UNIT-I

INTRODUCTION TO MEMS (MICRO ELECTRO MECHANICAL SYSTEMS)

1. INTRODUCTION

MEMS = silicon-based microelectronics + micromachining technology.

1. MEMS is a process technology used to create tiny integrated devices or systems that combine
Mechanical and electrical components. They are fabricated using integrated circuit (IC) batch.Processing
techniques and can range in size from a few micrometers to millimetres. These Devices (or systems) have
the ability to sense, control and actuate on the micro scale, and generate effects on the macro scale.
2. The interdisciplinary nature of MEMS utilizes design, engineering and manufacturing expertise from a
wide and diverse range of technical areas including integrated circuit fabrication technology, mechanical
engineering, materials science, electrical engineering, chemistry and chemical engineering, as well as
fluid engineering, optics, instrumentation and packaging.
3.MEMS can be found in systems ranging across automotive, medical, electronic, Communication and
defence applications. Current MEMS devices include accelerometers for airbag sensors, inkjet printer
heads, computer disk drive read/write heads, projection display chips, blood pressure sensors, optical
switches,microvalves, biosensors.
4. MEMS has been identified as one of the most promising technologies for the 21st Century and has the
potential to revolutionize both industrial and consumer products by combining silicon based
microelectronics with micromachining technology. Its techniques and microsystem based devices have
the potential to dramatically affect of all of our lives and the way we live. If semiconductor
microfabrication was seen to be the first micro manufacturing revolution.
MEMS is the second revolution.

2.HISTORY

The history of MEMS is useful to illustrate its diversity, challenges and applications. The
following list summarizes some of the key MEMS milestones

1950’s
1958:- Silicon strain gauges commercially available
1959 :-“There’s Plenty of Room at the Bottom” – Richard Feynman gives a milestone
presentation at California Institute of Technology. He issues a public challenge by
offering $1000 to the first person to create an electrical motor smaller than 1/64th
of an inch.

1960’s
1961 First silicon pressure sensor demonstrated
1967 Invention of surface micromachining. Westinghouse creates the Resonant Gate
Field Effect Transistor, (RGT). Description of use of sacrificial material to free
micromechanical devices from the silicon substrate.
1970’s
1970 First silicon accelerometer demonstrated
1979 First micromachined inkjet nozzle
1980’s
Early 1980’s: first experiments in surface micromachined silicon. Late 1980’s:
micromachining leverages microelectronics industry and widespread
experimentation and documentation increases public interest.
1982 Disposable blood pressure transducer
1982 “Silicon as a Mechanical Material” [9]. Instrumental paper to entice the scientific
community – reference for material properties and etching data for silicon.
1982 LIGA Process
1988 First MEMS conference
1990’s
Methods of micromachining aimed towards improving sensors.
1992 MCNC starts the Multi-User MEMS Process (MUMPS) sponsored by Defense
Advanced Research Projects Agency (DARPA)
1992 First micromachined hinge
1993 First surface micromachined accelerometer sold (Analog Devices, ADXL50)
1994 Deep Reactive Ion Etching is patented
1995 BioMEMS rapidly develops
2000 MEMS optical-networking components become big business

3. CONCEPTS OF MEMS

What is MEMS?
1. Micro-electromechanical systems (MEMS) is a process technology used to create tiny integrated
devices or systems that combine mechanical and electrical components.
2. They are fabricated using integrated circuit (IC) batch processing techniques and can range in size from
a few micrometers to millimeters. These devices (or systems) have the ability to sense, control and
actuate on the micro scale, and generate effects on the macro scale.
3. MEMS, an acronym that originated in the United States, is also referred to as Microsystems
Technology (MST) in Europe and Micromachines in Japan.
Regardless of terminology, the uniting factor of a MEMS device is in the way it is made. While
the device electronics are fabricated using ‘computer chip’ IC technology, the micromechanical
components are fabricated by sophisticated manipulations of silicon and other substrates using
micromachining processes.
4. Processes such as bulk and surface micromachining, as well as high-aspect-ratio micromachining
(HARM) selectively remove parts of the silicon or add additional structural layers to form the mechanical
and electromechanical components.
5.While integrated circuits are designed to exploit the electrical properties of silicon, MEMS takes
advantage of either silicon’s mechanical properties or both its electrical and mechanical properties
4.PRINCIPLES

1.In the most general form, MEMS consist of mechanical microstructures, microsensors, microactuators
and microelectronics, all integrated onto the same silicon chip. This is shown schematically in Figure 1.

Figure 1. Schematic illustration of MEMS components.

2.MEMS devices are very small; their components are usually microscopic. Levers, gears, pistons, as
well as motors and even steam engines have all been fabricated by MEMS

Figure 2. (a) A MEMS silicon motor together with a strand of human hair [1], and (b) the legs of a
spider mite standing on gears from a micro-engine

3. The designer applies the classical Lagrangian and Newtonian mechanics as well as electromagnetics
(Maxwell’s equations) to study conventional electromechanical systems and MEMS.
Microsensors detect changes in the system’s environment by measuring mechanical, thermal, magnetic,
chemical or electromagnetic information or phenomena. Microelectronics process this information and
signal the microactuators to react and create some form of changes to the Environment.
4.However, MEMS is not just about the miniaturization of mechanical components or making things out
of silicon. MEMS is a manufacturing technology; a paradigm for designing and creating complex
mechanical devices and systems as well as their integrated electronics using batch fabrication techniques.
MEMS is also referred to as MST, strictly speaking, MEMS is a process technology used to create tiny
mechanical devices or systems, and as a result, it is a subset of MST.

Figure 3. Classifications of microsystems technology

Micro-optoelectromechanical systems (MOEMS) is also a subset of MST and together with MEMS forms
the specialized technology fields using miniaturized combinations of optics, electronics and mechanics.
Both their microsystems incorporate the use of microelectronics batch processing techniques for their
design and fabrication. There are considerable overlaps between fields in terms of their integrating
technology and their applications and hence it is extremely difficult to categorise MEMS devices in terms
of sensing domain and/or their subset of MST. The real difference between MEMS and MST is that
MEMS tends to use semiconductor processes to create a mechanical part. In contrast, the deposition of a
material on silicon for example, does not constitute MEMS but is an application of MST.
5. APPLICATION AND DESIGN

MEMS has several distinct advantages as a manufacturing technology.

1. In the first place, the interdisciplinary nature of MEMS technology and its micromachining techniques,
as well as its diversity of applications has resulted in an unprecedented range of devices and synergies
across previously unrelated fields (for example biology and microelectronics).
2. Secondly, MEMS with its batch fabrication techniques enables components and devices to be
manufactured with increased performance and reliability, combined with the obvious advantages of
reduced physical size, volume, weight and cost.
3. Thirdly, MEMS provides the basis for the manufacture of products that cannot be made by other
methods. These factors make MEMS potentially a far more pervasive technology than integrated circuit
microchips.
APPLICATION
Today, high volume MEMS can be found in a diversity of applications across multiple markets.

