MEMS

Micro Electro Mechanical Systems

MEMS ACTUATORS
ELECTROSTATIC ACTUATION
• Electrostatic actuation is produced by the electric field of a
capacitor.
• Two basic configurations of a capacitor for electrostatic
actuation of a MEMS device:
• the parallel plate
• the interdigitated comb capacitor

• The interdigitated comb capacitor is dominated by the fringe

electrostatic field
• The parallel plate capacitor is dominated by the direct
electrostatic field
Parallel Plate Capacitor
• Capacitance
• In a variable parallel plate capacitor,
• the movable plate moves normally to the fixed plate as defined
by the coordinate, z

• Energy of a capacitor,

• Electrostatic force between the plates is determined by

differentiating the energy function, W with respect to the
coordinate in the direction of the force.
• As z increases and the
distance between the
plates goes to zero, the
electrostatic force
becomes very large.
• The equilibrium point for
a capacitor-spring
system for each voltage
can be plotted at the
intersection of the spring
constant and force curve,
which are marked with an
“o.”
Force balance between
• pull-in, occurs at the point the spring and the
at which the electrostatic electrostatic forces
force curve and the
spring constant curve are
tangent, the electrostatic
instability point
• Pull-in phenomenon will occur when the derivative of voltage
with respect to position is zero.

• Deflection at pull-in,
• Voltage at which pull-in occurs,

• The deflection at pull-in is a

• function of only geometric
• quantities (i.e., not a function
• of spring constant, area, etc.).
• The limited deflection before
• pull-in occurs is a significant
• design constraint that must be
• considered in the design of
• sensors and actuators.
• electrostatic spring softening.
• Taylor series expansion of force balance at pull in gives

• The linear term of the Taylor series is the slope of the

electrostatic force vs deflection curve
• It can also be viewed as a negative electrostatic stiffness, Kes,
that will cancel the elastic stiffness at pull-in.
• The electrostatic spring softening can be used to soften or tune
the spring stiffness electrostatically
Interdigitated Comb Capacitor
• A movable plate inserted between two fixed plates.
• The direct parallel plate electrostatic force and the
• fringe field force are the two types of electrostatic
• forces acting upon the movable plate.
• The motion of the movable plate is constrained so
• that motion can only be in the direction that inserts
• the movable plate between the two fixed plates
• and maintains the gaps between them.
• The width of the combs is w.
• Capacitance,

• Electrostatic force is not a function of displacement, and hence

is a more controllable force.
• Parallel plate electrostatic forces are more powerful than the
fringe field electrostatic forces of the combs.
• The constant force
contour lines are
horizontal, indicating that
they are not a function of
displacement.
• If this variable capacitor is
connected to an elastic
spring, the spring and
electrostatic force lines
uniquely intersect for each
voltage applied.
• There are no electrostatic
instability phenomena for
this configuration.
• Because electrostatics is only
• attractive, there are combs to
• actuate in both directions
• Stroke: 17 μM in both directions
• approximately a 25-μN force.
• The folded spring suspension
allows motion in the comb
meshing direction, but prevents
the combs from moving laterally
due to the parallel plate force
of the comb fingers.
• The actuation speed capabilities of electrostatic actuators are very high and limited
only by the natural frequency of the suspension and the electrical time constant of
the system.
• Parallel plate actuation can provide very high forces (~100 μN) with small stroke (~3
μm), but the force is highly nonlinear with instability within the displacement range.
• Interdigitated comb actuation provides a moderate level of force (~10 μN) with large
strokes (~20 μm) with a very controllable force (i.e., no instabilities).
• The pull-in instability of parallel plate electrostatics is a significant constraint
limiting the stable actuation range that electrostatic actuator designers have
attempted to overcome.
• Stable repeatable actuation (positioning) throughout the operational range is
required for devices such as analog optical mirrors.

• 1.Passive circuit components: The inclusion of a fixed capacitor in series

with the variable capacitor.
• The additional capacitor has the effect of extending the effective gap of the
actuator and its stable range of operation.
• Parasitic capacitance is an issue, but the stable range of actuation can be
extended
• The size of the additional capacitor is a complicating factor in actuator
design.

• 2.Closed loop control:. This approach required an integrated electronics and

MEMS fabrication process to implement.
• The utilization of this approach is very complicated and difficult.
• 3.Leveraged bending. It is a simple technique where a beam is
electrostatically actuated close to its anchor
• It utilizes the leverage of the beam length to obtain the actuation
deflection at another position on the beam.
• The stable deflection is obtained at the expense of greater beam
length and actuation voltage.
THERMAL ACTUATION
• Thermal actuation uses the thermal expansion of materials (solid, liquid, gas) to
achieve mechanical actuation.
• The thermal expansion of a solid material is characterized by coefficient of thermal
expansion (CTE), αT.
• CTE of a material increases with temperature
• Thermal expansion of a solid produce high forces if constrained but a low amount of
stroke.
• Mechanical actuation due to the thermal expansion of a solid material can be
accomplished by a device with two or more materials with different CTE, or by the
thermal gradients within a device made of the one material.
• A bimorph utilizes the CTE mismatch of materials to achieve mechanical deflection.
• Bimorph actuators are generally long thin beams with layers of materials with
different CTE
• CTE mismatch of the materials will produce a moment, M, in the beam when the
temperature of the beam is changed
• The moment in the beam is calculated by integrating the thermal expansion forces
through the thickness of the beam.
• Although this type of thermal actuator has limited stroke, it has been used in fluidic
control valve applications.
• Lateral thermal actuator consists of a thin hot leg and a
wide cold leg;
• These will create a difference in temperature due the
differing current densities and ability to transfer heat.
• The actuator is heated by Joule heating due to current
flow.
• The hot leg is heated to 400 to 600°C, and the actuator
will deflect laterally.
• These actuators can only move in one direction and are
not reversible (i.e., positive vs. negative direction).
• Bent beam thermal actuator is a symmetric structure consisting of a long
thin beam canted at a small angle.
• When Joule heating is applied to the device, the bent beam expands and
produces lateral motion.
• This beam is a mechanical amplifier of the small deflection produced by
thermal expansion.
• The bent beam devices may be cascaded by running in parallel as well as
by using additional bent beam structures for mechanical amplification
• A thermal actuator can produce high force (>100 μN) and a displacement of
10 μm or greater utilizing only low voltage ~ 10 V.
• However, these actuators can require significant power (~100 mW).
• The design of thermal actuators is a multiphysics problem requiring thermal,
structural, and electrical analysis.
• The heat transfer involved in a thermal analysis includes conduction and
convection.
• Radiation heat transfer is generally not significant.
• The upper practical limit for temperature in the thermal actuator is
approximately 600°C, above which material property changes such as
localized plastic yielding and material grain growth become an issue.
• Thermal actuators are also generally limited to a frequency response less
than 1000 Hz because of the time constants associated with heat transfer.

• Thermal actuation can be used for direct mechanical motion by thermal

expansion of a fluid or change of phase of the fluid.
3.LORENTZ FORCE ACTUATION
• Lorentz force is generated by a magnetic
• field, B, and current, i, in a conductor of
• length, L
• Microvalve : actuated by Lorentz force
• produced by the interaction of a DC current, and a magnetic
field, created by a permanent magnet mounted adjacent to the
valve.
• MEMS magnetometer: actuated by cyclic Lorentz force
produced by an AC current, i, interacting with a
• permanent magnet
• Can detect changes in magnetic fields and
• could be used to establish direction of
• movement of magnetic objects like hidden
• weapons or bombs
MEMS SENSING
1.CAPACITATIVE SENSING
• MEMS devices with parallel plate or interdigitated comb finger
capacitors that can move as a result of a physical variable
(acceleration, pressure) excitation
• Differential capacitors can be fabricated in such a manner that
the two capacitors have nearly equal capacitance in the
unperturbed position and change by equal and opposite
amounts as the structure is deflected.
• size of the capacitors utilized in MEMS devices is small —
generally a fraction of a picofarad.
• The variation of the nominal capacitance to be sensed to
provide the dynamic signal of interest is in the femtofarad range
or less.
2.PIEZORESISTIVE SENSING
• Piezoresistivity is a widely utilized phenomenon for MEMS
sensors.
• It is the basic operating principle for the metal and foil strain
gauges
• piezoresistive effect in silicon had following impacts
• Integration with MEMS devices and microelectronics is
possible due to material compatibility.
• Integration of the piezoresistive material and the MEMS
device allows good transmission of strain without
hysteresis or creep.
• The piezoresistive effect in silicon is over an order of
magnitude greater than in metals.
• MEMS fabrication processes allow good matching of
resistors utilized in the Wheatstone bridge-sensing
circuits.
• resistivity of a semiconductor
• number of charge carriers, N, and their mobility, μ.
• The resistivity of a semiconductor such as silicon can be
controlled by the concentration of an impurity that directly
controls the number of charge carriers N.
• The carrier mobility, μ, is a function of temperature and doping
concentration.
• Mobility decreases with increasing temperature because lattice
vibrations caused by increased temperature scatter the
electrons.
Piezoresistance in Single-Crystal Silicon
• Single crystal silicon has an anisotropic crystal structure of the
cubic family
• Electric field vector, {E},
• {J}, current density vector
• resistivity matrix, [ρ]

