JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 11, NO.

6, DECEMBER 2002

665

A Thermal-Bubble-Actuated
Micronozzle-Diffuser Pump
Jr-Hung Tsai and Liwei Lin, Member, IEEE

AbstractA thermal-bubble-actuated micropump by the principles of liquid/vapor phase transition and nozzle-diffuser flow regulation is successfully demonstrated. The micropump consists of
a resistive heater, a pair of nozzle-diffuser flow controller and a
1 mm in diameter, 50 m in depth pumping chamber. The actuation mechanism comes from periodically nucleating and collapsing
thermal bubbles. A net flow is generated from the nozzle to the
diffuser by the nozzle-diffuser flow controller. Two heater designs,
single-bubble and dual-bubble actuation mode, have been investigated. In the single-bubble pumping mode, a maximum flow rate
of 5 l/min is measured when the driving pulse is 250 Hz at 10%
duty cycle under an average power consumption of 1 W. A similar flow rate of 4.5 l/min is achieved in the dual-bubble pumping
mode, at the driving pulse of 5% duty cycle at 400 Hz with lower
average power consumption, 0.5 W. The static pumping pressure is
measured at a maximum value of 377 Pascal when the net volume
flow rate is zero. As an application example in a microfluidic device, this valve-less micropump is used in a microfluidic system to
enhance the fluid mixing by agitating the flows.
[742]
Index TermsBubble, microfluidics, mixing, phase change,
pump.

I. INTRODUCTION

ICROPUMPS have wide applications in MEMS devices, such as handling small and precise volume of
fluids for chemical, biological or medical systems. Many design
concepts and prototype micropumps have been demonstrated
over the past decade and they can be classified into following
types by means of actuation mechanisms: mechanical displacement [1][3], bubble [4], electrohydrodynamic [5], and
electrocapillary [6]. Among these actuation methods, bubble
actuation can be easily generated by resistive heaters and can
be applied to nonspecific liquids, thus is an attractive actuation
source in the micro scale. Existing device examples include
bubble-jet printers [7] and bubble powered micro mechanical
actuators [8].
On the other hand, micro valves are indispensable components in microfluidic systems for flow rate and flow direction
control such that they are usually integrated with micropumps
[3], [9][11]. Unfortunately, micro valves generally increase the
complexity of the fabrication process and the possibility of parManuscript received August 21, 2001; revised February 8, 2002. This work
was supported in part by an NSF CAREER AWARD (ECS-0096098) and a
DARPA/MTO/BioFlips grant. Subject Editor G. Stemme.
J.-H. Tsai was with Mechanical Engineering, University of Michigan,
Ann Arbor, MI USA. He is now with the Mechanical Engineering Department, University of California, Berkeley, CA 94720-1740 USA (e-mail:
[email protected]).
L. Lin is with the Mechanical Engineering Department, University of California, Berkeley, CA 94720-1740 USA.
Digital Object Identifier 10.1109/JMEMS.2002.802909

Fig. 1. Pumping principle illustrated from the top view of the pumping
chamber.

ticle clogging and contamination in the microfluidic pathways.

These shortcomings make valve-less pumping attractive as an
alternative way for the micropump design [2], [12], [13]. A
valve-less micropump by means of nozzle-diffuser design and
piezoelectrical actuation was developed previously [2]. The efficiency of this micropump depends largely on the deflection
magnitude and driving frequency of the mechanical membrane.
It becomes significantly difficult to further miniaturize these diaphragm-based pumps because the large force is required to deflect the diaphragms for pumping.
This paper applies the principle of thermal bubble nucleation
for pumping and the nozzle-diffuser flow controller design
to direct the induced oscillatory flow. As such, the pumping
efficiency can be improved because: 1) thermal bubbles can
be easily generated even when the pump is further miniaturized and 2) thermal bubble can sweep across most of the
pumping chamber for optimal efficiency. Furthermore, the
concern of complicated fabrication process for micro valves is
alleviated because there are no moving mechanical parts in the
micropump.
II. MICROPUMP DESIGN, FABRICATION,
EXPERIMENTAL SETUP

AND

A. Design
Fig. 1 illustrates the fundamental design concept of the micropump. Micro bubble is generated by thermal bubble nucleation in the micro chamber to create a pumping pressure source.
As the pumping bubble expands, the volume flow rate at the dif, is larger than the one at the nozzle,
. When the
fuser,
is smaller than
. Therefore,
pumping bubble collapses,
there is a net flow generated from the nozzle to diffuser each
cycle as shown. The boundary of the bubble works as a membrane that sweeps across the chamber to minimize the pumping
dead volume. Furthermore, because of the nozzle-diffuser design, no conventional mechanical valve is used in the operation
such that the fabrication process is simplified.