Automotive Electronics Medical Communications Defence

Internal Disk drive Blood pressure Fibre-optic Munitions
navigation heads sensor network guidance
sensors components
Air conditioning Inkjet printer Muscle RF Relays, Surveillance
compressor heads stimulators & switches and
sensor drug filters
delivery systems
Brake force Projection Implanted Projection Arming systems
sensors & screen pressure sensors displays in
suspension televisions portable
control communications
accelerometers devices and
instrumentation
Fuel level and Earthquake Prosthetics Voltage controlled Embedded
vapour pressure sensors oscillators (VCOs) sensors
sensors
Airbag sensors Avionics Miniature Splitters and Data storage
pressure analytical couplers
sensors instruments
"Intelligent" Mass data Pacemakers Tuneable lasers Aircraft control
tyres storage systems

i) Automotive airbag sensor

Automotive airbag sensors were one of the first commercial devices using MEMS. They are in
widespread use today in the form of a single chip containing a smart sensor, or accelerometer, which
measures the rapid deceleration of a vehicle on hitting an object. The deceleration is sensed by a change
in voltage. An electronic control unit subsequently sends a signal to trigger and explosively fill the airbag.
Initial air bag technology used conventional mechanical ‘ball and tube’ type devices which were
relatively complex, weighed several pounds and cost several hundred dollars. They were usually mounted
in the front of the vehicle with separate electronics near the airbag.
MEMS has enabled the same function to be accomplished by integrating an accelerometer and the
electronics into a single silicon chip, resulting in a tiny device that can be housed within the steering
wheel column and costs only a few dollars (Figures 4 and 5). The accelerometer is essentially a capacitive
or piezoresistive device consisting of a suspended pendulum proof mass/plate assembly. As acceleration
acts on the proof mass, micromachined capacitive or piezoresistive plates sense a change in acceleration
from deflection of the plates. The sense plates can be seen in Figure 4 .
 Figure 4. (a) The first commercial accelerometer from Analog Devices (1990); its size is less than 1
cm2 (left) [12], and (b)capacitive sense plates, 60 microns deep (right)

ii) Medical pressure sensor

Another example of an extremely successful MEMS application is the miniature disposable pressure
sensor used to monitor blood pressure in hospitals. These sensors connect to a patients intravenous (IV)
line and monitor the blood pressure through the IV solution. For a fraction of their cost ($10), they
replace the early external blood pressure sensors that cost over $600 and had to be sterilized and
recalibrated for reuse. These expensive devices measure blood pressure with a saline-filled tube and
diaphragm arrangement that has to be connected to an artery with a needle.

Figure 5. Schematic illustration of a piezoresistive pressure sensor.

The disposable sensor consists of a silicon substrate which is etched to produce a membrane and is
bonded to a substrate (Figure 6). A piezoresistive layer is applied on the membrane surface near the edges
to convert the mechanical stress into an electrical voltage. Pressure corresponds to deflection of the
membrane. The sensing element is mounted on a plastic or ceramic base with a plastic cap over it,
designed to fit into a manufacturer’s housing (Figure 7). A gel is used to separate the saline solution from
the sensing element.
New MEMS Applications
The experience gained from these early MEMS applications has made it an enabling technology for new
biomedical applications (often referred to as bioMEMS) and wireless communications comprised of both
optical, also referred to as micro-optoelectromechanical systems (MOEMS), and radio frequency (RF)
MEMS.
6.SCALING (PROPERTIES/ISSUES)

1. Scaling theory is a valuable guide to what may work and what will not. By understanding how
phenomena behave and change as their scale size changes, we can gain some insight and better
understand the profitable approaches.
2. Refer separate sheet attached

Micromachining Processes: MEMS fall into three general classifications

1.Bulk micromachining:- Bulk micromachining involves the removal of part of the bulk substrate. It is
a subtractive process that uses wet anisotropic etching or a dry etching method such as reactive ion
etching (RIE), to create large pits, grooves and channels. Materials typically used for wet etching include
silicon and quartz, while dry etching is typically used with silicon, metals, plastics and ceramics.
2. Surface micromachining:- Surface micromachining involves processing above the substrate,
mainly using it as a foundation layer on which to build. It was initiated in the 1980’s and is the newest
MEMS production technology. Material is added to the substrate in the form of layers of thin films on the
surface of the substrate (typically a silicon wafer). These layers can either by structural layers or act as
spacers, later to be removed, when they are known as sacrificial layers. Hence the process usually
involves films of two different materials: a structural material out of which the free standing structure is
made (generally polycrystalline silicon or polysilicon, silicon nitride and aluminium) and a sacrificial
material, deposited wherever either an open area or a free standing mechanical structure is required
(usually an oxide).
These layers (or thin films) are deposited and subsequently dry etched in sequence, with the sacrificial
material being finally wet etched away to release the final structure. Each additional layer is accompanied
by an increasing level of complexity and a resulting difficulty in fabrication. A typical surface
micromachined cantilever beam is shown in Figure 21. Here, a sacrificial layer of oxide is deposited on
the silicon substrate surface using a pattern and photolithography. A polysilicon layer is then deposited
and patterned using RIE processes to form a cantilever beam with an anchor pad. The wafer is then wet
etched to remove the oxide (sacrificial) layer releasing and leaving the beam on the substrate. More
complex MEMS structures can be achieved using structural polysilicon and sacrificial silicon dioxide,
including sliding structures, actuators and free moving mechanical gears. Figures 22 shows the process
flow for the fabrication of a micromotor by the commercially available Multi-User MEMS Process
(MUMPS).
The levels of complexity achievable with MEMS has already been shown in Figure 16. In this case, five
mechanical levels of micromachined polysilicon can be achieved using Sandia Ultra-Planar Multi-Level
Technology (SUMMiT).
The success of the surface micromachining process depends on the ability to successfully remove all of
the sacrificial layers to free the structural elements so that they can be actuated. This step is responsible
for curtailing the yield (percentage of the devices on a wafer that function properly) and reliability of
fabricated MEMS due to the phenomenon known as stiction. Stiction refers to the sticking of structural
elements either to the substrate or the adjacent elements. Capillary forces from rinsing liquids, as well as
electrostatic and van der Waals forces can also produce permanent adhesion after the system has dried.
Figure
3.High-aspect-ratio micromachining:- (HARM), which includes technology such as LIGA (a
German acronym from Lithographie, Galvanoformung, Abformung translated as lithography,
electroforming and moulding).
High-aspect-ratio micromachining (HARM) is a process that involves micromachining as a
tooling step followed by injection moulding or embossing and, if required, by electroforming to replicate
microstructures in metal from moulded parts. It is one of the most attractive technologies for replicating
microstructures at a high performance-to-cost ratio and includes techniques known as LIGA. Products
micromachined with this technique include highaspect- ratio fluidic structures such as moulded nozzle
plates for inkjet printing and microchannel plates for disposable microtitreplates in medical diagnostic
applications. The materials that can be used are electroformable metals and plastics, including acrylate,
polycarbonate, polyimide and styrene.

Substrates:-

The most common substrate material for micromachining is silicon.

