Flow boiling in microchannel heat sink offers an effective cooling solution for high power density micro devices. A three-dimensional numerical study based on volume of fraction model (VOF) coupled with evaporation condensation model accounting for the liquid-vapor phase change is undertaken to recreate vapor bubble formation in saturated flow boiling in wavy microchannel. Constant wall heat flux imposed at the bottom surface of the substrate while other faces are insulated. To understand the conjugate effects, simulations has been carried out for substrate thickness to channel depth ratio (δ sf ∼ 1–5), substrate wall to fluid thermal conductivity ratio (k sf ∼ 22–300) and waviness (γ ∼ 0.008–0.04). Bubble nucleation, growth, and departure of bubble plays a significant role in heat transfer and pressure drop characteristics in two-phase flow in wavy microchannel. The coolant (water) temperature at the inlet is taken to be 373 K while flow was at atmospheric pressure. This makes shorter waiting period of bubble nucleation, and the number density of bubbles on the solid surface increases. This results in enhancement of the boiling effect, and thus with the presence of bubbles, the mixing of laminar boundary layers improves and enhances the overall heat transfer coefficient. Channel amplitude play an important factor that can suitably reduce the friction factor and enhances the heat transfer coefficient.

Heat dissipation is a significant challenge for three-dimensional integrated circuits (3D IC) due to the lack of heat removal paths and increased power density. In this paper, a 3D IC system with an embedded microfluidic cooling heat sink (MFHS) is presented. In the proposed 3D IC system, high power tiers contain embedded MFHS and high-aspect ratio (23:1) through-silicon-vias (TSVs) routed through the integrated MFHS. In addition, each tier has dedicated solder-based microfluidic chip I/Os. Microfluidic cooling experiments of staggered micropin-fins with embedded TSVs are presented for the first time. Moreover, the lateral thermal gradient across a chip is analyzed with segmented heaters.

Polymer electrolyte fuel cells (PEFCs) are highly expected as a next-generation power supply system due to the purity of its exhaust gas, its high power density and high efficiency. The polymer electrolyte membrane is a critical component for the performance of the PEFCs and it is important to understand the nanostructure in the membrane to enhance proton transport. We have performed an atomistic analysis of the vehicular transport of hydronium ions and water molecules in the nanostructure of hydrated Nafion membrane by systematically changing the hydration level which provides insights into a connection between the nanoscopic and mesoscopic structure of ion clusters and the dynamics of hydronium ions and water molecules in the hydrated Nafion membrane. In this study, classical molecular dynamics simulations are implemented using a model of Nafion membrane which is based on DREIDING force field and newly modified and validated by comparing the density, water diffusivity, and Nafion morphology with experimental data. The simulated final density after the annealing procedure agrees with experiment within 1.3 % for various water contents and the trends that density decreases with increasing hydration level are reproduced. In addition to determination of diffusion coefficients of solvent molecules as a function of hydration level (from λ = 1 up to λ = 18), we have also calculated radial distribution functions and static structure factors not only to clarify the structure of water molecules and hydronium ions around the first solvation shell of sulfonate groups but also to validate the mesoscopic periodic structure among water clusters. The diffusion coefficient of water molecules increases with increasing hydration level and is found to be in good agreement with experimental data. The diffusion coefficient of hydronium ions has showed that general trends in the experimental data are reproduced by the simulations although the classical models have the limitation of probing hydronium dynamics. The static structure factors of liquid molecules at low wave length provide insights into the periodic structure of the inter-water clusters. These results are consistent with the Gebel’s model based on small-angle X-ray scattering that considers the dry membrane to be made of isolated spherical ionic clusters of radius ∼7.5 Å that swell with increasing hydration.

Temperature management in microsystems is a technical problem with an increasing importance: although the power consumption of integrated circuits is not increasing, due to further miniaturization the local power density is still increasing. Moreover, in the near future more and more micro components will be integrated in flexible system-in-foil (SIF) packages. These packages can contain ultra-thin (8–50 micron) flexible embedded silicon chips combined with polymer electronics, optical systems and microfluidic channels e.g. for point-of-care diagnostics. However, the low thermal conductivity of the polymeric package is aggravating the heat management problem. The life span of micro components, but also the performance of some micro components, like (O)LEDs, can be strongly temperature dependent. Therefore an adequate temperature control is required. The thermal management problems can potentially be addressed by embedding micro-channels containing a flowing cooling medium in close proximity and preferably directly underneath the electronic circuit. However, many applications do not allow for external pumps and therefore pumping needs to be integrated in these channels as well. In this paper some promising integrated micro pumping techniques, like AC-electro osmosis and ferrofluidic pumping, will be described and discussed. The multi-physics modeling approach will be presented and the numerical results will be analyzed and compared with flow fields that are measured by 3D astigmatism micro particle tracking velocimetry (3D micro-PTV).

