functional thin films and nanostructures for sensors. synthesis, physics, and applications

Functional thin films are being used in the manufacture of devices such as surface acoustic wave SAW devices for high-frequency telecommunications filtering, infrared detectors, pres-sur

Nanostructures for Sensors

Series Editor: Dr Radislav A Potyrailo

GE Global Research, Niskayuna, NY

For other titles published in this series, go to

www.springer.com/series/7427

Functional Thin Films and Nanostructures for SensorsSynthesis, Physics, and Applications

ISBN: 978-0-387-36229-8 e-ISBN: 978-0-387-68609-7

DOI: 10.1007/978-0-387-68609-7

Library of Congress Control Number: 2008944096

© Springer Science+Business Media, LLC 2009

All rights reserved This work may not be translated or copied in whole or in part without the written permission of the publisher (Springer Science+Business Media, LLC, 233 Spring Street, New York,

NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis Use in connection with any form of information storage and retrieval, electronic adaptation, computer software,

or by similar or dissimilar methodology now known or hereafter developed is forbidden.

The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights.

Printed on acid-free paper

springer.com

Anis Zribi

United Technologies Corporation

Fire and Security

Kidde Detection Technology Research

Development and Engineering

USA

Olena, Nadia

Michelle, Abi, Libby

In recent years, there has been a convergence of fundamental materials science and materials processing methods This convergence, although highly interdisciplinary

in nature, has been brought about by technologies such as bandgap engineering and related techniques that have led to application-specific devices such as lab-on-a-chip and system-on-a-chip The demand for reduced device size, device portability, and low power dissipation coupled with high speed of operation continues to dictate terms and conditions for the evolution of nanotechnology The present trend in approaches to systems manufacturing continues to focus on integration of multi-functionalities on the same chip These functionalities include, for example, on-board laser sources, sensors, and amplifiers Both the military and civilian markets continue to drive the research and development component In recent years, the emergency preparedness guidance systems have added excitement and curiosity to this expanding industry The outgrowth of technologies of interest for emergency preparedness includes the development of terahertz sources and detectors and sys-tems for detection of explosives and concealed weapons, among others

Sensors made from bulk materials have been around for a long time Enormous advances in the processing technologies of thin films have led to the ability to manufacture application-specific functional thin films These include transparent electrodes and antireflection films such as indium tin oxide, which serve as inter-face components between humans and electronic devices, or optical circuit ele-ments used in optical communication networks, or as contacts and antireflection coatings in solar cells Products are also being developed with magneto-optical, electrochromic, or UV material for their use as functional thin films in optics Photonic crystals contain a variety of functional thin films; they require processing

of thin films under very stringent control of their structure and properties

For microelectromechanical systems (MEMS), in addition to silicon-based nology, ferroelectric thin films are being used in the fabrication of microactuators and micromotors, capacitors, and other thin-film devices Functional thin films are being used in the manufacture of devices such as surface acoustic wave (SAW) devices for high-frequency telecommunications filtering, infrared detectors, pres-sure sensors, accelerometers, force sensors, vibration, thickness, and chemical sen-sors and biosensors The reduction in size from bulk to micro- and nanostructured transducers, while promising high sensitivity, high speed, and increased selectivity,

tech-vii

requires new design considerations that should consider factors such as integration with other devices and device lifetime.

Functional thin films offer an enormous infrastructure for a highly ary integration of inorganic/semiconducting, organic/bio, and electronic/optoelec-tronic sensor systems The field is constantly evolving and will continue to do so

interdisciplin-by absorbing novel materials approaches such as carbon nanotubes, high Tc conductors, ferroelectrics, and thermoelectrics

super-The chapters in this book are designed to give the reader the big picture, from the design phase to the implementation and realization of a transducer Every effort has been made to include the state-of-the-art in each chapter The intended audience

is scientists, researchers, and engineers, however, graduate students will find the book to be very useful in their research and understanding of sensors and beyond The editors and contributors are leading researchers in industry and academia in their subject areas

February 2008

In my career I’ve found that “thinking outside the box” works better if I know what’s “inside the box.”

Dave Grusin, composer and jazz musician

Different people think in different time frames: scientists think

in decades, engineers think in years, and investors think in quarters.

Stan Williams, Director of Quantum Science Research,Hewlett Packard Laboratories

Everything can be made smaller, never mind physics;

Everything can be made more efficient, never mind thermodynamics;

Everything will be more expensive, never mind common sense.