It has been successful in the microelectronics industry and will continue to be in areas of miniaturization
for several reasons:
i) silicon is abundant, inexpensive, and can be processed to unparalleled purity

ii) silicon’s ability to be deposited in thin films is very amenable to MEMS

iii) high definition and reproduction of silicon device shapes using photolithography are perfect for high
levels of MEMS precision

iv) silicon microelectronics circuits are batch fabricated (a silicon wafer contains hundreds of identical
chips not just one)

Other crystalline semiconductors including germanium (Ge) and gallium arsenide (GaAs) are
used as substrate materials due to similar inherent features, but silicon is distinguished from other
semiconductors in that it can be readily oxidized to form a chemically inert and electrically insulating
surface layer of SiO2 on exposure to steam.
The homogeneous crystal structure of silicon gives it the electrical properties needed in
microelectronic circuits, but in this form silicon also has desirable mechanical properties. Silicon forms
the same type of crystal structure as diamond, and although the interatomic bonds are much weaker, it is
harder than most metals. In addition, it is surprisingly resistant to mechanical stress, having a higher
elastic limit than steel in both tension and compression. Single crystal silicon also remains strong under
repeated cycles of tension and compression. The crystalline orientation of silicon is important in the
fabrication of MEMS devices because some of the). etchants used attack the crystal at different rates in
different directions (Figure 18

Figure . Low crystallographic index planes of silicon

Silicon is dominant as a substrate for MEMS but research and development is ongoing with other non-
semiconductor substrate materials including metals, glasses, quartz, crystalline insulators, ceramics and
polymers. The ability to integrate circuitry directly onto the substrate is currently the underlying issue
with today’s MEMS substrate materials; hence the success of silicon lithography,
Lithography

There are processes for creating material layers required in the fabrication of microsystems. The
successful development of microsystems involves successive steps of deposition and patterning of various
material layers. One of the key steps in patterning is the process of transferring a geometrical pattern on a
mask to a radiationsensitive material called a resist. This process is known as lithography .
 The geometrical pattern defines the area being protected or removed in a subsequent process step.
The required geometrical pattern of the mask is first prepared in digital form in a computer program.
Commands from the computer drive a pattern generator that photoengraves the particular pattern on an
optically flat glass or quartz plate covered with a thin chromium film. This plate containing the patterned
chromium layer for the whole wafer is known as a mask.
The opaque portion of the mask prevents the radiation from reaching the resist on the wafer. A
mask is used repetitively to expose selected regions of the resist on the wafer.
Since ultraviolet (UV) radiation (wavelength in the range of 350 to 450 nm) is most commonly used, the
process is known as photolithography. A mercury vapor lamp with radiation wavelength 300 to 400 nm is
a popular UV light source. As the resolution achievable is proportional to the wavelength of the incident
radiation, deep UV (<250 nm) and extreme UV (<100 nm) radiation are used to achieve higher resolution
than is possible with normal UV light.

WET/DRY ETCHING PROCESSES

Wet Etching
Wet etching describes the removal of material through the immersion of a material (typically a
silicon wafer) in a liquid bath of a chemical etchant. These etchants can be isotropic or anisotropic.
Isotropic etchants etch the material at the same rate in all directions, and consequently remove
material under the etch masks at the same rate as they etch through the material; this is known as
undercutting (Figure 19 a and b). The most common form of isotropic silicon etch is HNA, which
comprises a mixture of hydrofluoric acid (HF), nitric acid (HNO3) and acetic acid
(CH3COOH). Isotropic etchants are limited by the geometry of the structure to be etched. Etch rates can
slow down and in some cases (for example, in deep and narrow channels) they can stop due to diffusion
limiting factors. However, this effect can be minimized by agitation of the etchant, resulting in structures
with near perfect and rounded surfaces (Figure 19a)

Figure 19. Isotropic etching with (a) and without (b) agitation, and anisotropic wet etching of (100)
and (110) silicon

Anisotropic etchants etch faster in a preferred direction. Potassium hydroxide (KOH) is the most common
anisotropic etchant as it is relatively safe to use. Structures formed in the substrate are dependent on the
crystal orientation of the substrate or wafer. Most such anisotropic etchants progress rapidly in the crystal
direction perpendicular to the (110) plane and less rapidly in the direction perpendicular to the (100)
plane. The direction perpendicular to the (111) plane etches very slowly if at all. Figures 19c and 19d
shows examples of anisotropic etching in (100) and (110) silicon. Silicon wafers, originally cut from a
large ingot of silicon grown from single seed silicon, are cut according to the crystallographic plane. They
can be supplied in terms of the orientation of the surface plane.
Dopant levels within the substrate can affect the etch rate by KOH, and if levels are high enough, can
effectively stop it. Boron is one such dopant and is implanted into the silicon by a diffusion process. This
can be used to selectively etch regions in the silicon leaving doped areas unaffected.

Dry Etching

Dry etching relies on vapour phase or plasma-based methods of etching using suitably reactive gases or
vapours usually at high temperatures. The most common form for MEMS is reactive ion etching (RIE)
which utilizes additional energy in the form of radio frequency (RF) power to drive the chemical reaction.
Energetic ions are accelerated towards the material to be etched within a plasma phase supplying the
additional energy needed for the reaction; as a result the etching can occur at much lower temperatures
(typically 150º - 250ºC, sometimes room temperature) than those usually needed (above 1000ºC). RIE is
not limited by the crystal planes in the silicon, and as a result, deep trenches and pits, or arbitrary shapes
with vertical walls can be etched Deep Reactive Ion Etching (DRIE) is a much higher-aspect-ratio etching
method that involves an alternating process of high-density plasma etching (as in RIE) and protective
polymer deposition to achieve greater aspect ratios (Figure 20).

Figure . Deep Reactive Ion Etching (DRIE)

Etch rates depend on time, concentration, temperature and material to be etched. To date there are no
universally accepted master equations to predict etch performance and behaviour .

Deposition processes (Evaporation, sputtering, CVD, etc)

One of the basic building blocks in MEMS processing is the ability to deposit thin films of material.
In this text we assume a thin film to have a thickness anywhere between a few nanometer to about 100
micrometer. MEMS deposition technology can be classified in two groups:

1. Depositions that happen because of a chemical reaction:

o Chemical Vapour Deposition (CVD)
o Electrodeposition
o Epitaxy
o Thermal oxidation

These processes exploit the creation of solid materials directly from chemical reactions in gas
and/or liquid compositions or with the substrate material. The solid material is usually not the only
product formed by the reaction. Byproducts can include gases, liquids and even other solids.
2. Depositions that happen because of a physical reaction:
o Physical Vapor Deposition (PVD)
o Casting

Common for all these processes are that the material deposited is physically moved on to the
substrate. In other words, there is no chemical reaction which forms the material on the substrate.
This is not completely correct for casting processes, though it is more convenient to think of them
that way.

This is by no means an exhaustive list since technologies evolve continuously.

Chemical Vapour Deposition (CVD)

In this process, the substrate is placed inside a reactor to which a number of gases are supplied. The
fundamental principle of the process is that a chemical reaction takes place between the source gases.
The product of that reaction is a solid material with condenses on all surfaces inside the reactor.

The two most important CVD technologies in MEMS are

1. Low Pressure CVD (LPCVD)
The LPCVD process produces layers with excellent uniformity of thickness and material
characteristics. The main problems with the process are the high deposition temperature (higher than
600°C) and the relatively slow deposition rate.