In the present work, a chip embedded thin film current collector for vanadium fueled microfluidic fuel cells is proposed. Embedded current collectors are expected to increase the number of contact points with the porous electrodes and thereby reduce the overall contact resistance and ohmic losses of the cell. The micromachining based thin film process is compatible with the overall cell fabrication scheme, which is based on soft lithography, and does not require a substantial modification of the original cell design. The flow-through microfluidic fuel cell architecture with porous carbon paper electrodes is used as a proof-of-concept in this study. Our preliminary study shows that the peak power density of the cell with the current collector is increased by 79% at 300 μL min −1 compared to an otherwise identical cell without current collector, indicating that the contact resistance is significantly reduced. Electrochemical impedance spectroscopy analysis is carried out to estimate the overall cell ohmic resistance and confirms a 32% reduction using the current collectors.

The implementation of high power density, multi-core central and graphic processing units (CPUs and GPUs) coupled with higher clock rates of the high-end computing hardware requires enhanced cooling technologies able to attend high heat fluxes while meeting strict design constrains associated with system volume and weight. Miniature loop heat pipe (mLHP) systems emerge as one of the technologies best suited to meet all these demands. This paper investigates experimentally a mLHP system designed for workstation CPUs. The system incorporates a two-phase flow loop with capillary driving force. Since there is a strong demand for miniaturization in commercial applications, emphasize was also placed on physical size during the design stage of the new system. Hence system weight is reduced to around 450g, significantly smaller than that of commercial coolers consisting of copper heat sinks that weight around 782g. Experimental characterization shows that the system can reach a maximum heat transfer rate of 170W with an overall thermal resistance of 0.12 K/W. The heat flux is 18.9 W/cm 2 , approximately 30% higher than that of larger size commercial systems. To further miniaturize the evaporator module while maintaining the same heat flux, a new structure for the porous evaporator is proposed, which consist of a porous bi-layer, with nanopores at the top surface. The role of the nanoporous layer is to provide a larger surface area for phase-change, enhancing the evaporation rate.

Evaporation of liquids is of major interest for many topics in process engineering. One of these is chemical process engineering, where evaporation of liquids and generation of superheated steam is mandatory for numerous processes. Generally, this is performed by use of classical pool boiling and evaporation process equipment. Another possibility is creating mixtures of gases and liquids, combined with a heating of this haze. Both methods provide relatively limited performance. Due to the advantages of microstructure devices especially in chemical process engineering [1] the interest in microstructure evaporators and steam generators have been increased through the last decade. In this publication several microstructure devices used for evaporation and generation of steam as well as superheating will be described. Here, normally electrically powered devices containing micro channels as well as non-channel microstructures are used due to better controllability of the temperature level. Micro channel heat exchangers have been designed, manufactured and tested at the Institute for Micro Process Engineering of the Karlsruhe Institute of Technology for more than 15 years. Starting with the famous Karlsruhe Cube, a cross-flow micro channel heat exchanger of various dimensions, not only conventional heat transfer between liquids or gases have been theoretically and experimentally examined but also phase transition from liquids to gases (evaporation) and condensation of liquids. However, the results obtained with sealed microstructure devices have often been unsatisfying. Thus, to learn more onto the evaporation process itself, an electrically powered device for optical inspection of the microstructures and the processes inside has been designed and manufactured [2]. This was further optimized and improved for better controllability and reliable experiments [3]. Exchangeable metallic micro channel array foils as well as an optical inspection of the evaporation process by high-speed videography have been integrated into the experimental setup. Fundamental research onto the influences of the geometry and dimensions of the integrated micro channels, the inlet flow distribution system geometry as well as the surface quality and surface coatings of the micro channels have been performed. While evaporation of liquids in crossflow and counterflow or co-current flow micro channel devices is possible, it is, in many cases, not possible to obtain superheated steam due to certain boundary conditions [4]. In most cases, the residence time is not sufficiently long, or the evaporation process itself can not be stabilized and controlled precisely enough. Thus, a new design was proposed to obtain complete evaporation and steam superheating. This microstructure evaporator consists of a concentric arrangement of semi-circular walls or semi-elliptic walls providing at least two nozzles to release the generated steam. The complete arrangement forms a row of circular blanks. An example of such geometry is shown in Figure 8. A maximum power density of 1400 kW · m −2 has been transferred using similar systems, while liquid could be completely evaporated and the generated steam superheated. This is, compared to liquid heat exchanges, a small value, but it has to be taken in account that the specific heat capacity of vapor is considerably smaller than that of liquids. It could also be shown that the arrangement in circular blanks with semi-elliptic side walls acts as a kind of micro mixer for the remaining liquid and generated steam and, therefore, enhances the evaporation.