Tomas Hirschfeld, pioneer of industrial spectroscopy

Integrated Analytical Systems

Series Editor: Dr Radislav A Potyrailo, GE Global Research, Niskayuna, NY

The book series Integrated Analytical Systems offers the most recent advances in all

key aspects of development and applications of modern instrumentation for cal and biological analysis The key development aspects include: (i) innovations in sample introduction through micro- and nanofluidic designs; (ii) new types and methods of fabrication of physical transducers and ion detectors; (iii) materials for sensors that became available due to the breakthroughs in biology, combinatorial materials science, and nanotechnology; and (iv) innovative data processing and mining methodologies that provide dramatically reduced rates of false alarms

chemi-A multidisciplinary effort is required to design and build instruments with ously unavailable capabilities for demanding new applications Instruments with more sensitivity are required today to analyze ultratrace levels of environmental pollutants, pathogens in water, and low vapor pressure energetic materials in air

previ-ix

Sensor systems with faster response times are desired to monitor transient in vivo events and bedside patients More selective instruments are sought to analyze specific proteins in vitro and analyze ambient urban or battlefield air For these and many other applications, new analytical instrumentation is urgently needed This book series is intended to be a primary source of both fundamental and practical information of where analytical instrumentation technologies are now and where they are headed in the future.

Looking back over peer-reviewed technical articles from several decades ago, one notices that the overwhelming majority of publications on chemical analysis has been related to chemical and biological sensors and has originated from depart-ments of chemistry in universities and divisions of life sciences of governmental laboratories Since then, the number of disciplines has dramatically increased because of the ever-expanding needs for miniaturization (e.g., for in vivo cell analy-sis, embedding into soldier uniforms), lower power consumption (e.g., harvested power), and the ability to operate in complex environments (e.g., whole blood, industrial water, or battlefield air) for more selective, sensitive, and rapid determi-nation of chemical and biological species Compact analytical systems that have a sensor as one of the system components are becoming more important than indi-vidual sensors Thus, in addition to traditional sensor approaches, a variety of new themes has been introduced to achieve an attractive goal of analyzing chemical and biological species on the micro- and nanoscale

Anyone with the most cursory knowledge of sensors must have had a chance to use such devices at some point in their life or career Whether to collect data in a lab course, to automate an otherwise tedious process, to improve the efficiency of a delicately tuned process, or to do something as mundane as taking a family picture, sensors have become an integral part of our environment and our daily lives Charge-coupled devices also known as CCD photodetector arrays, for example, have revolutionized photography, astronomy, spectroscopy, X-ray diffraction, and medical imaging to name but a few A number of scientific discoveries have been enabled by CCDs including the possibility to determine molecular and lattice struc-tures at intermediate stages of a chemical synthesis or a structural transformation

At the core of the widespread adoption of sensors are their rapidly decreasing print and cost and increased functionality The miniaturization of solid-state devices

foot-in general and sensors foot-in particular was made possible thanks to significant formations and a large number of incremental and disruptive inventions in the area

trans-of thin-film and nanostructure science and fabrication technologies

Thin films and nanostructures can play multiple roles in a sensor including structural support, reliability enhancement, filtering, and transduction Thin films and nanostructures are called functional when they fulfill a function other than structural support These micro- and nanostructured materials have applications that extend far beyond sensing to data storage, lighting, displays, hydrophobic coat-ings, decoration, and a large number of other fields that are outside the scope of this book In this book, these materials are discussed in the context of transduction and how they contributed to the current sensor revolution

Sensor design and fabrication are multidisciplinary and require broad and deep knowledge in diverse areas of science and engineering such as materials science, physics, chemistry, biology, and mechanical and electrical engineering Covering a subject with so many roots in diverse scientific and engineering disciplines is undoubtedly a daunting task and any author who attempts it will do so with signifi-cant trepidation Aware of the challenge at hand, the editors of this book attempted, ambitiously, to cover in one volume an account of general sensor theory, design considerations related to the use of functional thin films and nanostructures, and specific case studies of functional thin films and nanostructure applications in sensing Part of our motivation in taking on this task is that no such work, to our

xi

knowledge, has been published Having said this, we are strongly familiar with the large body of publications in this area that we refer to in this book and we are keenly indebted to the works of many authors in putting this book together.This book is devoted to teaching the new sensor designer the key steps involved

in developing sound transducer technology from materials selection, to design for performance, to process development, and finally to integration Throughout the chapters, the authors emphasize and highlight the important role played by func-tional thin films in solving problems and discuss how to take advantage of such materials to build superior devices The book is also intended to provide the more experienced designers with a condensed summary of sensor design methodology and excellent references that will prove useful in future sensor design endeavors

