Abstracts

Many synthetic and natural colloids have heterogeneous surfaces and are highly nonspherical. The charge heterogeneity and the geometry of the particle play important roles in the processing of these colloids. Conventional methods in characterizing the surface charge of colloidal particles rely on the assumption that the particle has homogeneous surface properties. Recent theories predict that a particle with a distribution of zeta potential having a dipole moment will rotate into alignment with the applied field. In this thesis, an experimental method was developed to study particle motion; in particular, particle rotation and alignment caused by nonuniformity of the surface charge. The new experimental setup and techniques allowed recording of the experimental observations of particle motion on a videotape. The images were then digitized for image processing and analysis.

The alignment effect was studied through sedimentation studies which involved single particle measurement of sedimentation velocity (the component of the velocity parallel to gravity) before and after the application of an electric field for kaolinite and latex particles. The disk-like kaolinite and the spherical latex displayed very similar sedimentation behavior which was neither a simple, steady sedimentation nor an accelerating motion with a terminal velocity parallel to gravity. The sedimentation velocity after the application of an electric field was found to be lower than before the field was applied, but after a period of time the particle accelerated and approached its sedimentation velocity equal to the value before application of the field. An observed time-dependent component of the sedimentation velocity for kaolinite was attributed to natural convection because it was also observed with latex. The sedimentation velocity measured for latex before the application of the electric field was in agreement with the hydrodynamic predictions, based on Stokes law.

The combination of video microscopy and digital image analysis in the experimental setup allowed determination of the y-coordinate (i.e., in the direction of gravity) on the order of +1 μm, an accuracy superior to the classical electrophoresis apparatus. Such accurate data provides a means of testing hydrodynamic models of particle motion far more accurately then previous techniques. The “by-hand” tracking of single particle used in this thesis sets the stage for computer algorithms which simultaneously track many particles and allow a large number of data manipulations on the same particle.

In this thesis, we use x-ray and neutron reflectivity to probe the molecular and microscopic structure of surfactant monolayers on silicon oxide/silicon surfaces. A cationic surfactant, cetyltrimethylammonium bromide (CTAB), and an anionic surfactant, sodium dodecyl sulfate (SDS), are examined. We study the structure of these monolayers both after they are deposited from receding menisci of bulk surfactant solutions and as they exist in precursing films when reconnected to the bulk menisci. The self-assembly of the surfactant molecules at both the solid/liquid and liquid/vapor interfaces is driven by hydrophobicity; while at the solid surface, it is strongly influenced by the charge interaction between the surfactant head groups and the solid surface. Our x-ray reflectivity measurements allow us to deduce the fine details of the molecular structure of the as-deposited monolayers. On hydration, the CTAB monolayer does not desorb from the substrate nor does it restructure. In contrast, the SDS monolayer desorbs, forming a film with thickness on the order of 100 Å. The packing density of the SDS monolayer changes as the film evolves. The location and quantity of the water of hydration in the surfactant monolayer is found by combining x-ray and neutron reflectivity. Thus, our work reveals new information on the self-assembly mechanisms governing the deposition of soluble surfactants from solution. Further, we see the modification of the self-assembly due to the confined environment of the precursing film and have gained new insight into the microscopic mechanisms governing the wetting of oxide surfaces by surfactant solutions.

The nature of forces within colloidal doublets were studied using differential electrophoresis. Tangential forces have been found to exist between two non-touching colloidal spheres. Our study focused on the behavior of silica-polystyrene doublet particles under the effects of simultaneous electric and gravitational fields. Several doublets present in electrolyte solutions of different concentrations (3mM, 5mM and 7mM) were examined. The hydrodynamic state of a doublet has been used as a monitoring tool for identifying forces other than the forces already known to exist along the line of centers between the two colloid spheres. A rigid hydrodynamics state occurs as tangential forces between the spheres oppose the hydrodynamic forces which tend to rotate the spheres with respect to each other when an electric field is applied. We observed shifts between a rigid state and a freely-rotating state suggesting time varying tangential forces. “In-between” hydrodynamic states suggest the intermediate values of the tangential forces. The ability to break doublets that have been present in rigid states indicates that tangential forces are not likely to arise from the physical contact between the two spheres but are transmitted across the fluid gap.

Dynamic wetting is the displacement of one fluid by another immiscible fluid across a solid surface as it spreads. Such processes control many natural phenomena and technological applications. The spreading dynamics of macroscopic fluid bodies are dictated by the hydrodynamics in a microscopic region near the moving contact line. Analytical models have been devleoped to describe the interface shape and velocity field near the contact line. Using videomicroscopy, particle image velocimetry, and digital image analysis, we make simultaneous measurements of the fluid/fluid interface shape and fluid flow field within the first few hundred microns near a moving contact line. Our experiments establish the validity and limitations of these analytical models. This work extensively tests assumptions embedded in the models and sets up bounds on the parameter space in which the models are valid. The models uccessfully describe the hydrodynamics near the contact line up to a capillary number ~0.10 but break down at higher capillary number. We determine the origins of this breakdown. We also carefully probe those regions near the contact line where the interface shape and flow field are independent of the macroscopic geometry. Our experimental technique provides a means of obtaining such material-dependent, geometry-independent information about the system. Such information serves as boundary conditions transferable among different macroscopic geometries. It is an essential ingredient for numerical calculations of the spreading dynamics. The work reported in this thesis sets the stage for predictive modeling of dynamic wetting.

Dynamic wetting, the displacement of one fluid by another immiscible fluid, is important in many natural phenomena and technological applications. Complexity and modeling difficulties arise due to the unique hydrodynamics which operate in the microscopic region near the contact line and control the macroscopic spreading of a liquid. Most previous experimental work measuring the detailed hydrodynamics near contact lines has focused on the spreading of fluids over bare solid surfaces. However, in many situations, fluids spread over surfaces with pre-coated thin films. In this thesis, we have moved to regimes not treated before where thick fluid films and thin molecular scale films exist near the contact line at higher velocities. We study the hydrodynamics using video microscopy and digital image analysis to measure the liquid/vapor interface shape in a small region near a contact line advancing over a pre-existing film. In the process of our investigations, we have also developed techniques for dip-coating and characterizing uniform molecular scale and micron scale films. Several models, which describe the interface shape close to the moving contact line in the presence of a film, are experimentally examined. Our studies reveal the hydrodynamic behavior of the fluid in the film and near the moving contact line and give the dependence of dynamic contact angle on contact line speed and film thickness in the advancing case. We have also studied effects of possible complex material properties of molecular scale films (101 Å ~ 102 Å) on the macroscopic spreading of the liquid. We have begun to develop a unified picture of the wetting behavior of liquids on dry surfaces, molecular scale films, and thick fluid films.