2.  Plasma Enhanced CVD (PECVD).

The PECVD process can operate at lower temperatures (down to 300° C) thanks to the extra
energy supplied to the gas molecules by the plasma in the reactor. However, the quality of the films
tend to be inferior to processes running at higher temperatures. Secondly, most PECVD deposition
systems can only deposit the material on one side of the wafers on 1 to 4 wafers at a time. LPCVD
systems deposit films on both sides of at least 25 wafers at a time. A schematic diagram of a
typical LPCVD reactor is shown in the figure below.
 Figure Typical hot-wall LPCVD reactor.

Electrodeposition

This process is also known as "electroplating" and is typically restricted to electrically

conductive materials. There are basically two technologies for plating: Electroplating and Electroless
plating. In the electroplating process the substrate is placed in a liquid solution (electrolyte). When an
electrical potential is applied between a conducting area on the substrate and a counter electrode
(usually platinum) in the liquid, a chemical redox process takes place resulting in the formation of a
layer of material on the substrate and usually some gas generation at the counter electrode.

In the electroless plating process a more complex chemical solution is used, in which deposition
happens spontaneously on any surface which forms a sufficiently high electrochemical potential with
the solution. This process is desirable since it does not require any external electrical potential and
contact to the substrate during processing. Unfortunately, it is also more difficult to control with
regards to film thickness and uniformity.

The electrodeposition process is well suited to make films of metals such as copper, gold and
nickel. The films can be made in any thickness from ~1µm to >100µm. The deposition is best
controlled when used with an external electrical potential, however, it requires electrical contact to the
substrate when immersed in the liquid bath. In any process, the surface of the substrate must have an
electrically conducting coating before the deposition can be done.
 Figure : Typical setup for electrodeposition.

Epitaxy

This technology is quite similar to what happens in CVD processes, however, if the substrate is
an ordered semiconductor crystal (i.e. silicon, gallium arsenide), it is possible with this process to
continue building on the substrate with the same crystallographic orientation with the substrate acting
as a seed for the deposition. If an amorphous/polycrystalline substrate surface is used, the film will also
be amorphous or polycrystalline.

There are several technologies for creating the conditions inside a reactor needed to support
epitaxial growth, of which the most important is Vapor Phase Epitaxy (VPE). In this process, a number
of gases are introduced in an induction heated reactor where only the substrate is heated. The
temperature of the substrate typically must be at least 50% of the melting point of the material to be
deposited.

An advantage of epitaxy is the high growth rate of material, which allows the formation of films
with considerable thickness (>100µm). Epitaxy is a widely used technology for producing silicon on
insulator (SOI) substrates. The technology is primarily used for deposition of silicon. A schematic
diagram of a typical vapor phase epitaxial reactor is shown in the figure below.
 Figure 3: Typical cold-wall vapor phase epitaxial reactor.

This has been and continues to be an emerging process technology in MEMS. The process can be used
to form films of silicon with thicknesses of ~1µm to >100µm.

Thermal oxidation

This is one of the most basic deposition technologies. It is simply oxidation of the substrate surface in
an oxygen rich atmosphere. The temperature is raised to 800° C-1100° C to speed up the process. This
is also the only deposition technology which actually consumes some of the substrate as it proceeds.
The growth of the film is spurned by diffusion of oxygen into the substrate, which means the film
growth is actually downwards into the substrate. As the thickness of the oxidized layer increases, the
diffusion of oxygen to the substrate becomes more difficult leading to a parabolic relationship between
film thickness and oxidation time for films thicker than ~100nm. This process is naturally limited to
materials that can be oxidized, and it can only form films that are oxides of that material. This is the
classical process used to form silicon dioxide on a silicon substrate. A schematic diagram of a typical
wafer oxidation furnace is shown in the figure below.

Figure 4: Typical wafer oxidation furnace.

Evaporation

In evaporation the substrate is placed inside a vacuum chamber, in which a block (source) of the
material to be deposited is also located. The source material is then heated to the point where it starts to
boil and evaporate. The vacuum is required to allow the molecules to evaporate freely in the chamber,
and they subsequently condense on all surfaces. This principle is the same for all evaporation
technologies, only the method used to the heat (evaporate) the source material differs. There are two
popular evaporation technologies, which are e-beam evaporation and resistive evaporation each
referring to the heating method. In e-beam evaporation, an electron beam is aimed at the source
material causing local heating and evaporation. In resistive evaporation, a tungsten boat, containing the
source material, is heated electrically with a high current to make the material evaporate. Many
materials are restrictive in terms of what evaporation method can be used (i.e. aluminum is quite
difficult to evaporate using resistive heating), which typically relates to the phase transition properties
of that material. A schematic diagram of a typical system for e-beam evaporation is shown in the figure
below.

Figure 5: Typical system for e-beam evaporation of materials.

Sputtering

Sputtering is a technology in which the material is released from the source at much lower
temperature than evaporation. The substrate is placed in a vacuum chamber with the source material,
named a target, and an inert gas (such as argon) is introduced at low pressure. A gas plasma is struck
 using an RF power source, causing the gas to become ionized. The ions are accelerated towards the
surface of the target, causing atoms of the source material to break off from the target in vapor form
and condense on all surfaces including the substrate. As for evaporation, the basic principle of
sputtering is the same for all sputtering technologies. The differences typically relate to the manner in
which the ion bombardment of the target is realized. A schematic diagram of a typical RF sputtering
system is shown in the figure below.

Figure 6: Typical RF sputtering system.

Film Stress:

1. Stress is defined microscopically as the force per unit area acting on the surface of a differential
volume element of a solid body. Mechanical stress in thin films is an important reliability issue in
microelectronic devices and systems. The presence of large stresses can lead to the formation of defects
that can cause device failure. The ability to control the magnitude of stress during film formation is,
therefore, crucial to the fabrication of defect-free and reliable electronic devices and systems. However,
the origin of stress in thin films is still a subject of intense debate. The development of a detailed
understanding of the origin of stress hinges on our ability to make accurate stress measurements during
and after film deposition. To this end, two novel MEMS structures were developed to measure the stress
of thin films deposited using chemical vapor deposition (CVD).
2.Deposited films will often be under mechanical stress. A further factor limiting film thickness
and the structures that can be created is mechanical stress in the deposited films. Too much stress will
lead to the structure buckling or the films’ wrinkling or cracking. High-temperature nitride films have a
particular problem: they exhibit high tensile stresses within the film and cannot be deposited directly onto
silicon. A stress-relieving layer of oxide is required.
3. It is possible to control the stress in films by altering the deposition parameters or the
composition of the resulting film (using PECVD to deposit a hydrogen-contaminated silicon oxynitride
layer, for instance). The mechanical, electrical, and chemical (etching) properties of the film will all be
affected by the deposition parameters used, so it is necessary to carefully characterize and monitor each
process (a demanding and time-consuming job) or seek out a foundry that has experience with the
processes required for the device under development.
Nitride films have higher stress but are mechanically harder and chemically more resilient (to attach to
and with respect to diffusion of ions or moisture).
4. Signs of Stress: Stress in films may cause one of the following several problems:
• Cracking or wrinkling of the film
• Strings peeling off from sharp corners
• Twisting or buckling of structures (particularly, cantilever beams)
• Buckling of the silicon wafer (in extreme cases)

BUCKLING
Plane Stress
Plane stress is a special case that occurs very frequently in thin-film materials used in MEMS devices. It
refers to a thin film attached to a (relatively) rigid.