1057-7157/02$17.00 2002 IEEE

666

JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 11, NO. 6, DECEMBER 2002

Fig. 2. Schematic drawing of the nozzle-diffuser based bubble pump.

Fig. 4. Fabrication process.

pumping mode, respectively. In the single-bubble pumping

mode, the heater is designed to generate one bubble group in
the center of the pumping chamber. In the dual-bubble pumping
mode, the heater is designed to generate two bubble groups
simultaneously that can be barely observed in Fig. 3(b).
B. Fabrication

Fig. 3. Two types of heater patterns, (a) single-bubble mode and (b) dualbubble mode.

Based on the operation principle, the micropump is fabricated

with a micromachining process as schematically displayed in
Fig. 2. The pumping chamber, nozzle-diffuser flow controller
and channels, are fabricated on a silicon substrate by the same
depth of 50 m using DRIE (Deep Reactive Ion Etching).
Through holes for liquid inlet and outlet ports and electrical
interconnection are also fabricated on the silicon substrate by
using another DRIE process. A Pyrex glass wafer is used to seal
the microsystem and aluminum is evaporated and patterned on
it as shown in Fig. 2 as the resistive heater. A meander-shape
heater is selected rather than a solid rectangular plate-shape
heater (commonly used in bubble jet printers) such that bubble
nucleation in the pumping chamber can be observed from the
glass side. In the prototype trial, the selection of the dimensions is based on the previous report of a piezoelectrically
driven, nozzle-diffuser micropump [14]. The circular pumping
chamber is chosen as 1 mm in diameter and is connected with
a pair of nozzle-diffuser flow controller of 30 m in the narrow
neck, 274 m in the open mouth and the diverging angle is
14 . The fluid channels that connect the micropump to the fluid
inlet/outlet ports are 200 m in width. Two polyimide tubes
of 310 m in outer diameter and 275 m in inner diameter
are used to connect the micropump to the macro world via the
etched through liquid inlet/outlet ports on the silicon substrate.
Fig. 3(a) and (b) display two optical photos of fabricated
micropumps of single-bubble pumping mode and dual-bubble

The fabrication process is illustrated in Fig. 4. On the silicon substrate, a two-step DRIE process is used. The silicon
wafer is deposited with a 1- m-thick thermal oxide that is later
patterned and etched as the oxide mask to define the pumping
chamber, nozzle-diffuser pair and microchannels. A 9- m-thick
photoresist is spun on, patterned as the photoresist mask to define through holes for the liquid inlet/outlet ports and electrical
interconnections as shown. After these steps, Fig. 4(a) applies.
The first DRIE process is conducted afterwards to etch through
the whole silicon wafer and the photoresist mask is removed as
shown in Fig. 4(b). The second DRIE process is carried out with
the oxide mask and about 50 m deep of silicon is etched away.
Afterwards, the silicon wafer is dipped into hydrofluoric acid
(HF) to remove silicon dioxide and the silicon wafer is ready for
assembly as shown in Fig. 4(c). On the Pyrex glass substrate, recesses of about 1 m in depth for embedding aluminum heaters
are formed to prevent liquid leakage around the electrodes and
to ensure good anodic bonding result. Photoresist is used to define the patterns and buffered HF is used as the etchant to form
the recesses as shown in Fig. 4(d). Aluminum is then evaporated and patterned as resistive heaters by the lift-off process as
shown in Fig. 4(e) and (f), respectively. Finally, The silicon and
Pyrex wafers of 500 m in thickness are aligned and anodically
bonded at 380 C in Fig. 4(g).
C. Experimental Setup
In the proof-of-concept experiments, ceramic particles with
size of 315 m in diameter are mixed with isopropyl alcohol
(IPA) as the pumping liquid. IPA is chosen as the test liquid
for its easy access and low boiling temperature. By tracing the

TSAI AND LIN: THERMAL-BUBBLE-ACTUATED MICRONOZZLE-DIFFUSER PUMP

667

Fig. 5. Schematic drawing of the experimental setups. (a) Flow rate measure
under zero back pressure. (b) Pumping pressure under zero flow rate.

