Objective

System-Level Approach for Multi-Phase Nanotechnology-Enhanced Cooling of High-

Power Microelectronic Systems

Objective

Our objective is to develop a robust, low-temperature thermal management approach for

distributed, large-scale, high-power electronic systems using improved heat transfer technologies and system modeling/control concepts to reduce the total system thermal resistance so that chip

and device temperatures are maintained below 65°C.

Overall ChallengeProblem: Operation of a large-scale, distributed electrical/electronic system including devices with ultra-high heat fluxes that require low surface temperatures.

65°C

State-of-the-art and Our GoalsEstimates of Conductances in State-of-the-Art

Systems and Future Systems

Thermal Conductance State-of-the-art Target

Device substrate conductance (MW/m2-K) (hchip = 1/ARchip) 0.1 5.0

Interface conductance (MW/m2-K) (hTIM= = 1/ARTIM) 0.1-0.5 5.0

Microchannel/micro domain base conductance (MW/m2-K) (hmc-solid = 1/ARmc-solid)

0.1 5.0

Convective conductance of microchannel to coolant including area enhancement due to fins (MW/m2-K) (hmc-coolant = 1/ARmc-coolant)

0.05 0.20

Convective conductance of coolant to condenser (MW/K) (Hcoolant-condenser = 1/Rcoolant-condenser)

0.06 0.18

Wall conductance in condenser (MW/K) (Hcondenser wall = 1/Rcondenser wall)

1.65 4.95

Convective conductance of condenser to ocean (MW/K) (Hcondenser-ocean = 1/Rcondenser-ocean)

0.12 0.36

A Multidisciplinary Approach to a Multiscale Problem: Chip to Ship

nm

μm

mm

m

km

10-9

10-6

10-3

100

103

Nanoscale Energy Transport and Conversion:

Thermal Management of Electronics:

Macroscopic Energy Transport and Conversion:

System Scaling and Optimization:

Electrical Engineering, Material Science, Solid-State Physics, Statistical Mechanics

Heat Conduction, Material Science, Microfluidics, Mechanics, Fatigue

Convective Heat Transfer, Fluid Dynamics, Thermodynamics

Electrical Engineering, Control Systems

Length:Scale:

Primary ThrustsThrust 1 – Solid/Solid Interfaces

Thrust 2 – Solid/Liquid Interfaces

Thrust 3 – System Design and Optimization

Understanding the origins (interfacial chemistry, structure, and electron/phonon transport) of solid-solid interfacial thermal resistance, will lead to the development of a new thermal interface material with a target conductance of G~5 MW/m2-K.

Studying flow boiling and condensation at the micro-scale through novel noninvasive thermometry tools will guide the enhancement of convective heat transfer using advanced micro-domain configurations, engineered surface topography, and changes in chemical composition for control of interfacial physics.

Sensor feedback and modern control theory will be used to develop design and control methodology for the system and devices and the results of Thrusts 1 and 2 will be integrated in a modular and scalable testbed.

Integration

Thrust 1: Solid/Solid Interfaces• Primary Contributors – UVa: Pam Norris– ASU: Ravi Prasher– UC Berkeley: Costas Grigoropoulos

• Thrust Statement/Objectives– Increase the long-term stable conductance at

solid/solid interfaces– Understand the origins (interfacial chemistry,

structure, and electron/phonon transport) of solid-solid interfacial thermal resistance

– Use this knowledge to help develop new thermal interface materials

Thrust 1: Accomplishments for 1st Yearas of June 2008

• Theoretical models developed for thermal transport in mesoscopic nanowires/nanotubes (thermal conductivity, thermal boundary resistance, specific heat).

• Measured elastic modulus of vertically aligned CNT arrays under compression, and studied mechanism.