A typical microfluidic fuel cell is comprised of a Y- or T-shaped microchannel. The fuel and the oxidant streams are introduced from the two different inlets. The anodic and cathodic flows meet each other at the beginning of the main channel and start to travel together along the channel. Due to the fact that the viscous forces dominate the inertia forces in microchannels, the oxidant and the fuel streams establish a side-by-side co-laminar flow which makes the anolyte and catholyte flow together without turbulent mixing. Laminar flow in microfluidic fuel cells plays the role of the membrane in proton exchange membrane (PEM) fuel cells by maintaining the separation of the fuel and oxidant. This eliminates the need for the membrane and overcomes the membrane-related issues such as the ohmic overpotential and water management which are relevant to PEM fuel cells. In addition to the above advantage, the high surface-to-volume ratio of these micron-scale devices contributes to their high power density. This advantage is due to the fact that the electrochemical reactions in fuel cells are surface-based. The electrodes on which the electrochemical reactions are occurring are installed appropriately on the walls of the channel in a way that reacting flows are restricted to the proper electrodes. Since the flow is laminar the performance of the microfluidic fuel cell significantly depends on the device geometry. In this paper, different channel geometries and different electrode configurations are modeled and their performances are compared through the polarization curves. It has been found that the high aspect ratio provides the largest power density. In this work, the performance of the flow-through porous electrode was also modeled and compared against the conventional non-porous electrode microfluidic fuel cells. The flow-through porous electrode design is based on cross-flow of aqueous vanadium redox species through the electrodes into an exit channel, where the waste solutions meet and establish a co-laminar flow. This co-laminar flow of reacted species facilitates ionic charge transfer in a membraneless configuration. It has been found that the flow-through porous architecture provides an increased active surface area which contributes to a higher power density as opposed to the fuel cells with non-porous electrodes.

An approach towards practical application of microchannel cooling system is necessary as the demand of high power density devices is increasing. Colgan et. al. [1] have designed a unit known as Single Chip Module (SCM) by considering the practical issues for packaging a microchannel cooling system with a microelectronic device. The performance of the SCM has already been investigated by using water as working fluid by Colgan et. al. [1]. Considering the actual working conditions, water cannot be used in electronic devices as the working fluid because any leakage may lead to system damage. Alternative fluids like refrigerants were considered. In this research, the performance of SCM has been studied by using refrigerant R-123 as working fluid and compared with water cooled system. Cooling of 83.33 W/cm 2 has been achieved for a powered area of 3 cm 2 by maintaining chip temperature of 60°C. The heat transfer co-efficient obtained at a flowrate of 0.7 lpm was 34.87 kW/m 2 -K. The results obtained indicate that from a thermal viewpoint, R-123 can be considered as working fluid for microelectronic cooling devices.

In this article, ion selective nanochannels are studied to generate electric power from concentration gradient by reverse electrodialysis. When nanochannels bring into contact with aqueous solution, the surface of nanochannels acquires charges from ionization, ion adsorption, and ion dissolution. These surface charges draw counter-ions toward the surface and repel co-ions away. Therefore, when an electrolyte concentration gradient is applied to nanochannels, counter-ions are transported through nanochannels much more easily than co-ions, which results in a net charge migration of ions. Gibbs free energy of mixing, which forces ion diffusion, thus can be converted into electrical energy by using ion-selective nanochannels. Silica nanochannels with heights of 26 nm and 80 nm fabricated by glass-silicon anodic bonding were used in this study. We experimentally investigated the power generation from these nanochannels placed between two potassium chloride solutions with various combinations of concentrations. The power generation per unit channel volume increases when the concentration gradient increases, while it decreases as channel height decreases. The highest power density measured is 26 kW/m 3 . Our data also indicates that the efficiency of energy conversion and the ion selectivity increase with a decrease of concentrations and channel height. The best efficiency obtained is 24%. Compared with ion-selective membranes, nanochannels promise more reliable operation since they are readily compatible with standard CMOS process and do not shrink and swell in response to their environment. Power generation from concentration gradient in ion selective nanochannels could be used in a variety of applications, including micro batteries and micro power generators.

