Surface tension- Two experimental methods were proposed for measuring the effective surface tension of liquid marbles [47-49]- a) Capillary method, b) Wilhelmy plate method

Capillary method – It was based on the idea that powder particles on the liquid surface may significantly affect the radius of curvature. It captures the difference in height of the capillary rise from the marble and from the flat water surface. Equation based on marmur model as given by,

e

2 2 cos 0Rw w w wg h hr , was used to calculate the surface tension.

Where w is the liquid surface tension, θ is the liquid/capillary wall contact angle, and g are the liquid density and the gravitational constant, and Re and hw are the equilibrium values of the radius of the spherical liquid drop and the height of capillary rise, respectively.

Wilhelmy plate method - Here it was assumed that surface tension of a flat surface covered by a layer of inert hydrophobic powder should be same as that of water marble covered with the same powder. It is based on the force experienced by the plate when it is advanced or receded on a flat water surface covered by hydrophobic powder. Based on the measurement from the above two methods following observations were made regarding the properties of liquid marbles:


13

{1}No appreciable change in surface tension was observed for liquid surface covered with particles compared to bare liquid surface {2}The magnitude of surface tension for smaller particles is higher than the bigger size particles; same trend was observed for both the methods. {3}Capillary method is more accurate because use of coverslip in the wilhelmy plate method has disturbed the powder/water interface. {4}comparison of the surface tension values of liquid marbles with pure liquid suggests that comparing surface free energy values of the liquid and solid phases (through λSL) is not a valid predictive indicator for solid powder spreading over liquid surface.

2. Electro-wetting of liquid marbles- The apparent contact angle of the liquid marble decreases on application of electrical voltage and it further increases on reducing the magnitude of voltage [60]. This phenomena is reversible i.e there is no contact angle hysteresis in the case of liquid marble unlike the the case of electrowetting of sessile liquid droplet. This is because when a marble is placed on a flat solid surface and electrowetting performed it spreads but with the water remaining effectively encapsulated by the grains. It behaves like the system were a droplet of water on a surface consisting of solid posts. This is same to the technique where roughness or structure is created to make a super-hydrophobic surface which can reduce contact angle hysteresis (Cassie Baxter state). Here the solid surface structure providing the super hydrophobicity has been attached to the liquid rather than to the substrate. Thus this electrowetting property of liquid marble with no contact angle hysteresis suggests one potential of liquid marble. The liquid marble may provide one method


14

to overcome the limitation of contact angle hysteresis in droplet electrowetting, and indeed for lab on- a-chip applications, controllable movement would need to be complemented by development of methods to dispense, coalesce, separate, and mix liquid marbles and extract the liquid from the marble. If such control could be achieved, the liquid marble concept might thus be able to provide a route to reducing the surface fouling effects experienced in the electrowetting of biological fluids [60].

3. Evaporation rate of liquid marbles- Evaporation rates of PTFE marbles were compared with the rates of pure water droplets in terms of evaporation resistance, parameter in reference[60]. They found that PTFE marbles have longer life-time than water droplets. This is due to screening of water molecules to atmosphere by PTFE molecules. It was found that the presence of PTFE micro-powder on the liquid marble reduced the evaporation rate of pure water from 25 to 45% and the calculated evaporation resistance, values varied from 0.365 to 0.627 with the increase of RH. Further in another research study Dandan et al. [61] have shown that Graphite liquid marbles have longer lifetimes than the pure water droplets. This increased life-time of liquid marbles in terms of evaporation rate suggests many promising applications of liquid marbles in microfluidics, genetic analysis, antifouling, wear-free micro-machine, electromechanical actuator and valve fields under normal atmospheric conditions. However pertaining to evaporation of the liquid marbles, the significance of the particle size distribution on the evaporation rate and life-time of the liquid marble deserves further work.


15

4. Freezing and condensation rate of liquid marbles- The dynamics of freezing of liquid marble is remarkably different from that of freezing of a pure liquid droplets. Unlike frozen liquid droplets which have pointed tip, liquid marbles acquire a flying-saucer shaped morphology upon freezing. This is because in liquid marbles the particle coating acts as an icing-inhibitor by providing thermal energy to the adjacent liquid molecules. Thus the nucleation preferentially takes place on the inside of the marble near the substrate. Due to the nucleation there exists a gap the freezing front and the coating. Owing to temperature gradients, a surface tension driven flow also known as marangoni flow exists. The direction of marangoni flow is from the marble top to marble sides around the inner ice core. This flow helps transport liquid away from the top of the marble, and in turn contributes to the absence of pointy peaks in ice-marbles. Figure 4. Below shows the cross-section of the liquid marble with the proposed gap between the freezing front and the coating due to preferential nucleation reference Hashmi et al. (2012) [20]. Unlike liquid marbles, there are no preferential nucleation sites and no gap exists between the water-air interface and the freezing front in case of freezing water droplets. This results in the absence of a sideways convection in a liquid droplet. As a consequence, water in the vicinity is only able to see the freezing front and freezes which results in an upward expansion of the water droplet that finally converges to a pointy peak. The speed at which the freezing front propagates in case of a freezing water droplet is observed to be relatively higher than that in liquid marbles. Further the experimental observations of freezing patterns point to the fact that gravity has


16

a negligible effect on the freezing process of liquid marbles ref Hashmi et al. (2012) [20].

Figure 2.3. Showing the gap between the freezing front and the particle coating due to nucleation inside a liquid marble. The arrow indicates the flow of water to the gap due to Marangoni effects [20].

5. Viscosity of liquid marbles- The interfacial rheology aspect of liquid marble

formation has not been addressed to date. The vast majority of work examining liquid marble deformation has relied on the concept of an effective surface tension to describe how the presence of a particle-laden interface affects deformation behavior [47-49]. However, particle-laden interfaces cannot be described by surface tension alone; in fact, the surface tension of these interfaces is typically identical to the surface tension of the liquid until the interfaces become fully jammed. The study of rheological properties of liquid marbles will give a new insight for the liquid marble formation and deformation dynamics. Thus extensive research is required in this field.

Deformation dynamics of Liquid Marbles Impact of pure liquid droplets on solid surface- Several research has been done

until now on the impact of pure liquid droplets on the rigid surfaces. Studies of drop impacts are driven by several areas relevant both in nature and in industry. In nature, a drop of water hollows a stone, and splashing produces aerosols, causes erosion, and brings the


17

smell of the earth during rain ( Joung & Buie 2015). In industry, there are spray/wall interactions in coating, cleaning, cooling, and combustion. Inkjet droplets play an increasing role in fabrication, from the soldering of electronics to microarrays in biotechnology.

An important study of a drop impact dynamics on a solid substrate is the normalized ratio, maximum spreading ratio max max oD D ,where Dmax is the maximum spread dimeter of the liquid droplet after the impact and Do is the intial droplet diameter before impact. The knowledge of maximum spreading ratio, max has important applications in industry. The maximum spreading ratio is directly related with the performance of spray systems in industrial processes[?] or in water transport from a rain droplet to soil in nature. Also in inkjet printing (Minemawari et al. 2011) and forensic science (Hulse-Smith et al. 2005, Attinger et al. 2013) the study of maximum spreading ration is an important parameter of interest. Thus controlling or predicting maximum spreading is essential for many problems involving the deposition of an impacting drop. However the more crucial quantity is the residual spreading radius, which differ from the maximum spreading radius. Numerous relations between the maximum spreading ratio, max and the impact parameters have been established. A large number of parameters, such as drop size, impact velocity, liquid properties (density, viscosity and surface tension), surface roughness, and wettability, play a role in the maximal spreading achieved by a droplet after impact. Most of the existing models for calculation of maximum spreading ratio, max are based on two dimensionless parameters, Weber number (We) and Reynolds number (Re). The Weber number ( 2

i o lvWe V D ) is the ratio between the kinetic and capillary energy, Ek/Eγ, and


18

Reynolds number (Re = ρViD0/μ) is the ratio between the kinetic and viscous energy, Ek/Eμ, where ρ is the liquid droplet density, γLV is the surface tension, μ is the viscosity, and Vi is the impact velocity. Theoretically, the expression of maximum spreading ratio, max has commonly been estimated based on the balance between inertia and viscous and capillary contributions (Chandra & Avedisian 1991 [67], Pasandideh-Fard et al. 1996 [68], Clanet et al. 2004 [37], Ukiwe & Kwok 2005 [69], Roisman 2009 [70-71], Eggers et al. 2010 [72]).The most classical proposition to calculate the maximum spreading ratio, max was made by (Chandra & Avedisian 1991 [67]). They proposed that the kinetic energy of the impinging drop ( 3 2

o oD U ) is dissipated by viscosity during the impact (which scales as 3max( / )oU h D , where h is the thickness of the drop at maximum spread after impact). Now

applying volume conservation they [67] found that the maximum spreading ratio scales as 1/5

max Re . The maximum diameter thus increases as U1/5. In another research study [73] a simple scaling analysis comparing the initial kinetic energy with the surface energy at the maximum spreading radius. This yields the relation the maximum spreading ratio as ,

1/2max We . Clanet et al. (2004) [37] developed an alternative expression. They

considered that the drop during its impact experienced an effective acceleration, 2 /o oU D as its velocity after hitting a solid decreases from U0 to 0. Using a mass conservation argument based on the pancake thickness created by the impact leads to the scaling law, 1/4

max We

. However the available scaling laws involving weber numbers is valid in the regime in which there is no splash. A more detailed models have been included which involves corrections due to the initial surface energy, contact angle, or viscous dissipation. Table 1 presents the most commonly used formulas in the literature to determine the maximum


19

spreading ratio, max of droplet on impact with a rigid surface. However the scaling law by Clanet et al. [37] has had great success as it gives good quantitative agreement with experimental data for hydrophobic substrates.

At the maximum spreading state, the liquid flow changes its direction and recoils inward due to surface energy. The amount of retraction depends on several factors including the initial kinetic energy of the impacting drop, the surface energy of the liquid, and the interaction energy between the liquid and the surface. In some cases, the liquid will retract to the equilibrium position and stop. In other cases, the liquid will retract beyond the equilibrium position and rise in the region of the initial impact. Under appropriate conditions, rebounding will occur where the liquid separates from the surface, rises a short distance and returns to the surface. Figure 2.4 shows the spreading process of a water droplet on impact with a solid surface as a function of time.

Figure 2.4. Impact and spreading process of a pure liquid droplet onto a solid surface(a) before impact, (b) maximum spreading (c) maximum retract/rebound, and (d) equilibrium.


20

Table 2.1. Different models for the spreading factor βmax in literature as a function of the impact parameters ref [57].

On impact after initial maximum spread, the drop exhibits different behaviors

depending on the impact and substrate parameters, ranging from smooth spreading to splashing, fingering, and rebound. This knowledge of spreading dynamics is important in many applications in industry. For instance, for inkjet printing, the spreading- splashing transition is crucial for the printing quality.