• For an unstressed crystal in the cubic family such as silicon,

• The resistivity of a piezoresistive material is a function of stress

that is also direction dependent due to the anisotropic crystal
structure.
• The resistivity components of a stressed crystal is the sum of
the unstressed resistivity and the change in resistivity due to
stress
• The changes in resistivity, ρij, due
• to stress can be quantified by associating
• piezoelectric coefficients, πij, with
• every stress component
• The piezoelectric coefficients indicate the magnitude of the
piezoresistive effect for a material.
• These coefficients are properties of a material and are affected by the
temperature and doping of the material.
• The coefficients decrease with increasing temperature.
• The temperature sensitivity of the coefficients is a major concern for
piezoelectric sensors.
• The piezoelectric coefficients will also decrease with increasing
impurity (dopant) concentrations.
• Gauge factor, G, is a term frequently used to describe the sensitivity
of a piezoresistive sensor.
• The gauge factor is the ratio of the fractional change in resistance,
ΔR/R and the strain ε

• For single-crystal silicon G ≈ 100 and metal strain gauge G ≈ 2.

Piezoresistivity of Polycrystalline and Amorphous Silicon
• Polycrystalline silicon is composed of small silicon crystals separated by
grain boundaries.
• Polycrystalline materials may show texture, which is a statistical measure of
the crystal orientations within the polycrystalline material.
• Amorphous silicon has no crystalline structure at all.
• Their resistivity can be controlled by ion implantation with boron or
phosphorus
• The total resistance in a polycrystalline material is a combination of the
resistance of the grains and the resistance of the grain boundaries.
• They have high gauge factor compared to metal foil sensors;
• However, their gauge factors are
significantly lower than that of single-
crystal silicon and they strongly depend
upon processing parameters.
Signal Detection of piezoresistive devices
• Half- or full-bridge resistor bridges can provide
high sensitivity and first-order temperature
compensation.
• The three configurations of wheatstone bridge,
single-active bridge, half-active bridge, and
fully active bridge, utilize one, two, and four
variable resistances, respectively.
• For a half-bridge (i.e., voltage divider) circuit the output voltage will have a
DC offset of Vin/2 and a small nonlinearity due to the deleted terms from the
Taylor series expansion
• A linearized output voltage relationship for the single- and half-active bridge
circuits was obtained via a Taylor series expansion; therefore, both of these
configurations have a small nonlinearity in their response.
• The output voltage gain for the single active-bridge is the same as the
voltage divider, but the single active-bridge does not have an offset voltage.
• The half-active bridge circuit has an improved output voltage gain, but the
two variable resistors must be matched.
• MEMS fabrication processes will tend to minimize the variations in the
resistors.
• In fully active bridge, two resistors increase resistance with increasing strain,
and the other two decrease resistance with increasing strain.
• By careful design in placement and orientation of the resistors, these criteria
can be met with a single-crystal silicon piezoelectric approach.
• The fully active bridge has the largest output voltage gain and has a linear
response.
3. ELECTRON TUNNELING
• Used for infrared (IR), magnetometer, and accelerometer sensors.
• Electron tunneling can provide an extremely sensitive method of position
transduction.
• A current is passed across a narrow gap.
• A gap of finite size would pose a barrier to current flow.
• However, for sufficiently small gaps (~10Å), the probabilistic
• nature of quantum mechanics becomes apparent.
• In quantum mechanics, when a particle comes to a barrier that it does not
have enough energy to penetrate, the wave function dies off exponentially.
• However, for sufficiently small gaps (~10Å),the wave function will predict a
significant probability of finding the particle on the other side of the gap.
• Therefore, if the gap is small enough, a tunneling current will exist even
though a break has occurred in the circuit.
• The electron tunneling current I
• s = gap
• Φ= height of the tunneling barrier
• V = bias voltage (V << Φ)
• α = conversion factor,1.025Åo –1 ev–1/2)
• Typical values of Φ and s are 1 eV and 10 Åo.
• A typical tip bias voltage, V, is only a few 100 mV and a typical tunneling
current, I, is 1 nA.
• Electron tunneling is an extremely sensitive method of position transduction
in which the current can vary by a factor of three for each Å o change in gap
separation, s.
• The tunneling tip and opposing surface must be metalized with a thin layer
of metal such as 100 Åo of gold.
• The tunnel effect is not extremely sensitive to tip geometry.
• One reported tunneling tip was a 50-μm pyramid with a 1 to 5-μm radius of
curvature;
• Because the tunneling tip is so close to the surface of a moving mass or
membrane to be measured, the gap must be controlled by feedback during
operation.
• This can be accomplished by measuring the tunneling current and applying
correction signals an actuator to control the position of the tunneling tip or
the moving mass.
SENSOR NOISE
• Noise in a sensor can arise from many sources, ranging from
very small temperature-induced vibrations of the sensor atoms
to electronic noise produced by the discrete nature of electrons.
• A noise signal can be described in the time domain via plots of
amplitude vs. time or in the frequency domain via plots of
amplitude or phase vs. frequency
• Power spectral density function (PSD), S(f): is a function of
frequency and has units of amplitude squared per hertz.
• The PSD function can be describing a voltage signal that would
be a plot of V2/Hz vs. hertz.
• For an acceleration signal, the plot would be acceleration
• squared per hertz (i.e., g2/Hz).
• The PSD provides an indication of how the “power” of the signal
is distributed over frequency.
• A noise signal that has equal
• Frequency content over all frequencies
• is known as white noise;
• A signal that has limited frequency
• content is called band limited.
• As signals pass through different instruments such as amplifiers, filters,
actuators, or sensors, the frequency content of the signal is altered.
• The mean square of the signal,

• Mean square amplitude of the total signal, which is a combination of several

other signals:

• signal to noise ratio (SNR) is a metric of the relative amount of signal and
noise present in a sensor output.
Noise Sources
• Noise in a sensor can arise from a number of mechanical and
electrical effects.
• Shot noise is associated with direct current flow across potential
barriers present in devices such as p–n diodes and bipolar
transistors.
• The passage of each current carrier (electrons or holes) across the
p–n junction in diodes and transistors is a random event dependent
upon the carrier having sufficient energy.
• The continuous external current is actually a large number of discrete
pulses.
• The time constant associated with the passage of carriers across the
potential barrier is extremely small; therefore, the PSD of the shot
noise current can be modeled as white,
• Burst or popcorn noise, is sometimes found in integrated circuits
and discrete transistors.
• The mechanism is not fully understood, but is believed to be related
• Flicker noise is a low-frequency noise component arising from the
capture and release of charge carriers by trap sites produced by crystal
defects or contamination in semiconductors.
• The trap sites capture and release carriers randomly with time constants
that are primarily low frequency.
• The flicker noise current PSD shows a 1/f dependency in amplitude vs.
frequency; and hence called 1/f noise.
• Thermal noise is generated in electrical and mechanical components;
• it arises from the random motion of electrons and atoms, which is
directly proportional to the absolute temperature, T.
• Thermal noise is generated in these systems through their energy
dissipation mechanisms (i.e., resistance for electrical or damping for
mechanical).
• Thermal noise is also known as Johnson noise in electrical systems and
Brownian noise in mechanical systems.
• The PSD for thermal noise is white.
• This is the most fundamental noise limitation for a sensor.
MEMS PHYSICAL SENSORS
1. Accelerometer
• Accelerometers are one of the most frequently utilized physical sensors for
detecting and measuring motion.
• Accelerometers have found applications ranging from measurement and
control to inertial navigation.
• MEMS accelerometers have a large commercial market in automotive
airbag deployment
• The basic components of an accelerometer:
• Inertial mass
• Suspension
• Sensing element
• The suspension supporting the inertial mass will deflect under acceleration
• A sensing element will transduce the deflection of the suspension to an
electrical signal.
• The transduction can be accomplished by a number of means, but the most
common utilize piezoresistive, piezoelectric, or capacitance means.
• Relative displacement of the inertial mass relative
• to the housing, Z = X –Y
• Z is the variable that could be transduced to an
• electrical signal.
• Force balance on the inertial mass including
• spring, K, and damper, C, forces is given by

• Combining the equations,

• i.e. the acceleration input to the housing , is related to the dynamics of the
relative displacement of the housing and inertial mass, Z.