movement of the particles, the net flow speed is estimated. It is

observed that the liquid is oscillating in the same frequency as
the actuation pulse even in the interconnecting polyimide tubes.
When the micropump is operating at frequencies higher than
10 Hz, it becomes difficult to trace particle movement with a
regular image recorder of 25 frame/s.
The experimental setups for both the volume flow rate and
pumping pressure measurements are illustrated in Fig. 5. In both
experiments, the far ends of the polyimide tubes from the micropump are connected to syringes filled with IPA as liquid
reservoirs. In the volume flow rate measurement as shown in
Fig. 5(a), the two syringes are set at the same height for equilibrium pressure. The pumping flow rate is measured from the
averaged speed of an air slug that is placed into the polyimide
tube in advance. Volume flow rate data are calculated from the
average speeds of the air bubble travelling a distance of 1 cm
in each case. The syringes have inner diameter of 14 mm which
is much larger than the diameter of the polyimide tube. Therefore, the liquid level change in the syringe during the pumping
process is negligible and the measured volume flow rate is assumed to be under zero back pressure.
In the pumping pressure measurement shown in Fig. 5(b),
the back pressure that stops the net flow is recorded. The setup
is similar to the one for measuring volume flow rate except
that there is no syringe at the outlet port of the pump. Liquid
level in equilibrium is reached before the pump is turned on.
The pumping pressure is then calculated from the liquid level
change in the outlet polyimide tube when the maximum height
is reached.

Fig. 6. Two sequential pictures showing particle movement in 8 seconds under

an excitation frequency of 3 Hz.

Fig. 7. Volume flow rate of single-bubble pumping mode under pulse voltage
of 20 V.

III. RESULTS AND DISCUSSIONS

Fig. 6 shows two sequential pictures taken 8 seconds apart at
about 1 mm upstream from the nozzle during the particle tracing
experiment. Particles were observed oscillating coherently to
the bubble expanding/collapsing cycle at the pulse excitation
frequency of 3 Hz, duty cycle of 5% and average power consumption of 0.05 Watt. At the neck of the nozzle, particle traces
can hardly be observed during the bubble expanding process
due to the high flow speed. As time goes by, particles can be
observed to have different net displacements. It indicates the

nonuniform velocity profile across the microchannel. An averaged velocity of 6.9 m/s that corresponds to a volume flow rate
of 3.7 nl/min is estimated by tracing the circled particles in a period of 8 s.
When the driving frequency is increased, Fig. 7 shows the
measured flow rates in the single-bubble pumping mode actuated by three different square pulse duties, 5%, 10%, and 15%
over a frequency spectrum up to 500 Hz. The applied pulse
voltage is 20 V and current is 0.5 A. The maximum pumping

668

JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 11, NO. 6, DECEMBER 2002

Fig. 8. Volume flow rate by single- and dual-bubble pumping modes at 5%

duty cycle and 0.5 W average input power.

Fig. 9. Volume flow rate versus pulse width by single-bubble (SB) and
dual-bubble (DB) pumping mode under various average input power.

rate is found as 5 l/min when the pulse duty is 10% and the
pulse frequency is 250 Hz. According the model developed by
of nozzle-diffuser
Olsson et al. [15], the volume flow rate
pumps can be modeled as

(1)

where is the variation of liquid volume in the chamber per

and
are
pumping cycle, is the pulse frequency, and
the coefficients of pressure loss of the nozzle and diffuser, respectively. The thermal bubble based nozzle-diffuser pumping
system represents a rather complicated thermodynamic problem
including heat transfer in the microscale, bubble nucleation and
collapsing, fluidic reaction of the nozzle-diffuser design. A general discussion based on the experimental results of Fig. 7 is
made with the assistance of (1). First, it is assumed that the ratio
of coefficients of pressure loss of the nozzle and diffuser is constant during the operation. When the driving frequency is increased, the flow rate increases proportionally as indicated by
(1) and the experimental results follow the trend. However, this
prediction is based on the assumption that the bubble pumping
volume per cycle remains constant. Two factors contradict this
assumption. First, at very high driving frequency, the bubble
forms and persists like the result of a dc input. Without the
collapsing of bubbles at each cycle, the pumping volume per
cycle, , reduced dramatically and to zero when the chamber is
completely filled with a residual bubble. Second, when the duty
cycle is high and the pumping frequency is low, such as the case
of 15% duty cycle at frequencies lower than 200 Hz in Fig. 7.
It is observed that a large residual bubble can not be collapsed
during the operation, thus the pumping flow rate is reduced and
undetectable. This phenomenon is believed to be the result of
heat accumulation under high duty cycle and low operation frequency that prevents the bubble from collapsing.