When a single drop impacts a surface, the phenomena of fingering and splashing can occur as shown in figure 2.5. Fingering is a quasi-periodic unevenness of the boundary of the lamella, the fingers which break during splashing are referred to as secondary droplets. Whereas splashing is a breaking of such fingers from the lamella into the air or along the surface. As the impact velocity, Vi, is increased the drop will splash (i.e., break up and eject smaller pieces). Classically, the splashing threshold was


21

characterized by the so-called splashing parameter proposed by Stow & Hadfield (1981) [74] and Mundo et al. (1995) [75], which incorporated the inertia, viscous stress, and surface tension given as ReK We . For impacts on rigid surfaces at splashing parameter, 3000K , one can expect a splash. Morever extensive research is presently going on to find the critical values of K for various impact conditions (e.g., Roisman et al. 2015 [76]).

Figure 2.5. Showing a) Fingering b) Splashing pattern on drop impacts with rigid surfaces ref. [75]. Allen (1975) [77] suggested that fingering occurs due to the Rayleigh-Taylor instability of the decelerating edge of the lamella. This was supported by recent linear stability analysis of a liquid rim (Roisman et al. 2006 [78]). Marmanis & Thoroddsen (1996) [79] demonstrated that the number of fingers N scales with a modified impact

Reynolds number as, 3/41/2 1/4ReDN We , where ReD is the modified Reynolds number

for impact given as ReDUD with U as the impact velocity, D the drop diameter and

the kimematic viscosity of the liquid in the drop. The splashing pattern obviously depends on the surface properties. The wettability of the substrate is crucial for the splashing and breakup of the edge. Figure 6 shows the difference in splashing pattern at two different


22

surfaces of different wetting characteristics. One surface is hydrophilic (water loving) and other hydrophobic (water hating) in figure 2.6.

Figure 2.6. Splashing pattern of drop impact on two different patterned surfaces a) Hydrophilic b) Hydrophobic ref [42]. Rupture of thin liquid films: Rupture in radially spreading liquid films occurred through the formation and growth of holes inside water films thereby making them unstable. Kheshgi and Scriven [80] pointed out that in order to nucleate a hole in a macroscopic film (thickness ~ 100 μm) it has to be first thinned locally by some external disturbance, such as gravity or capillarity-induced flow, and surface asperities, to a thickness (~ 1 μm) where long range molecular forces become active and rupture the film. Thus to render droplet film thin enough for rupture it should have high impact velocity so


23

that resulting radially spreading film have low thickness sufficient for rupture. Changing the wettability of the substrate changed impact dynamics significantly. Therefore film rupture depends on the wettability of the substrate. Increasing surface roughness greatly promotes film rupture compared to a smooth surface.

Figure 2.7. Showing the rupture of a thin water film (thickness 25μm) produced by the impact of a 550 μm water droplet on a mirror-polished stainless steel surface at 40 m/s (a) Formation of holes, (b) growth of holes ref [80].

Impact of liquid marbles on solid surface – Not much study has been done yet on impact of liquid marbles on solid surfaces. Planchette et al. (2011) [54] investigate the properties of liquid marbles when impacting onto a solid substrate. They analyzed different behaviors during impact non bouncing, bouncing and rupture. The critical conditions for transitions between the three regimes were then proposed in this study ref [54]. Theoretically they derived a expression for transitional velocity of no bouncing to bouncing for liquid marbles on impact onto rigid surfaces. For this the average of translation oscillatory kinetic energy of the liquid marble before impact was equated to gravitational potential energy to lift the liquid marble by a certain distance, . Thus a critical weber

number, 41 46 c

DWe l was established for transition from no bouncing to bouncing. Impacts

with We> We* were characterized by bouncing, while if We<We* there is no bouncing.


24

They observed experimentally the transition velocity for no bouncing to bouncing increases with the droplet diameter D. Morever varying the particle mean diameter of a factor five from 32 μm to 159 μm does not significantly affect this transition. Morever in this study it was suggested that the maximum spreading diameter, Dmax for impact of liquid marble onto dry solid is given by

maxD D WeD (6)

Where 0.12< a < 0.33, For We>100, a=0.12 and We<1, a=0.23. No other research study has been done yet to study the impact of liquid marbles on solid surfaces which support the results by Planchette et al. (2011) [54]. Thus there is a need of extensive research in this area to study scaling law for maximum spreading ratio, max and other impact dynamics like splashing, fingering and rupture mechanism for liquid marbles just like the case of pure liquid droplets.

Elasticity of liquid marbles- Most of the applications of liquid marble concerns on its stability and mobility. Researches have been done study regarding elasticity of liquid marbles. Asher et al. (2015) [81] have shown that liquid marbles can sustain reversible elastic compression of upto 30% of their original size. The elasticity of the liquid marble is mainly due to the liquid meniscus formed between the coating particles. At lower deformation the spring constant is identical to that of a pure water droplet of the same size. At higher deformation the spring constant increases until the liquid marble bursts. The higher spring constant at higher deformation is due to additional capillary attraction forces which develops across the emerging cracks in the particle coating. The whole stress–strain curve of liquid marble is very similar to that for microcapsules filled with liquid. Thus it


25

can be concluded that the elasticity of liquid marbles under deformation within the linear elastic regime where spring constant is same as pure water droplet is mainly determined by the surface tension of the liquid and hence by the nature of liquid surface (interface). Application of Liquid Marbles

Liquid marbles are considered as a perfect non-wetting system. They behave as micro-reservoirs of liquid capable with high mobility without any leakage, because the hydrophobic particles on the liquid surface form nonstick droplet/substrate interfaces to reduce motion resistance. The encapsulating particles on the liquid surface reduce evaporation of the encapsulated liquids [61][62]. These desirable characteristics make liquid marbles ideal platforms for various applications as stated below: Liquid Marbles as sensors-

The stability of the particle layer in liquid marble can be used as a tool for sensing applications. Bormashenko et al. (2009)[82] used this principle to reveal water surface pollution. He used PVDF (size ~130 nm) water-based liquid marble and placed it on a bath of water. Marbles floated stably on the water surface, whereas they were destroyed on the water surfaces contaminated with silicon oil and kerosene. Thus liquid marble can give a simple visual indication of pollution by oils, solvents, petroleum, etc [82]. Dupin et al. designed liquid marbles using latex powders whose wettability switched from hydrophobic to hydrophilic as pH changed from alkaline to acidic. They float over water bath when it was at neutral or alkaline, but would immediately disintegrates on addition of acid [83]. In another study [84] liquid marbles were investigated for applications in biological assays like human blood typing. Here liquid marbles were used as sensors which changes color based on the pattern of haemagglutination reactions. Thus from resulting color pattern


26

blood sample were identified as being A+, A−, B+, AB+, O+ and O−.The liquid marble skin can be thin and gas permeable based on particle coating. Thus a liquid marble containing an indicator solution placed in contact with a gas source can be an effective sensor. Tian et al. [11] used this concept to show sensing of ammonia and hydrochloric gases. Liquid marble in microfluidic application-

Aussillous and Quere observed that liquid marbles could be moved under electrostatic or magnetic forces [4][5]. They found that placing a liquid marble in an electrostatic field through charged stick of teflon or by placing it between the plates of a capacitor, can create motion [4][5]. Liquid marbles can also be remotely controlled using controlled electric field or magnetic field for many potential applications. For example Elliott and Newton et al demonstrated that controlled motion in liquid marble can be achieved using a controlled field, such as produced by individually addressable strip electrodes [85].The liquid marble might also be actuated via the dielectropheretic response in an inhomogeneous electric field of particles in the skin of the liquid marble or via the core liquid [8].Bormashenko et al [86] created PVDF liquid marble with Fe2O3 nanoparticles in water to create a ferrofluidic marble which can be controlled by magnetic field. Zhao et al. [9] also created liquid marbles using hydrophobic Fe3O4 nanoparticles. They observed that this nanoparticles could be cleared using a magnet from the upper surface thus exposing the liquid [9] and when the magnet was removed the nanoparticle layer again re-covered in a similar manner to the self-coating process. Thus this study has shown that liquid marbles can be opened for the addition of other liquids, etc, and can be


27

closed again by iteslf. This controlled opening and closing suggests various applications of liquid marbles in reactors etc. In Microreactors- Recently the application of liquid marbles as miniature reactors has been explored extensively for its unique properties. Liquid marbles requires use of relatively small amounts of chemical reagents and solvents. They provide much confined micro-environment for reactions and can be precisely controllable. In this area Miao et al. (2014) [87] fabricated an Ag nanoparticle-based catalytic liquid marbles for heterogeneous degradation of methylene blue. The catalytic efficiency for the reaction was close to 100%.Thus they demonstrated that particles adsorbed at the air–liquid interface on liquid marbles could act as catalysts to catalyze the chemical reaction. Xue et al. [10] showed that liquid marbles coated with fluorinated decyl polyhedral oligomeric silsesquioxane and magnetic Fe3O4 powder can be used as remotely-controllable chemical miniature reactors for chemiluminescence reactions. However, in most of the current applications of liquid marbles as miniature reactors were limited because the encapsulating shells only act as inert isolating layers to provide a confined environment and do not itself participate in the reactions. Sheng et al. (2015) [88] fabricated a silica based liquid marbles where the encapsulating particle shells not only acts as inert isolating layers to provide a confined environment, but it also provides the reactive substrate surfaces in regulating the classical silver mirror reaction. Commercial Applications

Most of the present research is concerned on production, properties and application of single macroscopic liquid marbles. However most commercial applications consists of


28

liquid as continuous phase consisting of many small particle-encapsulated liquid droplets formed by mixing liquid with powder. The first patent on this dates back to 1968 which describes the use of hydrophobic fumed silica. In a patent dated 1976 a mixture using hydrophobic silicon dioxide and water was used to create drilling fluid composition. The encapsulation of volatile organic liquids for controlling the release of vapors was a patent dated 1984 .The recent patents includes the possible applications of hydrophobic fumed silica based liquid marble compositions for cosmetic (antiperspirants, skin-conditioning, moisturizing, etc) and pharmaceutical applications (delivery of active compounds to the skin) and hydrophobic nanoparticle encapsulations of detoxification reagents for chemical and biological agents. 2.2 Research Objectives The objectives of this research are:

Systematic study of the impact dynamics of a liquid drop onto a hydrophobic powder bed, which constitutes the first step in the formation of a liquid marble.

To assess the critical conditions for the formation of liquid marbles, where the drop becomes completely encapsulated by the powder, and arrested shapes where the drop cannot regain its spherical shape by varying both the size of the drop and impact speed coupled with the size range of the powder.

To investigate how a lower particle mobility can promote liquid marble formation at lower impact kinetic energies by using different hydrophobization agents.

Systematic study of impact of liquid marbles on rigid surfaces to determine the relation between the maximum spreading ratio, max and the impact parameters.


29

To compare the spreading process for liquid marbles upon impact on rigid surface to that pure water droplets.

To study the splashing and fingering pattern for liquid marbles on surfaces with different wetting characteristics.

To investigate the rupture dynamics of liquid marbles on normal impingement with a rigid surfaces of different wettability.

2.3 Organization of thesis Chapter 1 provides an in‐depth introduction of the topic of the research with reference to previous related work. Chapter 2 dealt with the background and motivation for the present research as well as identified existing studies done on the subject and mentioned their important conclusions. Chapter 3 describes the experimental methodology used. It first describes in detail the apparatus that was designed and built to investigate the formation of liquid marbles on normal impact of water droplets on a powder bed surface. It then depicts the apparatus that was used to study the normal impact of liquid marbles on solid surfaces. This is followed by a detailed description of the different glass beads used in the present research along with explanation of the manner in which different hydrophobization treatment were done. The chapter 4 discusses the experimental observation of the present research, both for the liquid marble formation study as well as for the liquid marble deformation study. Chapter 5 presents the key conclusions of this research work, and lists recommendations for future work in this area. At the end, references, and appendices containing important derivations and calculations are given.