To put all of the shared design and fabrication knowledge into perspective and add

a touch of reality to the concepts discussed in Chapters 1 through 4, Chapters 5 through 8 are completely dedicated to putting the theory into practice and demon-strating the whole design process using a number of concrete applications

Jeffrey Fortin

Anis Zribi is the manager of the Detection Technology Research, Development

and Engineering group at Kidde UTC Fire and Security Prior to joining UTC, he was a senior scientist and a principal investigator at the Global Research Center (GRC) of General Electric where he (leads) led research in the area of Microsystems and microfluidics for chemical and biological detection He received an M.S.E

in physics from the Polytechnic Institute of Engineering (1996 France), an M.S in materials physics from Chalmers University of Technology (1998 Sweden) and a Ph.D in materials science from the State University of New York (2002 NY) Since joining GRC in 2002, Dr Zribi has contributed to and led several projects including the Nanotechnology Advanced Technology Program, the Photonics Advanced Technology Program, and a number of MEMS sensors and actuators projects His research interests and activities at GRC include MEMS spectrometers, chemical and biological sensing, magnetic field sensing, medical parameters sensing, fouling detection, and micro- and nanotransducers Dr Zribi holds 7 patents and over 40 pending patent applications in MEMS, photonics, and sensors He authored or co-authored more than 32 articles in peer-reviewed journals and conference pro-ceedings and two book chapters

Jeff Fortin is the manager of the Microsystems and Microfluidics Lab at GE Global Research in Niskayuna, NY His team’s charter is to develop and deliver innovative micro- and nanosystems and microfluidics via the development and integration of MEMS and NEMS sensing, actuation, and microfluidic technologies, driving miniaturization, increased performance, portability, and low cost Jeff holds

a Ph.D in engineering science from Rensselaer Polytechnic Institute, an M.S in physics from RPI, and a B.A in physics from the University of Southern Maine

He has over ten years of experience in semiconductor technology, MEMS, and microsensors He joined GE GRC in 2000 and since this time his research has focused on MEMS and microsystems design and fabrication for a variety of microsensor and microactuator applications for GE He holds ten patents and has co-authored over 12 refereed journal articles in the area of MEMS, thin polymer film development, and chemical vapor deposition as well as eight conference pub-lications He is also the co-author of a text on chemical vapor deposition polymer-ization of parylene and is a member of the MRS

xiii

1 Sensor Design Guidelines 1Anis Zribi

2 Transduction Principles 17

Jeffrey Fortin

3 Growth and Synthesis of Nanostructured Thin Films 31Yiping Zhao

4 Integrated Micromachining Technologies

for Transducer Fabrication 65Wei Cheng Tian

5 Applications of Functional Thin Films

and Nanostructures in Gas Sensing 85Audrey Nelson

6 Chemical Sensors: New Ideas for the Mature Field 103Radislav A Potyrailo

7 Applications of Functional Thin Films for Mechanical Sensing 145Chang Liu

8 Sensing Infrared and Terahertz Regions by Functional Films 167Magnus Willander, Victor Ryzhii, and Qingxiang Zhao

Index 211

xv

Jeffrey Fortin

GE Global Research Center

Micro and Nano Structures Technologies

GE Global Research Center

Micro and Nano Structures Technologies

United Technologies Corporation Fire and Security, Kidde Detection

Technology Research, Development and Engineering Colorado Springs,

CO, USA

Sensor Design Guidelines

Anis Zribi

Abstract This chapter focuses on introducing fundamental design principles of transducers, familiarizing readers who are new to this field with the common vocab-ulary used in describing transducer performance, and providing a succinct historical background about the implementation of thin films and nanostructures in sensors and analytical instruments A systematic methodology and a sequence of guiding steps to follow in designing a transducer beginning with a concept, through materials selection, and transducer design and fabrication are presented These steps are covered

in more detail in subsequent chapters with concrete examples

The Big Picture

Sensors are ubiquitous in our environment and play essential roles in our everyday life Our own view of the world is defined by our senses that enable us to perceive stimuli from the environment through a network of biological sensors Tiny hairs in our inner ears detect the deflection of a membrane as it vibrates in response to acoustic waves and make it possible for us to hear; photoreceptors in our eyes enable us to see objects and discern their colors; chemical receptors on the tongue (known as taste buds) allow us to differentiate between salty, sweet, bitter, and sour This fascinating network of biological sensors caters to our organs’ needs to control certain biological processes and our needs for security and safety

