ME 381R Fall Lecture 24:

1
ME 381R Fall Lecture 24 Micro-Nano Scale
Thermal-Fluid Measurement Techniques
Dr. Li Shi Department of Mechanical Engineering
The University of Texas at Austin Austin, TX
78712 www.me.utexas.edu/lishi lishi_at_mail.utexas.
edu
2
Visualization of Microflows

• Caged fluorescence
• Micro Particle Image Velocimetry (mPIV)

• References
• A particle image velocimetry system for
  microfluidics, Santiago, J.G et al. Experiments
  in Fluids, 25, pp. 316-319. (1998)
• 2. PIV measurements of a microchannel flow,
  Meinhart et al. Experiments in Fluids, 27, pp.
  414-419 (1999)
• 3. J.I. Molho, A.E. Herr, T.W. Kenny, M.G.
  Mungal, P.M. St.John, M.G. Garguilo, P.H . Paul,
  M. Deshpande, and J.R. Gilbert, "Fluid Transport
  Mechanisms in Microflui dic Devices",
  Micro-Electro-Mechanical Systems (MEMS), 1998
  ASME International Mechanical Engineering
  Congress and Exposition (DSC-Vol.66) 

3
Caged Fluorescence

• Fluorescent dye chemically locked in a stable
  molecule until hit with NdYAG laser which
  uncages it.
• Uncaged dye is pumped with Microblue diode pumped
  laser.
• Fluorescence is imaged with CCD camera.
• (Molho. Et.at. 1998)

4
Results
Experiment matches prediction for uniform plug
flow for some cases studied. No discernable
boundary layers, but some diffusion. http//microf
luidics.stanford.edu/caged.htm
5
More Results
In other cases though, flow looks very much like
a pressure-driven Poiseuille flow
Electro-Kinetic Flow can actually induce a
pressure gradient in a capillary flow and thus
alter the basic flow structure http//microfluidic
s.stanford.edu/caged.htm
6
Comparison with CFD
Electro-Osmotic flow is relatively simple to
model with standard CFD solvers. For pressure
driven micro-capillary flow, CFD predicts flow
field remarkably well, as shown in comparison of
experimental and computational results at
left. (Molho et.al. 1998)
7
Particle Image Velocimetry (PIV)
Cross-correlation
Velocity vector
Interrogation windows 32x32 pixels, 0.6 x 0.6 mm
Images from Tsurikov and Clemens (2002)
Particle fields 1024 x 1024 pixels 21 x 21 mm
Raw velocity field
Mean velocity subtracted Turbulent velocity field

• Seed flow with particles
• Dont affect fluid characteristics
• Accurately follow the flow
• Illuminate flow at two time instances separated
  by ?t (e.g. using NdYAG laser)
• Record images of particle fields (e.g. CCD
  camera)
• Determine particle displacement
• Calculate velocity as V? ?x/ ?t

8
The Need for ?-PIV

• The physics is not very clear in micro flows
  (e.g. surface tension)
• Typical length scales of 1-100 ?m, traditional
  flow diagnostics cannot be employed
• Most micro-flow measurements were limited to bulk
  properties of the flow like wall pressure and
  bulk velocity
• PIV enables measurements of velocity field in two
  dimensions

9
Other efforts

• Particle streak imaging by Brody et al. (1996)
• Less accurate than pulsed velocimetry
  measurements
• Lanzilloto et al. (1997) used X-ray micro-imaging
  of emulsion droplets
• Emulsion is deformable, large and not a good
  tracker of the flowfield
• Optical Doppler Tomographic imaging by Chen et
  al. (1997) using Michelson interferometry
• Single point measurement

10
?-PIV

• Particles used must be small enough to
• Follow the flow
• Should not clog the device
• They must also be large enough to
• Emit sufficient light
• Sufficiently damp out Brownian motion
• Particles are tagged with a fluorescent dye
  hence actually imaging the fluorescence
• Elastic scattering measurements are more
  difficult to employ in the micro-scale
• Inelastic scattering like fluorescence can be
  readily filtered out

11
?-PIV

• Errors in measurement due to Brownian motion when
  measuring velocities of 10 ?m/sec
• Error induced by Brownian motion sets a lower
  limit on the time separation between the images

12
First ?-PIV system
From Santiago et al. (1998)

• Essentially a microscope imaging fluorescence
  from the seed particles

13
State of the art ?-PIV system
From Meinhart et al. (1999)

• http//microfluidics.stanford.edu/piv.htm

14
Demonstration of ?-PIV

• Hele-Shaw flow (Re3e-4)
• used the first ?-PIV system discussed before
• Micro-channel flow
• Uses the laser based system

15
Velocity fields Hele-Shaw
From Santiago et al. (1998)

• Shows instantaneous and average images
• Effect of Brownian motion goes away on averaging
• Spatial resolution 6.9 ?m x 6.9 ?m x 1.5 ?m

16
Velocity Fields in a Micro-channel

• Shows mean velocity profiles in a micro-channel
• Measurements agree within 2 to analytical
  solutions

From Meinhart et al. (1999)
17
Comparison to analytical solution
From Meinhart et al. (1999)
18
Thermometry of Nanoelectronics
Techniques
Spatial Resolution
Infrared Thermometry
1-10 mm Laser Surface Reflectance
1 mm Raman Spectroscopy
1 mm Liquid Crystals
1 mm Near-Field
Optical Thermometry lt 100 nm
Scanning Thermal Microscopy (SThM) lt 100 nm
Diffraction limit for far-field optics
19
Scanning Thermal Microscopy
Atomic Force Microscope (AFM) Thermal
Probe
Laser
Deflection Sensing
Cantilever
Temperature sensor
Sample
X-Y-Z Actuator
20
Microfabricated Thermal Probes
Pt Line
Tip
Pt-Cr Junction
Laser Reflector
SiNx Cantilever
Cr Line
Shi, Kwon, Miner, Majumdar, J. MicroElectroMechani
cal Sys., 10, p. 370 (2001)
21
Thermal Imaging of Nanotubes
Multiwall Carbon Nanotube
Topography
Topography
3 V
m
88
A
m
m
1
m
1
m
Spatial Resolution
V)
m
50 nm
Thermal signal (
Distance (nm)
Shi, Plyosunov, Bachtold, McEuen, Majumdar, Appl.
Phys. Lett., 77, p. 4295 (2000)
22
Metallic Single Wall Nanotube
Topographic
Thermal
DTtip
A
B
C
D
2 K
0
1 mm
23
Metal-Oxide-Semiconductor Field-Effect Transistor
(MOSFET)
24
Ideal MOSFET
VGgt0
25
Pinch-Off IV
26
Thermal Circuit
Particle transport theory
Fouriers law of heat conduction
27
Joule Heating inHigh-Field Devices
Localized heat generation near the pinch-off point
28
Future Challenge Temperature Mapping of
Nanotransistors
SOI Devices
SiGe Devices

• Low thermal conductivities of SiO2 and SiGe
• Interface thermal resistance
• Short (10-100 nm) channel effects (ballistic
  transport, quantum transport)
• Phonon bottleneck (optical-acoustic phonon
  decay length gt channel length)

• Few thermal measurements are available to verify
  simulation results