Nanoscale thermal management (NSTM) will become imperative as power density of IC’s increases further or as 3-dimensional IC’s are introduced. Such NSTM solutions must be integrated with the IC’s to be effective. It can be envisioned that an ideal NSTM solution will involve two-phase flow, liquid flow and gas flow. This paper focuses on gas flow, both as a fundamental thermal management technique in any future NSTM solution, and also as a basis for understanding more complex liquid and two-phase flow techniques that may also be involved in any NSTM solution. Heat removal by gaseous flow in any NSTM solution may be broken up in four fundamental processes: (1) bulk flow through micro-scale ducts where there may or may not be any heat transfer through the side-walls, (2) interaction of gaseous molecules with duct walls which may be at the same or a different temperature, (3) interaction between gas molecules and any enhanced surfaces such as carbon nanotubes (CNTs) as pin fins, and (4) thermal conduction and distribution in the device substrate. This paper touches upon all four fundamental process of heat removal by gaseous flow in an NSTM solution. Preliminary results from the first three processes are presented whereas a preliminary design and fabrication solution is presented for the fourth process. Preliminary computational results for pressure-driven flow of helium as the coolant gas through a micro-channel duct with a backward facing step are presented as an example for the first process. Backward facing step simulates any change in flow cross-section that may be unavoidable or desirable in any practical NSTM solution. Computation is based on direct simulation Monte Carlo (DSMC), where statistical noise due to low-speed flow is greatly eliminated through the use of the IP technique. Pressure boundary conditions are used in this simulation as they would be more realistic to represent an NSTM application. Preliminary results for computation of tangential momentum accommodation coefficient (TMAC) for helium-wall collisions are presented as an example for the second process. Molecular dynamics (MD) simulation of collisions between helium atoms and copper atoms in a crystallographically perfect copper wall are performed for different ratios between kinetic energy of helium atoms and Lenard-Jones energy of He-Cu interactions. The results are compared to limited experimental results that are available. Preliminary computational results for energy transfer between hydrogen molecules/atoms and a single wall carbon nanotube (SWNT) are presented as an example for the third process. Here MD simulation is employed where high temperature hydrogen molecules are allowed to collide with a SWNT that is initially at a lower temperature. As time progresses, the amount of net energy transfer from the hydrogen molecules to the SWNT is monitored. The result is normalized with the nominal surface area of the SWNT and the driving temperature differential to form coefficient for thermal energy transport (CTET). The values obtained are compared against macroscopic correlations. Deposition of nano-crystalline diamond is proposed as a solution for thermal spreading in the fourth process. Some examples of deposition and the corresponding atomic structure are presented.

Surface structure modification and wettability control are very crucial for the advancement in microfluidic systems. The wettability control of the surface is achieved by depositing a Self-Assembled Monolayer (SAM) on the surface. The technology of SAM has been advanced by coating the selective interior regions of silicon wafer. This was accomplished using a mask wafer with open slots attached to define the areas that need to be coated. Besides the capability of controlling the wettability, structure modification such as controlling hole size leads to the magnification of the influence of surface tension force. As the size of the hole gets smaller, the surface tension could become a dominant force. This paper utilizes a unique approach to improve the water and air management at the cathode of a micro Direct Methanol Fuel Cell (DMFC). Both structure modification and local surface wettability control are utilized. A two-inch silicon wafer is formed of alternate strips of hydrophobic and hydrophilic zones with arrays of holes of different sizes. Water will be guided along the hydrophilic wetting zones with large hole openings; while the air goes into the cathode from the hydrophobic dry areas with smaller holes. This process ensures that the cathode is not flooded by water while air passes easily. The excess water at the cathode could be pumped back by a micro-pump to the anode so that the size of water storage will be minimized. Therefore, most of the storage reservoir space is used for the pure methanol, which enables the achievement of a high power density of the system. In this study, a silicon-based membrane is built accordingly to observe the water and air management. Images of the CCD camera showed clearly that the water drained from the big holes without blocking the air passages.