The main objective of this thesis is to find out the short term spreading behavior of surfactant solutions spreading on a polymeric solution subphase. Two different experiments were done simultaneously to investigate the research. Spreading experiments were done to quantify the motion of the surfactant solution. Surface tension measurements were made to determine the surface tension difference between the subphase and the initial condition of the surfactant solution.

The major findings of this thesis are: 1.) The initial spreading rate varies monotonically with surface tension difference. If the difference is negative the spreading will not begin. 2.) The spreading behavior has two different regimes, above and below CMC. 3.) The end of the spreading event cannot be accounted for by all surfactant staying on the surface of the subphase and the surface tension becoming uniform across the entire subphase surface.

Other preliminary work is also reported here: 1.) a study of the appearance of an apparent instability during spreading; 2.) an examination of the time scale for mixing of the surfactant solution with the polyacrylamide subphase; 3.) a comparison of spreading behavior on simple and complex fluid subphases; and 4.) a measurement of time of the onset of movement of surface tracer particles and its correspondence to the speed of the shock wave moving across the surface.

A quantitative model of contact angle hysteresis is necessary for engineering surfaces with specified wetting properties and for better use of contact angles for surface characterization. We examine issues pertinent to developing such a model. We measure wetting characteristics of heterogeneous surfaces produced by monolayer surfaces altered by UV/ozone treatment. These surfaces mimic the wetting of a large class of ambient surfaces common in nature and technology. We characterize microscopic contact line structure and dynamics as well macroscopic contact angles and contact angle hysteresis. Qualitative energy functionals provide a useful langu11:ge for examining and describing the results. Our most notable results include the approximate power-law scaling of contact line roughness with length scale, the lack of connection between macroscopic contact angle hysteresis and contact line roughness, and the partial or complete mitigation of hysteresis from vibrational noise. We identify where theoretical and laboratory experimental models succeed or fail to correctly describe the wetting of ambient surfaces. Experimental microfabricated models with defects on a backing fail to properly demonstrate the essential features of wetting on ambient surfaces. We show the need for new theoretical models that treat the regime of strong, dense surface heterogeneity. Also, the time scale of relaxation of contact angles due to vibrations compared to the time scale of typical measurement processes must be incorporated into the models.

While steric stabilization of colloidal suspensions has been extensively studied, redispersion from the aggregated state for polymer-coated particles in low shear conditions has not. Many colloidal aggregates can be redispersed by high shear agitation, but we are motivated specifically by those systems where high shear is technologically infeasible (such as particulate drug dispersion from pharmaceutical tablets in the gastrointestinal tract). In such systems, the driving force for redispersion derives entirely from surface chemistry. Swelling of interfacial polymer films in the gaps between aggregated particles provides the driving force for redispersion. Sterically stabilized suspensions can be caused to reversibly aggregate under poor solvent conditions for the adsorbed polymer and to redisperse as good solvent quality conditions are restored.

In this thesis, we investigate the aggregation and redispersion of colloids that are sterically stabilized by adsorbed polymers. We examine four specific questions: 1) How fast do aggregates of polymer coated particles redisperse when the solvent quality is changed from a poor solvent to a good solvent? 2) How does the structure of the adsorbed polymer film control redispersion kinetics? 3) Are the redispersion kinetics dictated solely by the conditions of the polymer films, or does the overall structure of the aggregate play a role? 4) What qualitative generalities exist in redispersion kinetics behaviors across varying polymer solution compositions and conditions?

We examine the dynamics of aggregation and redispersion for colloidal polystyrene with adsorbed polymer layers that have a lower critical solution temperature (LCST). Two different types of polymer systems are used: a triblock co-polymer, poly(ethylene oxide)-block-poly(propylene oxide)-block-poly(ethylene oxide) (Pluronic F108) and a homopolymer, poly(N-isopropylacrylamide) (PNIPAM). The investigation involves changing the solvent quality of the stabilizing polymer by controlling the temperature above and below the theta temperature. Each particle-polymer system is investigated in various solvent systems by changing the electrolyte. The electrolyte in each system contributes to modifying the theta temperatures at which particle interactions change between attractive and repulsive. We use small angle light scattering (SALS) to monitor the kinetics for the aggregation and redispersion. The effects of the polymer surface coverage, free polymer concentration, degree of cooling, and the aggregate structure are examined to determine their role on the redispersion kinetics.

The redispersion kinetics are found to display one or two regimes where one regime can be described by a single exponential decay, exp(-t/τ1), and the other by a complex exponential decay, exp{-(t/τ2)β}. The redispersion in both polymer systems shows a decrease in the time constant for the single exponential decay regime and a tendency for more of the total redispersion to occur in the single exponential decay regime with increasing polymer concentration. The overall time for redispersion is relatively constant within each experimental system except at the lower concentrations with PNIPAM which redisperses much more slowly. The onset temperature of each regime is also probed by cooling to specific temperatures below the theta temperature. The complex exponential decay regime begins at higher onset temperatures (better solvent quality) than the single exponential decay regime. The aggregate structure has no impact on the redispersion kinetics.

We have developed hot-stage fluorescence microscopy to examine the temporal evolution of single polymer-bound particle contacts. The microscopy shows that PNIPAM coated particles that are attached to PNIPAM coated glass surfaces undergo complex motions in-plane and out-of-plane prior to the release from the surface. The timescales for the release of the particles are on the same order as the SALS experiments. The presence of lateral traps in these experiments suggests that tangential forces between the particles in the colloidal aggregates are important in the redispersion process.

Traditionally, systems are said to exhibit one of three wetting conditions: complete wetting, where the contact angle of the capillary body is zero; partial wetting, where the contact angle is nonzero and the there is no fluid on the substrate beyond the macroscopic contact line; and pseudo-partial wetting, where the contact angle is nonzero, but there is a film in equilibrium with the capillary body. Using energy minimization and a force balance formulation, these wetting states are re-examined and the differences between them are elucidated in detail. The conditions leading to complete, partial, and pseudo-partial wetting in various geometry are determined. Moving beyond previous discussion, we clearly differentiate between partial wetting which can have a “foot” at the contact line and pseudo-partial wetting. We contrast the phenomenon of autophobing, in which a fluid has a nonzero contact angle on top of its own monolayer, with pseudo-partial wetting. Complete wetting and pseudo-partial wetting are examined in cases where the film terminates on the substrates and cases where the film is infinite in extent. Infinite substrates and systems where the end of the substrate is important in determining the wetting state are considered. Further, evidence that a single material system can exhibit complete wetting in one geometry and pseudo-partial wetting in another geometry is shown, demonstrating that such transitions in wetting behavior occur when the size of the system changes. This re-examination of wetting produces a complete and unified picture of the classic states of wetting.