Fig. 10. Pumping pressure at zero flow rate and volume flow rate at zero back
pressure under dual-bubble pumping mode, 5% pulse duty and 0.5 W averaged
input power.

In Fig. 8, the flow rate in the single-bubble pumping mode is

compared with the flow rate in the dual-bubble pumping mode.
In both cases, bubbles are roughly distributed in one and two
groups in the single and dual-bubble pumping modes, respectively, as illustrated in Fig. 3. The peak flow rate generated by
the dual-bubble pumping mode is higher than the single-bubble
one under the same average driving power. The possible reason
is that the design in the dual-bubble pumping mode alleviates
the overheating and residual bubble problem such that the maximum flow rate is increased and achieved at a higher pumping
frequency of 400 Hz.
The pumping flow rate relies on the many factors, such as
the input power, duty cycle and driving frequency. Fig. 9 shows
the measured flow rates with respect to pulse widths that is the
product of the pulse duty multiplied with the reciprocal of pulse
frequency. It is observed that when the pulse width is around 0.4
ms for the single-bubble pumping mode and 0.13 millisecond
for the dual-bubble pumping mode, the pumping flow rates

TSAI AND LIN: THERMAL-BUBBLE-ACTUATED MICRONOZZLE-DIFFUSER PUMP

Fig. 11.

669

Schematic drawing of the exemplary integrated system of a micropump and a mixing channel.

reaches maximum values, respectively. The observation implies

that pulse width plays the dominant role in achieving maximum
pumping flow rate.
On the other hand, it is difficult to evaluate the effect of input
power without establishing a three-dimensional (3-D) heat
transfer model that involves phase change, thermodynamics and
fluid reaction in the micro scale. The observation from Fig. 9 is
that the increment in input power helps increasing the pumping
flow rate before the input power is so high that generates
residual bubbles (such as the case of single-bubble pumping
mode with average input power of 1.5 W). It is also observed
that the dual-bubble pumping mode has higher maximum
pumping rates when compared with the single-bubble pumping
mode under a similar average input power. Furthermore, only
a smaller pulse width is required to generate the maximum
pumping flow rate in the case of the dual-bubble pumping
mode. A plausible explanation is that two groups of bubble are
generated under the dual-bubble pumping mode such that it is
easier for thermal bubbles to fill up the chamber for effective
pumping.
The pumping pressure at zero flow rate with respect to
various driving frequencies is measured and plotted in Fig. 10
based on the dual-bubble micropump mode and the flow rate
at zero back pressure is shown together for comparison. The
maximum measured pressure is 377 Pa at 5% pulse duty under
an averaged input power of 0.5 W. In experiments, at each
driving frequency, the liquid is still oscillating in the polyimide
tube and the maximum pressure is recorded when the maximum
fluid height at outlet port is achieved. These experimental data
have been gathered from the prototype devices. There are
two general areas to be further investigated for the optimal
design and operation of the bubble-powered, nozzle-diffuser
micropump: 1) the geometric optimization of the structure such
as the pumping chamber and nozzle-diffuser designs and 2)
the bubble generation mechanisms including heat transfer and
fluidic analyzes. Olsson et al. [14], [15] have reported detail
studies on the modeling and optimization of nozzle-diffuser
designs. Although they have used a piezoelectrically-driven,
mechanical diaphragm as the actuation source, the analytical
and experimental approaches will be extremely useful toward
the optimization of the micropump presented in this work.
In the area of microscale bubble formation, there have been
publications on the static and transient bubble formation

characterizations [16], [17] addressing key issues in microscale

resistive heating and bubble nucleation. It is believed that the
optimization of the thermal-bubble-actuated, nozzle-diffuser
micropump can be achieved with further analytical and experimental studies and with the help of these previous works.

