Edward P. Furlani
614 Furnas Hall
Research in the Furlani group involves multidisciplinary modeling for emerging applications in the fields of micro and nanoscale science and technology. The main thrust of this work is the development of mathematical methods and models to enable the development of innovative materials and devices with design features and functionality that are engineered at the nanometer to micrometer length scale. Current research interests span the areas of: microfluidics, broad applications with an emphasis on biomedical devices; nanophotonics, metamaterials, plasmonics, biosensing applications; and magnetic particle applications, transport, assembly and bioapplications, magnetic drug delivery, magnetic-assisted gene transfection (magnetofection) and bioseparation.
- Multidisciplinary modeling of materials, devices and systems
- Computational fluid dynamics — multiphase and inkjet systems, heat and mass transfer, nanofluids
- Microfluidics and Nanofluidics— devices and processes — design and simulation
- BioMEMS — point-of-service diagnostic applications
- Optofluidics — biosensors
- MOEMS/Photonics — design and simulation
- Nanophotonics — metamaterials, plasmonics, waveguides, resonators
- Nanomagnetics — bioseparation, drug delivery, magnetofection, self-assembly
- MEMS — design and simulation
Microfluidics and nanofluidics are interdisciplinary fields that involve the science and technology of fluid flow through materials and systems with micro to nanoscale features. Research and applications in these fields have proliferated in recent years due to unique advantages of small-scale fluidic processes combined with rapid advances in materials development and system integration. Microfluidic and nanofluidic systems enable highly efficient, repeatable and rapid processing of small fluid samples for applications that can involve integrated sequential or multiplexed processes such as chemical reactions, fluid heating, mixing and sensing. Research in the Furlani group involves modeling and simulation towards the development of novel processes and devices. Much of this work emphasizes the use of state-of-the-art computational fluid dynamic (CFD) analysis for studying fluidic phenomena involving Newtonian and Non-Newtonian fluids, conjugate heat transfer, phase change analysis, free-surface and multiphase analysis, fluid media interactions, flow through porous media, fully coupled fluid-structure and particle-fluid interactions and colloids.
Human pluripotent stem cells (hPSCs) are capable of differentiating into all somatic cell types and hold great potential for future clinical applications. Recent studies on cell culture in stirred tank bioreactors indicate that shear forces resulting from the interaction of the cells with the turbulent eddies significantly affect the differentiation propensity of stem cells and, furthermore, excessive shear stress can cause mechanical damage to the cells. Therefore, bioreactors used for the culture of these cells must limit the intensity of shear while still providing adequate mixing for uniform dispersion of cells throughout the reactor. The intensity of shear depends on the size of eddies that exist in the bioreactor relative to the microcarrier particles. It has been reported that cell damage can be avoided if the particle size is smaller than the size of the smallest eddy as characterized by the Kolmogorov length scale for turbulence. Thus, the Kolmogorov length scale is a key determinant of the differentiation outcome of cultured stem cells. Because there is significant spatial variation in shear intensity in stirred vessels, CFD modeling is essential to predict the precise shear conditions experienced by cells.
We use a synergistic combination of computational fluid dynamic (CFD)-based simulations and experiments to study the effects of the turbulent shear stress on the cell culture performance. The effects of parameters such as the impeller speed, culture medium fluid properties and cell size on the steady-state shear stress acting on the cell-laden microcarrier particles in the bioreactor are studied using CFD analysis. This is used to predict the precise shear conditions experienced by cells and identify optimum operating conditions that prevent turbulent shear damage of the cells. In addition, the effect of shear on the pluripotency of hPSCs is studied by determining the percentage of cells carrying the pluripotency markers Oct4, Sox2 and Nanog using flow cytometry and quantitative PCR. The cell cycle of hPSCs under different shear stress conditions is studied to determine the doubling time and the length of the G phase of the cells.