• Studied the growth mechanism of vertically aligned CNT arrays.• Fabricated TIM with CNT array with G ~ 1 MW/m2-K through In bonding technique.• Compared the interface thermal conductance of glass-CNT-Si sandwich structures

with/without In bonding, and with Ti wetting layer as well.• Investigated the adhesion problem between In film and Ti film.• Investigated the possibility of enhancing contact between In and Ti by adding Au

bonding layer.• Examined effects of atomic mixing at Cr/Si interface on G – developed new model

to take into account multiple scattering events from disorder.• Studied phonon scattering processes at interfaces – developed model to account

for inelastic scattering and separated role of elastic and inelastic scattering in G.• Experimentally investigated electron-phonon equilibration in thin Au films subject

to short pulsed heat in the presence of the film/substrate interface.• Determined that interfacial scattering can affect electron-phonon equilibration

time – developed new model based on ballistic electron transport and ballistic-diffusive approximation to the Boltzmann Transport Equation.

Thrust 1: Accomplishments for 2nd Yearas of June 2009

• Characterized the packing density of vertically aligned CNT array using high resolution SEM.

• Investigated the effect of hydrogen pretreatment on the CNT array density.• Comprehensively studied the sensitivity of transient opto-thermoreflectance

technique on ultra-small thermal interface resistance measurement.• Improved the experimental technique to allow high resolution thermal

measurement with improved sensitivity.• Investigated the effect of variation of packing densities on G. • Investigated the influence of interstitial metallic layers on G at metal-nonmetal

interfaces.• Emulated the behavior of metal-CNT contact by making thermal measurements

on metal-graphite samples.• Identified most probable material systems for ultimate TIM package assembly.• Preliminary development of the multidimensional diffuse mismatch model.• Refined UVA TTR setup to allow for measurement on more diffuse surfaces.• Fabricated CNT-array-based TIM with G ~ 1.5 MW/m2-K.

Thrust 1: Accomplishments for 3rd Yearas of August 2010

• Developed a theoretical model for thermal transport between isotropic films and anisotropic substrates considering contributions of inelastic scattering.

• Measured boundary conductance between Au films and sp2 carbon structures, investigating the role of interface structure and chemistry.

• Investigated the role of subconduction band excitations on thermal conductance at metal-metal interfaces.

• Reviewed the assumptions made in modeling diffusive transport in nanostructures and their application to specific systems.

• Increased the CNT array packing density to 11% volume fraction.• Fabricated CNT arrays of various lengths (~250, 200, 150, 100, and

50 μm) by decreasing growth duration of water-assisted synthesis from 3 minutes to 15 seconds, and with nearly fixed CNT array volume fraction of 3%.

• Measured thermal interface (CNT – bonding surface) conductance of high density (11%) sample to be over 5 MW/m2-K.

Thrust 2: Fluid/Solid Interfaces• Primary Contributors – RPI: Michael Jensen, Yoav Peles– UIUC: David Cahill, Steve Granick

• Thrust Statement/Objectives– Reduce convective thermal resistances in heat sinks

and heat exchangers– Understand flow boiling and condensation at the

micro-scale– Enhance heat transfer through modifications of

surface chemistry, surface topography, and micro-domain configurations

Thrust 2: Accomplishments for 1st Yearas of June 2008

• 95%+ complete on construction of three flow loops (boiling, condensation, microjet).

• 99% complete on test sections for all three loops.• Calibrations of components in three loops done except for test sections; data

acquisition and reduction programs complete.• Developed a program to optimize different microdomain geometries

(microchannels, pin fins, etc.) that takes into account heat transfer and pressure drop characteristics; this will be used for guidance on developing enhanced boiling and condensing geometries.

• Completed construction and testing of an IR (1.55 micron) femtosecond pump-probe measurement system capable of probing through the back side of Si wafers with low doping levels.

• Completed first-generation construction of homebuilt apparatus for surface plasmon resonance imaging of bubble nucleation events.

Thrust 2: Accomplishments for 2nd Yearas of June 2009

• Obtained data for critical heat flux and boiling heat transfer data for R134a in microtubes.