• The relative displacement, Z, is a quantity that can be measured through

means such as the displacement of a beam, deflection of a piezoelectric
crystal, or capacitance change of the mass relative to the housing
• The design of an accelerometer depends upon the system damping.
• The amount of damping will determine the dynamic response and the
Brownian noise and thus the noise floor of the sensor.
• Macroscale accelerometers are designed to have a damping ratio of , ζ =
0.7.
• This type of system will have a fast response with very small overshoot.
• MEMS accelerometer designs, the natural frequency of the inertial mass-
suspension system is at least an order of magnitude higher than the highest
frequency signal to be sensed.
• Mechanical sensitivity of an accelerometer, SM, is the relationship between
the relative deflection of the inertial mass and case, Z, and the input
acceleration,
• Because the operating range of the accelerometer is at low frequency,
where ω = 0, an acceleration input to the system may be approximated by a
constant acceleration balanced by the suspension
Gyroscope
• Gyroscope: an inertial instrument capable of
sensing rotation
• The first gyroscope (Leon Foucault, 1852) was
based on the angular momentum of a spinning
wheel. H = Iω
• According to Newton’s laws of motion, the angular momentum of a body will
remain unchanged unless acted on by a torque.
• If a torque is applied in the same axis as the angular velocity, the effect is to
accelerate or decelerate the rotating body
• If the torque is applied orthogonal to the spin axis, the rotating body will
precess, Ω

• The cross-product in the second term generates the interesting gyroscopic

effects (i.e.,Ω , H, and T are related by the right-hand rule).
• Precession or the moments generated by precession are the effect utilized by
this form of gyroscope as a measure of angular rate.
• The spinning wheel gyroscope is used to implement a class of high-
performance gyroscopes for inertial navigation
• Because the fabrication of this type of gyroscope requires precision bearings,
machining, drive motors, and electronics, it is very costly.
• The development of a MEMS spinning mass gyroscope was
initially inhibited due to the lack of low-friction bearings and the
significant stiction and adhesion forces at the microscale.
• However, promising research on the development of an
electrostatic levitated spinning mass MEMS gyroscope is
proceeding
• This is an ambitious approach because of the necessity of
closed-loop control to stabilize the levitation, in addition to
driving the spinning mass and sensing its deflections due to
precession.
• This approach is currently in the research stages and no MEMS
spinning mass gyroscope commercial products are available.
• Optical gyroscope utilizing the Sagnac effect can be implemented with two
counter-rotating light beams circulating around an optical path of radius, R,
• The optical path is rotating with angular velocity, Ω
• The Sagnac effect can be observed by the time difference, Δt, between the
clockwise and counterclockwise beams striking a detector that is in and
rotating with the optical path.
• If the optical path is not rotating, the optical signal traveling in either direction
will complete the path at the same time.
• If the optical path is rotating clockwise the optical signal traveling in the
same direction as the rotation will have a slightly longer distance to travel
than the optical signal traveling in the opposite direction.
• The tangential speed of the rotating optical path: v = ΩR.
• The initial separation of the start and end points of the optical signals is 2πR
or, if one allows the signal to circulate N times around the path, N2πR.
• The time difference in the arrival of the signals due to the Sagnac effect

• c is the speed of light in the optical path medium.

• The Sagnac effect time interval is very small.
• The measurement of the Earth’s rotation rate (i.e., 15°/h) with a 1-km long
optical path will produce a Sagnac effect of only Δt = 3.3 × 10–9 s.
• A short time interval such as this can be resolved by phase shift effects of
the optical signals.
• The basic configuration can be implemented with fiber optics with multiple
turns (N) to increase path length and the Δt or phase shift measured as an
indication of rotation rate, Ω.
• The Sagnac effect is the basis of a number of optical rotation rate sensors
such as the interferometric fiber-optic gyro (IFOG).
• Vibratory gyroscope: the mechanism utilized by biological systems
such as a fly’s ability to sense angular rotation
• Vibratory gyroscopes are based on Coriolis acceleration, which is an
acceleration produced due to the changing direction in space of the
velocity of the body relative to the moving system.
• Coriolis acceleration, Acoriolis, produced on a body moving around an
axis with a fixed angular velocity,Ω , and moving radially with a
velocity V: Acoriolis = 2Ω x V
• The detection of the deflection of an object due to Coriolis
acceleration is the basis for a vibratory gyroscope.
Coriolis acceleration

Whenever a point is moving on a path and the path is

rotating, there is an extra component of the
acceleration due to coupling between the motion of
the point on the path and the rotation of the path.
This component is called Coriolis acceleration.

46
• A vibratory gyroscope is composed of a resonator that will
oscillate a body along one axis and measure the orthogonal
movement or force on the body due to Coriolis acceleration.
• The plate is driven along the x axis, the rotation rate to be
measured; Ω is along the z axis and the Coriolis acceleration
response is sensed along the y axis.
• Force balance for the body in the drive (x) and sense (y) axes

• The physical mechanism for a vibratory gyroscope is the

transfer of energy from one resonator axis to another via the
Coriolis acceleration coupling.
• The relative positioning of the suspension natural frequencies is a
gyroscope design decision.
• The sense direction natural frequency, ωy, is approximately 10% less than
the drive direction natural frequency, ωx.
• This will provide a modest mechanical gain without significant bandwidth or
phase shift reductions.
• The damping ratio of the mass in the x and y axes depends on the
orientation of the mass relative to the substrate, which will determine the
damping mechanism involved (e.g., squeeze film vs. lateral shear damping).
• The implementation of the gyroscope will require the mass to be driven in
the x axis by the force Fx.
• For many MEMS designs, Fx is electrostatic, such as an interdigitated
electrostatic comb drive.
• The drive amplitude, x, must be maintained very accurately because any
variation will directly contribute an error into the sense direction amplitude
and the gyroscope output.
• For this reason, the drive axis amplitude is controlled by an automatic gain
• The velocity, of the drive signal (the input to the Coriolis term in y-dirn
force balance equation) is simple harmonic motion,
• It will be zero at the extremes of motion of the driven mass and a maximum
as the mass passes through the undeflected position.
• The mass x displacement and the Coriolis force, which contains an x
velocity term, have a 90° phase difference
• The y displacement due to the Coriolis force will also have a 90° phase
difference.
• These signals are said to be in quadrature.
• This will lead to an oval deflection path (symmetric about the x axis) of the
mass when the gyroscope is subject to a constant rotation rate.
• With a zero rotation rate, the mass deflection pattern will not deflect in the y
direction and oscillate entirely along the x axis
• If mass or stiffness imbalances (2nd and 4th term in the equation)
exist in the system dynamics, the mass deflection pattern will
be as shown in Figure c.