It is typically true that a thin film deposited or formed on a substrate has some in-plane stress, arising
either from the details of the deposition process or from mismatches in thermal expansion between the
film and the substrate. In regions that are more than about three film thicknesses from the edge of the
film, all of the stresses lie in the plane, since the top surface is stress-free. In the edge regions, however,
the situation is more complex (see below).
It can be shown from the fundamentals of elasticity that it is always possible, in the case of plane stress,
to define a coordinate system in which there are two components of in-plane normal stress, and no in-
plane shear stresses.
These coordinates are called the principal axes. Expressed in principal-axis coordinates, the constitutive
equations of plane stress are
with all other stress components equal to zero. A further special case, called biaxial plane stress occurs
when the two in-plane stress components are equal to each other. In this case, the and components of
strain are also equal to each other. That is

Other Sources of Residual Thin-Film Stress

Thermal expansion mismatch is only one possible source of in-plane stress. Chemical reactions occurring
far from equilibrium, such as the oxidation of silicon and the LPCVD deposition of silicon nitride, can
result in films that are highly stressed, far beyond what one would calculate from thermal expansion
mismatch. In addition, modifications to a film’s equilibrium structural properties by substitutional doping,
addition of atoms by ion implantation, mismatch of lattice spacing during epitaxial growth, and rapid
deposition processes such as evaporation or sputtering can all result in thin-film stress. We call any stress
in a thin film that is present after deposition a residual stress. We tend to decompose it into two
components: the thermal mismatch stress, already discussed, and anything not explained by thermal
mismatch, which is called intrinsic stress.
Residual stresses can be modified by thermal processes such as annealing. For example, in polycrystalline
silicon, annealing at temperatures above the deposition process can result in grain growth and a
corresponding reduction in compressive stresses, even conversion to tensile stress.
Exotic Processes:

Exotic Materials can include plastics, superalloys, semiconductors, superconductors, and ceramics.

Work in materials other than silicon for MEMS applications has typically been restricted to metals and
metal oxides instead of more “exotic” semiconductors. However, group III-V and II-VI semiconductors
form a very important and versatile collection of material and electronic parameters available to the
MEMS and MOEMS designer. With these materials, not only are the traditional mechanical material
variables (thermal conductivity, thermal expansion, Young’s modulus, etc.) available, but also chemical
constituents can be varied in ternary and quaternary materials. This flexibility can be extremely important
for both friction and chemical compatibility issues for MEMS. In addition, the ability to continually vary
the bandgap energy can be particularly useful for many electronics and infrared detection applications.
However, there are two major obstacles associated with alternate semiconductor material MEMS. The
first issue is the actual fabrication of non-silicon devices and the second impediment is communicating
with these novel devices. We will describe an essentially material independent fabrication method that is
amenable to most group III-V and II-VI semiconductors. This technique uses a combination of non-
traditional direct write precision fabrication processes such as diamond turning, ion milling, laser
ablation, etc. This type of deterministic fabrication approach lends itself to an almost trivial assembly
process.
Mechanical Transducers:

There is a tremendous variety of direct mechanical sensors that have been or could be micromachined
depending on their sensing mechanism (usually piezoresistive, piezoelectric or capacitive) and the
parameters sensed (typically strain, force and displacement).
i) Piezoresistive sensors
As a result of the piezoresistive effect (defined as the change in resistivity of the material with applied
strain), changes in gauge dimension result in proportional changes in resistance in the sensor. The
piezoresistive effect in semiconductors is considerably higher than in traditional metals, making silicon an
excellent strain sensor. MEMS piezoresistors are readily manufactured using bulk silicon doped with
p-type or n-type impurities.
ii) Piezoelectric sensors
Piezoelectric sensors utilize the piezoelectric effect in which an applied strain (or force) on a
piezoelectric crystal results in a potential difference across the crystal. Similarly, if the crystal is
subjected to a potential difference, a displacement, or strain, is produced. The effect can be used to
sense mechanical stress (i.e. displacement) and as an actuation mechanism, although displacements are
small even for large voltages. Common piezoelectric materials used for MEMS applications include
quartz, lead zirconate titanate (PZT), polyvinylidene fluoride (PVDF) and ZnO, PVDF and ZnO being the
most common. Silicon is not piezoelectric; hence a thin film of a suitable material must be deposited on
the devices.

iii) Capacitive sensors

Capacitive (or electrostatic) sensing is one of the most important (and widely used) precision sensing
mechanisms and includes one or more fixed conducting plates with one or more moving conducting
plates. Capacitive sensing relies on the basic parallel-plate capacitor equation shown below. As
capacitance is inversely proportional to the distance between the plates, sensing of very small
displacements is extremely accurate.

iv) Resonant sensors

MEMS resonant sensors consist of micromachined beams or bridges which are driven to vibrate at their
resonant frequency. They can be attached to membranes or designed to adhere to a particular substance
(as in the case of a biosensor). Movement of the membrane or increased build-up of the binding
substance will affect the resonant frequency and can be monitored using implanted piezoresistors.
Types of mechanical sensor include:
a) Strain gauge - a strain gauge is a conductor or semiconductor that is fabricated on or bonded directly
to the surface to be measured. An example of a polysilicon strain sensor unable to be fabricated by any
other method than MEMS is an implantable piezoresistive strain gauge to measure forces in heart and
brain tissue.
b) Accelerometer - accelerometers sense acceleration by using a suspended proof mass on which
external acceleration can act (Figure 7). Upon acceleration (or deceleration), a force (F=ma) is generated
on the proof mass resulting in displacement. The force or displacement is usually measured by
piezoresistive and capacitive methods.

Figure 7. Suspended proof mass in a piezoresistive accelerometer (not to scale).

c) Gyroscope – a gyroscope is a device that measures the rotation rate and detects inertial angular
motion. As a result it can be found, for example, in transportation, navigation and missile guidance
applications. It relies on measuring the influence of the Coriolis force on a body in a rotating frame.
MEMS gyroscopes typically use vibrating structures because of the difficulty of micromachining rotating
parts with sufficient useful mass.
d) Pressure sensor - MEMS pressure sensors are usually based around thin membranes with sealed gas
or vacuum-filled cavities on one side of the membrane and the pressure to be measured on the other
side. Piezoresistive and capacitive membrane deflection measurement techniques are most commonly
used in commercial pressure sensors.
Transduction methods:-

The process of converting one form of energy to another is known as transduction.A transducer is a

device that converts energy from one form to another.

Accelerometers:

An accelerometer is a device that measures proper acceleration. Proper acceleration, being

the acceleration (or rate of change of velocity) of a body in its own instantaneous rest frame, is not the
same as coordinate acceleration, being the acceleration in a fixed coordinate system. For example, an
accelerometer at rest on the surface of the Earth will measure an acceleration due to Earth's gravity,
straight upwards (by definition) of g ≈ 9.81 m/s2. By contrast, accelerometers in free fall (falling toward
the center of the Earth at a rate of about 9.81 m/s2) will measure zero.