30

CHAPTER 3 EXPERIMENTAL METHODOLOGY

3.1 Liquid marble formation experiment set up The experimental setup, as shown schematically in Fig.3.1. , consists of a small

container filled with fine glass beads. We release a drop of pure water from heights 7.5rh cm up to 40 cm using hydrophobic glass capillaries to achieve highly repeatable

droplet sizes, D0 = 0.8–3 mm. The falling drops thus impact vertically with speed 2o rV gh which varies from 0.61 to 2.6 m/s in our case. The impact, spreading and

rebound of the impinging droplets are all captured in a single video sequence using a high-speed video camera (Phantom V711, Vision Research Inc.) equipped with a Nikon 60 mm micro-lens, which yielded a range of effective pixel sizes of 29–41 m /px. Frames rates of up to 10,000 fps were used.

Fig. 3.1. Schematic of the experimental setup used.


31

3.1 Liquid marbles impact experiment set up

Figure 3.2. The experimental set up for liquid marbles impact study The experimental setup for the liquid marbles deformation study (refer Figure 3.2.)

consists of a glass plate (coated/uncoated) at the base onto which we release the liquid marble through an inclined channel from varying heights. The falling liquid marble thus impact vertically with speed 2i iU gh varying from (0.79-2.34) m/s onto the plate. The liquid marble impact, spreading and rupture are all captured both from bottom and side view in two synchronized video sequences by two high-speed video cameras, Phantom V711, Vision Research Inc., and Phantom M310. The phantom V711 is placed at the side and Phantom M310, at the bottom of the rigid glass plate as shown in figure 10. The effective pixel sizes of bottom camera is 17.5 μm/px and that of side camera is 20 μm/px. Frames rates of 5,000 fps were used in both the cameras. The detailed schematic with a closed view of all the parts used in the experiment is shown in figure 3.3.


32

Figure 3.3. The schematic view of the experimental set up for liquid marbles impact study 3.3 Hydrophobizing agent and powder treatment method

For hydrophobization two commercial agents were used – namely – Glaco mirror coat ‘zero’ (Soft 99 Co.) and Ultra every dry (UltraTech International Inc.). The former is an alcohol-based suspension of silica nanoparticles, which form micron and sub-micron sized roughness elements when heat-cured [29]. Traditionally, glaco mirror coat ‘zero’ is used in aerosol‐form, which can be sprayed onto car wing mirrors to repel water droplets. However, we had the product in a pure solution form, which we use to yield hydrophobic powders. The Ultra ever dry is a two layer coating. The bottom layer binds to the surface where the coating is applied, and another layer binds to the base layer and has an exposed layer that displays superhydrophobicity and oleophobicity, due to the formation of nano-pillar structures. To prepare the beads, we first washed them by immersing in a water–ethanol mixture and placing in an ultrasonic cleaner for at least 15 min. The beads are then


33

immersed in the hydrophobizing agent and sonicated again for a further 15 min before drying for several hours on a hot plate. 3.4 Particle sizing and shape determination

Particle sizing was performed on an API Aerosizer (TSI Inc.), where the particle diameter is calculated based on the time of flight of the particles. The particle size distributions allowed us to extract the mean size, d50, as well as corresponding upper and lower percentiles d10 and d90. The particle shape was characterized by a circularity parameter based on two-dimensional projections using microscopy imaging, and subsequent image analysis, an example of which is shown in Fig.3.4. The expression for circularity is given by, 24 /C A P , where A is the projected area and P is the perimeter of the glass beads. In all cases we found C > 0.9, indicating a high degree of sphericity.

Fig.3.4. Microscope image (left) and binarized image (right) of particles ( 488pd m ) used for measurement of circularity. The scale bar is 2 mm long. 3.5 Contact angle measurement

The contact angle was measured using the sessile drop method, which was critically evaluated previously by [46]. The glass beads were firmly attached to a microscope slide using double-sided adhesive tape. A water droplet is then gently placed over the glass slide. Using a Nikon D90 camera equipped with a microscope objective, the apparent contact angle between the water droplet and the beads attached on the glass-slide can be estimated,


34

as shown in images in Fig.3.5. The contact angle of a water droplet placed on the adhesive tape is 57o, which increased to between 96o and 150o when the hydrophobized glass beads were attached.

Fig.3.5. Images of water drops on (a) adhesive tape, (b) glass beads with 148pd m , and (c) glass beads with 488pd m used to derive contact angles. The scale bar is 2 mm long.

3.6 Powder properties

The primary particles used for liquid marbles preparation for both the study were glass beads (Potters Industries Inc.), with a total particle diameter range of 25 500pd m . A summary of the powder properties is given in Table 3.1. Table 3.1: Summary of the properties of the particles used. Contact angles stated are based on the sessile drop technique.

Coating Mean Size(micron) d10 (micron) d90 (micron) Contact angle Glaco MCZ 25 8 56 138 ± 4 Glaco MCZ 62 39 103 100 ± 5 Glaco MCZ 118 62 186 147 ± 5 Glaco MCZ 148 81 235 102 ± 2 Glaco MCZ 170 110 269 115 ± 4 Glaco MCZ 204 107 309 135 ± 6 Glaco MCZ 488 442 523 150 ± 1

Ultra-Ever Dry 170 110 269 104 ± 3


35

3.7 Video capture and analysis The analysis was mainly done manually using Photron Fastcam Viewer (PFV)

Software or through image processing tools in Matlab. In Matlab we wrote a script using image processing tools for determine the spreading of water droplets/liquid marbles on impact. The algorithm consisted of two parts: (i) determine initial drop diameter, Do, and impact velocity, Vo. (ii) determine spread diameter as a function of impact time and maximum spread diameter (Dmax). The inputs for the code included the file name, effective pixel size, frame rate and frame number where the drop begins to impact the surface. The initial parameters, initial drop diameter, Do, and impact velocity, Vo were computed by processing the image so that only a white drop with a black background appears, as in Fig.3.6. The area and centroid of the drop are found and used to calculate the diameter and displacement of the drop as it moves through the first 10 frames. Just prior to impact the area was used for calculation of drop initial diameter.

Figure 3.6. Processed image for drop impact study. In the image the centroid of the drop is marked as green.


36

The spreading values were a bit more difficult to obtain. The extrema points of the drops/liquid marbles were extracted and the spreading diameter, Dspread, was taken to be the difference between the left-top and right-top values as shown in figure 3.7. For the liquid marble impact experiment the maximum spread was however calculated from the bottom view images. The images in the side view were not clear as powder was flying from the liquid film in all directions which was making it difficult to identify the edge of the film at the instant of the maximum spread. The matlab code for the video analysis is included in the appendix.

Figure 3.7. Processed image for a liquid marble (coated with 204 m particles) on impact with a hydrophilic plate. In the image the extrema points are marked as green.

3.8 Instrumentation 3.8.1 High‐Speed Camera

For liquid marble formation experiment a Phantom V711, Vision Research Inc., high‐speed camera was used that can record up to one million frames per second. However, most of the video sequences for these experiments are recorded at 10,000 frames per second with an effective pixel sizes of 29–41 m /px. However the liquid marble impact, spreading

Dmax


37

and rupture are all captured both from bottom and side view in two synced video sequences by two high-speed video cameras. Phantom V711, Vision Research Inc., was placed at the side and Phantom M310, Vision Research Inc., at the bottom of the rigid glass plate. The effective pixel sizes of side camera is 20 μm/px and that of bottom camera is 17.5 μm/px. Frames rates of 5,000 fps were used in both the cameras. The total video sequence in both the experiments is cropped to the frames of interest, which means several frames before impact until the drop/marble comes to complete rest. Note that the recording duration was sufficient to allow for auto triggering. From the video sequences, it was then possible to extract basic data such as the drop size, drop impact speed, maximum spread diameter, and the times associated with these features. The analysis was then done manually using Photron Fastcam Viewer (PFV) Software or through image processing tools in Matlab. 3.8.2 Objective Lens

The lenses used in the experiments were either a Nikon 60 mm micro-lens, or a Nikon 50 mm micro-lens, fitted to the high speed cameras directly or via extension mount for enlarged view of the image. The spatial resolutions i.e. highest lp/mm grid lines that we can resolve using Nikon 60mm micro lens is 8 lp/mm whereas for Nikon 50 mm microlens it is 1.6 lp/mm . The lenses have variable aperture settings, with maximum aperture as f/2.8 and minimum of f/36. The minimum focal distance for Nikon 60 mm lens is 0.219m and that of 50 mm lens is 0.45m. The detailed specifications of these lenses can be available online. 3.8.3 Light sources

Due to the short exposure times used to capture the events at high‐speed, we typically require a high‐intensity light source. As such, for the liquid marble impact


38

experiment we employed LED science LS-S6-20 series 6 LED light with 20 degree medium accubeam . One light was placed directly opposite the side camera, with a diffuser screen infront, whilst the other was directed down on top of the glass plate, to have a clear view from the bottom camera for this experiment. For the liquid marble formation since there was only one camera for the side view we placed the light source directly opposite the camera with a diffuser screen in front of it. Figure 3.8 shows the images of the light sources used in the two experiments.

Figure 3.8. Images of light source directed towards the surface of the impact. A) For the side view camera B) For the bottom camera

A) B)


39

CHAPTER 4 RESULTS AND DISCUSSION

4.1 Overview of impact dynamics for liquid marble formation Dimensionless groups for drop impact

The main dimensionless parameter used to characterize the impact is the Weber number, We 2

o oWe D V , whilst the deformation due to gravity can be expressed by

the Bond number, Bo 2 /oBo gD and the role of viscosity can be evaluated by the Ohnesorge number, Oh oOh D . Note that the largest drop size of approximately

3 mm exceeds the capillary length for water g =2.7 mm, however, given that all Bond numbers herein are (0.1) (1)Bo O O , and that 1We O , we assume the impact dynamics to be inertia dominated and thus we neglect gravitational deformation. Furthermore, in the absence of viscous forces and also 6(10 )Oh O , the shapes during impact are determined by the competition between inertia and surface tension, i.e. the Weber number. Therefore, we expect a priori that impact kinetic energy plays a dominant role in the formation of liquid marbles.

Previously it was shown [18] that the impinging droplet survives impact without shattering onto the powder bed if We < 1000 and Oh > 0.05. For our experiments, we used pure water with 0.001 Pa s, whereby 0.05Oh for all cases as indicated in Table 4. In this regime, the droplets survived impact only for We < 100. This critical weber number holds for the range of the liquid droplet diameters in our experiment for (a) 1oD mm,

(b) 2oD mm and (c) 3oD mm. The role of the packing fraction of the bed on droplet


40

shattering has already been investigated by[21]. However the effect of other key factors on droplet splashing like the nature of coating of the glass beads and the ratio o pD d has yet

to be studied. As shown in Table 4, for 1oD mm the value of Bond number is around 0.13 which would imply a spherical shape. Still in some cases, as shown herein we found that the shape of the liquid marble formed is not spherical. This naturally poses the question: what determines the final shape of the liquid marble formed if not the Bond number (i.e. surface tension) of the impinging liquid droplets? Table 4.1: Weber, Bond and Ohnesorge numbers for different drop diameters ( 1oD , 2 and 3 mm) as they impact the powder bed for speed V0 = 0.6–2.6 m/s.