Surfactant driven “Marangoni” flows can be used to cause material transport along interfaces. One potential use of this is to cause post deposition transport of aerosolized drug used to treat lung diseases such as cystic fibrosis. The work in this thesis explores the underlying science of such a treatment by studying the transport properties in three different ways. First, I explore the ability of the lung’s natural surfactants (lipids) to transport drug across a mucus-like subphase. This work shows that lipid can drive such spreading and can even do so in low surface tension environments, such as the lung. Second, I develop an apparatus to study the effects of subphase height and surfactant amount on spreading and non-surface active material transport. The effects on flow behavior are mapped into an operating diagram and other transport properties are tracked and related to three distinct flow regimes. Lastly, I simulate the interactions of two spreading disks of surfactant to bring research closer to the case of aerosol deposition where multiple sources of Marangoni stress are present. From this work, I explore the effects of two surfactant sources on flow behavior. This thesis provides powerful insights into surfactant driven drug delivery and offers several future research directions which will advance the field.

The behavior of a physically entangled, semi-dilute polymer solution at a few times its overlap concentration in contact with its solvent is not very well understood. The thesis provides a physical picture of how the dissolution occurs.

The local viscoelastic properties of a simple uncross-linked flexible polymer of 2.1 wt% 1,000,000 Da Polyethylene oxide solution (PEO) which is at 14 times its overlap concentration in 50%-50% H2O-D2O mixture are measured. The PEO solution is placed in contact with pure solvent and the local viscoelastic properties of the solution (both on the PEO solution and H2O-D2O mixture sides of the interface formed) are measured by characterizing the thermal motion of micron sized carboxylate modified fluorescent polystyrene beads as a function of distance from the interface and time across the solution using one point microrheology.

Two major conclusions are drawn. (1.) The PEO phase swells due to water flux into the polymer network. The swelling is monotonically increasing with time at a fixed distance from the interface and monotonically decreasing with distance from the interface in the PEO phase at a fixed time after the interface is formed. (2.) Little or no polymer escapes from the entangled network of the polymer matrix and moves across the water region above the interface.

A drop of liquid, when deposited on a subphase of another immiscible liquid ideally exhibits two types of spreading behavior: it may either spread completely to a thin film on the underlying fluid, or it may reach an equilibrium shape of a static lens. The equilibrium spreading coefficient, which depends on the three interfacial tensions, subphase-air, drop-subphase and drop-air, determines this. In some cases, we notice a static lens even though a positive spreading coefficient predicts complete spreading. Also, we always observe a surface tension change at considerable distances outside the lens boundary. Surfactant solutions and surfactants alike show this behavior. This thesis investigates this abnormal behavior and we deduce that surfactants move across the contact line of a lens in the form of a monolayer and reduce the surface tension. This explains why the lens resists spreading on the monolayer thereby leading to autophobing.

Hexanol is used as a cosurfactant in a wide variety of consumer products and industries manufacturing surfactant based formulations. Surfactants present in these products adsorb at interfaces and cause interfacial tension driven Marangoni flows. Presence of a medium chain alcohol additive in conjunction with the surfactant is expected to effect the Marangoni driven flow at the liquid/liquid interface. Alcohol molecules are surface active, they not only compete with the surfactant to adsorb at the interface, but also are incorporated into the surfactant micelles at concentrations greater than the CMC, thus changing the micellar size and shape. This work shows the effect of hexanol on the sodium dodecyl sulfate driven Marangoni transport at the oil/water interface. Interfacial particle trajectories are recorded using optical microscopy. Presence of hexanol along with SDS is found to increase the peak Marangoni velocity of the interfacially adsorbed microparticles at concentrations both above and below the CMC during the adsorption stage. An increase in the displacement during the adsorption stage is also observed. This is only significant as low SDS concentrations. At the higher SDS concentrations tested in this study, the displacement for the multi-component SDS and hexanol system is not found to be statistically different from the single-component SDS system. Hexanol can be used to alter the lag time for Marangoni flow in the desorption stage since it lowers the CMC of the surfactant solutions.

For obstructive pulmonary diseases, such as cystic fibrosis (CF), aerosol drug delivery is a viable option to deliver drugs to infected airways. Current aerosol drug delivery techniques rely only on aerodynamics for drug dispersal inside of the lungs; however with limited ventilation and unusual aerodynamics in obstructed airways, various regions in the lungs are often left untreated. The mucus in such lungs can be highly viscous and unclearable, and can harbor bacterial infections in the absence of proper drug dispersal. Such infections can spread throughout the lung creating distress for patients and eventually may lead to mortality. We propose the use of surfactants to enhance pulmonary drug delivery. By enhancing surface transport via surface tension driven “Marangoni flows” along the airways surfaces after delivery, we expect surfactant-laden aerosols to produce a more uniform distribution of delivered drug in the lungs.

The research reported in this thesis determined the effects of different aerosol parameters on the maximum spreading of a deposited dye (a drug mimic) on entangled polymer solution subphases. The entangled polymer solution subphases served as mimics for the airway surface liquid (ASL) layer inside of the lungs. We roughly modeled deposition at a lung bifurcation by depositing the aerosol onto the subphase confined in a tube. The aerosol was direction perpendicular to the axis of the tube. We utilized a humidified bias gas ow to direct the aerosol to one end of the tube. This allowed us to distinguish direct and post deposition transport. We used fluorescence microscopy to determine the extent of spreading, and absorbance of the dye to determine the amount of aerosol formulation delivered. Aerosol droplet diameters were (1- 4 µm) within the range of sizes used clinically (1 – 5 µm).