Highly sensitive accelerometers are components of inertial navigation systems for aircraft and missiles.
Accelerometers are used to detect and monitor vibration in rotating machinery. Accelerometers are used
in tablet computers and digital cameras so that images on screens are always displayed upright.
Accelerometers are used in drones for flight stabilisation. Coordinated accelerometers can be used to
measure differences in proper acceleration, particularly gravity, over their separation in space; i.e.,
gradient of the gravitational field. 

Conceptually, an accelerometer behaves as a damped mass on a spring. When the accelerometer

experiences an acceleration, the mass is displaced to the point that the spring is able to accelerate the mass
at the same rate as the casing. The displacement is then measured to give the acceleration.

In commercial devices, piezoelectric, piezoresistive and capacitive components are commonly used to

convert the mechanical motion into an electrical signal. Piezoelectric accelerometers rely on
piezoceramics (e.g. lead zirconate titanate) or single crystals (e.g. quartz, tourmaline). They are
unmatched in terms of their upper frequency range, low packaged weight and high temperature range.
Piezoresistive accelerometers are preferred in high shock applications. Capacitive accelerometers
typically use a silicon micro-machined sensing element. Their performance is superior in the low
frequency range and they can be operated in servo mode to achieve high stability and linearity.

Modern accelerometers are often small micro electro-mechanical systems (MEMS), and are indeed the
simplest MEMS devices possible, consisting of little more than a cantilever beam with a proof mass (also
known as seismic mass). Damping results from the residual gas sealed in the device. As long as the Q-
factor is not too low, damping does not result in a lower sensitivity.
Another, relatively new type of MEMS-based accelerometer is a thermal (or convective) accelerometer
that contains a small heater at the bottom of a very small dome, which heats the air/fluid inside the dome,
producing a thermal bubble that acts as the proof mass.

Gyroscope – a gyroscope is a device that measures the rotation rate and detects inertial angular motion.
As a result it can be found, for example, in transportation, navigation and missile guidance applications. It
relies on measuring the influence of the Coriolis force on a body in a rotating frame. MEMS gyroscopes
typically use vibrating structures because of the difficulty of micromachining rotating parts with sufficient
useful mass.
Pressure sensor - MEMS pressure sensors are usually based around thin membranes with sealed gas or
vacuum-filled cavities on one side of the membrane and the pressure to be measured on the other side.
Piezoresistive and capacitive membrane deflection measurement techniques are most commonly used in
commercial pressure sensors.
MEMS microphones

The application of MEMS (microelectro-mechanical systems) technology to microphones has led to the
development of small microphones with very high performance. MEMS microphones offer high SNR,
low power consumption, good sensitivity, and are available in very small packages that are fully
compatible with surface mount assembly processes. MEMS microphones exhibit almost no change in
performance after reflow soldering and have excellent temperature characteristics.

Mechanical structures:

1. These are non-moving structures, such as microbeams and microchannels.

Micromachining is used commercially to produce channels for microfluidic devices and also to fabricate
systems referred to as ‘‘labs on a chip’’ for chemial analysis and analysis of biomedical materials .
Usually such channels are made on plastics or glass substrates. Figure 1.9 (a) shows one such device
reported in the literature . Micromachining is also used to make a variety of mechanical structures. Figure
1.9 (b) shows an SEM image of a silicon nanotip fabricated using a ‘‘bulk micromachining’’ process.
Such microtips find applications in atomic force microscopy (AFM) technology and field emission array
for futuristic vacuum electron devices capable of operation in terahertz frequency range .
Miniature mechanical structures showing (a) polymer mesopump; (b) silicon nano tip fabricated using
bulk micromachining.
Actuators

Sensors are transducers that convert mechanical, thermal, or other forms of energy into electrical
energy, actuators do the exact opposite.
An actuator is a device that is responsible for moving or controlling a mechanism or system. It is
controlled by a signal from a control system or manual control. It is operated by a source of energy, which
can be mechanical force, electrical current, hydraulic fluid pressure, or pneumatic pressure, and converts
that energy into motion. An actuator is the mechanism by which a control system acts upon an
environment. The control system can be simple (a fixed mechanical or electronic system), software-based
(e.g. a printer driver, robot control system), a human, or any other input.
Bidirectional transducers convert physical phenomena to electrical signals and also convert
electrical signals into physical phenomena. An example of an inherently bidirectional transducer is
an antenna, which can convert radio waves (electromagnetic waves) into an electrical signal to be
processed by a radio receiver, or translate an electrical signal from a transmitter into radio waves. Another
example is voice coils, which are used in loudspeakers to translate an electrical audio
signal into sound and in dynamic microphones to translate sound waves into an audio signal.
Over the last few years, micromachined actuators such as RF switches, micropumps, and
microvalves , which can be actuated using the electrostatic or piezoelectric effect, have been reported. In
addition, electrostatically actuated tiny micromirror arrays acting as optical switches have been developed
by Lucent Technologies for fiber-optic communication and as digital micromirror devices (DMDs) by
Texas Instruments in projection video systems. In this section, two examples of actuators are presented to
highlight the wide range of applications of microactuators.
First, a schematic diagram of the electrostatically actuated bulk micromachined silicon
micropump is shown in Figure.
The diaphragm deflects upward when an actuation voltage is applied as shown and the inlet check valve
opens due to a fall in chamber pressure, thereby letting in the fluid flow into the pump chamber. When the
actuation voltage drops to zero, the diaphragm moves back to its equilibrium position. As the chamber is
now full of fluid, the pressure inside is higher than that outside. This forces the inlet valve to close and the
outlet valve to open, thus letting the fluid flow out.

The second example is the famous electrostatically actuated micromirror array consisting of tiny
mirrors as shown in Figure 1.12. Each mirror is 0.5mm in diameter, about the size of the head of a pin.
Mirrors rest 1mm apart and all 256 mirrors are fabricated on a 2.5 cm square piece of silicon. The figure
on the left-hand side shows how the tilted mirrors switch the optical signals from one fiber to another
fiber. Thus, there is no need of making an optoelectronic conversion. This arrangement gives 100-fold
reduction in power consumption over electronic switches.
The application of MEMS (microelectro-mechanical systems) technology to microphones has led
to the development of small microphones with very high performance.  MEMS microphones
offer high SNR, low power consumption, good sensitivity, and are available in very small
packages that are fully compatible with surface mount assembly processes.  MEMS microphones
exhibit almost no change in performance after reflow soldering and have excellent temperature
characteristics.

Figure 1 Top port and bottom port MEMS microphones

 

MEMS microphone acoustic sensors

 

MEMS microphones use acoustic sensors that are fabricated on semiconductor production lines
using silicon wafers and highly automated processes.  Layers of different materials are deposited
on top of a silicon wafer and then the unwanted material is then etched away, creating a
moveable membrane and a fixed backplate over a cavity in the base wafer.  The sensor backplate
is a stiff perforated structure that allows air to move easily through it, while the membrane is a
thin solid structure that flexes in response to the change in air pressure caused by sound waves.  
 