Droplet Diameter (mm) Weber Number (-) Bond Number (-) Ohnesorge Number (-) 1 5.1-92.6 0.13 43.2 10 2 10.2-185.2 0.54 42.2 10 3 15.3-277.8 1.21 41.8 10

Overview of impact dynamics Figs. 4.1 and 4.2 show images from high-speed video sequences of the impact of

water droplets of diameters 1oD , 2 and 3 mm onto powder beds of hydrophobic glass beads with 25pd m for a range of impact Weber numbers and Bond numbers (see captions for details). In Fig. 17, the droplets are fully encapsulated and all regain their spherical shape after the initial impact with the powder bed, i.e. they form liquid marbles. In contrast, Fig. 4.2 shows the same size droplets impacting at higher speed whereby the surface of the droplets become jammed with particles and do not regain a spherical shape. This observation of deformed marbles was first reported by [29,30] and it is precisely this stark difference in the resulting form of the liquid marble that is a key motivation for this


41

study. As such, we have discussed the impact dynamics and different observed features pertaining to liquid marble formation in detail in the next subsections.


42

Fig. 4.1. Image sequences showing the formation of fully encapsulated spherical marbles during the impact of water droplets onto hydrophobic glass beads with 25pd m . (a) 1oD mm, V0 =1.3 m/s and We = 23; (b) D0 = 2 mm, V0 = 1.3 m/s, and We = 46; (c) D0 = 3 mm, V0 = 1.1 m/s and We = 49. The scale bars are all 2 mm long.

Fig. 4.2. Image sequences showing the formation of deformed marbles with jammed interfaces during the impact of water droplets onto hydrophobic glass beads with 25pd m . (a) 1oD


43

mm, V0 =2.6 m/s and We = 152; (b) D0 = 2 mm, V0 = 1.9 m/s, and We = 101; (c) D0 = 3 mm, V0 = 1.42 m/s and We = 83. The scale bars are all 2 mm long. 4.2 Maximum spread

Fig. 17 provides example images of droplets at their maximum spread (see image 3, image 2 and image 2 in Fig. 17a, b and c, respectively). The maximum spread provides us with a single quantitative measure of the impact dynamics which, as per previous arguments [36, 37], should be reconciled with the impact Weber number via an appropriate scaling law. As such, In Fig. 19, we have presented the full results of the maximum spread determined by high-speed video observations. Given that each data point is the average of several repeat trials, this data incorporates over 1100 experimental trials. Across nearly three orders of magnitude in the Weber number, we find that the single-valued exponent which best describes the entire data range is approximately 0.32, i.e. 0.32

max 0D D We , represented by the red line. Noting that this data comprises droplet diameters of 1, 2 and 3 mm and a large range of particle diameters, this single scaling is remarkably universal.

We now consider an energy-based argument to describe the spreading and compare this to our empirical scaling law. An expression for the maximum spread in the inviscid case can be found by simply balancing the surface and kinetic energies of the drop both prior to the impact and at the time of maximum spread. Since the contact angle of the impinging drop with the powder bed is generally high, we make the overly simplistic assumption of negligible contact between the particle bed and the drop, thereby assuming that surface tension acts of both the top side and underside of the drop and around the periphery. Note also that several authors have determined surface tensions of particle-laden interfaces to be very similar to that of the pure liquid [47–49], so that this assumption appears justified. As a first approximation we neglect the influence of the compliance of


44

the powder bed. Finally, we approximate the shape of the maximally deformed droplet as

a cylinder, where the height, h, can be found by volume conservation as 30

2max

23

Dh D . We

can thus write the balance of surface and kinetic energies as: 23 2 2 max0 0 0 max

1 ( )2 2DD V D D h (4)

Which, upon substitution for h and some rearranging, yields the following equation for the normalized maximum spread:

3max max

0 0

42 03 3D DWeD D

(5)

Solving Eq. (5), we thus find the maximum spread scales as 1max 3

0

D WeD for low Weber

numbers with 10We . In the high Weber number limit, neglecting the constant term, we

find the 90o trivial solution max 0 2 / 6D D We , i.e. we recover the scaling 1/2We

for high Weber numbers ( 100We ). This high-Weber number approximation simply recovers the original 1/2We scaling by [36] in the limit of negligible surface energy prior to impact.

The data in Fig. 6 with the power-law exponent of 0.32 certainly gives support to our simple energy-based model for the low-Weber number regime (We < 100) but does not follow the model in the high-Weber number range (We > 100). We attribute this to two principal causes – namely – bed compliance [33–35] and kinetic energy. Examining the data sets for 1 mm, 2 mm and 3 mm droplets independently, we find a slightly weaker


45

dependence on the Weber number for larger droplets, which may be partly be explained by the larger energy dissipation through crater formation, i.e. the larger the impact kinetic energy, the larger the crater that will be formed. Thus the initial assumptions in the energy balance, where all other forms of dissipation, including crater formation are neglected, are only likely to hold for smaller droplets, i.e. low Weber numbers, where indeed the best agreement is found.

Since our data in Fig. 4.3 covers the full range of particle properties shown in Table 2, we conclude that if the powder is hydrophobic, i.e. contact angle 90o , the maximum spread is independent of both particle size and contact angle.

Fig. 4.3. Normalized maximum spread diameter (Dmax/D0) versus the impact Weber number (We) for all droplet sizes and particles sizes used herein. The data points are the average of multiple trials, however error bars are not shown for display purposes. 4.3 Liquid marble formation


46

Criteria for complete marble formation have been reported previously [14,18], where it was found that the principal prerequisite for complete particle coverage on the original droplet is that the ratio of droplet-to-particle diameter, D0/dp, must exceed a threshold of 25. However, our findings of liquid marble formation, summarized in Table 4.2, indicate that liquid marbles can be formed with ratios as low as 6.8, i.e. 0 6.8pD d .

The total size ratio tested herein was from 0 2 120pD d . We highlight the fact that different hydrophobization techniques and particle size distributions could lead to differences in contact angle and particle mobility, which may be the cause of the discrepancy in terms of the critical ratio D0/dp between our study and those of [14,18]. Table 4.2. Summary of marble formation for all drop and particle diameters.

Droplet diameter D0 (mm)

Mean particle diameter, dp ( m )

Ratio D0/dp (-) Marble formation

(Yes/No) 1 25 39.8 Yes 1 62 16.2 Yes 1 118 8.5 No 1 148 6.8 Yes 1 170 5.9 No 1 204 4.9 No 1 488 2 No 2 25 79.6 Yes 2 62 32.4 Yes 2 118 17.0 Yes 2 148 13.5 Yes 2 170 11.8 Yes 2 204 9.8 Yes 3 488 6 No 3 25 119.5 Yes 3 62 48.7 Yes 3 118 25.5 Yes 3 148 20.3 Yes 3 170 17.6 Yes 3 204 14.7 Yes 3 488 6 No


47

Fig. 4.4 plots the maximum spread diameter against the impact speed for water droplets of diameter 1 mm, 2 mm and 3 mm, respectively. As shown in Fig. 20, monotonic increase in maximum spread with impact speed is observed until the onset of splashing with the ejecta being lifted off the surface (e.g Fig. 18 b and c). The different symbols in the Fig. 20 correspond to different observed features :( Squares) not fully encapsulated, (Circles) Full and mobile spherical liquid marble, (Diamonds) frozen marble. In light of these three different states, we now elucidate to the factors which control the particle encapsulation over the liquid droplet. Whilst the kinetic energy has previously been given significant attention [14], we infer that this effect is more neatly summarized by the relationship between the maximum spread and impact Weber number, noting that the Weber number can be thought of as the balance between impact kinetic energy, 3 2

0 0KE D V , and

surface energy 20SE D . As such, with this influence being evident from the data in Fig.

4.4, we now proceed to evaluate factors specific to the particle which may influence the formation of liquid marbles and arrested shapes. In particular, particle mobility on the surface of the droplets.


48


49

Fig. 4.4. Plots of maximum spread versus impact speed highlighting the transitions from incomplete coverage (open squares) to complete coverage (filled circles) to jammed interfaces (filled diamonds). Drop sizes are (a) 1 mm, (b) 2 mm, (c) 3 mm. The vertical dashed lines mark the onset of the deformed marbles and splashing, which largely coincide. The colors represent different particle diameters. (For interpretation of the references to color in this figure refer legend) 4.4 Particle mobility and liquid marble formation

One key factor in determining whether liquid marbles can be formed with such low size ratios is particle aggregation on the surface. An example of this is shown in Fig. 4.5 (a) where we observe that particles form aggregates or ‘‘rafts” [24]. This particular feature itself is a manifestation of the hydrophobization, evident upon comparison of Fig. 4.5 (a) and (b) both for the same particle diameter. In Fig. 4.5 (a), the particles were hydrophobized with Ultra Ever dry and we observe the formation of particle rafts, which essentially stabilize networks of particles against gravity, with patches of free-surface still visible. Such particle rafts are known to induce elasticity and, furthermore, interfacial rheology [22,23,50]. By contrast, in Fig. 4.5 (b) the particles were coated with Glaco MCZ and we do not observe the formation of such rafts, rather the particles appear to move independently of each other and settle under gravity around the periphery of the drop. Thus


50

the particles coated with Glaco MCZ, exhibit a higher apparent mobility on the surface, meaning that particles will move more freely over the free surface, as seen in Fig. 4.5 (b).

Fig. 4.5. Impact of 2.1 mm water droplets onto glass beads with mean diameter dp = 170 m coated with (a) Ultra ever dry and (b) Glaco MCZ. The impact speed in both cases is V0 = 0.94 m/s with We = 25. Images in (a) are taken 10 ms apart, whilst in (b) images are taken at t = 0, 10, 40, 55, 70, 115,145 and 200 ms from impact. The scale bars are 2 mm long.

To extend the raw observations of raft formation, we plot the formation of liquid marbles for these two coatings for both 2 and 3 mm droplets most clearly manifested in the Dmax vs. V0 parameter space, shown in Fig. 4.6. The particles coated with Ultra every dry form complete liquid marbles at impact speeds of approximately 0.9 m/s for both 2 and 3 mm droplets, whereas those coated with Glaco MCZ do not form marbles until impact speeds of approximately 1.35 and 1.85 m/s for 3 mm and 2 mm drops respectively. The fact that liquid marbles form at lower speeds for one particular hydrophobization treatment


51

is therefore due to the aggregation of particles into elastic rafts. We postulate that this is attributed to microscale interactions between the surface structures formed during the hydrophobization, but this is yet to be conclusively determined. Note that the contact angle measured with the sessile drop method is lower for the Ultra ever dry coating which, following previous observations [51], should lead to increased propensity for raft formation. Therefore, cohesive forces between particles due to the specific treatment could be a significant factor in determining individual particle mobility, the mesostructure of the interface, and ultimately whether a liquid marble will form or not.