Driven by the need to better understand our world, to increase the productivity

of industrial processes and machines, and to improve our quality of life, scientists and engineers constantly seek to develop the necessary measurement tools These tools are sensors and instruments that are often inspired by biological sensors and their functioning principles Such devices have become essential for advancing

A Zribi and J Fortin (eds.), Functional Thin Films and Nanostructures for Sensors, 1

Integrated Analytical Systems,

DOI: 10.1007/978-0-387-68609-7_1, © Springer Science + Business Media, LLC 2009

metrology and science, optimizing and controlling processes, providing security against known and unknown threats, law enforcement, and health monitoring The miniaturization of sensors and analytical instruments is a continuing trend that finds its roots in nature and its beginnings in miniaturizing mechanical, optical, and recently electronic devices The driving forces for sensor miniaturization are numerous and keep increasing as we encounter and develop new applications for devices that have become virtually invisible and intangible Sensor cost and porta-bility are major incentives for miniaturization but additional reasons include faster response, integration of multiple functions, reduction of occupied space, and lower device-to-device variability

Electronic miniaturization, starting with the invention of the transistor in 1947, played a key role in developing the fabrication processes that stand at the heart of the new leap in sensor miniaturization In fact tremendous progress has been made in the past 20 years in high vacuum technologies, ultra purification processes for raw mate-rials, self-assembly, and material deposition and etching processes with various degrees of selectivity, high precision, and small feature-patterning techniques This progress produced dramatic advances and control over device quality and induced a technological evolution from the bulk crystalline age to the age of thin films, thin film multilayers, and nanostructured materials In this new age, the properties and perform-ance of submicron devices are dominated by surface and interfacial phenomena This paradigm shift produced materials with novel or enhanced transduction properties known as functional thin films and functional nanostructures

Thin films and nanostructures play an increasingly important role in state-of-the-art sensors and actuator technologies both as transducers (functional materials) and struc-tural materials microelectro mechanical (MEM) systems provide a good example of the growing use of materials confined to submicron dimensions to fulfill numerous and versatile functions in advanced devices Doped silicon/polysilicon thin films, for example, have been implemented as strain gauges in MEM pressure sensor devices

transceivers (cMUT; Jin et al 1998 ), active alignment actuators in high-accuracy optic aligners (Petersen 1982) , and the list goes on The growing interest in functional thin films and nanostructures is not only driven by device miniaturization but also by the novel and unique set of physical and chemical properties that materials confined to submicron dimensions exhibit

fiber-In his famous talk before the audience of the 1959 annual meeting of the American Physical Society, Richard Feynman predicted many of the now proven advantages of “manipulating and controlling things on a small scale” (Feynman 1959) Since Feynman’s talk, the advantages of scaling devices and materials to submicronic dimensions have been proven to go beyond a dramatic increase in data storage density, faster and more intelligent computing, faster heat removal, and higher natural resonance frequency

Today, functional thin films and nanostructures are being used and are under development for use in a wide range of devices, sensors, and actuators Such devices include: accelerometers (air bag devices), force sensors, shutters, optical switches,

optical computation, micromotors, chemical and biological sensors; MEMS devices such as piezomicroactuators and sensors; very small scale microreaction vessels for chemical and biological (lab-on-a-chip) sensing; substrates for plasmon resonance-based signal enhancement and amplification, electro-optical devices for thermal imaging based on the pyroelectric effect, display technologies; pressure mats (piezo-electric thin film); surface acoustic wave (SAW) devices for telecommunications, chemical and biological detection, and more