We used previously established lung models and deposition probabilities to determine the aerosol deposition fluxes in the first sixteen airway generations. Comparing the results of these calculations to our experimental results, we determined that the aerosol delivery fluxes attained in our experiments spanned the deposition expected in 0 – 8 airway generations (large airways). When surfactant-laden aerosols are delivered onto the entangled polymer solution subphases at these fluxes, they provided enhanced spreading compared to their surfactant-free counterparts. The majority of the spreading occurred at short times post-deposition and plateaued at long times. We analyzed the dependence of spreading on various independent (surfactant concentration, ow rate and time of delivery) and dependent (surfactant mass and total delivered volume) variable. At the deposition rates in our experiments, aerosol droplets deposited on top of each other, forming a larger fluid body, In fact, the spreading of the fluid body formed from the multiple aerosol droplets was similar to the spreading of microliter scale drops deposition on the same subphases in the same geometry.

Two phenomena were studied in this thesis: 1) the relative motion between two or more colloidal particles on an electrode experiencing an ac potential with aqueous electrolyte solutions, and 2) the movement of ions and particles in non-polar liquids in one and two dimensional electric fields. The unifying theme of the work is the use of electric fields to position particles over micron length scales.

In the first part of the thesis, I discuss aggregation or separation of pairs of negatively charged polystyrene latex particles deposited on an electrode under ac polarization. I studied the aggregation experimentally by varying the zeta potential of the particles, the electrolyte composition and concentration, and the frequency of the electrode potential. Trajectories of two adjacent particles were recorded by video-imaging, and analyses were performed to determine the relative velocity between the particles. The relative velocity did not depend on the zeta potential of particles, implying the motion was not electrokinetic in nature. The two-particle dynamics were distinctly different between electrolytes containing bicarbonate ion versus hydroxyl ion; pairs of particles aggregated in bicarbonate solutions at frequencies between 30 and 500 Hz but separated at 1,000 Hz, while the pairs separated with hydroxyl ions at all frequencies up to 1,000 Hz. In all cases the relative velocity between a pair of particles was relatively independent of electrolyte concentration and the cation of the electrolyte. The lack of dependence of the two-particle dynamics on the zeta potential of the particles suggests that the phenomenon of aggregation/separation is driven by electrohydrodynamics resulting from nonuniform electric fields at the electrode surface, but no model is able to quantitatively fit the experimental results.

The second part of the thesis discusses manipulation of particles in non-polar solutions with dc electric fields. The goal was to understand ion conduction in these fluids and characterize particle motion in response to two-dimensional fields established in a thin film of the fluid. First ion conduction was studied with solutions of OLOA 371 (primarily the amphiphile poly(isobutylene succinimde)) and dodecane confined between two parallel electrodes. The conductivity of the solutions was measured with a conductivity probe and found to be proportional to the OLOA concentration. The applied dc potential across the parallel electrodes caused an initial current that was proportional to the applied potential, and the conductivity determined from this proportionality agrees with that measured with the conductivity probe. The current decayed to about 10% of its initial value within 10 seconds of applying the potential and then remained nearly constant up to 10 hours. This long-time residual current could be due to continuous polarization of the electrode if the capacitance of the double layer was to increase with time, but it could also be due to charge transfer across the electrode/fluid interface. I applied the Gouy-Chapman model of the diffuse double layer to analyze the transient current within 1 second of the applied potential and obtained the equivalent ionic strength of the solution. Using this ionic strength and the known solution conductivity, I calculated the size of the charge carrying species to be 10 nm in radius for OLOA concentrations over the range 0.5 – 3.5 weight percent. This result is consistent with the notion that the charge carrying species are micelles of PIBS with only one charge per 105 PIBS molecules. I conclude that OLOA/dodecane solutions represent a very weak electrolyte system with charge carriers about 100 times the size of simple ions in aqueous electrolyte solutions.

Conduction of OLOA/dodecane solutions in two-dimensional fields was also studied in a cell consisting of electrode strips on one of the two glass slides confining the solution. A dc potential was applied between the center strip and the two outer strips, one on each side. Theoretical calculations of the electrode cell constant gave results that agree with the experimental determinations from the initial current versus applied potential difference. As with the planar electrode configuration, the initial current decayed in seconds to about 10% of its initial value, and the residual current persisted for hours. Carbon black particles about 1 μm in diameter were allowed to sediment onto the lower glass slide containing the strip electrodes. Their trajectories were video-taped and analyzed during the residual current phase. The particles moved against the direction of the electric field between the strips, indicating they were negatively charged. In the region midway between two strips, the velocity of the particles was nearly constant as expected since the electric field in this region is very uniform. The electrophoretic mobility determined from these trajectories is in reasonable agreement with data for this system published in the literature. The particles accelerated when they approached the more positive strip, and appeared to stall near the lower potential strip. This anomalous motion cannot be explained by dielectrophoresis nor by the higher fields near the edges of the strips. I suggest that electrohydrodynamic flows might be affecting the particle motion near the edges. These flows could arise, for example, by uneven polarization of the strip electrodes.

Surfactant solutions exhibit a wide variety of wetting and dewetting behaviors on high energy surfaces. These behaviors are driven by surfactant self-assemblies at the moving contact line. To probe these self-assemblies, we have undertaken a study of surfactant structure at the three interfaces near a receding contact line. We immerse a hydrophilic silica surface in aqueous solutions of polyethyleneglycol monododecyl ether (C12En, 1≤n≤8) below the critical micelle concentration. The substrate is withdrawn from solution at a speed, U<Ucrit, the critical velocity for pulling a macroscopic film on the solid surface, so that a receding contact line moves across the surface. We determine the area per molecule adsorbed at the solid-liquid and liquid-vapor interfaces, and the structural details of the monolayer deposited to the solid-vapor interface at the receding contact line. We also describe in detail a new technique which we have developed for objectively interpreting data from x-ray reflectivity measurements, our primary tool for probing structure at the solid-vapor interface. We find that the adsorbed amount at the solid-liquid interface is a small-to-negligible contribution to the monolayer deposited at the solid-vapor interface for all n. The primary source of the deposited surfactant is the self-assembled layer at the liquid-vapor interface. The density of the deposited monolayer is substantially less than the density at the liquid-vapor interface. Conservation of mass demands a dividing streamline in the bulk, along which surfactant from the liquid-vapor interface is returned to solution. We note a transition at n=6 from reversible to partially irreversible adsorption, suggesting the ethylene oxide (EO) head groups begin to behave like PEO polymer for n≥6. At the liquid-vapor interface the area per molecule increases monotonically with n, suggesting increasing disorder in the head group region. The deposited monolayer at the solid-vapor interface shows a more complicated, non-monotonic dependence on n. Substantial rearrangement of molecules takes place as surfactant is deposited from the liquid-vapor to the solid-vapor interface. Processes at the receding contact line and the structure of the deposited monolayer show marked transitions at n=3, indicating a significant interaction between head group and substrate for n>3.