Figure 2 Cross-section diagram of a MEMS microphone sensor

Figure 3 A typical MEMS microphone sensor viewed from above 

Changes in air pressure created by sound waves cause the thin membrane to flex while the
thicker backplate remains stationary as the air moves through its perforations.  The movement of
the membrane creates a change in the amount of capacitance between the membrane and the
backplate, which is translated into an electrical signal by the ASIC.
 
MEMS microphone ASICs 
 
The ASIC inside a MEMS microphoneuses a charge pump to place a fixed charge on the
microphone membrane.  The ASIC then measures the voltage variations caused when the
capacitance between the membrane and the fixed backplate changes due to the motion of the
membrane in response to sound waves.  Analog MEMS microphones produce an output voltage
that is proportional to the instantaneous air pressure level.  Analog mics usually only have 3 pins:
the output, the power supply voltage (VDD), and ground.  Although the interface for analog
MEMS microphones is conceptually simple, the analog signal requires careful design of the PCB
and cables to avoid picking up noise between the microphone output and the input of the IC
receiving the signal.  In most applications, a low noise audio ADC is also needed to convert the
output of analog microphones into digital format for processing and/or transmission. 
 
As their name implies, digital MEMS microphones have digital outputs that switch between low
and high logic levels.  Most digital microphones use pulse density modulation (PDM), which
produces a highly oversampled single-bit data stream.  The density of the pulses on the output of
a microphone using pulse density modulation is proportional to the instantaneous air pressure
level.  Pulse density modulation is similar to the pulse width modulation (PWM) used in class D
amplifiers.  The difference is that pulse width modulation uses a constant time between pulses
and encodes the signal in the pulse width, while pulse density modulation uses a constant pulse
width and encodes the signal in the time between pulses.
 
In addition to the output, ground, and VDD pins found on analog mics, most digital mics also
have inputs for a clock and a L/R control.  The clock input is used to control the delta-sigma
modulator that converts the analog signal from the sensor into a digital PDM signal.  Typical
clock frequencies for digital microphones range from about 1 MHz to 3.5 MHz.  The
microphone’s output is driven to the proper level on the selected clock edge and then goes into a
high impedance state for the other half of the clock cycle.  This allows two digital mic outputs to
share a single data line.  The L/R input determines which clock edge the data is valid on.
 
The digital microphone outputs are relatively immune to noise, but signal integrity can still be a
concern due to distortion created by parasitic capacitance, resistance, and inductance between the
microphone output and the SoC.  Impedance mismatches can also create reflections that can
distort the signals in applications with longer distances between the digital mic and the SoC.
 
Although codecs are not required for digital MEMS microphones, in most cases the pulse density
modulated output must be converted from single-bit PDM format into multibit pulse code
modulation (PCM) format.  Many codecs and SoCs have PDM inputs with filters that convert the
PDM data into PCM format.  Microcontrollers can also use a synchronous serial interface to
capture the PDM data stream from a digital mic and convert it into PCM format using filters
implemented in software.

 
MEMS microphone packages
MEMS microphone have hollow packages that consist of a substrate with pads that can be
soldered to a circuit board or flex circuit, and a lid that creates a cavity where the acoustic sensor
and the ASIC are located.  Most MEMS microphones use separate die for the MEMS sensor and
the interface ASIC, which allows the MEMS process to be optimized for creating moving
structures while a using a standard CMOS process to fabricate the ASIC.  The ASIC is wire
bonded to sensor and the substrate, and a lid is then placed over them and sealed to the substrate.
 
 Figure 4 A MEMS microphone with its lid removed
 

MEMS microphones need to have a hole in their package to allow sound to reach the acoustic
sensor.  The sound inlet can be located either in the lid (top port) or on the bottom next to the
solder pads (bottom port).  Bottom port microphones also require a hole in the circuit board they
are mounted on to allow sound to reach the sound inlet.  The choice of whether to use a top port
or bottom port microphone is usually determined by factors such as the location of the
microphone in the product and manufacturing considerations, to mention a couple.  Performance
can also be a major factor in microphone port selection since top port microphones have
traditionally had poorer performance than equivalent bottom port microphones.  However, the
introduction of high performance top-port mics such as ST’s MP34DT01 means this is no longer
necessarily true.
 

The membrane of the acoustic sensor divides the interior of a MEMS microphone into two
sections.  The area between the sound inlet and the sensor membrane is generally referred to as
the front chamber, and the section on the other side of the membrane is known as the back
chamber (Figure 5 ).  The sensor in bottom port microphones is usually placed directly over
the sound inlet which provides several benefits.

Figure 5 Cross-section diagram of a typical bottom port MEMS microphone

 
The sensitivity of most MEMS microphones increases at higher frequencies.  This increase in
sensitivity is caused by the interaction between the air in the sound inlet and the air in the front
chamber of the microphone.  This interaction creates a Helmholtz resonance, which is the same
phenomenon that creates sound when blowing into a bottle.  As with bottles, smaller air volumes
create higher resonant frequencies and larger air volumes create lower resonant frequencies.  The
microphone sensor is mounted directly over the sound inlet in most bottom port microphones,
which results in a relatively small front chamber and a high center frequency for the Helmholtz
resonance.  Because the Helmholtz resonance is normally located in the upper part of the audio
band, increasing the resonant frequency leads to a flatter frequency response.
 
Placing the sensor directly over the sound inlet also creates a relatively large back chamber.  A
larger volume of air in the back chamber makes it easier for the membrane to move in response
to sound waves, which improves the sensitivity of the microphone and leads to higher SNR.  A
large back chamber also improves the microphone’s low frequency response. The construction of
top port microphones has traditionally been very similar to bottom port microphones, with the
sensor and the interface IC mounted on the substrate with a hollow lid enclosing them. 
Traditionally, the only real difference between top port and bottom port microphones is that the
sound inlet is located in the microphone lid instead of in the substrate.  For these microphones,
moving the sound inlet to the lid turns what was previously the front chamber into the back
chamber and vice versa.

Figure 6 Cross-section diagram of a traditional top port MEMS microphone

 
The smaller air volume in the back chamber of traditional top port MEMS microphones makes it
more difficult for the membrane to move, which hurts the sensitivity of the sensor and leads to a
lower SNR.  In addition, the larger air volume in the front chamber between the sound inlet and
the membrane lowers the resonant frequency, hurting the microphone’s high frequency
response.  This combination of lower SNR and poorer frequency response at both high and low
frequencies is the reason that most top port microphones have poorer performance than an
equivalent bottom port microphone.
 