IV. APPLICATION IN MICROFLUIDIC MIXER

This bubble powered nozzle-diffuser micropump has been
used in a microfluidic system to assist mixing of liquids [18].
With proper microchannel design, the oscillatory flows generated by the micropump can be used to induce a wavy interface
between two liquids to enlarge the contact surface and accelerate the mixing process.
Fig. 11 illustrates a microfluidic system including a micropump and a mixing channel. This mixing channel is
constructed together with the micropump process as shown
in Fig. 4 such that no extra manufacturing step is required.
Liquids are supplied and pumped from two inlet ports by means
of gravity force to supply a flow rate of 6.5 l/min and the
micropump is used to agitate the liquid interface to accelerate
the mixing process. Fig. 12 shows the experimental results.
The dark color is the blue dye testing liquid coming from
liquid inlet port 1 and moving upward in the microchannel that
is 200 m in width and of 50 m in depth. The bright color
is IPA passing through micropump and coming from liquid
inlet port 2. The micropump is connected to the IPA injection
channel as shown in Fig. 11. Before the micropump is turned
on, there is little mixing effect as shown in Fig. 12(a) where
tow liquids are flowing with a speed of about 10 mm/sec.
When the micropump is turned on with a driving frequency of
5 Hz, Fig. 12(b) shows the mixing result when the bubble is
expanding. It appears that IPA is pushed out into the blue dye
and the well-defined boundary in Fig. 12(a) is broken. During
the bubble collapsing process, the boundary of IPA moves back
a bit at the intersection of the injection channel and the mixing
channel. Hence, a wavy liquid boundary is generated. This
effect increases the contact area and assists the mixing effect.
Fig. 13 shows a series of pictures, in which the micro pump is
driven under various pulse frequencies of, 5, 50, 100, 150, and
200 Hz from left to right. As it can be seen, the wavelength
of the wavy patterns becomes shorter as the driving frequency

670

Fig. 12.

JOURNAL OF MICROELECTROMECHANICAL SYSTEMS, VOL. 11, NO. 6, DECEMBER 2002

Wavy liquid interface and interlacing injection effect caused by the micropump agitation at 5 Hz driven frequency.

are necessary to improve the reliability and performance of

thermal-bubble-actuated micronozzle-diffuser pumps.
ACKNOWLEDGMENT
The authors would thank the Professor Liepmanns research
group at the University of California (UC), Berkeley, testing and
discussions. These devices are fabricated in the microlab at UC
Berkeley.
REFERENCES

Fig. 13. Mixing of IPA and blue dye near the initial encountering point under
various micropump driving frequencies, 5, 50, 100, 150, and 200 Hz from left
to right.

increases and the color of mixing liquids becomes uniform. It

indicates that better mixing results are accomplished.
V. CONCLUSION
A family of thermal-bubble-actuated micronozzle-diffuser
pumps, including single-bubble and dual-bubble pumping
modes, are demonstrated. The thermal bubbles generated in
the pump are driven by pulsive electrical signals with driving
frequencies up to 500 Hz. The measured pumping flow rate
can reach 5 l/min and the highest pumping pressure is 377
Pa. Optimal pulse widths of 0.4 ms and 0.13 ms are characterized experimentally for highest pumping rates for single
and dual-bubble pumping designs, respectively. It is found that
residual bubbles in the pumping chamber reduce the pumping
efficiency and pressure head and the design of dual-bubble
pumping mode can alleviate the problem. One application is
demonstrated in a microfluidic system to use the bubble pump
to generate oscillatory flow for microfluidic mixing applications. In summary, this work accomplished the principles of
thermal bubble pumping with the help of nozzle-diffuser flow
regulation. Further investigations in the areas of microscale
heat transfer, phase change, fluid dynamics in the nozzle-diffuser designs and bubble nucleation and collapsing processes