These new applications for functional thin films and nanostructures are enabled by newly discovered physical and chemical phenomena that underlie the transduction mechanisms in these materials In fact, size confinement and the associated symmetry breaking, interfacial interactions, high surface-to-volume ratio, structural disorder, and induced entropy have been extensively covered in recent publications The fundamen-tals of these effects and their applications in transduction are the topic of numerous current investigations Quantum dots, for example, have been demonstrated to exhibit

a lateral quantum confinement that enables direct coupling of normally incident light with the intraband electronic excitations (Kuo et al 200 1 ) The geometry of these nanostructures (height and radius) can be tuned to ensure that the lowest energy of electronic transition from ground state to first excited state falls within a specific optical spectrum band It has also been demonstrated that the confinement significantly reduces the electronic tunneling rate of these nanostructures These properties indicate the potential of quantum dots to be used as low dark current and hence high signal-to-noise ratio photodetectors especially in the near infrared to infrared part of the spec-trum where current detectors are prone to thermally induced electronic tunneling This and other transduction benefits that emanate from the physical confinement of material structures are discussed in later chapters of this book

Sensor Architecture

The basic function of a sensor is to selectively identify and measure a physical, chemical, or biological parameter such as pressure, light intensity, gas concentration,

or the presence and concentration of a biological analyte The typical architecture of

a sensor encompasses a transducer or multiple transducers (operating in series or in parallel) directly exposed to the measurand, acquisition and conditioning electronics,

a power source, a processor, a storage medium, and a display The transducer plays a central role in the operation of the sensor: it is essentially an energy converter where the input energy (mechanical, optical, chemical, biological, electrical) is converted into an electrical signal most of the time The electrical signal is then acquired by the electronics, conditioned, and noise filtered out before processing (e.g., interpolation using a calibration curve) and finally the data (typically the magnitude of the meas-urand) are either stored in memory and displayed or routed for action (e.g., alarm) or simply displayed

Fig 1.1 shows a general block diagram that highlights the main subsystems of

a sensor system and the flow of data among the various blocks The focus of this

book is on the transducer block and the role that functional thin films and

Depending on the type of transducer, two basic types of sensors can be guished: quantitative (analog or digital) and threshold or binary The two are quite different in function and in application A quantitative sensor produces an output value that is a direct and continuous function of a measurand value For example, a thermocouple might have a potential differential of 10 mV at room temperature and

distin-a potentidistin-al differentidistin-al of 20 mV 10° distin-above room temperdistin-ature Any differentidistin-al potential value between these two is possible depending on the particular tempera-ture to which the sensor is exposed Threshold sensors, on the other hand, have only two states, often called “on” and “off” Perhaps the most familiar example of a threshold sensor is a smoke detector which is triggered to the on position if a fire erupts and the signal is used to trip an alarm

Sensor Figures of Merit/Performance Attributes

The wealth of sensor and transducer technologies available to measure the same

perform-ance of sensor devices Therefore, it is critical to define a set of performperform-ance criteria that the designers can use to develop sensors that meet customer specifications and the user can use to appraise and contrast the various options Table 1.1 summarizes the list of performance attributes that are most commonly used to assess sensor tech-nologies and a more detailed description of these parameters is provided later in this chapter

Fig 1.1 Sensor architecture block diagram

Value N

Power Block

Processor Storage Display

Transducer Amplifier Filter A/D Sensing

Material

Input Dynamic Range

The dynamic range of a sensor is the span of measurands that constitute the overall operating domain for the device Within this interval, the sensor is supposed to maintain its properties and reliability characteristics

Response Curve

Every quantitative sensor is characterized by a response curve that represents the

output of the sensor N versus the measurand M applied to its input The transducer

response can be linear or nonlinear as shown in Fig 1.2 , but in most cases the sor electronics are designed to linearize the response curve of the overall sensor in

Table 1.1 Sensor Performance Attributes Attribute Denotation Input dynamic range DM

order to simplify the calibration procedures Also, it is worth noting that ily the response of the sensor is normalized with respect to reference values (e.g., resistance at room temperature, resonance frequency of the membrane under known

affect the sensor output, the sensor response N can be expressed as

• Transducer sensitivity

• Amplification

• Analog filter sensitivity

transduceri j,

j i

N S

M

∂

=

from the energy transformation performed by transducer j Transducers in a detector

can operate in parallel or in series Typically, transducers that operate in parallel are used for imaging applications, which we are not concerned with within this manuscript

Assuming the p transducers that make up the transducer block operate in series, the

The amplification A is an intrinsic property of the amplifier and the analog filter

sensitivity is defined in a similar manner to the transducer block by Equation (1.3):