By direct microscopic measurements of the meniscus shape in the immediate vicinity of the moving contact line, we quantify the effects of viscous forces on the interface shape. Single parameter fits to the interface shapes predicted by theory give excellent results. The existence of the geometry independent region on the interface has been verified. In analogy to the static contact angle, the interface shape in this geometry independent region forms a well defined characterization of the dynamic wettability of a materials system. We have thus developed and demonstrated an instrument and method for measuring a dynamic contact angle boundary condition. With this technique other measures of dynamic wettability can be properly interpreted and predictive models of spreading can be implemented for the first time. Our results put strong restrictions on the asymptotic functional form of the interface shape predicted by any model which attempts to describe the hydrodynamics occurring on the microscopic scale near a moving contact line. With our methods, the velocity dependence of the parameters which govern the flow in this inner region can be investigated.

In this thesis, we both address two longstanding questions in the wetting literature and reveal new quantitative features of surface diffusion in pure and alloyed metallic systems. By studying equilibrium wetting and precursing film growth kinetics experimentally in Pb, Bi and a Pb-Bi alloy on Cu(111), and by atomistic computer simulation for Ag on Ni(100), we have given the first clear recognition of pseudo-partial wetting – a thin film in equilibrium with a macroscopic body having a non-zero contact angle – in the literature and show that the mass transport mechanism during the film growth is diffusive. In addition, our detailed, spatially and temporally resolved compositional studies of the films, allows us to reconstruct the diffusion coefficient as a function of concentration for the pure Pb and Bi on Cu(111) systems. We connect those concentration dependence variations with surface structure. In the case of Pb-Bi on Cu(111), such reconstruction is not yet possible, but we observe the clear impact of the interaction between the species on the diffusive transport in the film.

The UHV apparatus and our sample geometries provide us with unique capabilities to study wetting and diffusion. It affords us both a clean environment and electron beam based diagnostic tools. The cleanliness of the UHV environment allows us to study wetting phenomena without the complexities of contact angle hysteresis and of solutal Maragoni flows that likely occur due to contamination in an ambient environment. Due to the exceedingly low vapor pressure of our systems, evaporation is negligible, simplifying both the film and contact angle behaviors. In our Pb-Bi alloy system, the low vapor pressures also remove the possibility of differential evaporation of various components which can cause instabilities in contact lines or dominate transport in precursing films. The Auger electron based spectroscopy characterization tools allow measurement of composition and density within the precursing film at submicron resolution. Thus, we can characterize the film growth by tracking the full compositional profile as it evolves with time. Finally, the nonzero contact angle of the systems investigated here ensures that the advancing contact line of a spreading drop is not influencing the film growth.

Wetting on ambient surfaces is history dependent. The macroscopic manifestation of this history dependence is contact angle hysteresis. The hysteresis arises from the chemical and physical heterogeneity of these surfaces. Accompanying this macroscopic phenomenon is unsteady motion of the contact line as it traverses the heterogeneous surface. In fact, this unsteady motion provides the dissipation innate to the hysteretic process.

The emphasis of this thesis has been to develop a logical and feasible way of studying the unsteady contact line motion on heterogeneous, ambient surfaces. Unsteady motion of the contact line has been tracked and characterized. Important lengths and times in the problem have been identified. Our results substatiate models of the metastable states which trap the system, and a common misconception concerning the irreproducible nature of the contact line motion on such surfaces has been dispelled. For the first time the role played by noise in the problem has been unambiguously identified. Our microscopic investigation places the measurement of macroscopic contact angle hysteresis on a firmer footing.

This project investigates the effect of surfactant mixtures on Marangoni transport of colloidal particles adsorbed at the oil/water interface using a narrow-gap, radial stagnation point flow cell. Marangoni transport is created by the transient interfacial tension gradient established when surfactants adsorb from a solution in a laminar flow to an initially clean oil/water interface. The magnitude of Marangoni velocity for surfactant mixtures was compared with the Marangoni velocity for single component surfactants which comprise the surfactant mixtures. The surfactant pair is cationic surfactant cetyltrimethylammonium bromide (CTAB) and anionic surfactant sodium dodecyl sulfate (SDS), and the oil is olive oil. To achieve a certain peak Marangoni velocity, the lowest concentration required was for CTAB as a single component surfactant among all the tested surfactant solutions including surfactant mixtures at molar ratios SDS/CTAB = 10/90 and 90/10, pure SDS, and pure CTAB. Since in a system with Marangoni synergism a particular peak Marangoni velocity can be achieved by a mixture having lower total surfactant concentration than the concentration of either pure surfactant needed to produce that same peak Marangoni velocity, none of these surfactant mixtures yields a positive Marangoni synergism in my measured peak velocity range. Instead, for SDS/CTAB = 90/10 composition, SDS/CTAB mixtures yield negative synergism, where the larger concentration of mixture than either pure surfactant is required to produce a same peak Marangoni velocity within my measured velocity range. After measuring the corresponding interfacial tension for each solution, we found that interfacial tension and Marangoni transport synergism do not occur at the same compositions. Further anionic fatty acids in olive oil form complexes with CTAB at the interface, which alter the Marangoni transport.

Cystic fibrosis is one of the leading causes of death in the United States for which effective treatment that is less harsh for the patients is not readily available. Dry powder inhalation formulations offer many advantages over the traditional liquid or pressurized delivery systems. The spray drying technique is a simple and effective method for preparing dry powder formulations. Powders consisting of lactose, fluorescein di-sodium salt and varying concentrations of sodium dodecyl sulfate (SDS) were spray dried at a temperature well above the boiling point of water. The powder morphology was observed using scanning electron microscopy (SEM). At low surfactant concentrations, the spray drier produced spherical and homogeneous solid particles with a smooth surface. Increasing the SDS concentration was found to induce irregularities in the particle morphology such as formation of hollow shells, wrinkles and collapses in the particle shell, and fragile particle structure. The effect of ambient and saturated humidity on the particles was studied using SEM. The size measurement of particles having different SDS concentrations showed that there was no significant dependence of the compositions considered in this work on the particle size distribution.