• Obtained data for single-phase heat transfer with single micro-jet and with micro-jet arrays.

• Designed new experiment for multi-jet array impinging on enhanced surface.• Performed optimization study for single-phase flows with various enhanced

heat transfer techniques in micro-domains.• Invented two-photon thermometry for ultra-high space and time resolution

of thermal transport.• Invented surface plasmon thermometry for imaging local nucleation events.• Invented pump-probe ellipsometry for molecular understanding of heat

transfer from solid to liquid on the picosecond time scale.

Thrust 2: Accomplishments for 3rd Yearas of August 2010

• Completed a detailed experimental study of heat transfer coefficients and CHF condition for flow boiling of R134a in circular microtubes.

• Completed a parametric study of area-averaged heat transfer coefficient of a microjet array varying area ratio and using air and water.

• Converted the experimental microjet array apparatus to a closed, pressurized, refrigerant system charged with R134a, and conducted flow boiling experiments with R134a.

• Designed and constructed test sections to measure condensation heat flux, and, using these, completed condensation heat transfer experiments on square mini-channels.

• Designed micro devices for studying single-phase and flow boiling heat transfer of jet impingement on micro pin-fin structures; microfabricated the micro devices using MEMS microfabrication technologies.

• Conducted heat transfer and pressure drop experiments using water as coolant for jet impingement on micro pin fins.

• Discovered that time-averaged heat transfer during water droplet impacts can reach 500 W/cm2.

• Observed thermal accommodation between a solid and a condensing vapor for the first time.

• Directly visualized heat transfer in the nucleation of single bubbles for the first time.

Thrust 3: System Design and Optimization• Primary Contributors – RPI: J. Wen, M. Jensen, Y. Peles– ASU: P. Phelan, R. Prasher

• Thrust Statement/Objectives– Integrate results of Thrusts 1 and 2 into a modular

and scalable refrigeration system• Demonstration of 1 kW/cm2 on single module • Demonstration of scalability to multi-module systems

– Develop design & control methodology for system and devices using sensor feedback and modern control theory

Thrust 3: Accomplishments for 1st Yearas of June 2008

• Development of steady-state vapor compression cycle modeling and optimization tool (in MATLAB) including CHF consideration. Extensibility to multiple evaporators, multiple loops, pump cycle.

• Establishment of the two-loop architecture for system design.

• Initial design and setup of a fully instrumented vapor compression cycle testbed capable of operating at a wide range of conditions.

• Initial design and equipment purchase for the high-heat-flux, single-module testbed.

• Initial development of dynamic modeling.

Thrust 3: Accomplishments for 2nd Yearas of June 2009

• Developed optimization module for steady-state system design at multiple operating conditions for a single vapor compression cycle.

• Developed initial steady-state design tool for two-loop system analysis. • Completed the first phase RPI test facility beds and started model

identification for expansion valve and compressor.• Nearly (99%) completed construction of ASU test facility.• Obtained initial transient stability result using singular perturbation.• Developed Ledinegg instability model based on data from Intel.• Developed preliminary pressure drop instability model based on data from

Intel.

Thrust 3: Accomplishments for 3rd Yearas of August 2010

Macro-scale Systems• Validated steady-state vapor compression cycle models (compressor, valve, heat

exchangers) with experimental test facility at RPI.• Developed steady-state design and optimization module for two-loop cooling system

at multiple operating conditions.• Completed dynamic identification for empirical modeling and controller design for

single evaporator case.• Completed evaluation of linear controller design using expansion valve opening to

control evaporator wall temperature.• Investigated the effects of transient heat load on two-phase flow instabilities in

electronics cooling systems.Micro-scale Systems• Studied micro-thermal-fluid transients, developed new HTC correlation, and combined

extremum-seeking and thermal-fluid stabilizing controller.• Developed pressure drop flow oscillation model and active instability control

strategies for microchannel boiling system.• Proposed non-identical structures for parallel-channel flow instability control.