Y

∂

=

The overall sensitivity of the device to measurand M i is defined by the following equations,

times of the various blocks, then the total response time of the sensor is:

ttotal =ttransducer+tfilter+t A D/ (1.8)

ttotal =ttransducer+telectronics (1.9)

domi-nant response time in Equations (1.7) and (1.8) Therefore, we typically can ignore the contributions from electronics and physical transducers However, this may not

be the case in thin-film or nanostructure-enabled sensors as the response time of the material is considerably reduced because of the designed nanomorphology of the material This simplification is also not an option for physical sensors where the response time of the transducer is on the same order of magnitude as the other sen-sor subsystems

The response time of the whole device is usually estimated as the time required for

a transient output signal to reach a fraction (e.g., 70%) of its steady-state change

Clearly for nonlinear sensors, the resolution is dependent on the operating point and can vary if operating conditions vary Maximum resolution is attained at maximum sensitivity and minimum noise

Accuracy

The accuracy of a transducer is the maximum deviation of its output from the value

of the unknown measurand as determined by a gold standard technique It is a quantitative indication of the degree of conformity of the sensor to a standard Accuracy represents a systematic error or a bias in the sensor and measuring it and correcting for it are difficult Transducer calibration is required to account for this type of error and partially correct it Long-term changes in the performance of the transducer because of materials’ aging and wear will affect its accuracy

Precision

The second type of measurement uncertainty associated with a transducer is random errors or what is often referred to as precision Precision is often mistakenly confounded with accuracy It quantifies statistical fluctuations in the measurement and is attributed to variability in the measurement conditions and the limitations of the selectivity of the sensor towards the measurand of interest Precision represents the repeatability of the measurement given the same sample

Hysteresis and Drift

Hysteresis characterizes the lagging of the sensor response behind the variation of the measurand This can be attributed to the sensing material memory and/or to the trans-ducer properties As a performance specification, hysteresis is defined as the maximum difference between the upscale and downscale readings on the same artifact during a full-range traverse in each direction It is often reported as the ratio (or percentage) of the difference between the upscale and downscale readings to the full scale

Drift can be defined as the slow unpredictable change of the sensor output at constant input Drift can affect both the signal and noise levels and it can emanate, for example, from residual stress relaxation, residual diffusion, material aging, and degradation Drift in sensors is defined for a specific time interval of interest

Selectivity

It is typical of a sensor designed to detect variations of a given measurand M to

different natures, physical or chemical, as well as biological In general, selectivity may be quantified by

=∏ ransducer 1

1 ransducer1

,

j p t

j

S Se

S

(1.11)

sensitivities to the various measurands are interdependent therefore second-order and cross-terms should be included in the expression of the selectivity

Sensor Design Considerations

Sensor design is one of the most interdisciplinary technical areas requiring both breadth and depth of knowledge in materials science, physics, chemistry, biology, mathematics and statistics, electronics, and packaging The successful design of a sensing system requires very good communication between scientists and engi-neers of different backgrounds

Much as in any design activity, it is very difficult and undesirable to bind creativity

by a set of design rules or a design methodology However, past a first stage, focused

on brainstorming, transduction mechanism and materials down-selection and bility analysis, a successful design team must have a clear objective and a guiding design methodology to steer their effort The typical sensor design steps include:

• Collection and analyses of the device specifications

• Transducer selection/invention

• Materials selection/invention

• Sensor design and modeling

• Prototyping

• Measurement of materials properties

• Prototype testing and model calibration

• Design iteration(s)

• Final device fabrication

• Technology transition to manufacturing

• Fabrication process scaleup

The most challenging and least regulated steps are the first three and they are least covered by the literature In the next sections, we provide some guidance, concepts, and ideas to help the new sensor designer make faster progress towards his or her ultimate goal

Selection and/or Invention of the Transduction Mechanism

Assuming that the design team is armed with a clear set of specifications for the desired sensor performance and the operating and storage conditions, the next steps

ought to target the selection or invention of the transducer(s) that can convert the measurand into an electrical/optical signal The questions that need to be answered

by the team at this stage are:

1 What specific physical, chemical, and biological properties does the measurand possess that differentiate it from other potential sensor inputs?

2 For each of the identified characteristics, what are the potential confounding inputs, which in the future will be identified as noise sources?

3 What type of energy conversion is required to transform the input signal cal, chemical, biological, electromagnetic) to a device output signal (optical or electrical)?