In this dissertation, we discuss the investigation of two problems in dynamic wetting: the hydrodynamics of dip-coated, finite-length films of evaporative fluids and the surfactant structures controlling the spontaneous dewetting of a surfactant solution.

While films pulled from non-volatile fluids on a vertical substrate are essentially infinite in length, films pulled from volatile fluids have a finite length. We examine such finite films using three well-controlled oligomer liquids as well as surfactant solutions. We find that the finite length of the film is controlled by a global balance between mass lost by evaporation and mass input by viscous forces. While the attendant thermally driven Marangoni flows have small impact on the mass balance, they do alter the velocity field in the film in the direction parallel to the substrate. Using measured film profiles, we have developed a novel method to determine the combined effects of evaporation and Marangoni flow on velocity and pressure fields in the film. This method is independent of any specific model of the evaporation process. In preliminary studies with surfactant solutions, we observed strong effects of solutal Marangoni flows on dip-coated films.

For the second problem, we examine the structures of self-assemblies left on a solid as a contact line spontaneously retreats across a surface during an autophobing event. We find that surfactants of a continuous structural gradient are deposited: from molecules lying down on the surface with low packing densities in a region never touched by the solution, to molecules standing up with higher packing densities in a region where the contact line has moved slowly. Despite significant free volumes within the self-assemblies, we see no evidence of clustering of molecules. We see a clear correlation between contact line speed and the surfactant structures. We show that the dynamics during at least a later period of the autophobing event is dominated by the time evolution of Young’s force dictated by the self-assembly near the contact line.

Our studies on a series of non-Newtonian fluids suggest looking at dynamic wetting from a new perspective. Flow in the wedge-like region near the moving contact line demands that shear rates increase to very high values as the contact line is approached. Thus, somewhere in the flow field, molecular relaxations in the fluid will be unable to relax, leading to non-Newtonian behavior on some small scale near the contact line. This is crucial because in dynamic wetting, behavior at a very small, micron scale has impact on wetting behavior at the submillimeter and millimeter scale where fluid behavior must be precisely controlled in many applications. We find that non-Newtonian fluids dominated by shear thinning reduce viscous bending near the contact line while fluids dominated by elasticity increase the bending, compared to a Newtonian fluid with the same zero shear viscosity. These changes in the viscous bending can, in turn, impact the dependence of the dynamic contact angle on capillary number (Ca), changing it from the classic ωo ~ Ca1/3 exhibited by Newtonian fluids. We find that even fluids with such short relaxation times and undetectable non-Newtonian behavior in their rheological characterization, do not show dynamic wetting described by models that assume Newtonian behavior. Our work suggests a modified flow field with a region of non-Newtonian behavior intervening between the region at shortest length scales where the classical contact line singularity is alleviated and the region far from the contact line where viscous forces just begin to become significant.

Gradients in the surface excess concentration of surfactants adsorbed at a liquid/vapor interface create surface tension gradients that drive Marangoni flows. This phenomenon may be used to enhance the uniformity of aerosol drug delivery in the lungs of patients with obstructive pulmonary diseases, such as cystic fibrosis. We are investigating the potential use of surfactant-generated Marangoni flows to develop self-dispersing aerosols with improved uniformity of drug delivery in the treatment of partially obstructed lungs. For liquid aerosol systems, an active agent is formulated with an aqueous surfactant solution to enhance the spreading of aerosol droplets after deposition on the airway surface liquid (ASL).

Experiments with deposition of surfactant-laden aqueous drops on model subphases show significantly enhanced drop spreading relative to surfactant-free controls. We find that the surfactant escapes the drop’s contact line and spreads across the liquid subphase surface faster than the contact line. Our work with deposition of surfactant-laden aerosol droplets shows that the resulting Marangoni flows not only cause greater spreading of individual aerosol droplets on the liquid subphase surface but, more importantly, drive a large scale convective expansion of the entire field of deposited droplets across the subphase. The distances spanned by the aerosolized surfactant droplet field expansion are comparable to the lengths of airways deep in the lung where sub-monolayer droplet deposition is expected during aerosol inhalation.

Surfactant-induced Marangoni flows can also be used to enhance dispersal of dry powder medications in lung airways. We find that a physical mixture of solid surfactant and a model drug powder also disperses much further than the surfactant-free powder on the same ASL mimic liquid subphases. We characterized the forces acting on a single moving particle at the interface under the influence of Marangoni flows.

These spreading results suggest that the addition of surfactants to aerosol formulations could enhance aerosol drug delivery by increasing the extent of post-deposition spreading in lungs.

The long-term efficacy of inhaled aerosol medications depends largely on the uniformity of pulmonary drug deposition. Aerosol antibiotics are currently used for treatment of lunginfections associated with cystic fibrosis (CF). Due to the decreased ventilation caused by airway obstructions associated with CF, drug is not deposited uniformly and certain regions of the lung go untreated. Previous in vitro studies have shown that when the deposited surfactant solution has a surface tension lower than that of the mucin solution, surface tension gradient driven or “Marangoni” flows are created [1]. It has been observed that surfactant enhances aerosol spreading on hydrated mucus surface as compared to saline [2]. We study the spreading of aerosolized aqueous surfactant solutions on entangled, aqueous subphase solutions, used to mimic the airways surface liquid (ASL) of the lung. The goal of this in vitro study is to determine the optimal surfactant and concentration that can be used as a carrier for pulmonary aerosol drug delivery in a future in vivo study. Experiments were carried out to compare the spreading of three surfactants as a function of their concentration on 1 % w/v Polyacrylamide solution. Amongst the three surfactants that were compared, the most effective surfactant as potential aerosol drug carrier, Tyloxapol, was then deposited on a subphase of Porcine Gastric Mucin to study its spreading efficacy as a function of its concentration. The effect of type of subphase on spreading behavior was also studied for Tyloxapol by carrying out experiments keeping all parameters constant except the subphase type.

It has long been known that drops of soluble surfactant solution induce Marangoni flows at air-liquid interfaces. These surfactant drops create a surface tension gradient, which causes an outward convective ow at the fluid interface. In this thesis, we show that aqueous phospholipid dispersions may be used for this same purpose. In aqueous dispersions, phospholipids aggregate into vesicles that are not surface-active, so these dispersions do not initiate Marangoni ow. However, aerosolization of these dispersions causes the vesicles to shear open, allowing access to the surface-active lipid monomer within. Deposition of lipid via aerosolization leads to surface tensions as low as 1 mN/m on water and can induce spreading on entangled polymer subphases even in the presence of pre-deposited phospholipid layers.