High performance single-phase Si microchannel coolers have been designed and characterized in single chip modules in a laboratory environment using either water at 22°C or a fluorinated fluid at temperatures between 20 and −40°C as the coolant. Compared to our previous work, key performance improvements were achieved through reduced channel pitch (from 75 to 60 microns), thinned channel bases (from 425 to 200 microns of Si), improved thermal interface materials, and a thinned thermal test chip (from 725 to 400 microns of Si). With multiple heat exchanger zones and 60 micron pitch microchannels with a water flow rate of 1.25 lpm, an average unit thermal resistance of 15.9 C-mm 2 /W between the chip surface and the inlet cooling water was demonstrated for a Si microchannel cooler attached to a chip with Ag epoxy. Replacing the Ag epoxy layer with an In solder layer reduced the unit thermal resistance to 12.0 C-mm 2 /W. Using a fluorinated fluid with an inlet temperature of −30°C and 60 micron pitch microchannels with an Ag epoxy thermal interface layer, the average unit thermal resistance was 25.6 C-mm 2 /W. This fell to 22.6 C-mm 2 /W with an In thermal interface layer. Cooling >500 W/cm 2 was demonstrated with water. Using a fluorinated fluid with an inlet temperature of −30°C, a chip with a power density of 270 W/cm 2 was cooled to an average chip surface temperature of 35°C. Results using both water and a fluorinated fluid are presented for a range of Si microchannel designs with a channel pitch from 60 to 100 microns.

Currently, the heat exchangers allowing the cooling of the low temperature fuel cells (PEMFC) are integrated in the bipolar plates and constituted of a network of straight channels. The flow regime is laminar, and thus, unfavorable to an intense convective heat transfer. In order to increase the power density of the fuel cells, the use of chaotic geometries in the cooling system is envisaged to intensify high convective heat transfer. In this numerical study, several chaotic three-dimensional mini-channels of rectangular section (2 millimeters × 1 millimeter) are evaluated in terms of heat transfer efficiency, mixing properties and pressure losses. Their performances are compared to those of a straight channel geometry currently used in the cooling systems of the PEMFC, and a serpentine 2-D channel. Hydrodynamic and thermal performances of these geometries are computed using the commercial CFD code Fluent©. At the inlet section, the velocity profile is hydrodynamically established. The thermophysical properties of the fluid are constant and equal to those of water at 300 K. The Nusselt number is evaluated for a Reynolds number equal to 200 and with a uniform density flux imposed on the walls and equal to 10,000 W/m 2 . For the calculation of the mixing rate, a condition of adiabatic wall is imposed. The inlet section is horizontally divided into two parts. Water in the higher part is at the temperature of 320K and in the lower part is at the temperature of 300K. The calculation of the mixing rate is made for Reynolds numbers equal to 100 and 200. The present study shows that a 3-D chaotic channel geometry significantly improves the convective heat transfer compared to regular straight or serpentine channels. Among all the studied geometries, one of them induces the higher heat transfer intensification (mean Nusselt number equal to 20) with a strong pressure loss. With an alternative geometry, we obtained a better compromise between high heat transfer and reduced pressure loss. However, all the chaotic geometries present similar mixing rate for the two studied Reynolds number. To confirm the performances of the selected geometries, an experimental study is currently undertaken. The final aim is to realize and test a prototype of chaotic heat exchanger in a bipolar plate of PEMFC.

Demands for increased functionality from mobile electronic products have led to an increase in the amount of heat that needs to be dissipated. The authors propose a novel mini-heatsink to meet the growing power density needs of the miniature electronics market. This paper looks at the optimisation of a miniature radial flow heatsink with laminar forced convection flow originating from the center. Two single length scale fins positioned at different angles were used to simulate a segment of the heatsink. A Labview PID program was used to monitor and maintain the temperature of these fins. The optimum angle at which maximum heat transfer occurred was determined as a function of pressure drop. The design was varied using the multiscale approach by inserting fins of a smaller length scale into regions of underutilised flow at the exit to achieve the highest heat transfer rate while simultaneously minimising the viscous drag. This method is in contrast to previous work where investigators inserted smaller length scales into the inlet of the original architecture in the regions of unused (isothermal) flow. Non-dimensional results are presented which show increased heat transfer with the addition of smaller length scales. This paper concludes that the multiscale design feature increased the heat transfer rate.