4 What are the candidate transduction schemes that can be used for the target measurand?

The final transduction scheme may involve one or multiple transducers operating

in series or parallel (array of detectors) Considering the following criteria can further refine the list of candidate transduction mechanisms

1 Simplicity of the transducer and robustness to failure

2 Materials requirements dictated by operating environment

3 Development time

4 Cost

Selection of Transducer Material

The transducer material properties are important from performance and sensor ability perspectives The suite of functional materials available to sensor and micro-/nanoinstrument designers is rapidly expanding and numerous techniques have been developed to integrate a large variety of organic and inorganic materials and their alloys The wealth of options puts the designer face to face with the chal-lenge of selecting the best-suited materials given a transducer design concept Numerous considerations come into play when selecting a transducer material and they fall into three categories: performance, process compatibility, and reliability Although it is very difficult to discuss the performance selection criteria of a trans-ducer in general terms, process compatibility and reliability criteria are common to all transducers and they are of chemical (chemical resistance, photo definability, adhesion) and physical nature (rheology, mechanical, dielectric) Depending on the transducer length scale, whether it falls in the bulk (>100 m m), micro (<100 m m),

reli-or nano range (<100 nm), materials properties and their behavireli-ors and responses to environmental conditions and stimuli are significantly different

Over the years, bulk materials properties have been well documented and often standardized, however, submicron materials properties are still not as well docu-mented and many are being investigated (Srikar and Spearing 200 3 ) Some of the

challenges associated with the measurement of micro- and nanomaterials properties are related to the lack of analytical instruments for this regime but many are related

to the high anisotropy of materials properties and to the strong dependence of these properties on the sample geometry and synthetic technique used to prepare the sample In certain cases, it will prove essential to perform materials characteriza-tion to provide a sound sensor system design; in other cases initial guesstimates of the materials properties using bulk materials data and physicochemical laws com-bined with system-level testing, failure analyses, and design iterations will be suf-ficient to develop a robust sensor

The following section is by no means a rigorous treatment of physical and chemical confinement effects on materials The objective is to give the reader a flavor of some of the expected effects of size confinement on various materials properties and provide references where further details can be found

Physical and Chemical Considerations

As device dimensions scale down to hundreds of microns and single digit microns, structural and functional materials’ dimensions that make up the device are approaching the size of a few atomic layers and molecules, a range now familiar to most of us as the nanoregime (<100 nm) In this nanoworld, the surface-to-volume ratio is extremely high and diverges as dimensions shrink down In this dimensional range, surface effects (curvature, surface energy) and structural defects dominate materials’ properties This can be further explained by the fact that intermolecular forces such as Van Der Waals, London dispersion forces, ionic interactions, hydro-gen bonds, and dipole–dipole interactions prevail at the nanoscale In liquids con-fined to nanodimensions, short-range order has been already observed and reported

in numerous studies and liquid thin-film thickness begins to take discrete values The implication is that the physical and chemical properties of nanomaterials differ greatly from their bulk counterparts and they are often strong functions of thermal fluctuations These properties include the effective viscosity, diffusion coefficients, melting point, glass transition temperature, refractive index, mechanical properties (elastic modulus), and the thermal conductivity

Melting Point

A number of studies established that the melting point of a geometrically confined material is different from its bulk melting point (Alcoutlabi and McKenna1 20 05 )

depression of a particle of diameter d follows the Gibbs–Thomson equation (Alcoutlabi and McKenna1 2005) :

This relationship is only applicable if certain conditions related to the isotropy

of the surface tension and invariability of the bulk enthalpy of fusion and density are satisfied These conditions are often violated in the nanoscale but the Gibbs–Thomson relationship is still a good approximation of the melting point depression

of nanomaterials and will give a close estimate of the real melting temperature of thin films and nanostructures More complex melting mechanisms have been and are still being developed and the reader is encouraged to consult this existing large

Coombes 1972 ; Lai et al 199 8 ; Zhang Z et al 2000)

Glass Transition Temperature

The glass transition temperature is by definition the temperature below which molecules have very little relative mobility Different theories predict contradictory effects of confinement on the glass transition temperature of partially or wholly amorphous materials Experimental measurements, however, indicate a glass tran-sition depression accompanying the size reduction of materials (Alcoutlabi and McKenna1 2005) Currently, there are no readily available theories to explain this reduction and no formalism that enables the prediction of the confinement effects