Most methods for introducing a surfactant solution to a liquid subphase involve dropwise deposition, either by pipette or by aerosol. In order to better understand the behavior of these miscible drop/subphase systems, we study drops of solvent as they slowly diffuse into their polymer solution subphase. Previous work has shown that, even when two fluids are completely miscible, they can maintain a detectable “effective interface” for long times. Effective interfacial tension has been probed using a number of methods, but that work has not extended to the three-phase system of a uid drop on top of a miscible pool. By observing drop shapes, we show that these drops obey immiscible wetting conditions, that their shapes obey the augmented Young-Laplace equation, and that the low density difference of these systems provides a unique ability to probe very small pressures across the drops.

The ability to use phospholipids naturally found in the lung as spreading agents on low surface tension surfaces of macromolecular solutions, in tandem with an understanding of the effective interfaces controlling drop shape, yields a deeper understanding of how we might improve surfactant-driven delivery of therapeutic agents along the complex airway surfaces in the lung.

Intricate and fascinating physics control how liquids spread on solids. Wetting is also of major importance in many technological processes. Dynamic wetting is an intrisically complex problem with difficulties arising due to the unique hydrodynamics which must operate in the microscopic region near the contact line and control the macroscipoc streading of a liquid. Models and experiments on wet and dry substrates exist, but all are in the slow flow regime where viscous forces very near the contact line are especially important. In this thesis, we have moved to new regimes where thin films exist near the moving contact line and hwere inertia is becomiing important. We examine the hydrodynmics using video microscopy and digital image analysis to measure the liquid/vapor interface shape in a microscopic region near the contact line. We have developed and experminatally verified an analytical model describing the interface shape close to the moving contact line in the presence of a film. We have observed that molecular scale and micron scale films have different effects on the hydrodynamics near the contact line. We have also investigated the effects of inertia on the hydrodynamics close to contact lines. We look at the case wehre steady state motion of the contact line leads to moderate Reynolds numbers. We have established that inertia decreases the dynamic curvature and the apparent contact angle as the Reynolds number increases.

Dynamic wetting is an intrinsically complex problem because of the unique hydrodynamics in the microscopic region near the contact line. While the region is small in extent, it essentially controls macroscopic wetting phenomena. Both the fluid motion and the material properties of the liquid and the solid surface influence the dynamic wetting behavior. We have expanded the regime treated by previous research in wetting to new regimes: unsteady wetting and wetting by a specific class of non-Newtonian fluids, shear-thinning fluids. Our basic experimental technique is the precise measurement of the liquid/vapor interface shape using video microscopy. Our main analysis tool is comparison to hydrodynamic models based on Stokes flow. We have also developed a quasi-steady model to describe the contact angle relaxation during the unsteady spreading process and a two-region shear-thinning model to predict the shear thinning effects on steady state interface shapes. Deviations from these models help us understand the influence of unsteady flow and non-Newtonian behavior on wetting. In our unsteady wetting studies, we observe both unsteady and quasi-steady behaviors. The unsteadiness in our moderately high viscosity system arises from temporal relaxation of the contact angle rather than inertial effects or momentum diffusion. In our study of non-Newtonian fluids, the shear-thinning properties reduce the viscous force in the fluid flow and lead to a less curved dynamic interface than that of a Newtonian fluid at the same contact line speed and zero-shear viscosity.

Almost any treatment of wetting in natural or technological settings must deal with fluids containing surface active materials. While previous work identified Marangoni flows and surfactant transport as important in wetting by surfactant solutions, it did little to quantify the impact of these phenomena on the dynamic wetting of soluble surfactant systems

In this thesis, we have experimentally investigated the hydrodynamics of receding contact lines and surfactant deposition from soluble surfactant solutions in two velocity regimes: at low substrate withdrawal speeds where the surface emerges dry, and at higher withdrawal speeds where a film is entrained on the surface. We have developed a technique to measure the fluid velocity on the liquid-vapor interface and this enables us to quantitatively study the Marangoni stresses and surfactant transport near the contact line. We determine the Marangoni stresses near the contact line. While they are the least dominant stresses, they do alter the flow fields compared to those of a pure fluid. The surfactant-surface interactions at the contact line govern the Marangoni stresses and thereby control the hydrodynamics. In the two velocity regimes examined, we observe two different mechanisms of surfactant deposition on the solid-vapor interface. We also see a sharp transition in the boundary between these two wetting regimes with surfactant concentration. We find that this transition is universal across several surfactant-surface systems.

Many industrial and natural processes involve colloidal suspensions, and a major challenge in the engineering of colloidal processes is understanding the stability and flow behavior of the suspension. Both properties depend ultimately on the forces between colloidal particles. These forces are seldom known, and the classical theory of colloidal forces, DLVO theory, has gone largely untested for colloidal particles at close separations. This thesis has sought to answer the following questions: 1) Does DLVO theory correctly predict the normal forces between two polystyrene latex particles or between a polystyrene latex particle and a silica particle? 2) Are the tangential forces and restraining torques between colloidal particles zero, as DLVO theory predicts, and if not, do they depend upon the normal forces?

In order to measure the forces between colloidal particles, we have developed the experimental technique of differential electrophoresis. The forces are measured by interpreting experimentally observed particle trajectories with solutions of the hydrodynamic and electrostatic equations for two particles. Both normal and tangential forces (and restraining torques) can be measured with sub-piconewton resolution.

The following are the primary findings of this research: 1) DLVO theory is inadequate to describe the normal forces between polystyrene latex particles or between silica and polystyrene latex particles in aqueous suspension. The data show that the normal forces holding these colloidal particles together were sometimes greater than 50 pN, which is at least an order of magnitude higher than what DLVO theory predicts. 2) Tangential forces play a significant role in the dynamics of colloidal doublets. Doublets can act as single rigid bodies, as loosely coupled spheres (freely rotating doublets), or something in between. Tangential forces from 0 to 4 pN were measured, and these were not always correlated with the normal forces between two particles. This rich behavior of tangential forces and restraining torques could impact the way colloid scientists view the flow behavior of dense colloidal suspensions. The presence of surface heterogeneity provides one hypothesis for why the large discrepancies exist between the force data and the DLVO predictions, and recent theoretical papers support this position.