Fig. 4.6. Plot of maximum spread versus impact speed, with arrows indicating the onset of complete marble formation for different hydrophobization agents (red = Glaco MCZ, black = Ultra ever dry). Square symbols indicate incomplete coverage, circles indicate formation of full liquid marbles and diamonds indicate arrested shapes. (For interpretation of the references to color in this figure refer legend)

To further quantify particle mobility, we tracked individual grains located at the interface of the partially coated droplets. An example image sequence highlighting this tracking process is shown in Fig. 4.7 (a) and the corresponding digitized particle locations


52

are shown in Fig. 4.7 (b). Whilst the drop is still undergoing small oscillations, the drop shape in Fig. 4.7 (b) is represented as a perfect sphere for ease of display.

Fig. 4.7.(a) Image sequence of a partially encapsulated droplet (D0 =1mm, dp =170 m ) with arrows indicating the location of an individual particle rolling across the interface. The scale bar is 2 mm long. (b) Plot representing the motion of the individual particle in (a). The initial location, shown by the blue square is taken 20 ms after impact and subsequent data points (black circles) are taken 5 ms apart. (For interpretation of the references to color in this figure refer legend)

a)


53

By performing such tracking for different droplet diameters and calculating the displacement frame by frame, the derived average velocities of the particles around the droplets were 13.7 m /s for D0 = 3 mm, 23.4 m /s for D0 = 2 mm, and a maximum of about 37.5 m /s for D0 = 1 mm. We can rationalize these increasing velocities simply with the ratio 0 / pD d . The larger the drop, the lower the curvature and the more planar-like the surface becomes. A higher curvature (i.e. smaller drop diameter) renders particles relatively more mobile. To our knowledge, the only study to date examining the effect of liquid interfacial curvature on particle contact angles is that in [52], where it was found that higher curvature resulted in lower receding contact angles whilst the advancing angle was unaffected. Thus, based on a higher contact angle hysteresis, this would imply that higher curvature results in a lower degree of mobility, which would therefore seem at odds with our observations. We note, however, that the sign of curvature relative to the particle could be a factor that has not been considered, see Fig.4.8.

Fig. 4.8. Sign of curvature relative to particle. In our experiments, where we have a positive curvature (based on the convention shown in Fig. 4.8) we estimate the particle contact angle by examining images of particles at the interface, such as that shown in Fig.4.9. Based on the protrusion length,

1 cospl r , we find 120o , in reasonable quantitative agreement with the contact angle measurements using the sessile drop method. We have not observed or measured the


54

advancing or receding (i.e. dynamic) contact angles for these particles and while study of dynamic contact angles of hydrophobic particles moving across the droplet surface would certainly complement our work herein and [52], it is considered beyond the scope of the present study.

Fig.4.9. Image of an isolated particle dp =148 m at a drop surface. The protrusion length is l =113m .

4.5 Transition to jammed interface Prior observations of ‘‘frozen” droplet shapes upon rebound from a hydrophobic

powder surface [29,30] found a critical impact speed *0 1.6V m/s for 0 2D mm.

Herein, we have also observed a critical impact speed above which the droplets do not regain a spherical shape. As evident from the data in Fig.4.4, the critical impact speed is dependent primarily on the drop diameter and does not exhibit any quantifiable dependence on the particle size. For our experiments, we find critical impact speeds of *

0V 2.1, 1.5 and 1.3 m/s for drop diameters of D0 = 1, 2 and 3 mm, respectively. This means that we can describe the transition to a jammed interface in terms of a critical Weber number with

*We 60–70. One important factor to bear in mind when discussing the formation of


55

jammed interfaces is the onset of splashing: The vertical dashed lines in these plots represent the threshold in terms of impact speed for the onset of the deformed marbles and splashing, respectively. These two thresholds largely coincide, which is not unexpected as the detachment of satellite drops leaves a smaller volume to encapsulate. Based on the total detached volume of all satellite drops, we estimate the surface area of the remaining volume to be 0

20.7 D . If we further assume the total contact area of the drop during the spreading

phase to be 2max / 4D , we can then derive a critical condition Dmax > 1.67D0 to fully

cover the droplet. Thus, with reference to Fig. 19 , given that the maximum spread beyond the splashing threshold is typically max 0 02 3D D D , the drop attains more than enough powder to become fully encapsulated, thus increasing the propensity to yield a jammed interface. Precisely how much powder is required to yield a jammed interface for liquid marbles remains an open question, however, we assert that the liquid free surface must vanish entirely [30], which is consistent with measurements of effective surface tensions [49] for liquid marbles where the surface tension was observed to decrease as the surface coverage of particles increased. 4.6 The most deformed shapes and liquid marble formation

We end our analysis for liquid marble formation experiment by examining the most deformed shapes, which were given by the smallest particle sizes used dp = 25 m . Fig. 4.10 presents the reader with three example sequences of jammed interfaces, which result in arrested shapes that are clearly far from a spherical equilibrium that would normally be observed under the influence of surface tension. What these shapes clearly demonstrate is that a simple interfacial (surface) tension is no longer adequate to describe the properties of the interface and that we should instead seek both mechanical and rheological properties


56

[22–24,50] to characterize deformation of liquid marbles. Such work needs to be addressed in future works. One final observation is that the arrested shapes result from higher impact speeds, leading to higher speeds of retraction from maximum spread in the horizontal direction and in the vertical axis upon rebound (e.g. image 3 in Fig. 4.10 (a)). As such, it appears that the rate of change of surface coverage could influence the formation of arrested shapes.

Fig.4.10. Arrested shapes formed during the impact of 2 mm water droplets with beds of hydrophobic 25 m particles. 4.7 Maximum spread of liquid marbles on impact

Figure 4.13. Shows the spreading process of a liquid marble coated with glass beads (Glaco MCZ) with mean diameter dp = 170 m on impact with a hydrophilic surface as a function of time from impact. Similar to the case of impingement of pure water droplet with a solid surface, when the liquid marble touches the surface, the diameter increases until it reaches the maximum of the spread indicated by the maxima on the curve as shown


57

in figure 4.13. At the maximum spreading state, the liquid flow changes its direction and recoils inward due to surface energy. The amount of retraction depends on several factors including the initial kinetic energy of the impacting drop, the surface tension of the liquid, and the interaction energy between the liquid and the surface. In cases of impact with hydrophilic surfaces, the liquid marbles retracted to a position different from the initial maximum spread and stop. However for the cases of impact on hydrophobic surfaces it rebounds where the liquid separates from the surface, rises a short distance and returns to the surface. Figs. 4.11 and 4.12 a) and b) show images from high-speed video sequences of the impact of liquid marble coated with glass-beads (Glaco MCZ) with mean diameter dp = 170 mm onto rigid hydrophilic and hydrophobic surface respectively. See captions in the figure for details of impact conditions.

a)


58

Fig. 4.11. Images taken from side showing impact of liquid marbles coated with particles of mean diameter dp = 170 m with (a) Hydrophilic plate (b) Hydrophobic plate. The impact speed in both cases is V0 = 0.99 m/s with We = 46. Images are taken 20 ms apart.

b)

a)


59

Fig. 4.12. Images taken from bottom showing impact of liquid marbles coated with particles of mean diameter dp = 170 m with (a) Hydrophilic plate (b) Hydrophobic plate. The impact speed in both cases is V0 = 0.99 m/s with We = 46. Images are taken 20 ms apart.

Figure 4.13. Impact and spreading process of a liquid marble coated with glassbeads (Glaco MCZ) with mean diameter dp = 170 m onto a hydrophilic surface. The impact velocity, V0 = 0.65 m/s and weber number for impact, We = 25.

b)


60

Extensive research has been done studying the impact of pure liquid droplets to the rigid surfaces because of its relevance both in nature and in industry. Numerous relations between the maximum spreading ratio, and the impact parameters have been established based on balance, in terms of kinetic energy and surface energy prior to impact and surface energy along with viscous dissipation at its maximum spreading. Recently more detailed models have included corrections due to the initial surface energy, dynamic contact angle, or viscous dissipation [53][69].

In fig.4.14 and 4.15 we have presented the full results of the maximum spread of liquid marbles determined by high-speed video observations on hydrophobic and hydrophilic surfaces respectively. Given that each data point is the average of several repeat trials, however we have not included error bars for display purposes. Across nearly two orders of magnitude in the Weber number, we find that the single-valued exponent which best describes the entire data range is approximately 0.25, i.e. 0.25

max 0D D We , represented by dotted black line for impact onto hydrophilic surfaces. However as shown in figure 31 for impact onto hydrophobic plate impact we found slightly higher dependence of approximately 0.33, i.e. 0.33

max 0D D We , on impact weber number compared to hydrophilic substrate in some cases. From the exponent describing normalized maximum spread diameter (Dmax/D0) as a function of the impact weber number (We) for impact on hydrophilic/hydrophobic for all particles sizes used herein we found that type of surface slightly affect the dynamic wettability and maximum spreading at low impact velocity as pointed by Lee et al.(2016). Moreover we found that varying the particle mean diameter of a factor five from 25 μm to 204 μm does not significantly affect the spreading process i.e maximum spreading ratio for impact on hydrophilic surfaces as shown in figure 30.


61

However for hydrophobic surfaces we observed slight dependence on particle diameter as shown in figure 4.15.

In figure 4.16 we have shown the sequence of images from high-speed video observations to compare the spreading onto hydrophobic surface for liquid marbles with two different sized particle diameters, mean diameter dp = 62 m and mean diameter dp = 204 m respectively.

From the observations in figure 32 we postulate that for smaller sized particles inter-particle cohesive force is strong compared to bigger ones. See in image 2 and 3 in figure 4.16 a) and b) 62 m particles are moving in clusters whereas 204 m are moving freely as single particle. Hence liquid marbles coated with bigger particles has effective surface tension close to pure water whereas those coated with smaller ones has higher effective surface tension than water hence they spread less. However we suggest future work and measurement of surface tension for different sized liquid marbles to support our observation conclusively.

Our experimental data thus gives good quantitative agreement with the scaling law by Clanet et al. [37] for impact on hydrophilic substrates. However for hydrophobic substrate the data fit our empirical scaling law established based on simple energy-based argument for impact of liquid droplets on hydrophobic bed resulting in marble formation [37].In that experiment we found the single-valued exponent describing the impact of liquid droplets on hydrophobic bed was approximately 0.32. The derivation was made with the overly simplistic assumption of negligible contact between the particle bed and the drop, thereby assuming that surface tension acts of both the top side and underside of the drop and around the periphery while deriving the scaling law. For the case of liquid marbles


62

the impact of impinging liquid marble is with rigid substrate not a hydrophobic bed especially for the case of hydrophilic surface. Hence during the impact the drop experiences an effective acceleration, 2

0 0V D much more intense than the gravity field, which flattens it and fixes its extent as suggested by Clanet et al. (2004) [37]. This holds true for wettable surface i.e hydrophilic surface but for super-hydrophobic substrates the higher contact angle of the impinging liquid marble contribute to minimize viscous dissipation. Thus the 1/4We scaling law fits out marble impact data better than 1/3We in case of impact on more wettable surface.


63

Fig. 4.14. Plots of normalised maximum spread versus impact weber number. The colors represent different particle diameters. Diamond symbols indicate impact on hydrophilic surface, (For interpretation of the references to color in this figure refer legend).