• These subtle imbalances in the vibration of the sense mass

produce a deflection in the y direction known as quadrature
error, which contaminates the Coriolis signal, which is the
measure of rotation rate.
• The effects of quadrature error can be negated by a quadrature
error cancellation scheme involving the use of electrostatic
actuators with properly phased signals to cancel the imbalance
• The one configuration that has
been employed for macroscale
and MEMS vibratory gyroscopes is
the tuning fork gyro (TFG)
• The TFG consists of two plates
that are driven in an antiphase
manner (i.e., both plates move
outward and inward relative to the
center axis).
• The rotational field will cause the
plates to move perpendicular to
the substrate in opposite
directions.
• This configuration enables
differential sensing, which will
allow common mode signals such
as external accelerations to be
rejected.
• The two masses may have
coupled or separate suspensions.
Pressure Sensors
• The greatest commercial application for a MEMS pressure sensor is for
pressure greater than atmospheric regime
• The method of implementation of a vacuum sensor depends on the
operating range.
• In the 1 to 2000 mtorr range, vacuum may be sensed by measuring the
thermal conductivity of the ambient gas.
• At vacuums of 1 mtorr to 10–8 torr, ionization of the gas may be used as a
measure of vacuum.
• A Pirani gauge, which measures vacuum, can be implemented by
measuring the thermal conductivity of the ambient gas.
• Thermal conductivity proportional to pressure can be implemented by a
heated filament suspended in a gas
• The resistance of the suspended filament is a
function of temperature, which is related to the
surrounding pressure and thermal conductivity.
• At very low pressures (1 mtorr to 10–8 torr), gas ionization may be used as
a measure of vacuum.
• An Ionization gauge emits electrons from a cathode; these are accelerated
toward an anode plate.
• Positive ions are created by the electron–gas collisions.
• The current on the anode is proportional to the absolute pressure (vacuum)
of the gas.
• Pressure sensors for greater than atmospheric pressure utilize a deformable
• diaphragm.
• The deflection of the diaphragm is the measure of pressure, which can be
sensed by capacitative, piezoresistive, or optical means.
• The pressure sensor may measure absolute pressure that has a vacuum or
a reference pressure on one side of the diaphragm.
• Alternatively, a gauge pressure or differential pressure sensor would have
one side of the diaphragm vented to atmosphere or to another pressure that
would be a reference for the
measurement.
• The pressure diaphragm is generally rectangular or circular in
shape.
• Pressure sensors fabricated with bulk micromachining methods
are generally rectangular
• Motorola MPX200 pressure sensor has a square, single-crystal
diaphragm that is 1448 μm in length and 26.5 μm thick
• The shape of surface micromachined pressure sensors is not
restricted because they are photolithograhically defined.
• A typical surface micromachined pressure sensor can have a 2-
μm thick diaphragm of 200 μm diameter.
• Circular diaphragms may have an advantage over square or
rectangular ones due to the absence of stress concentrations at
the corners.
• The applied pressure for the diaphragm-based pressure
sensors is determined by the deflection of the diaphragm.
• The pressure on the diaphragm is directly proportional to the
applied pressure for rectangular and circular diaphragms.
• For the case of a diaphragm with large built-in stress or large
deflections, the direct proportionality is no longer true.
• In general, it is desirable to use a diaphragm with a linear
relationship with pressure because calibration and
measurement are simpler.
• Deflection of the diaphragm may be sensed via capacitance or
piezoresistive sensing
• Pressure sensors are generally used for low-frequency pressure
measurement.
• Microphones are similar to pressure sensors, but microphones sense a
dynamic pressure signal.
• The frequency response and mechanical sensitivity of ordinary pressure
sensors are inadequate for use as a microphone due to the acoustic
resistance and squeeze film damping between the diaphragm and stationary
cavity of the sensor.
• The damping effects can be mitigated by perforating a stationary plate or
venting so that the air can escape to a larger chamber
• These modifications will allow a flat frequency response over a broader
range.

Microphone with a perforated backplate

CHEMICAL SENSORS
• The chemical to be sensed, analyte, interacts with a chemically
sensitive layer, which will produce an effect that can be
transduced to an electrical signal.
• The chemical reaction that occurs between the analyte and the
chemically sensitive layer may be reversible or irreversible.
• Some reversible chemical reactions may be reversed simply by
removing the chemical to be sensed, thus causing it to
dissociate from the sensitive layer.
• Others may be reversed through the addition of heat, which will
cause the analyte to detach with no net change in the
chemically sensitive layer.
• An irreversible reaction will cause the sensitive layer to be
consumed, which limits the sensor lifetime.
• Biological sensors are similar in approach to chemical sensors
except that the sensitive coating may include biological
materials such as an antigen.
• The effect produced by the chemical
reaction between the analyte and the
chemically sensitive layer
• Electrochemical: changes in charge, electric potential, or current; can be
directly transduced to an electrical output.
• Eg: chemically sensitive resistors, FETs, and capacitors.
• Heat: Measures heat produced by the analyte reacting with the chemically
sensitive layer.
• The heat produced by the reaction is directly related to the analyte
concentration.
• Eg: Calorimetric sensors for the detection of glucose,gases, etc.
• Optical: utilize optical absorption, transmission, or luminescence resulting
from the analyte interacting with the chemically sensitive layer.
• They can be highly sensitive but can be limited by the optical properties at
the wavelength of interest.
• Eg:to sense pH, oxygen, glucose, etc.,
• Mass change: The mass change resulting from the chemical reaction of the
analyte and the chemically sensitive layer may be detected by highly
sensitive acoustic wave devices
• These devices generate a high-frequency wave on the surface of a
piezoelectric crystal, which is coated with the chemically sensitive material.
• As the wave passes through the material with adsorbed analyte molecules,
their effect on the velocity of the acoustic wave is detected.
• The acoustic wave devices are split into two categories: bulk acoustic wave
(BAW) and surface acoustic wave (SAW).
• SAW devices generally operate at frequencies above 50 MHz.
• A well-known member of the BAW device category is the quartz
microbalance (QMB), which is generally 10 to 15 mm in size, operates
between 10 and 30 MHz, and can detect mass changes as low as 10 –9 to
10–10 g/cm2.

• A unique application of chemical sensing is the electronic nose

• This is a combination of multiple chemical sensor detection and pattern
recognition to achieve the recognition of complex odors.
• Appln: food processing industry for the recognition of food freshness,
beverages, and perfumes.
Taguchi Gas Sensor
• Taguchi gas sensor (TGS), a very successful commercially
available sensor for a wide variety of gases, was first developed
by N. Taguchi in 1971.
• Appln:alcohol breath analyzers; automatic cooking controls;
combustible gases (methane, propane, CO, hydrogen, etc.);
volatile organic vapors (alcohol, ketone, esters, benzols, etc.);
• This sensor is a solid-state sensor composed of a sintered
metal oxide (SnO2), which detects gases through an increase
in electrical conductivity when reducing gases are adsorbed on
the sensor’s surface.
• The sensing material, tin dioxide (SnO2), is a polycrystalline
material consisting of crystals (grains) embedded in an
amorphous matrix of the material.
• The metal oxide is heated to 300 to 400°C in air.
• Oxygen is adsorbed on the crystal surface with a negative charge.
• Donor electrons in the crystal are transferred to the adsorbed oxygen, which
results in positive charges in a space charge layer in the crystal.
• A surface potential that is a potential barrier to electron flow (i.e.,resistance
increases) is produced.