Dynamic wetting poses a complex fluid mechanical problem of both fundamental and technological importance. This complexity arises because the fundamental geometry of the wedge-like region near the moving contact line requires that fluid elements experience ever increasing deformation rates as they pass into and out of that region. Thus, somewhere in the flow field, relaxation modes caused by segmental and chain motions in a viscous polymeric fluid may be unable to relax, leading to nonlinear stresses and possible changes in the shape of the deformable free surface. In this thesis, we investigate the wetting behavior of model viscoelastic fluids, Boger fluids, and the viscous oligomeric fluids which are the base solvents in these Boger fluids. We determine that the dominant relaxation mode in Boger fluids has only small effects on both viscous bending and dynamic contact angles even approaching the air entrainment limit. A lubrication analysis, used to examine the impact of the linear elasticity due to dominant relaxation mode, also shows only small effects on the interface curvature and on the dynamic contact angle. The viscous oligomeric base fluids, which have only very short relaxation times, also show deviations from the viscous bending that a Newtonian fluid exhibits near the contact line. Our experiments prove that these deviations do not arise due to the interactions between the fluid and the solid surface. It is very possible that this behavior arises from the influence of the non-Newtonian behavior at a small length scale on the Newtonian hydrodynamics at larger distances from the contact line. Finally, a new relaxation regime has been discovered as the final static contact angle is approached in the spontaneous relaxation of the viscous fluids with non-zero static contact angles. This new regime may be due to slow rearrangements of the polymeric molecules altering the static force balance at the contact line. predictions, and recent theoretical papers support this position.

Dynamic Wetting is the process wherein a fluid body spreads across a solid surface while displacing a second immiscible fluid. Such phenomena are ubiquitous in both natural and industrial settings. The work presented here has been done in the pursuit of two goals. The first goal has been to produce the first systematic study of the relation between material properties and the dynamic wetting of a material system. This stands in analogy to the work of Zisman, who examined the static wetting of a variety of systems to understand what materials characteristics controlled static wetting phenomena. The second goal has been to assess the applicability of theoretical models of the unique hydrodynamics which occur near the contact line. This is analogous to later attempts to test models of the chemical interactions at interfaces which could account for the static wetting behavior.

We characterize the dynamic wetting of a viscous bulk polymer liquid by the shape of the liquid-vapor interface near the moving contact line as a function of the contact line speed. We show that the dynamic wetting behavior is strongly affected by several properties of the material system. These properties are the surface energy of the solid substrate, the chain length and end termination of the polymer, and the presence or absence of a multi-molecular, mobile water layer at the solid surface. The dynamic wetting is not a universal function of the capillary number and we show that the parameters describing the hydrodynamics near the contact line must be functions of the contact line speed. Our measurements are more consistent with slip models of the hydrodynamics in the region near the contact line than with models which depend on precursing thin films.

We investigated the dynamics of thin films driven by Marangoni stress on the inner surface of a glass cylinder. The surface of the cylinder is rigid and stationary. The axis of the cylinder is horizontal. The inner wall of the cylinder was coated with a very thin layer of 1 wt% aqueous poly(acrylamide) solution (PA). Surfactant drops were then placed on the PA either on the bottom or top inner surface of the cylinder, and their spreading was monitored. As the drops spread on the PA, the radius of the surfactant front, the radius of the contact line, and spreading area were recorded. Different radii of cylinders were also tested. We found that gravity plays an important role in the spreading process. For the surfactant front, gravity increases the spreading velocity down the sides of the wall (arc direction) and decreases the velocity along the horizontal direction (axis direction). Gravity also increases the total spreading area of the contact line when the surfactant drop is added at the top of the cylinder as compared to a drop added at the bottom. However, gravity does not influence the spreading area of the surfactant front. Therefore, we believe that total spreading area of the surfactant front may be determined by surfactant inventory alone.

We also compared the spreading of a surfactant drop placed on a flat plate with spreading of a drop placed at the bottom of a horizontal cylinder. In this case, the spreading area of the contact line and the surfactant front are the same, however, the spreading pattern of the surfactant front changes significantly. On a flat plate, the drop spreads in a circle, however, for a small radius cylinder, the spreading in the axis direction moves much more quickly than in the arc direction.

With the observation of spreading on a flat plate and in a cylinder, we conclude that the spreading mainly has three features: the contact line, the surfactant front, and a ridge of accumulated liquid around the contact line. We believe that this third feature is the liquid dewetting from the contact line area and that it is composed primarily of solvent from the surfactant drop.

We also explored the influence of solvent inventory and surfactant inventory on spreading. We found that increased the surfactant inventory will increase the spreading area of the contact line and the surfactant front. However, increased solvent inventory does not have influence on the spreading area of the contact line or the surfactant front.

In this project, the effect of powder with different amounts of sodium dodecyl sulfate (SDS) on Marangoni flow is investigated. The powder is made by using a spray dryer, then it is added on a thin water sub-phase by a scoop. The powder causes the fluid to move outward along the surface tension gradient. In the sub-phase solution, erythrosine is added and its absorbance change is used to indicate the height change of sub-phase, and the absorbance of fluorescein disodium salt (FDS) in the powder is used to calculate the amount of powder added. The results show that the powder composition and SDS mass added would change the Marangoni rate, outflow volume, shape of flows and number of Marangoni waves.

Coalescence of liquid drops is important in many natural and industrial processes, such as raining, inkjet printing and coating applications. The coalescence for sessile drops is more complicated due to the additional interplay between the drops and solid surface. This work examines the impact of gravity, interfacial tensions and wetting properties on both the static and dynamic aspects of the coalescence of sessile drops. In the presence of gravity, seven dimensionless parameters are identified to describe the axisymmetric configuration of a compound sessile drop after coalescence. A stability criterion is established based on the perturbation of Laplacian shape and the stability criterion is numerically evaluated in the zero Bond number limit. Surface Evolver simulations and experiments are performed for compound sessile drops at small and intermediate Bond numbers. Both simulations and experiments agree closely with the zero Bond number analysis, exhibiting a small discrepancy at intermediate Bond number. For the dynamics of sessile drop coalescence, experiments are performed for miscible fluids with similar surface tensions but different densities and viscosities. The coalescence behavior shows three distinctive stages with well separated timescales: an initial stage of fast bridge healing process, an intermediate stage of advective motion for fluids with different densities, and a final stage of diffusion. A dimensional analysis shows that the flow behavior for the advective motion resembles gravity current. A more detailed analytical model based on the lubrication approximation is conducted and demonstrates good qualitative agreement with the advective motion during the sessile drop coalescence.