The first animation above shows a 3D CFD simulation of the Corning’s bench-scale stirred-flask bioreactors running at 60 rpm. A state-of-the-art multiphysics CFD program Flow-3D (www.flow3d.com) was used for this analysis. In the computational model, the reactor is filled with 50 ml of cell culture media and loaded with micro carrier particles that are 150 microns in diameter in a concentration of 0.5 g/ml. The simulation takes into account fully-coupled particle/fluid interactions wherein the particles move in response to viscous drag imparted by the flow field and their motion, in turn, alters the flow. These interactions can be visualized in the second animation, which provides a close up view of an inside portion of the reactor. Here, the particles are colored by their x-velocity. The third animation shows the distribution of shear stress acting on the particles in a cross-sectional plane of the reactor. Regions that experience stress that could cause cell damage can be readily identified. The fourth animation shows the Kolmogorov length scale distribution inside the reactor at the same plane. This length scale is a key criterion in determining fate of the stem cells for differentiation. It has been reported that cell damage can be avoided if the particle size is smaller than the size of the smallest eddy as characterized by the Kolmogorov length scale for turbulence. The average eddy size can be controlled by changing the stirring speed. In this way, we identify the optimum conditions to prevent the turbulent shear damage.
Nanophotonics is an interdisciplinary field that involves the study and application of light-matter interactions at the nanometer length scale. Artificial nanostructured photonic materials can be designed to confine, enhance, slow down, filter and generally manipulate light to enable unpresidented control that cannot be achieved using naturally occurring materials. Nanophotonic materials are having a transformative impact in the fields of imaging, sensing, display, communications and computing. Research in the Furlani group involves the use of computational electromagnetic modeling for fundamental understanding of nanophotonic phenomena and for the development of novel nanostructured materials and integrated microdevices for applications broadly in the areas of sensing, imaging and biomedical therapy. Current projects include the development of chiral materials and planar metametaterials for controlling the polarization of light and applications of plasmonics for sensing, imaging, photothermal therapy and all optical manipulation of colloidal nanoparticles.
The interest in nanoscale photothermal phenomena has grown steadily in recent years along with new applications in fields such as analytical and material chemistry, nanophotonics and biomedicine. One of the most promising areas of research in this field involves the use of plasmonics wherein laser light is used to remotely heat sub-wavelength noblel metal (e.g. Au and Ag) nanoparticles. Such particles have unique optical properties that make them well-suited for photothermal heating, most notably they exhibit localized surface plasmon resonance (LSPR).
At plasmon resonance, there is a coherent oscillation of free electrons within the particles that gives rise to intense absorption and scattering of incident light, as well as highly localized field enhancement. The absorbed photon energy is efficiently converted to heat, which is ultimately transferred to the surrounding medium.
We use computational models to investigate fundamental photothermal effects associated with nanosecond-pulsed, laser-heated colloidal plasmonic nanoparticles.* We use a combination of numercal photonic and computational fluid dynamc (CFD) analysis to simulate energy conversion within the nanoparticles at plasmon resonance, heat transfer from the particle to the surrounding fluid and phase change of the fluid leading to homogenous bubble nucleation. We study various nanoparticle geometries such as spheres, rods, rings and tori. and show that process parameters such as the laser intensity, incident wavelength, pulse duration and shape of the nanoparticles can be tuned to optimize the photothermal process.
We use 3D full-wave time-harmonic field theory to study the absorption spectra of the metallic nanoparticles as a function of their dimensions, dielectric properties and orientation relative to the polarization of the incident field. A typical computational model is shown in Fig. 1. The illustrates an Au nanotorus being illuminated with a linearly polarized planewave with teh E field parallel to the x-axis. Fig. 1b shows the absorption spectra for this nanoparticle, which indicates a plasmon resonance of approximately 780nm.
We use CFD analysis to study the thermofluidic behavior of laser-heated nanoparticles in fluid. This is used to predict thermal, pressure and f low effects including the temperature rise in the particle, heat transfer from the particle to the fluid, phase change with in the fluid leading to homog eneous bubble nucleation, the dynamic behavior of the bubble as it expands and collapses, and the temperature, pressure and flow throughout the fluid during the entire process. The Flow-3D CFD program (www.flow3d.com) was used for this analysis.