Fig.4.15. Normalized maximum spread diameter (Dmax/D0) versus the impact Weber number (We) for liquid marble impact on hydrophobic surface for all particles sizes used herein. The data points are the average of multiple trials, however error bars are not shown for display purposes.


64

Fig.4.16. Images taken from bottom showing impact onto hydrophobic glass plate of liquid marbles with particles (Glaco MCZ coated) of mean diameter (a) dp = 204 m (b). dp = 66 m . a) The impact velocity, V0 = 1.53 m/s and weber number for impact, We = 101, b) The impact velocity, V0 = 1.53 m/s and weber number for impact, We = 108.1. Images are taken 10 ms apart.

We observed in the liquid marble formation experiment, the liquid marbles form at lower speeds for one particular hydrophobization treatment (Ultra ever dry) due to the aggregation of particles into elastic rafts. Following that observation we had postulated that

a) b)


65

for Ultra ever dry coating there is microscale interactions between the surface structures formed during the hydrophobization. Due to which the interparticle cohesive force is strong compared to Glaco MCZ.

During the liquid marble impacts too we observed as shown in figure 4.17 a) and b) the particles coated with Glaco MCZ, exhibit a higher apparent mobility while spreading on impact, meaning that particles are moving freely without clustering. Whereas as shown in image 5, image 6, image 7 and image 8 in Fig.4.17 a) the particles coated with Ultra ever dry tend to move in clusters and hence with relative less apparent mobility as the liquid marble spreads on impact.

From this observations we can thus conclude that our postulation that due to microscale interactions between the surface structures formed during hydrophobization the cohesive forces between the particle changes is somewhat correct. However future work is needed to determine this more conclusively.


66

Fig. 4.17. Images taken from bottom showing impact onto hydrophilic glass plate of liquid marbles with particles of mean diameter dp = 170 m coated with (a) Ultra ever dry (b) Glaco MCZ. The impact speed in both cases is V0 = 1.1 m/s with We = 47. Images are taken 10 ms apart from impact with last image at 1000 ms. 4.8 Comparison of spreading process of liquid marbles and pure droplets

For completeness we performed trials for the impact of pure water droplets onto both hydrophilic and hydrophobic plates. Figures 4.18 show the relation between the

a)

b)


67

impact velocity or the impact weber number and the maximum spread of the water drops impact on hydrophilic and hydrophobic surface respectively. The relation is almost linear in both the cases. However for impact on hydrophilic plate the data collapses when scaled as 1/4

max 0/D D We and for hydrophobic data it scaled more accurately as 1/3max 0/D D We

. This fit is determined by polyfit regression analysis in matlab. Previously Clanet et al. [37] has shown that the single-valued exponent which best describes the impact of water droplets on rigid surface is approximately 0.25, i.e. 1/4

max 0D D We , irrespective of the wettability of the substrate. The dependency is very clear for the data of impact onto hydrophilic plate. However for impact onto hydrophobic plate we observed slightly higher dependence. This is in line with the empirical scaling law we have obtained for impact of liquid marbles on hydrophilic/ hydrophobic surfaces respectively.

Fig.4.18. Plots of normalised maximum spread versus impact weber number. Diamond symbols indicate hydrophilic surface, circles indicates impact on hydrophobic surfaces (For interpretation of the references to color in this figure refer legend).


68

In fig. 4.19 and 4.20 we have presented the high-speed video observations of the maximum spread of liquid marbles and pure water for comparison on hydrophilic and hydrophobic surface respectively. Each data point is the average of several repeat trials, however we have not included error bars for display purposes. As shown in figure 4.19, we find that the single-valued exponent which best describes the entire data range for liquid marbles impacts and water droplets impact on hydrophilic surface is approximately 0.25, i.e. 1/4

max 0D D We as represented by the black line, across nearly two orders of magnitude in the Weber number. However for impact onto hydrophobic substrates as shown in figure 4.20, in both the cases we observed slightly higher dependence, approximately 0.33, 1/3

max 0D D We in agreement with the universal scaling law which we had proposed for liquid marble formation experiment.

Fig.4.19. Plots of normalised maximum spread versus impact weber number of liquid marbles with all particles sizes used herein and water droplets for comparison. Diamond symbols indicate impact onto hydrophilic surface (For interpretation of the references to color in this figure refer legend).


69

Fig.4.20. Plots of normalised maximum spread versus impact weber number of liquid marbles with all particles sizes used herein and water droplets for comparison. Circles indicates impact on hydrophobic surfaces (For interpretation of the references to color in this figure refer legend).

Figure 4.21. Compares the spreading process of a liquid marble coated with glassbeads (Glaco MCZ and Ultra ever dry) of mean diameter dp = 204 m on impact with a hydrophilic surface as a function of time from impact. In both the cases on touching the surface, the diameter increases until it reaches the maximum of the spread indicated by the maxima on the curve as shown in figure 4.21. At the maximum spreading state, the liquid flow changes its direction and retracts to surface tension to a equilibrium position and stop. This occurs when the impact is with hydrophilic surfaces as shown in figure 4.21. For the cases of impact on hydrophobic surfaces the droplet/liquid marbles rebounds and rises a short distance and returns to the surface. On comparing the spreading process of liquid marbles with pure water droplets on rigid surfaces we found that retraction velocity of liquid marbles is comparable to that of water droplets as shown in figure 4.21. We calculated the displacement frame by frame after the maximum spread till the droplets/liquid marbles reaches equilibrium. The derived average retraction velocity for liquid marble was 76.9 m /s .Whereas for water droplet was about 82.3 m /s for impact


70

conditions as shown in figure 34. However trapped particles in case of liquid marble should increase the mobility of the water like pillars in a cassie baxter case which is not in this case. Thus we can state that the spreading process of liquid marbles resembles to that of water droplets. The presence of trapped particles does not alter the spreading process of pure water.

Figure 4.21. Comparison of impact and spreading process of a liquid marble coated with glass beads (Glaco MCZ and Ultra ever dry) with mean diameter dp = 170 m with a water droplet onto a hydrophilic surface. The impact velocity, V0 = 0.89 m/s and weber number for impact, We = 34. 4.9 Splashing behavior of liquid marbles

We studied the splashing behavior over the range of particle sizes for liquid marbles on rigid hydrophobic/hydrophilic substrates. The splashing behaviour of liquid marbles with particles was different from that observed in the case of pure liquids and it was dependent on the substrate wettability. The effect of particles on the surface of the droplet on the spreading process was dependent on size and substrate as illustrated in table 4.3. We

Liquid marble dp 170 um(Glaco MCZ) Liquid marble dp 170 um (Ultra Dry)

Pure water droplets


71

found that for hydrophobic substrates there is occurrence of splashing, receding breakup and rupture whereas on the hydrophilic surface there is only splashing i.e ejection of secondary droplets from the lamella as explained in figure 4.22 ref. Grishaev et al. (2015). We calculated the splashing parameter, ReK We for impact at each case. Typically [74], impacts on rigid surfaces 3000K indicates the onset of a splash depending on the wettability of the substrate. We found that the K value for hydrophobic substrate is around 3000 but for the case of impact on hydrophilic surfaces the splashing threshold value of K is much greater than 3000 even for pure water droplets as shown in table 6. Figure 4.23 shows the onset of splashing of liquid marble coated with particles of mean diameter dp = 25 m onto hydrophilic and hydrophobic surfaces respectively. We observed that varying particle sizes in case of liquid marble slightly affect the splashing threshold as shown in table 4.3 and figure 4.24. Liquid marble encapsulated with bigger particles splashes for lower value of K (splashing threshold) than those made with smaller particles. The influence of particle size can largely be described by the effective viscosity: as particle size is increased the volume fraction of particles decreases, hence viscosity decreases; as a result, the liquid marble laden with bigger particles breaks into secondary droplets easily on impact. The presence of particles caused early splashing, receding breakup, and rupture for impact on hydrophobic surfaces and early splashing for hydrophilic surfaces as shown in figure 4.25 which compares the splashing behavior of liquid marble and water droplets for impact onto hydrophilic surfaces.


72

Table 4.3. Summary of splashing behavior for liquid marbles of all particle diameters and water droplet onto hydrophilic/hydrophobic surface.

Coating Particle Size(um) Impact Plate Splashing Threshold

ReWe Liquid Marble (Glaco MCZ) 25 Hydrophillic 12249.03 Liquid Marble (Glaco MCZ) 62 Hydrophillic 14034.63 Liquid Marble (Glaco MCZ) 170 Hydrophillic - Liquid Marble (Glaco MCZ) 204 Hydrophillic -

Liquid Marble (Ultradry) 170 Hydrophillic 5445.37 Pure water Hydrophillic >25000

Liquid Marble (Glaco MCZ) 25 Hydrophobic 3412.69 Liquid Marble (Glaco MCZ) 62 Hydrophobic 2594.17 Liquid Marble (Glaco MCZ) 170 Hydrophobic 1209.31 Liquid Marble (Glaco MCZ) 204 Hydrophobic 1164.13

Liquid Marble (Ultradry) 170 Hydrophobic 1886.12 Pure water Hydrophobic 3555.84

a)

b)

c)


73

Figure 4.22.Showing splashing behavior on impact onto hydrophilic and hydrophobic surfaces respectively. a) Hydrophilic surface- Prompt splash b) Hydrophobic surface- Receding break up c) Hydrophobic surface- drop rupture ref. Grishaev et al. (2015).

Figure 4.23.Showing spreading and onset of splashing of a liquid marble coated with glass beads (Glaco MCZ) with mean diameter dp = 25 m onto a) hydrophilic surface b) hydrophobic surface. a) The impact velocity, V0 = 1.88 m/s and weber number for impact, We = 157, b) The impact velocity, V0 = 1.08 m/s and weber number for impact, We = 56.

Figure 4.24. Showing spreading and onset of splashing of a liquid marble coated with glass beads a) (Ultra ever dry) with mean diameter dp = 170 m b) (Glaco MCZ) with mean diameter dp = 25m on impact with hydrophilic surface. a) The impact velocity, V0 = 1.47 m/s and weber number for impact, We = 84, b) The impact velocity, V0 = 1.88 m/s and weber number for impact, We = 157.