• An analyte molecule reacts with the adsorbed oxygen to release electrons;

this decreases the resistance.
• This is a reversible reaction that is subsequently repeated.
• The sensor incorporates a heater to heat the metal oxide to 300 to 400°C,
which consumes a few hundred milliwatts.
• The temperature increases the chemical reaction rate and speeds the
sensor response time.
• The metal oxide can be doped to improve
the sensitivity and selectivity of the sensor
to specific gases
Optical Pick-Off Pressure Sensors
• Optical devices offer particular advantages in speed and remote
sensing,
• As the sensors can be addressed and read with light, they are
potentially useful for hazardous environments.
• The measurement of high-speed turbulent air-flow on aeronautical
surfaces requires remote sensing, and good spatial and temporal
resolution.
• An approach is to correlate pressure fluctuations in the ambient to
mechanical vibrations of the surface.
• An array of pressure-sensitive
• cavities are used in conjunction
• with an external near-infrared laser
THERMAL SENSORS
• Thermal sensors are sensors that measure a primary thermal quantity, such
as temperature, heat flow, or thermal conductivity.
• Other sensors may be based on a thermal measurement; for example, a
thermal anemometer measures air flow.
• The most important thermal sensor is the temperature sensor.
• Material properties like Young's modulus, shear modulus, heat capacity,
thermal conductivity, etc. - vary with operating temperature.
• The effect of temperature can sometimes be minimised by choosing
materials with a low temperature coefficient of operation
• Many nonthermal microsensors (and MEMS devices) have to operate either
at a constant temperature - an expensive and power-intensive option when
requiring heaters or coolers, - or in a mode in which the temperature is
monitored and real-time signal compensation is provided.
• Temperature sensors are commonly found embedded in microsensors,
microactuators, MEMS, and even in precision microelectronic components,
such as analogue-to-digital converters.
Resistive Temperature Microsensors
• The bulk electrical resistivity ρ, and hence resistance R, varies with absolute
temperature T.
• For metal chemoresistors,
• ρo, Ro: at a standard temperature
• αT: linear temperature coefficient of resistivity or resistance (TCR)
• Platinum is the most commonly used metal in resistive temperature sensors
because it is very stable when cycled over a very wide operating
temperature range of approximately —260 to +1700°C, with a typical
reproducibility of better than ±0.1 °C.
• Platinum temperature sensors are very nearly linear, with αT: of +3.9 x 10-
4
/K and βT:—5.9 x 10-7/K2.
• Thermistors are resistors formed from semiconducting materials, such as
sulfides, selenides, or oxides of Ni, Mn, or Cu, and Si
• They have highly nonlinear temperature-dependence.
Microthermocouples
• Works on Seebeck effect
• The metals have a different thermoelectric
power or Seebeck coefficient P;
• The thermocouple is a linear device,
• The voltage output (at zero current) being given by

• The Seebeck coefficient of single-crystal silicon varies with both

temperature and doping concentration (p-type)
• Doping reduces the temperature variation of the coefficient
• Hence, the response of a silicon-based
thermocouple becomes more linear.
• As a variety of doping levels are possible
in a planar IC process, a Seebeck coefficient
ranging from +0.5 to +5 mV/°C is achievable.
• Seebeck coefficient of a doped semiconductor is given by

• kB is the Boltzmann's constant, q is the carrier charge,

• Nc and Nv are the density of states at the bottom of the
conductance band and top of the valence band,
• n and p are the donor and acceptor concentrations,
• s is a parameter related to the mean free time between
collisions and the charge carrier energy and its value varies
between —1 and +2 depending on whether the carriers can
move freely or are trapped,
• Φ is a phonon drag term for the carrier.
• Seebeck coefficient can be readily estimated from the silicon
resistivity rather than the carrier concentrations
• A silicon thermocouple can be made in an IC process with doped
silicon and a standard metal contact, for example, aluminum.
• As the absolute Seebeck coefficient of p-type silicon is positive (e.g.
+1 mV/K at 300 K) and that for aluminum is negative (i.e. —1.7 μV/K
at 300 K), an output on the order of n millivolts per degree can be
achieved from a thermopile.
• Polysilicon/gold thermocouples have output of about +0.4 mV/K in
which the n-type (phosphorous) polysilicon has a lower Seebeck
coefficient of -176 μV/K and the gold has a standard value of +194
μV/K.
Thermodiodes and Thermotransistors
• Simplest and easiest way to make an integrated temperature sensor
is to use a diode or transistor in a standard IC process.
• The I-V characteristic of a p-n diode is nonlinear
• IS is the saturation current, typically 1 nA and λ is an empirical scaling
factor, λ = 0.5 for an ideal diode.
• Diode voltage
• When the diode is operated in a constant
current I0 circuit the forward diode voltage
Vout is directly proportional to the absolute
temperature and the voltage sensitivity ST is a constant depending on
the drive current:
• Overall temperature sensitivity of the diode depends on the relative
size of the drive current and saturation.
• When the drive current is set to a value well above the saturation
current,
• For a forward current of 0.1 μA and diode saturation current of
approximately 1 nA; the expected temperature sensitivity is
+0.2 mV/K.
• In practice, the temperature-dependence of a diode depends on
the strong temperature-dependence of the saturation current
itself.
• The actual value obtained experimentally from the temperature-
dependence of the forward junction voltage of a silicon diode is,
-2 mV/°C, and therefore Vf0 α T.
• An n-p-n transistor in a common-emitter configuration and
constant current circuit can be used as a temperature sensor.
• The base-emitter voltage VBE is proportional to the absolute
temperature and simply related to the collector current Ic by

• AE is the area of the emitter, Js is the saturation current density,

and IC0 is the reverse saturation current.
• It is possible to integrate resistive temperature sensors such as the
platinum Pt-100.
• But, the deposition of platinum or the thermistor oxide is a nonstandard IC
process and therefore requires additional pre- or post-IC processing steps.
• The inclusion of nonstandard materials during a CMOS process, which is
'intermediate' CMOS, is generally regarded as highly undesirable and
should be avoided if possible
• It is possible to fabricate silicon resistors in standard silicon IC process
• The resistivity of a single crystal of silicon varies with temperature and
doping level, and the lightly doped silicon provided the highest TCR.
• In practice, it is difficult to make single-crystal silicon with an impurity level
below ~1012 cm-3;
• Hence, it will not behave as an intrinsic semiconductor with a well-defined
temperature dependence because the intrinsic carrier concentration is about
1010 cm-3 at room temperature.
• In highly doped silicon resistors (~1018 cm-3), the temperature-dependence
approximates well to the second-order polynomial
• The temperature-dependence of a silicon resistor is nonlinear and
depends upon the exact doping level,
making it less suitable for use as a
temperature sensor
• Therefore, the preferred approaches
are to make a microthermocouple
out of Si or use the inherent temp.
sensitivity of a silicon diode or
transistor.
SAW Temperature Sensor
Surface Acoustic Wave
• Useful for the remote wireless
sensing of temperature –
perhaps on a rotating part
• SAW sensor consists of a thin lithium
• Niobate piezoelectric layer on top of a
• ceramic, glass, or silicon substrate
• A thin aluminum film is patterned using
• optical lithography to form a pair of interdigital electrodes connected to a
small microwave antenna and a pair of reflectors.
• A frequency modulated (FM) electromagnetic signal is transmitted remotely
and is picked up by the small antenna, which then drives the SAW via the
interdigital electrodes down to the reflectors.
• The reflectors return the wave that then drives the FM antenna and sends
back two signals to the remote location.
• The time-delays of the two signals are measured to obtain the phase
differences from the reference signal
• Phase angle is linearly related to the temperature by the two time delays τ1
and τ2
• FM signal has a frequency (2πωo) equal to 905 MHz, the time-delays of the
reflected signals are 1 μs and 1.1 μs at room temperature, and the
temperature coefficient of lithium niobate is 9.4 x 10-5/°C.
• The temperature sensitivity of the SAW IDT microsensor is calculated to be
3.1 degrees/°C.
• The resolution of the sensor is about 1 degree of angle or 0.33 °C.
• This temperature microtransducer has great practical value:
 sensing part is passive, that is, it requires
no power supply;
 it is wireless and therefore can be
embedded in rotating or moving parts,
such as car tyres, turbine blades,
helicopter rotors,
 it is very inexpensive to make.
RADIATION SENSORS
• Radiation microsensor detect electromagnetic radiation with
energies or wavelengths from the ultraviolet-to-near-infrared
(UV-NIR) region, which includes visible, through the NIR and
thermal-infrared region and into the microwave and radio
regions.

• The most important regions are the visible light region and the
NIR region because these are the wavelengths at which signals
are transmitted down fibre-optic cables in modern
telecommunication systems.