An exception to this rule is STMicroelectronics’ MP34DT01 top-port digital MEMS
microphone.  ST’s proprietary packaging technology makes it possible to mount the MEMS
sensor and the interface IC on the bottom side of the lid of the MP34DT01, directly beneath the
sound inlet (Figures 7 and 8 ).  This results in a small front chamber and a large back
chamber and allows the MP34DT01 to achieve the same level of performance as the bottom port
version of this mic, the MP34DB01.
 Figure 7 Cross-section diagram of ST’s MP34DT01 top port MEMS microphone
 

Figure 8  The MEMS sensor and ASIC mounted to the bottom side of the lid
 of the MP34DB01 top port MEMS microphone.
Measuring MEMS microphone performance
 
The SI unit for pressure is the pascal (Pa), which is a linear measure of force per unit area (1 Pa =
1 N/m  ).  However, logarithmic scales are more convenient when discussing sound pressure
2

levels due to the large dynamic range of the human ear, which can detect sounds from as low as
20 micropascals to more than 20 pascals.  Because of this, the key measures of microphone
performance are normally expressed in decibels (dB).  0dB SPL is equal to 20 µPa and 1 Pa is
equal to 94dB SPL.  The following parameters are normally the most important indicators of
microphone performance:
 
Signal-to-noise ratio (SNR)
The signal-to-noise ratio (SNR) is the most important measure of microphone performance in
most applications.  The signal-to-noise ratio is the difference between a microphone’s
sensitivity and its noise floor and is expressed in dB.  The SNR of current MEMS
microphones ranges from about 56 dB to about 66 dB.
 
Sensitivity
The sensitivity of a microphone is a measure of its response to a given sound pressure level. 
Sensitivity is normally specified at a frequency of 1 kHz and at 94 dB SPL (1 Pa).  The
sensitivity of analog microphones is usually expressed in decibels relative to 1 volt RMS
(dBV) while the sensitivity of digital microphones is normally expressed in decibels relative
to the microphone’s full scale output (dB FS)
 
Noise floor
The noise floor of a microphone is the amount of noise on its output in a perfectly quiet
environment.  Both the sensor and the interface ASIC contribute noise to the output of a
microphone.  The noise contributed by the sensor is created by the random Brownian motion
of air molecules while the noise from the ASIC is created by the preamplifier and, in the case
of digital microphones, the delta-sigma modulator.  The noise floor is measured across the
full audio band and A-weighting filters are used to provide a better measure of the noise level
as perceived by human ears. 
 
The noise floor is not always specified in microphone datasheets but it can be calculated by
subtracting the microphone’s SNR from its sensitivity, providing a result in dBV or dB FS.
The noise floor can be expressed as an equivalent input noise in dB SPL by subtracting the
SNR from the sound pressure level that the sensitivity is measured at (usually 94 dB SPL). 
 
Distortion (THD)
Distortion is a measure of how accurately a microphone can capture sound.  Distortion is
usually specified at about 94 dB – 100 dB SPL in order to provide a good indication of a
microphone’s audio quality at normal sound levels.
 
Acoustic Overload Point (AOP)
Distortion normally does not increase much as the sound pressure level increases until the
sound pressure level starts to approach the acoustic overload point of the microphone.  When
this happens the distortion starts rising rapidly.  The acoustic overload point of a microphone
is normally defined as the sound pressure level where the distortion reaches 10%.
 
Frequency response
The frequency response of a MEMS microphone refers to the change in its sensitivity at
various frequencies.  The frequency response of a microphone is usually set to 0 dB at 1 kHz
to normalize the results.  The sensitivity of most MEMS microphones falls off below 100 Hz
and starts rising between about 4 kHz – 6 kHz due to the Helmholtz resonance.  This is the
reason that many MEMS mics only specify their frequency response between 100 Hz and 10
 kHz.  However, high performance MEMS microphones are available with a relatively flat
frequency response over the full audio band from 20 Hz to 20 kHz.
 
Power supply rejection (PSR)
The power supply rejection of a microphone is a measure of its ability to prevent noise on the
microphone’s power supply input from appearing on its output.  PSR is usually specified
with a 217 Hz square wave to simulate the TDMA noise generated by GSM cellular radios
and/or a swept sine wave across the audio band.
 
Future trends
The desire for better audio quality is pushing MEMS microphones to higher performance
levels.  Many products are also starting to apply digital signal processing techniques to arrays
of two or more mics in order to reduce noise and/or focus the microphone sensitivity in a
particular direction.
Higher SNR
MEMS microphone performance continues to improve.  SNRs have increased from 55 – 58
dB a few years ago to 63 – 66 dB today, resulting in cleaner audio capture and allowing
microphones to be used at greater distances with the same level of clarity.  High SNR levels
are needed by automatic speech recognition algorithms to achieve good word accuracy rates. 
 
Higher sound pressure levels
Many microphone users are also requesting higher acoustic overload points to prevent
distortion in loud environments.  Distortion caused by clipping at sound pressure levels
above the acoustic overload point can make recordings made in loud environments such as
rock concerts unusable.
 
Smaller package sizes
MEMS microphone package sizes are also shrinking as consumer demand for thinner, lighter
products continues to increase.  Early MEMS microphones had package sizes of 3.76mm x
4.72mm x 1.25mm while today 3mm x 4mm x 1mm and 2.95mm x 3.76mm x 1mm packages
are common.  Newer MEMS microphones are available in 2.5mm x 3.35mm x 0.98mm and
2.65mm x 3.5mm x 0.98mm packages.  This trend is likely to continue, although smaller
microphone packages make it more difficult to maintain or improve audio quality due to the
shrinking size of the microphone’s back chamber.
 
Ambient noise reduction
 Many smartphones and tablets are starting to use more than one microphone to enable
features such as video recording.  Another common way in which multiple microphones are
used is for ambient noise reduction.  Many smartphones use a microphone located on the top
or the back of the phone to detect noise in the surrounding environment and subtract it from
the output from the voice microphone(s) to help improve intelligibility.  Microphones whose
primary purpose is video recording are frequently also used for ambient noise reduction.
 
Beamforming
Arrays of two or more microphones are also being used to perform beamforming, which
processes the outputs from a microphone array to increase the sensitivity in a particular
direction while at the same time rejecting sounds from other directions.  Most microphones
have omnidirectional outputs, i.e., the sensitivity is the same in all directions, but in many
cases, it is desirable to focus the sensitivity in a particular direction and reduce the sensitivity
in other directions in order to improve intelligibility.  Beamforming uses the phase
differences of sounds arriving from different directions to focus the sensitivity of the
microphones in a particular direction.  Beamforming can also be used to locate the direction a
sound is coming from.  Beamforming is particularly useful in applications where the
microphone is not close to the person speaking such as living rooms, conference rooms,
automobiles, etc.  It can also be very useful when using speakerphones or videoconferencing
in noisy environments.
 
Tighter control of sensitivity
The performance algorithms used to perform functions such as noise cancellation and
beamforming usually assume that the sensitivity of the microphones being used is the same,
so variations in sensitivity between the microphones in an array hurt the performance of the
algorithms.
 
This has created a demand for tighter sensitivity matching.  MEMS microphones typically
have a ±3 dB tolerance on sensitivity, but this can be tightened to ±1 dB by screening
microphones to a tighter tolerance (binning) and/or trimming of the microphone ASICs to
compensate for normal variation in microphone parameters.
 
The use of MEMS microphones is increasing rapidly.  IHS predicts that the usage of MEMS
microphones will almost double over the next four years, rising from about 2.7 billion units in
2013 to about 5 billion units in 2017.  The growth of applications such as tablets and
smartphones is a major contributor to the growth in the use of MEMS microphones.  The growth
in the usage of MEMS microphones is also being fueled by the increasing use of two or more
microphones in products to enable features such as video recording, active noise cancellation,
and beamforming.