Figure 4.25. Comparing spreading and onset of splashing of a) water droplet of same diameter with b) a liquid marble coated with glass beads (Glaco MCZ) with mean diameter dp = 25 m on impact with a hydrophilic surface. a) The impact velocity, V0 = 2.35 m/s and weber number for impact, We = 272, b) The impact velocity, V0 = 1.88 m/s and weber number for impact, We = 157.

a)

b)

a)

b)

a)

b)


74

4.10 Study of rupture in liquid marbles As pointed in the previous section 4.9 for impact onto the hydrophilic surface there

is only splashing i.e ejection of secondary droplets from the lamella as shown in figure 4.26. However on hydrophobic substrates there is occurrence of splashing, but usually receding breakup and rupture as in figure 4.27. The rupture occurs due to the formation of holes at the impact and their subsequent growth as shown in figure 4.27. These holes often indicated as dry spots form due to the break of air bubbles trapped between the impacting drop and the substrate. The rupture depends on impact weber numbers. In our case for the impact on the super-hydrophobic substrate, the rupture has been observed to occur starting at Re* = 5800 and * 100We . For lower values of Weber number and Reynolds number the rupture does not happen. In such cases, they are often reported to have break up into secondary droplets while receding often termed as “receding break up”. However rupture mechanism at the microscopic scale is still an issue as pointed out by Planchette et al. (2012). We observed that when droplet was spreading upon impact, particles were dragged along too to the periphery by the divergent flow. While at the maximum spread when the film was thinned to a thickness (~ 1 μm) where long range molecular forces become active, holes of typical size ‘l’ nucleate in the film. Slowly the holes grow and reaches a critical size when the rupture happens as shown in figure 4.27. To study how the rupture depends on the particle diameter, we studied the rupture varying the particle diameter as shown in figure 4.27. For smaller particles rupture through formation of holes begins at comparatively higher weber number compared to liquid marbles coated with bigger sized particles. As shown in figure 4.27 there is formation of a single hole for liquid marble coated with glass beads (Glaco MCZ) with mean diameter dp = 25 m whereas for liquid


75

marble coated with glass-beads (Glaco MCZ) with mean diameter dp = 204mm there is three holes on impact onto hydrophobic surface at similar impact conditions respectively. See captions in the figure for details of impact conditions. We also carried out systematic studies to determine whether the rupture depends on the interactions between particles. For particles of mean diameter dp = 170 m coated with Glaco MCZ there is rupture at higher impact weber number. Whereas for Ultra ever dry we observed no rupture even at higher impact weber number. Here particle tend to move in clusters and hence we assume stronger interaction between the particles compared to particles coated with Glaco MCZ. The particles coated with Glaco MCZ, are moving freely without clustering exhibiting a higher apparent mobility while spreading on impact particles. This study shows the rupture depends on the interactions between particles. However we need further studies to prove this conclusively.


76

Figure 4.26. Comparing impact of liquid marble coated with glass beads (Glaco MCZ) with mean diameter a) dp = 25 m b) ) dp = 204 m with a hydrophilic surface. a) The impact velocity, V0 = 1.88 m/s and weber number for impact, We = 157, b) The impact velocity, V0 = 1.88 m/s and weber number for impact, We = 141.

a) b)


77

Figure 4.27. Comparing impact of liquid marble coated with glass beads (Glaco MCZ) with mean diameter a) dp = 25 m b) ) dp = 204 m with a hydrophobic surface. a) The impact velocity, V0 = 2.35 m/s and weber number for impact, We = 247, b) The impact velocity, V0 = 2.35 m/s and weber number for impact, We = 292.

a) b)


78

CHAPTER 5 CLOSURE

5.1 Conclusion We have conducted an experimental investigation of the spreading of water droplets

and subsequent formation of liquid marbles during impact onto hydrophobic powder surfaces. Both the maximum spread and marble formation were subject to detailed analysis, with specific emphasis on the role of droplet to- particle diameter ratio and resulting particle mobility. We have shown that, in addition to impact kinetic energy of the droplet, the degree of cohesive forces between the particles as well as the droplet-to-particle diameter ratio both influence the marble formation in terms of minimum impact speed required to form a fully encapsulated marble and the transition towards a jammed interface or ‘‘deformed” marble.

We found that the maximum spread of the liquid droplets onto the powder bed is independent of the size of the particles and the degree of inter-particle cohesive force and can be described by the scaling law 0.32

max 0/D D We with (1,100)We for

0 1 3D mm. This is in relatively good quantitative agreement with a 1/3 exponent

found by a simple energy balance approach. However, for larger droplets, 0 2D mm, a slightly weaker dependence was observed and is thought to be due to energy losses during to the crater formation, i.e. the compliance of the powder bed. We observed that droplets remained intact only for We < 100 and propose future studies to investigate the influence of packing fraction in this phenomena, as previously indicated [21] as well as the role of other key factors like the nature of coating of the glass beads which


79

affects the inter-particle cohesive forces. In addition we have observed the transition between different regimes in the liquid marble formation process – namely – (i) partial coverage, (ii) full encapsulation with spherical liquid marble, and (iii) frozen deformed shape across a broad range of parameters.

We found that the transition to a deformed (arrested) shape is characterized by a critical impact speed, V*, consistent with previous results [29 , 30], leading to a critical impact Weber number * 60 70We and maximum spread max 02D D . We note that the arrested shapes occur during the retraction-rebound phase, which may imply that the rate of change of surface area and therefore surface coverage plays a dominant role in this phenomenon. Interfacial properties beyond a simple surface tension need to be explored for liquid marbles and are the subject of an ongoing investigation.

Further the impact of liquid marbles onto solid surface of different wettability is being investigated. We found that the single-valued exponent which best describes the impact on hydrophilic surface is approximately 0.25. However for impact on hydrophobic surfaces we found slightly higher dependence of 0.33, i.e.,

0.33max 0D D We . Varying the particle mean diameter of a factor five from 25 μm to

204 μm does not significantly affect maximum spreading ratio for impact on hydrophilic surfaces, However for hydrophobic surfaces we observed slight dependence on particle diameter. We found that the spreading process of liquid marbles resembles to that of water droplets. The presence of trapped particles does not alter the spreading process of pure water. The splashing behaviour of liquid marbles with particles was different from pure liquids and was dependent on the substrate wettability.


80

The presence of particles caused early splashing, receding breakup, and rupture for impact on hydrophobic surfaces and early splashing for hydrophilic surfaces. At large impact velocity, rupture of the liquid marbles occurs for impact onto super- hydrophobic surfaces at Re* = 5800 and * 100We . We observe the mechanism of rupture consists of formation of holes and its subsequent growth for impact on the super-hydrophobic substrate. 5.2 Recommendations for future work

Future work can be focused on experiments how particle mobility depends on the curvature when we have positive curvature on the droplet surface relative to particle during the formation of liquid marbles on impact. To study this we need to measure the advancing or receding (i.e. dynamic) contact angles for hydrophobic particles across the droplet surface.

Also extensive research to study regarding the dynamics of the liquid marbles after their formation is needed. Examples include experiments comparing study the properties of liquid marbles like surface tension and viscosity. In addition the impact of liquid marbles on liquids and other different types of powders can be investigated. Liquid marbles are an ongoing research topic of study which can be broadened and explored more for widening their potential applications numerous areas, such as in the pharmaceuticals industry.


81

References [1] Quere,D., Lafuma, A. & Bico, J., Slippy and sticky microtextured solids, Nanotechnology (14), 1109–1112 (2003). [2] Quere, D., Non-sticking drops, Reports on progress in physics (68), 2495–2532 (2005). [3] Lafuma, A. & Quere, D., Superhydrophobic states, Nature materials (2), 457–460 (2003). [4] Aussillous, P. & Quere, D., Liquid marbles, Nature 411, 924–927 (2001). [5] Aussillous, P. & Quere, D., Properties of liquid marbles, Proceedings of royal society A 462, 973–999 (2006). [6] McHale, G. & Newton, M.I., Liquid marbles: principles and applications, Soft Matter 7, 5473–5481 (2011). [7] Bormashenko, E., Liquid marbles: properties and applications, Current opinion colloid interface science 16 (4), 266–271 (2011). [8] Cha, D. & Wang, P., Remotely controlled liquid marbles, Advanced materials (24), 4756–4760 (2012). [9] Zhao, Y., Fang, J., Wang, H., Wang, X. & Lin, T., Magnetic liquid marbles: manipulation of liquid droplets using highly hydrophobic Fe3O4 nanoparticles, Advanced materials (22), 707–710 (2010). [10] Xue, Y., Wang, H., Zhao, Y., Dai, L., Feng, L., Wang, X. & Lin, T., Magnetic liquid marbles: a precise miniature reactor, Advanced materials (22), 4814–4818 (2010). [11] Tian, J., Arbatan, T. & Shen, W., Liquid marble for gas sensing, Chemical communication (46), 4734–4736 (2010).


82

[12] Mele, E., Bayer, I.S., Nanni, G., Heredia-Guerrero, J.A., Ruffilli, R., Ayadi, F., Marini, L., Cingolani, R. & Athanassiou, A., Biomimetic approach for liquid encapsulation with nanofibrillar cloaks, Langmuir 30, 2896–2902 (2013). [13] Chu, Y., Wang, Z. & Pan, Q., Constructing robust liquid marbles for miniaturized synthesis of graphene/Ag nanocomposite, Applied material interfaces 6, 8378–8386 (2014). [14] Eshtiaghi, N., Liu, J.S., Shen, W. & Hapgood, K.P., Liquid marble formation: spreading coefficients or kinetic energy?, Powder technology 196, 126–132 (2002). [15] Eshtiaghi, N., Arhatari, B. & Hapgood, K. P., Producing hollow granules from hydrophobic powder in high-shear mixer granulators, Advanced powder technology 20, 558–566 (2009). [16] Hapgood, K.P. & Khanmohammadi, B., Granulation of hydrophobic powder, Powder technology 189, 253–262 (2009). [17] Eshtiaghi, N., Liu, J. S. & Hapgood, K.P., Formation of hollow granules from liquid marbles: small scale experiments, Powder technology 197, 184–195 (2010). [18] Estiaghi, N. & Hapgood, K. P., A quantitative framework for the formation of liquid marbles and hollow granules from hydrophobic powders, Powder technology 223, 65–76 (2012). [19] Bhosale, P.S. & Panchagnula, M.V., Sweating liquid micro-marbles: dropwise condensation on hydrophobic nanoparticulate materials, Langmuir 28, 14860–14866 (2012).


83

[20] Hashmi, A., Strauss, A. & Xu, J., Freezing of a liquid marble, Langmuir 28, 10324–10328 (2012). [21] Marston, J.O., Thoroddsen, S.T., Ng, W.K., & Tan, R.B.H., Experimental study of liquid drop impact onto a powder surface, Powder technology (203), 223–236 (2010). [22] Fuller, G.G. & Vermant, J., Complex fluid–fluid interfaces: rheology and structure, Annual review of chemical and biomolecular engineering 3, 519–543 (2012). [23] Monteux, C., Jung, E. & Fuller, G.G., Mechanical properties and structure of particle coated interfaces: influence of particle size and bidisperse 2D suspensions, Langmuir 23, 3975–3980 (2007). [24] Vella, D., Aussillous, P. & Mahadevan, L., Elasticity of an interfacial particle raft, Europhysics letters 68, 212–218 (2004). [25] Subramaniam, A.B., Abkarian, M., Mahadevan, L. & Stone, H.A., Non-spherical bubbles, Nature 438, 930 (2005). [26] Subramaniam, A.B., Abkarian & M., Stone, H.A., Controlled assembly of jammed colloidal shells in fluid droplets, Nature materials 4, 553–556 (2005). [27] Pawar, A.B., Caggioni, M., Ergun, R., Hartel, R. W. & Spicer, P. T., Arrested coalescence in pickering emulsions, Soft Matter 7, 7710(2011). [28] Kotula, A.P. & Anna, S.L., Probing timescales for colloidal particle adsorption using slug bubbles in rectangular microchannels, Soft Matter 8, 10759–10772 (2012). [29] Marston, J.O., Zhu, Y., Vakarelski, I.U. & Thoroddsen, S.T., Deformed liquid marbles: freezing drop oscillations with powder, Powder technology 228, 424–428 (2012).