• Radiation microsensors can be distinguished by their underlying

operating principle, namely, photoconductive, photovoltaic (or
Photoelectric), pyroelectric, and microantenna.
Photoconductive Devices
• The radiation excites a number of electrons from the valence band of a
semiconductor material into its conduction band
• This creates both electrons and holes that can contribute to the conduction
process.
• Photoconductive effect is the dominant process when the energy of the
radiation is above the band gap energy of the semiconducting material.
• If the radiation produces Nt carriers per second in a slab of material of length
/, width w, and depth d, its change in electrical conductivity Δσ and change
in electrical conductance G is

• μn and μp :mobilities of the electrons &

holes and τn and τp are their lifetimes.
• The conductance I/Vo can be measured
in a constant voltage Vo circuit with the
cell resistance falling almost linearly with
illuminance from megaohms in the dark
to a few ohms.
• The response of a photoconductive cell to radiation is
determined by the choice of semiconducting material.
• Cadmium sulfide is commonly used to make a photoconductive
cell for the visible region (0.4 to 0.7 u.m) because it is
inexpensive and easy to process,
• Other materials used for IR photoconductors are, PbS with a
peak response at 2.2 μm, PbSe with a peak response at 2.2
μm, and HgCdTe (MCT) with a response tailored within the
range of 12 to 16 μm.
• Photoconductive cells are commercially available at low cost
and are commonly employed in a wide range of applications;
• light-activated switches for night lights, dimmers, and children's
toys.
Common semiconducting materials used in radiation microsensors and their
dynamic range within the UV-to-IR spectrum
Photovoltaic Devices
• Photovoltaic effect: the radiation induces a voltage across a
semiconductor junction;
• The materials most commonly used to make photovoltaic
sensors are Si for the visible/NIR region, and Ge, InGaAs, InAs,
or InSb for the NIR-to-IR region.
• Advantage of a photovoltaic cell over a photoconductive cell is
that it is compatible with a bipolar process (e.g. Si or GaAs).
• The reduced size and integrated electronics lead to a higher
sensitivity, faster response time (μs instead of ms), and better
stability.
• Photosensitive diodes, known as photodiodes, can be made in
a standard vertical bipolar process such as a p-n diode, or
variations on this process, such as a p-Si/insulator/n-Si PIN
diode, Schottky-type diode, and silicon avalanche diode
• Photodiodes: the photon creates an electron-hole pair in the space-charge
region of the junction.
• These charges are then separated by the local field to the different doped
regions, and they modify the diode voltage Vd.
• The diode voltage is the open circuit voltage Voc and can be measured by
reverse-biasing the diode and finding the voltage dropped across a high
external load resistor RL.