[1] H. Mizoguchi, M. Ando, T. Mizuno, T. Takagi, and N. Nakajima, Design and fabrication of light driven micropump, in Proc. 1992 IEEE
Micro Electro Mechanical Systems Workshop, Travemunde, Germany,
1992, pp. 3136.
[2] E. Stemme and G. Stemme, A vavleless diffuser/nozzle-based fluid
pump, Sens. Actuators, vol. A39, pp. 159167, 1993.
[3] B. Bustgens, W. Bacher, W. Menz, and W. K. Schomburg, Micropump
manufactured by thermoplastic molding, in Proc. 1994 IEEE Micro
Electro Mechanical Systems Workshop, Oiso, Japan, 1994, pp. 1821.
[4] T. Jun and C. J. Kim, Microscale pumping with traversing bubbles in
microchannels, in Proc. IEEE Solid-State Sensor Actuator Workshop,
Hilton Head Island, SC, 1996, pp. 144147.
[5] S. F. Bart, L. S. Tavrow, M. Mehregany, and J. H. Lang, Microfabricated electrohydrodynamic pumps, Sens. Actuators, vol. A2123, pp.
193197, 1990.
[6] J. Lee and C. J. Kim, Liquid micromotor driven by continuous
electrowetting, in Proc. 1998 IEEE Micro Electro Mechanical Systems
Workshop, Heidelberg, Germany, 1998, pp. 538543.
[7] N. J. Nielsen, History of ThInkjet printerhead development, HP
Journal, vol. 36, no. 5, pp. 410, 1985.
[8] L. Lin and A. P. Pisano, Thermal bubble powered microactuators, Microsyst. Technol. J., vol. 1, no. 1, pp. 5158, 1994.
[9] H. T. G. Van Lintel, F. C. M. van de Pol, and S. Bouwstra, A Piezoelectric micropump based on micromachining of silicon, Sens. Actuators,
vol. 15, pp. 153167, 1988.
[10] M. Esashi, S. Shoji, and A. Nakano, Normally closed microvalve and
micropump fabricated on a silicon wafer, Sens. Actuators, vol. A20, pp.
163169, 1989.
[11] J. G. Smits, Piezoelectric micropump with three valves working peristaltically, Sens. Actuators, Phys. A, vol. 2123, pp. 203206, 1990.
[12] S. Matsumoto, A. Klein, and R. Maeda, Development of bi-directional
valve-less micropump for liquid, in Proc. 1999 IEEE Micro Electro
Mechanical Systems Workshop, Orlando, FL, 1999, pp. 141146.
[13] F. Forster, R. L. Bardell, M. A. Afromowitz, N. R. Sharma, and A. Blanchard, Design, fabrication and testing of fixed-valve micro-pumps, in
Proc. ASME Fluids Eng. Div., 1995, pp. 3944.
[14] A. Olsson, Valve-Less Diffuser Micropumps,, Royal Institute of Technology, Stockholm, Sweden, 1998.

TSAI AND LIN: THERMAL-BUBBLE-ACTUATED MICRONOZZLE-DIFFUSER PUMP

[15] A. Olsson, G. Stemme, and E. Stemme, A valve-less planar fluid pump

with two pump chambers, Sens. Actuators, vol. A4647, pp. 549556,
1995.
[16] L. Lin, A. P. Pisano, and V. P. Carey, Thermal bubble formations on
polysilicon micro resistors, ASME J. Heat Transfer, vol. 120, pp.
735742, Sept. 1998.
[17] J. H. Tsai and L. Lin, Transient thermal bubble formation on a
polysilicon micro-resister, ASME J. Heat Transfer, vol. 124, no. 2, pp.
375382, 2002.
, Active microfluidic mixer and gas bubble filter driven by thermal
[18]
bubble micropump, Sens. Actuators , vol. A9798, pp. 665671, 2002.

Jr-Hung Tsai received the M.S. degree in power mechanical engineering from
the National Tsing Hua University at Taiwan in 1996 and the Ph.D. degree in
mechanical engineering from the University of Michigan, Ann Arbor, in 2001.
He is currently with Kumetrix, Inc., CA, as a a research staff member. His
major research interests are in design, modeling, and fabrication of microfluidic
components, microbiosensors, and microactuators in microfluidic systems.

671

Liwei Lin (S92M93) received the M.S. and

Ph.D. degrees in mechanical engineering from the
University of California, Berkeley, in 1991 and
1993, respectively.
From 1993 to 1994, he was with BEI Electronics,
Inc., in research and development of microsensors.
From 1994 to 1996, he was an Associate Professor in
the Institute of Applied Mechanics, National Taiwan
University, Taiwan. From 1996 to 1999, he was an
Assistant Professor at the Mechanical Engineering
and Applied Mechanics Department at the University of Michigan. In 1999, he joined the University of California at Berkeley
and is now an Associate Professor at Mechanical Engineering Department and
Co-Director at Berkeley Sensor and Actuator Center, NSF/Industry/University
research cooperative center. His research interests are in design, modeling and
fabrication of microstructures, microsensors, and microactuators as well as
mechanical issues in microelectromechanical systems including heat transfer,
solid/fluid mechanics, and dynamics. He holds seven U.S. patents in the area
of MEMS.
Dr. Lin is the recipient of the 1998 NSF CAREER Award for research in
MEMS Packaging and the 1999 ASME Journal of Heat Transfer best paper
award for his work on microscale bubble formation. He led the effort in establishing the MEMS subdivision in ASME and is currently serving as the Vice
Chairman of the Executive Committee for the MEMS subdivision.