84

[30] Marston, J.O., Zhu, Y., Vakarelski, I.U. & Thoroddsen, S.T., Freezing drops with powders, Physics of fluids 25, 091107 (2013). [31] Walsh, A.M., Holloway, K.E., Habda, P. & de Bruyn, J.R., Morphology and scaling of impact craters in granular media, Physical review letters 91, 104301(2003). [32] Umbanhowar, P. & Goldman, D.I. Granular impact and the critical packing state, Physical review E 82, 010301(2010). [33] Katsuragi, H., Morphology scaling of drop impact onto a granular layer, Physical review letters 104, 218001(2011). [34] Katsuragi, H., Length and time scales of a liquid drop impact and penetration into a granular layer, Journal of fluid mechanics 675, 552–573 (2011). [35] Zhao, R., Zhang, Q., Tjugito, H. & Cheng, X.,Granular impact cratering by liquid drops: understanding raindrop imprints through an analogy to asteroid strikes, PNAS 112, 342–347(2015). [36] Renardy, Y., Popinet, S., Duchemin, L., Renardy, M., Zaleski, S., Josserand, C., Drumright-Clarke, M.A., Richard, D., Clanet, C. & Quere, D., Pyramidal and toroidal water drops after impact on a solid surface, Journal of fluid mechanics 484, 69–83(2003). [37] Clanet, C., Beguin, D., Richard, D. & Quere, D., Maximal deformation of an impacting drop, Journal of fluid mechanics 517, 199–208(2004). [38] Aytouna, M., Bartolo, D., Wegdam, G., Bonn, D. & Rafai, S.,Impact dynamics of surfactant laden drops: dynamic surface tension effects, Experimental fluids 48, 49–57(2010).


85

[39] Eggers, J., Fontelos, M.A., Josserand, C. & Zaleski, S., Drop dynamics after impact on a solid wall: theory and simulations, Physics of fluids 22, 062101(2010). [40] Laan, N., de Bruin, K.G., Bartolo, D., Josserand, C. & Bonn, D., Maximum diameter of impacting liquid droplets, Physical review applied 2, 044018(2014). [41] Tran, T., Staat, H.J.J., Prosperetti, A., Sun, C. & Lohse, D., Drop impact on superheated surfaces, Physical review letters 108, 036101(2012). [42] Josserand, C. & Thoroddsen, S.T., Drop impact on a solid surface, Annual review of fluid mechanics 48, 365–391(2016). [43] Nefzaoui, E. & Skurtys, O.R., Impact of a liquid drop on a granular-medium: Inertia, viscosity and surface tension effects on the drop deformation, Experimental thermal fluid science 41, 43–50(2012). [44] Marston, J.O., Sprittles, J.E., Zhu, Y., Vakarelski, I.U. & Thoroddsen, S.T., Drop spreading and penetration into pre-wetted powders, Powder technology 239, 128–136(2013). [45] Nowak, E., Robbins, P., Combes, G., Stitt, E.H. & Pacek, A.W., Measurements of contact angle between fine, non-porous particles with varying hydrophobicity and water and non-polar liquids of different viscosities, Powder technology 250, 21–32(2013). [46] Nowak, E., Combes, C., Stitt, E.H. & Pacek, A.W., A comparison of contact angle measurement techniques applied to highly porous catalyst supports, Powder technology 233, 52–64(2013). [47] Arbatan, T. & Shen,W., Measurement of the surface tension of liquid marbles, Langmuir 27 ,12923–12929(2011).


86

[48] Bormashenko, E., Pogreb,R., Whyman,G., Musin, A., Bormashenko, Y. & Barkay, Z., Shape, vibrations, and effective surface tension of water marbles, Langmuir 25, 1893–1896(2009). [49] Bormashenko, E., Musin, A., Whyman, G., Barkay, Z., Starostinc, A., Valstifer, V. & Strelnikov, V., Revisiting the surface tension of liquid marbles: measurement of the effective surface tension of liquid marbles with the pendant marble method, Colloids surfaces A 425, 15–23(2013). [50] Barman, S. & Christopher, G.F., Simultaneous interfacial rheology and microstructure measurement of densely aggregated particle laden interfaces using a modified double wall ring interfacial rheometer, Langmuir 30, 9752–9760(2014). [51] Horozov, T.S., Aveyard, R., Binks, B.P. & Clint, J.H., Structure and stability of silica particle monolayers at horizontal and vertical octane–water interfaces, Langmuir 21, 7405–7412 (2005). [52] Senbil, N., He, W., Demery, V. & Dinsmore, A.D., Effect of interface shape on advancing and receding fluid contact angles around spherical particles, Soft Matter 11, 4999–5003(2015). [53] Lee, J, B., Dominique,D., Guyer,R. & Carmeliet,J., A.D. Dinsmore., Modeling the maximum spreading of liquid droplets impacting wetting and nonwetting surfaces, Langmuir (32), 1299-1308 (2016). [54] Planchette, C., Biance, A, L. & Lorenceau, E., Transition of liquid marble impacts onto solid surfaces, Europhysics letters, 14003(1-6) (2012).


87

[55] Stow, CD., & Hadfield, MG., An experimental investigation of fluid flow resulting from the impact of a water drop with an unyielding dry surface, Proc. R. Soc. A (41), 373-419 (1981). [56] Mundo,C ., Sommerfeld, M. & Tropea C., Droplet-wall collisions: experimental studies of the deformation and breakup process, Int. J. Multiphase Flow (21), 151–73 (1995). [57] Josserand, C. & Thoroddsen, S, T., Drop impact on a solid surface, Annual review fluid mechanics (48), 365-391 (2016). [58] Nguyen, T, H., Eshtiagi, N., Hapgood, K, P. & Shen, W., An analysis of the thermodynamic conditions for solid powder particles spreading over liquid surface, Powder Technology (201), 306‐310 (2010). [59] McEleney, et al., Liquid marble formation using hydrophobic powders, Chemical engineering journal (147), 373–382(2009). [60] McHale, G., Herbertson, D.L., Elliot, S.J., Shirtcliffe, N.J. & Newton, M.I., Electrowetting of nonwetting liquids and liquid marbles , Langmuir, 918-924 (2007). [61] Tosun, A. & Erbil,H,Y., Evaporation rate of PTFE liquid marbles, Applied Surface Science (256), 1278–1283(2009). [62] Dandan, M. & Erbil,H,Y., Evaporation rate of graphite liquid marbles: comparison with water droplets, Langmuir 25(14), 8362–8367 (2009). [63] Joung, YS. & Buie, CR., Aerosol generation by raindrop impact on soil, Nature Communication, 6:6083 (2015).


88

[64] Minemawari, H., Yamada, T., Matsui, H., Tsutsumi, J., Haas, S., et al., Inkjet printing of single-crystal films, Nature 475, 364–67 (2011). [65] Hulse-Smith, L., Mehdizadeh, NZ. & Chandra, S., Deducing drop size and impact velocity from circular bloodstains, Journal of forensic Science (50), 54–63 (2005). [66] Attinger, D., Moore, C., Donalson, A., Jafari A. & Stone, HA., Fluid dynamics topics in bloodstain pattern analysis: comparative review and research opportunities, Forensic science International (231), 375–96 (2013). [67] Chandra S. & Avedisian CT., On the collision of a droplet with a solid surface, Proceedings of royal society of London A, 432:13–41(1991). [68] Pasandideh-Fard, M., Qiao, YM., Chandra, S. & Mostaghimi, J., Capillary effects during droplet impact on a solid surface, Physics of fluids (8), 650–59 (1996). [69] Ukiwe, C. & Kwok, D., On the maximum spreading diameter of impacting droplets on well-prepared solid surfaces. Langmuir (21), 666–73 (2005). [70] Roisman IV., Berberovi, E. & Tropea C., Inertia dominated drop collisions. I. On the universal flow in the lamella, Physics of fluids (21), 052103 (2009). [71] Roisman IV., inertia dominated drop collisions. II. An analytical solution of the navier-stokes equations for a spreading viscous film, Physics of fluids (21), 052104 (2009). [72] Eggers, J., Fontelos, M., Josserand, C. & Zaleski, S., Drop dynamics after impact on a solid wall: theory and simulations, Physics of fluids (22), 062101(2010). [73] Richard, D., Clanet, C. & Quere, D, Contact time of a bouncing drop, Nature (417), 811 (2002).


89

[74] Stow, CD. & Hadfield, MG., An experimental investigation of fluid flow resulting from the impact of a water drop with an unyielding dry surface, Proceedings of royal society of London A 373, 419–41(1981). [75] Mundo, C., Sommerfeld, M. & Tropea, C., Droplet-wall collisions: experimental studies of the deformation and breakup process, International journal of multiphase flow 21, 151–73(1995). [76] Roisman IV., Lembach, A. & Tropea, C., Drop splashing induced by target roughness and porosity: The size plays no role, Advances in colloid and interface science (222), 615–21(2015). [77] Allen, RF., The role of surface tension in splashing, Journal of colloid and interface science (51), 350–51(1975). [78] Roisman IV., Horvat, K. & Tropea C., Spray impact: rim transverse instability initiating fingering and splash and description of a secondary spray, Physics of fluids (18), 102104(2006). [79] Marmanis, H. & Thoroddsen, ST., Scaling of the fingering pattern of an impacting drop, Physics of fluids (8), 1344–46(1996). [80] Kheshgi, H. S. & Scriven, L. E., Dewetting: nucleation and growth of dry regions, Chemical engineering science (46), 519-526 (1991). [81] Asher, S., Connor, J. N. & Sedev, R., Elasticity of liquid marbles, Journal of Colloid and Interface Science (449), 341-346 (2015). [82] Bormashenko, E. & Musin, A., Revealing of water surface pollution with liquid marbles, Applied surface science (255), 6429-6431 (2009).


90

[83] Dupin, D., Armes, S. P. & Fujii, S., Stimulus-responsive liquid marbles, Journal of the American chemical society (131), 5386-5387 (2009). [84] Arbatan, T., Li, L., Tian, J. & Shen, W., Liquid marbles as micro-bioreactors for rapid blood typing, Advanced healthcare materials (1), 80-83 (2012). [85] Newton, M. I., Herberston, D. L.,Elliott, S. J.,Shirtcliffe N. J. & McHale, G., Electrowetting of liquid marbles, Journal of physics D: Applied physics (40), 20-24(2007). [86] Bormashenko, E., Pogreb, R., Bormashenko, Y., Musin, A. & Stein, T., New investigations on ferrofluidics: ferrofluidic marbles and magnetic-field-driven drops on superhydrophobic surfaces, Langmuir (21), 12119–12122 (2008). [87] Miao, YE., Lee, H. K., Chew, W. S., Phang, I. Y., Liu, T. & Ling, X. Y., Catalytic liquid marbles: Ag nanowire-based miniature reactors for highly efficient degradation of methylene blue, Chemical communication (50), 5923—5926 (2014). [88] Sheng, Y., Sun, G., Wu, J., Ma, G. & Ngai, T., Silica-based liquid marbles as microreactors for the silver mirror reaction, Angewandte chemie international edition (54), 7012-7017 (2015).

supakar-thesis-2016 (1)

Documents