• IR is the photocurrent and is proportional

to the intensity of incident radiation
• Is is the reverse saturation current.
Pyroelectric Devices
• The radiation heats up the surface of a pyroelectric crystal
(usuallyLiTaO3) and reduces the normal polarisation state of
the crystal, thereby inducing the charge to flow off its surface
and creating a voltage.
• Pyroelectric sensors are not very sensitive to the wavelength of
the radiation and cover a wide dynamic range in the IR
spectrum
• Their dynamic response depends on the thermal time constant
of the device and the electrical time constant of the associated
circuit.
• Pyroelectric sensors are best-suited for detecting short-term
changes in the IR signal,
• detect the slow movement of a human body in a burglar alarm
system.
• The use of a band-pass filter removes sensor drift caused by
changes in the ambient conditions.
Microantenna
• Microwave or short-wavelength radio waves can be detected using a
small metal strip patterned using UV lithography onto a planar surface.
• The miniature antenna can detect low-energy microwave signals with a
suitable design of the loop.
• The signal can then be used to generate a SAW in a piezoelectric
material for a wireless mechanical sensor or to sense the electrical
signal and pass it onto a decoder
• As the microstrip can be made of aluminum, it is compatible with
standard microtechnology and can be deposited along with the
aluminum interconnects.
• The microantenna can also be used as a transmitter, in which case it is
acting as a radiation microactuator.
• The integrated microwave antenna may prove to be a very useful tool in
which a human operator can communicate with and remotely control a
small MEMS structure implanted in some inaccessible environment,
such as inside the human body
MECHANICAL SENSORS
• The most important class of
microsensor because of both the
large variety of different
mechanical measurands and their
successful application in mass
markets, such as the automotive
industry
• All mechanical microsensors to
date is a subset of only six
• Acceleration/deceleration
• Displacement
• Flow rate
• Force/torque
• Position/angle
• Pressure/stress
• Four of the most important types of mechanical microsensors
are:
• Pressure microsensors
• Microaccelerometers
• Microgyroscopes
• Flow microsensors
Flow Microsensors
• The measurement of the flow rate of a gas (or liquid) is important in
automotive, aerospace and chemical industries.
• The concept of a thermal flow sensor was first postulated by Thomas in
1911.
• The heat transferred per unit time (Ph) from a resistive wire heater to a
moving liquid is monitored at two points via thermocouple temperature
sensors.
• At steady state, the mass flow rate Qm is related to the difference in the
temperatures (T2 — T1)
• cm is the specific heat capacity of the fluid, assuming that there is no heat
loss from the wall of the tube
• The placing of the heating coil and temperature sensors within the walling of
a pipe makes more practical sense and the embodiment is the so-called
boundary-layer
flow meter.
• Johnson and Higashi reported in 1987 on the use of bulk micromachining to
make a pair of silicon microbridges.
• There is a thermistor on top of each rnicrobridge and a resistive heater is
split between the two
• The silicon rnicrobridge structure is advantageous because it is thermally
isolated and, therefore, the power loss of the heater is minimised.
• 1 mW of power produces a 15 °C rise in temperature in a gas.
• The temperature difference is measured in a Wheatstone bridge
• The device can be used to measure flow velocities of up to 30 m/s (approx
2.25 mg/min for air) in a 5 μm by 250 μm channel.
Micro Mechanical Electric Switch
• Electric switches (or relays) are very important devices in
communication and control systems.
• The state-of-the-art technology uses solid-state switches such
as GaAs MESFETs and PFN diodes.
• However, these solid-state switches have a large insertion loss
in the "on" state, and a poor electrical isolation in the "off” state
when the signal frequency is in GHz level.
• As micro mechanical switches make use of mechanical contact
and separation for the "on" and "off states, its performance will
be much better than solid-state switches.
• The switch uses a suspended silicon dioxide cantilever beam
as a moveable arm.
• When there is no voltage applied between the top and bottom
electrodes, the signal line-in and the signal line-out are not
connected electrically.
• The switch is called in an "off state.
• When an electric voltage is applied between the top and bottom
electrodes, the electrostatic force applied on the top electrode
pulls the cantilever down so that the contact metal bridges the
signal line-in and the signal line-out electrically.
• Thus, the switch is turned to an "on" state.
• For an "off state, electrical isolation of the switch depends on
the capacitive coupling between the two signal lines.
• As the coupling between the two signal lines is mainly through
the substrate, semi-insulating GaAs or ceramic substrate is
often chosen for RF switch instead of silicon substrate.
Micro Mechanical Optical Switch With Torsion Bar Actuator
• The incident light beam from an optic fiber is reflected towards position 1 by
the metal-coated refractive plate supported by two torsion bars when the
plate is at its original, horizontal position.
• This corresponds to an "off state.
• When a driving voltage V is applied between the refractive plate and the
metal electrode on the substrate under one side of the plate, the plate will
turn and hit the substrate at its one end.
• Then the incident light beam will be conducted towards position 2,
corresponding to an "on" state of the optical switch.
Micro Mechanical Optical Switch With Comb Drive Actuator
• In an original state (or, an "off state), the incident light beam is reflected
towards position 1by the vertical micro mirror attached to the movable
electrode.
• When a driving voltage is applied between the stationary electrode and the
movable electrode, the movable electrode and the mirror move away from
the light beam by the operation of the come drive actuator, and the light
beam goes to the position 2.
• The switch is thus turned "on".
• The size of the mirror has to be large enough to completely cut
off the beam light in the "off” state
• The travel distance of the mirror should be large enough to let
the light beam pass in an "on" state,
• Hence, the structure is usually processed using one of the high
aspect ratio technologies, such as the deep reactive ion etching
technology.
• For applications where switching speed is important, the natural
vibration frequency of the structure, the damping effect and the
magnitude of the electrostatic force play important roles.
Micro Mechanical Motors
1. Electrostatic Step Motor
• The rotor of the motor has 4 vanes and measures about 100μm in diameter.
• The stator around the rotor consists of 12 stationary electrodes.
• The motor is composed of mechanical parts in several polysilicon levels
using a surface micro machining technology.
• Each polysilicon layer is about 2 μm thick and the gaps between the
polysilicon layers are 1-2 μm.
• The gaps are controlled by the thickness of the sacrificial layers (e.g.,
phospho-silicate glass -PSG) temporarily filling the gaps in the process and
later removed by selective etch when the rotor is to be released.
• The rotor of the motor is driven into rotational movement by the electric force
between the stationary electrodes and the vanes.
• To do so, the rotor is grounded and the electric voltage is applied on some
selected stationary electrodes.
• For an original rotor position as shown in Fig, the voltage is applied on
electrodes 1, 4, 7 and 10 so that the electrostatic forces on the vanes
unanimously drive the rotor to rotate counter-clockwisely.
• Once the vanes of the rotor have moved to the position in alignment with the
electrodes 1,4,7 and 10, the voltage is switched to the next electrodes, i.e.
the electrodes 2, 5, 8 and 11.
• Thus, the electrostatic forces combined with the momentum of rotation, drive
the rotor to move on.
• The motor keeps running if the above step is
repeated continuously.
• The main difficulties with step motors are the small torque and relatively
large friction and stiction.
• The torque T on the rotor, T = nfr,
• where n is the vane number of the rotor, r the radius of the rotor and f the
tangential force on each vane.
• When the width of the vane is much larger than the gap distance d between
the vanes and the stators, the tangential force on each vane is
• t is the thickness of the rotor, εo the permitivity of free space, V the driving
voltage applied between the rotor and the stators.
• For a specific fabrication process and a given driving voltage, the only way
to increase the torque is to increase the vane number of the rotor.
• But the vane number n can hardly be larger than 10 for conventional
designs
• Stiction: two micro mechanical parts tend to stick together once they are
brought into contact and can only be separated by a relatively large force.
• Friction coefficient for micro mechanical parts is larger than that for
conventional mechanical parts
• Hence a voltage larger than expected has to be applied to keep the motor
running
2. Comb Drive Vibratory Motor
• Motor:"a machine that supplies power or cause motion" (The
Advanced Learner's Dictionary of Current English, 2nd edition,
Oxford University Press, Amen House, London).
• A motor is not necessarily a rotational one
• As each side of a movable finger of the comb drive structure
provides a tangential force as a vane does in a rotational step
motor and the finger number in a comb drive structure can be
increased without limit in principle, the electrostatic driving force
of a comb drive motor can be increased greatly.
• The movable parts are suspended over the substrate without a
direct contact with the stationary fingers or substrate which
greatly alleviated stiction and friction
• Hence, the comb drive micro motor is a more reasonable
design than a rotational micro motor for micro mechanical
world.
• As the comb structure can provide "back-and-forth" movement,
it is a vibratory motor.
• The "back-and-forth" movement can
also be transformed into a rotational
motion using a transformation
mechanism with cranks, pin joints,
wheels and gears
• Two comb drive micro motors are used
to provide two orthogonal forces: one
for driving in the x-direction and the
other for the j-direction.
• With some control on the phase
difference between the two motors, the
driving forces in the two directions can
be incorporated to drive the wheel into
a rotational movement.
• In this case, stiction and friction appear
again, but the problem is now
alleviated by the increased driving
force by using a large number of
fingers.
MICRO PUMPS
Mechanical Micropumps
• Mechanical actuation usually involves coupling mechanical deformation of a
moving boundary to an increase or decrease in the fluid pressure to drive
flow.
• Actuators that perform this function may be classified as either external or
integrated actuators
• External Actuators require either external components or coupling to the
micropump.
• As these actuators tend to be larger than the microfluidic components, they
increase the total size of the micropump.
• However, these actuators have the advantage of being capable of producing
large forces and stroke displacement.
• External Actuators: Electromagnetic Actuation, Piezoelectric Actuation,
Pneumatic Actuation, Shape Memory Alloy
• Integrated actuators are micromachined as a pump component during the
microfabrication process.
• As they are produced during the fabrication process, integrated actuators
usually exhibit a fast response time and better force coupling to the pump.
• Integrated actuators: Electrostatic Actuation, Thermopneumatic Actuation,
Bimetallic Thermal Actuation
Electromagnetic Actuation
• It works in a similar manner to a
solenoid valve
• An energized solenoid coil
produces a magnetic field, which
produces a force on a
ferromagnetic plunger.
• The plunger is coupled to a
membrane so that when the coil is
de-energized the plunger returns
to its original position.
• This is a fairly large structure due
to the size of the solenoid coil.
• However, large strokes can be
achieved
• It may be tailored by the coil
current and number of wire turns
within the coil.
Piezoelectric Actuation
• In a piezoelectric material such as lead zirconate titanate (PZT),an applied
voltage produces a strain and mechanical deformation.
• The piezoelectric element may be a piezoelectric disk which is glued directly
onto the pump membrane.
• A disk-style actuator has a fast response time and large stroke displacement
but the pump pressure is relatively small.
• Greater pressure may be obtained by using a piezoelectric stack actuator
Pneumatic Actuation
• It uses a high-pressure gas source that is coupled to the pump membrane
• Within a pneumatic chamber the gas pressure is controlled through the use
of two solenoid valves, which may either open to pressurize or close to
depressurize the pump chamber.
• Advantage : very large pressures may be generated.
• Disadvantages: the requirement of a pressurized gas source, and the
response time is limited by the speed at which the solenoid valves open and
close.
Shape Memory Alloy
• Shape memory alloys (SMA) may be used as actuation sources by coupling
the SMA with a bias spring
• The SMA is initially shaped and memorized into an expanded coil.
• The coil is compressed when placed into the actuator.
• When the SMA temperature is raised to the critical memory temperature by
running a current through the coil, the coil expands and produces a force on
the actuation membrane.
• As the coil cools, a bias spring produces an opposite force, which
compresses the coil back to its initial state.
• By cycling coil expansion and compression a pump stroke is obtained.
Electrostatic Actuation
• It is based on capacitive charging of two parallel electrodes; a fixed
electrode and a movable electrode in an electric field
• As the electrodes charge, a force pulls on the moveable electrode to deform
the pumping membrane.
• This actuation method generally has very low power consumption, but the
electrode displacement is usually very small because with constant applied
voltage, the force is inversely proportional to the square of the electrode
spacing.
Thermopneumatic Actuation
• It is based on gas expansion within a sealed actuation chamber to deform a
pump membrane.
• Thermopneumatic actuation may be produced:
• by gas expansion
• by liquid-vapor phase change (boiling) due to resistive heating
• by direct electrolysis of water to produce oxygen and hydrogen gas
bubbles.
• Thermal bubble generation may also be used directly within a microchannel
as bubble expansion acts as a piston to displace fluid.
• It produces very large forces and stroke volumes.
• But, resistive heating requires a large amount of energy.
Bimetallic Thermal Actuation
• It is based on thermomechanical actuation due to differences in the thermal
expansion coefficient of two layers of material
• When the structure is heated by resistive power, dissipation in each material
causes expansion and the difference in expansion in the two materials
causes deformation.
• The actuator displacement depends, in an approximately linear manner, on
the temperature change, which may be well controlled by the power
dissipated within the structure.
Electroosmotic Flow Micropumps
• It is a subset of electrokinetic phenomena related to the movement of
electric charges in an applied electric field.
• Charge movement, in turn, produces a shear on the surrounding fluid to
produce flow
• Electroosmotic pumps are used in small channels without a need for high
pressures
• Most solid surfaces develop a spontaneous electrical double layer (EDL)
when immersed in an electrolyte solution
• When an electrolyte is brought into contact with a solid surface, the surface
develops an electric charge.
• Countercharges from the electrolyte associate with these surface charges to
form a strongly associated charge layer known as the Stern layer.
• However, the Stern layer does not completely shield the surface charge and
so a more weakly associated diffuse layer of countercharges forms adjacent
to the Stern layer.
• The Stern layer is relatively immobile while the charges in the diffuse layer
are free to move.
• These two features are know as the electrical double layer, which is a local
region close to the solid surface where charge neutrality is not maintained.
• When a microchannel is produced
from an insulating material such as
glass or plastic, the surface becomes
negatively charged when filled with an
aqueous solution.
• As the microchannel is insulating,
a tangential electric field may be
applied down the length of the channel
to produce an electroosmotic flow.
• In the presence of the electric field, positively charged ions will move toward
the negatively charged cathode, and the negatively charge ions will move
toward the positively charged anode.
• In the bulk region of the microchannel, charge neutrality is maintained so
there is no net charge movement.
• However, in the electric double layer, the diffuse layer will move within the
electric field.
• These ions in turn will drag fluid with them
MICRO MIXERS
Introduction
• Mixing is of significant importance to realizing lab-on-a-chip
microscale reactors and bioanalysis systems
• the reactions carried out on the micro- or even nanoscale in
such devices require the on-chip mixing of samples and
reagents.
• Require modestly fast mixing in batch mode, should the mixing
be performed on-chip.
• The bottleneck lies with the thermal cycling reactions where a
specific volume of mixture is usually needed to be produced
• The method of mixing should consume low levels of the
resource used to drive the chip,
• i.e. low pressure if the device is pressure driven or low voltage
if the device is electrokinetically driven.
• This is critical in process where several levels of on-chip mixing
are necessary
• In continuous flow fast reaction kinetics experiments mixing
must be faster than reaction rates with timescales down to
microseconds.
• The emphasis is on localized mixing with very small timescales
• Low levels of the driving force to operate the chip are desired,
but can be sacrificed in exchange for speed, so long as the chip
is not structurally compromised.
• Another requirement of microscale mixers could be demand for
longer-term continuous production
• In this case the emphasis would be on maximizing the mixture
flow rate delivered, while keeping the driving force for delivery
at low levels.
• An application is in situ, designer drug delivery, whereby a
particular drug cocktail would be synthesized on-chip from
constituents at proportions specific to a particular patient, and
directly delivered at the point of care.
Basic Considerations
• Microscale mixing is driven by diffusion, and diffusion processes are
inherently slow
• On a macroscale, mixing can be made more effective through turbulent
transport, which reduces transport timescales by orders of magnitude
relative to diffusive ones,
• To generate turbulent flow (e.g., Re = 5000) in a microchannel 100 μm by
100 μm in cross-section, one requires a velocity of approximately 48m/s.
• Then the pressure drop in the channel is more than 4 atmospheres per
millimeter of length, which is prohibitive.
• In addition to the requirement of rapid and effective mixing, one has to be
vigilant with respect to the required pressure to drive the microfluidic chip.
• High pressure pumping impose higher loads on microfluidic chip
components and make it more prone to leakage if not breakage and/or
debonding of bonded surfaces.
• Increasing the microchannel cross-section alleviates the high pressure
requirement, but increases the volume of the device.
• In the absence of turbulent transport, the only recourse to
achieving effective mixing is the reduction of the molecular
diffusion length.
• the diffusive time, tD = δD2/D12,
• When the mass diffusion coefficient is very small (<O [10–10
m2/s]) the only way to cut down on the mixing time is to reduce
this diffusion length.
• lamination micromixers pursue the creation of several
alternating narrow layers of the compounds to be mixed, so as
to cut down on the diffusion length;
• micromixers based on chaotic advection (chaotic stirring)
achieve it by kinematically folding the interfaces between the
compounds multiple times
Active and Passive Micromixers
• Active mixers may effectively provide rapid mixing,
• But the additional mechanical and electronic devices, both on-
and off chip, often add undesirable complexity
• The additional devices need extra energy, space, and if on-
chip, may also be difficult to fabricate and integrate to form a
cost-effective and compact lab-on-a-chip.
• The electrical fields and heat sometimes generated by active
control may damage biological samples.
• The passive mixers are easier to fabricate and simpler in
design than active mixers.
• This is more so for pressure-driven devices than electrically
driven ones.
Passive Micromixer