Figure 2 shows the temperature of the torus throughout the photothermal process along with corresponding images of th e bubble dynamics. Initially the nanotorus is at ambient temperature. After 0.2 ns it is illuminated and its temperature begins to rise. During the first 3.4 ns of heating, its temperature gradually increases (Fig 2a) to the superheat temperature, at which point a bubble is nucleated around it. Once this occurs the torus is surrounded by vapor and its temperature increases rapidly as it is still absorbing energy. It reaches a peak temperature of approximately 1 000K, which occurs at the end of the heat pulse (4.1 ns), at which point it is completely surrounded by vapor (Fig. 2b). As soo n as the bubble has nucleated, it e xpands and reaches its maximum size at 5.4 ns after the onset of heating. At this time the bubble has a spherical shape, approximately 80nm in radius.
An interesting feature of this process is the residue of an isolated drop of heated fluid that forms in the middle of the torus during the bubble expansion as seen in Fig. 2c. Eventually, 8.7 ns after the onset of heating, the nanobubble collapses, bringing fluid back in contact with the torus (Fig. 2d). Consequently, it slowly cools to the ambient temperature as more of the fluid comes in contact with it (Fig. 2e). It is instructive to note that the capillary force that acts to collapse the bubble is relatively weak because of the relatively large radius of curvature that defines the fluid vapor interface as it gets closer to the torus. Thus, the nanobubble requires a substantial amount of time to completely collapse, compared to other geometries.
Figure 2. Photothermal heat cycle of a nanotorus (cross-sectional view): plot of nanotorus temperature vs. time, pulse duration indicated by red arrow and dashed line and inset plots showing various phases of the thermal cycle; (a) initial heating, (b) nanobubble formation, (c) nanobubble (maximum size), (d) nanobubble collapse, (e) cooling.
Figure 3. CFD simulation of bubble dynamics
Figure 3 shows a CFD simulation of pulsed-laser bubble generation and dynamics shown in Fig. 2.
*E. P. Furlani and I. Karampelas and Q. Xie, “Analysis of Pulsed Laser Plasmon-assisted Photothermal Heating and Bubble Generation at the Nanoscale,” Lab on a Chip, 7;12(19):3707-19, DOI 10.1039/c2lc40495h, 2012.
The interest in compact, portable and inexpensive biosensors has grown dramatically in recent years due in part to advances in microfluidics, especially lab-on-a-chip technology. A relatively new and promising approach to biosensing involves optofluidicswhere optic and fluidic functionality are integrated into a microsystem to leverage their combined advantages . Microfluidic functionality enables compact and rapid processing of small biofluid samples, and optical functionality enables high detection sensitivity of target biomaterials within these samples. To date, various optofluidic sensing devices have been developed.
Applications of magnetic micro and nano particles are proliferating, especially in the fields of microbiology, biotechnology, nanomedicine and lab-on-a-chip technology. This is due, in part, to the fact that biocompatible magnetic particles can be functionalized to selectively bind to (label) biomaterials from proteins to whole cells, and can be manipulated (targeted and sorted) noninvasively using an external magnetic field. We are developing computational models for predicting field-directed transport and assembly of magnetic particles, dynamics of particle microstructures and bioapplications of magnetic particles including drug delivery for cancer therapy, microfluidic-based bioseparation and sorting, and magnetic field-assisted gene transfection (magnetofection).
Figure 4. Magnetofection system and modeling
Magnetic particles are used to deliver gene vectors to target cells for uptake in a process known as magnetofection. Magnetic particle-based gene delivery has been successfully demonstrated for all types of nucleic acids and across a broad range of cell lines. It is well suited for multiwell culture plate systems wherein magnetic particles with surface-bound gene vectors are introduced into culture wells, and a magnetic force provided by rare-earth neodymium iron boron (NdFeB) magnets beneath and aligned with the wells attracts the particles to the cells for uptake. Magnetofection is performed using conventional multiwell culture and magnet plates in which target cells are located at the bottom of each well. Rare-earth (neodymium iron boron NdFeB) magnets beneath the wells provide a magnetic force that attracts the particle-gene vector complex towards the cells (Fig. 4a,b,c). We have developed models to predict the magnetic force (Fig. 3d), particle transport, and accumulation on the cells (Fig. 4e).*
*E. P. Furlani and X. Xue, “Field, Force and Transport Analysis for Magnetic Particle-based Gene Delivery,” Microfluidics and Nanofluidics, 13 (4), 589-602, 2012.
Magnetic Particle Dynamics and Self-Assembly
The interest in magnetic nanoparticles and ferrofluids has grown substantially in recent years as their applications continue to proliferate. Current applications include field directed transport of biomolecules and therapeutic drugs, enhanced gene transfection, bioseparation and sorting, high-density magnetic data storage, ferrofluidic seals and pumps, microfluidic mixers, and highly sensitive magnetoresistive-based sensors, among many others. However, despite the widespread and growing use of magnetic nanoparticles, there are many fundamental aspects of their behavior that remain unknown. Areas of particular interest are use of field-directed assembly to form micro- and nano-structured magnetic media and the controlled manipulation of assembled structures using time-varying fields.
We are developing computational models for predicting the assembly of magnetic particles in high-gradient fields and the dynamics of particle-based microstructures in time-dependent fields. The model involves the numerical integration of a Langevin equation that accounts for interparticle dipole-dipole effects, (that drive particle assembly), viscous drag, fluid-mediated (hydrodynamic) particle-particle interactions and a stochastic force to account for Brownian motion. We use dynamic time-stepping and analytical expressions for the magnetic force to greatly accelerate the computations. We discuss the model in detail and demonstrate its use in the analysis of both high-gradient and time-varying magnetic particle systems.
Coupled Particle-Fluid Dynamics and Self-Assembly in a Uniform Magnetic Field
Fugure 5. Coupled Particle-Fluid Dynamics
In a uniform magnetic field, magnetic particles become magnetized and assemble into chain-like microstructures due to dipole-dipole interactions. The microchains tend to align with the direction of the external field as shown in the simulation of Fig. 5. In this analysis, a uniform field is applied upward in the z-direction and the particles assemble into discrete chain-like structures that repel one another. This analysis takes into account fully-coupled particle fluid interaction wherein the fluid provides a viscous drag on particle motion and the moving particles, in turn, impart motion to the fluid. The Flow-3D CFD program (www.flow3d.com) was used for this analysis.
Assembled Microstructure Dynamics in Rotating Magnetic fields
Time-varying magnetic fields can be used to manipulate magnetic particles for applications such as micromixing. Figure 6 shows a simulation of the behavior of a magnetic particle chain in the presence of a rotating magnetic field. The stability of the chain-like structure depends on the strength and frequency of the external magnetic field, viscosity of the surrounding fluid, and the properties of particles, etc. When the frequency is low, the chain is stable and there is a slight delay between the orientation of the chain and the external magnetic field, which depends on the length of the chain.
Fugure 6. Assembled Microstructure Dynamics in Rotating Magnetic fields
However, in a high frequency magnetic field, the chain breaks into two parts that rotate independently with the external field without delay. These chain segments temporarily reassemble into a longer chain and then break apart again in a time-wise periodic fashion as shown in the simulation.
The assembly of magnetic particles into desired patterns can be controlled using template structures. Figure 7 shows the field-directed assembly of magnetic particles using a soft-magnetic ring structure template embedded in a substrate. When a unifom external field is applied to this system the ring becomes magnetized and produces a localized high-gradient magnetic field that attracts the magnetic particles to the surface of the ring where they assemble. An interesting feature of the assembly process is that the particles form a ring-like structure and are evenly spaced due to a repulsive dipole-dipole force between neighboring particles that exists once the particles reach the substrate.
Particles assemble in a ring-like structure with an even separation between neighboring particles due to a repulsivemagnetic dipole-dipole force between neighboring particles.
Last Updated:Jan 2014