• No external energy source is required as an input for

enhancing the mixing mechanism.

1. Pressure driven

2. Electrically driven
Pressure Driven
• Schwesinger designed micromixers for emulsifying immiscible
fluids
• A two-layer microchip mixer in silicon successfully produced
emulsions by successively layering the fluids in microchannels
of a triangular cross-section and typical size of 150 to 200 μm
• This was achieved by splitting and recombining the mixing
channels several times on two levels, so as to proliferate
alternating layers of the different fluids in the channel cross-
section.
• The layering succeeds in reducing the diffusion length by a
factor of 2n if the original stream is split and recombined ‘n’
times.
• The main complexity in the micromanufacturing process, is the
need for alignment of the two chip layers.
• Knight et al. demonstrated very rapid local mixing on a silicon
chip
• They used a single 10 μm–wide mixing channel of unitary
aspect ratio fed by three tributaries in a cross (Ψ) configuration.
• The middle one was shaped as a nozzle terminating in a 5:1
aspect ratio exit, with the intent of further facilitating the rapid
formation of a narrower middle diffusion layer.
• It generated very rapid (O[μs]) local mixing of very small
volumes (O[nL]) at supply pressures as high as 3 atmospheres.
• Bends in channels generate secondary flows at modestly high,
to high Reynolds numbers in the laminar regime.
• This has been employed by several investigators, and mixers
with bends have been used in integrated chips.
• Arrays of modest-aspect-ratio (2) single-level, serpentine
(zigzag) microchannels combined with simple Ψ-junction mixers
were introduced and used by Kamidate et al.
• The potential of attaining improved mixing by placing obstacles
in microchannels has also been explored.
• At very low Reynolds numbers obstacles cannot generate
secondary flows such as recirculations
• It is possible to achieve this at higher Reynolds numbers.
• Swirl chambers have also been employed on the microscale to
achieve effective mixing
• It was a very effective mixing device, but required very high
pressures (15 atmospheres) and a costly manufacturing
process.
Electrically Driven Passive Micromixers
• Application in microfluidic devices that handle biological samples and
perform biological assays.
• He et al. realized a design combining the effects of electroosmotic flow,
convective effects due to bends, and flow-stream splitting and
recombination on a quartz wafer
• Voltages on the order of Kilo Volts were applied to produce velocities of
order 0.3 mm/s
• Yager et al proposed a microscale mixing channel with electrodes (A and B)
at each end
• When a potential difference is applied between electrodes A and B, the fluid
will be set into motion near the wall region with a velocity in proportion to the
electroosmotic mobility of the wall material
• Because the ends of the mixing channel (between A and B) cannot be
penetrated by the flow, conservation of mass will dictate a flow inside the
core of the cross-section in the opposite direction to that near the walls.
• Thus a recirculating flow will be established within the mixing channel with
shear layers in the proximity of the walls.
Active Micromixers
• A broad variety of excitation forces have been used, including pressure
pulsations, electrical, magnetic, capillary, thermal, and acoustic.
• Evans et al. introduced pulsed-flow/pressure micromixer
• It was a chamber design realized on silicon with a quartz cover.
• The actuation was provided by thermally driven bubble valves/pumps
powered by integrated polysilicon heaters.
• It was complex and had a sizeable dead volume,
• This device is the first to incorporate the idea of improved mixing through
chaotic-advection-inducing flow pulsations
• A swirl-chamber mixer micromilled in PMMA was proposed by Chung et al.
• The swirling of the fluids was achieved by forward and backward pumping.
• The mixing chamber was fitted with two opposing channels of unit aspect
ratio tangent to the circular chamber.
• The pulsed flow and pressure micromixers are continuous-flow devices and
have been shown to improve mixing compared to their steady-state
counterparts leading to shorter mixing-channel lengths for Reynolds
numbers of order one or less.
• Lee et al made micromixers utilizing unsteady electrical fields
• They demonstrated a pressure driven, continuous-flow device
with periodic electrical excitation introduced in a chamber on
the flow path.
• They took advantage of dielectophoretic forces induced by the
inhomogeneous electrical field to improve mixing of dielectric
microparticles.
• Magnetohydrodynamic (MHD) forcing to improve mixing on the
microscale was demonstrated by Solomon et al.
• They performed experiments comparing long-range chaotic
mixing of miscible and immiscible impurities in a time-periodic
flow by producing an alternating magnetic field that generated
alternating vortex structures due to MHD instability.