Elasticity and Plasticity

The presence of a higher fraction of atoms near surfaces and interfaces is teristic of thin films and nanostructures The proximity of atoms to a surface or an interface creates an atomic environment, different from the bulk, where surface free energy plays a bigger role in the elastic and plastic behavior of materials Film thickness and grain size, for example, have been proven to affect the deformation mechanisms in metallic films significantly (Lilleodden et al 200 1 ) Metallic films with a thickness of a micron or less exhibit very different plastic behaviors depend-ing on the grain size Initially, films with larger grains exhibit pronounced hardness whereas fine-grained films show a soft behavior When exposed to a load producing

charac-displacements on the same order of magnitude as the characteristic length scales of the film or nanostructure, fine-grained films harden and large-grained films soften due to strain–gradient plasticity It has been established that the large number of grain boundaries in fine-grained structures act as a continuous source of disloca-tions responsible for the plastic behavior and the continuous load-displacement behavior observed in these structures and absent from large-grained structures

A theoretical treatment of the effect of surface free energy on the elastic modulus

of nanostructures can be found in Dingreville et al (200 4 ) In this publication, the authors demonstrated that at nanoscales, the contribution of the surface energy to the elastic modulus is not negligible anymore The effective elastic bulk and shear moduli of an isotropic spherical particle could be estimated using Equation (1.13):

1

1 shear shear

23

expansion of the surface energy density as a function of the surface strain, and L ,

M , and N are the third-order elastic constants of the material Equation (1.13)

clearly indicates that the contribution of the surface energy to the elastic modulus

is inversely proportional to the characteristic length of the structure This term increases the elastic modulus of materials in the nanoregime and is responsible for the high stiffness of nanomaterials An immediate consequence of high stiffness is that nanostructures possess much higher mechanical resonance frequencies than bulk structures This property has been implemented by numerous investigators to develop resonators for various applications including trace chemical and biological detection (Calleja et al 200 5 ), high sensitivity pressure calibrators, and others

Viscosity

Functional thin films and nanostructures can be solid or liquid phase In the micro- and nanosize regime, fluidic droplets and films differ from bulk fluids because of their pronounced inhomogeneity Micro- and nanofluids have been reported in the literature (Pozhar 2000) to exhibit various rheological behaviors that are often contradictory These behaviors are based on fluidic models with scarce experimental data and numerous assumptions regarding the nature and magnitude of interactions

of the fluid molecules with their surrounding

One of the most accurate models of the viscosity of confined, inhomogeneous molecular fluids has been formulated by Pozhar and Gubbins and thus named the

PG model The theoretical viscosity of an inhomogeneous fluid with the

inhomo-geneity in the z -direction can be estimated by using the PG expression shown in

diameter of the fluid molecule specific to the fluid–fluid hardcore intermolecular

propor-tional to the viscoelastic relaxation time

This model, described in detail in the literature (Pozhar 2000) , predicts that the average viscosity of a nanofluid confined to a volume that is a few times the molecular diameter increases up to four times that of the bulk fluid for a given fluid type, density, and temperature The viscosity is dependent on the location within the confining volume and approaches the average value as the critical dimensions

of the confining volume approach ten times the molecular diameter Although not yet supported by pertinent experimental data, this result is very valuable and is in good agreement with measurements conducted on complicated fluids (Pozhar 2000) It constitutes a good approximation of fluid behaviors in the nanoscale and can be used with caution to design transducers based on functional liquid-based thin films or droplet

Optical Properties

In addition to thermodynamic, mechanical, rheological, and electronic size effects,

a nanometer-range-confined material exhibits different optical properties from the bulk counterpart Numerous publications reported shifts in absorption bands (Huang and Lue 199 4 ), narrowing of absorption bands, surface plasmon resonance (SPR), and attenuation of absorption bands (Truong and Courteau 198 7 ) in metallic, polymeric, ceramic, and composite materials at the nanoscale These new effects are attributed to quantum confinement of electrons and the resulting changes in the electronic transitions between energy levels as well as collective conduction-band electron plasma oscillations

The optical properties of thin films and nanomaterials are not only challenging

to measure but also difficult to model These properties are governed by the dielectric function which consists of a real and an imaginary part For structures

with dimensions larger than 10 nm, electrons still behave as do particles and the classical size effect based on Drude’s model is applicable For these length scales, the dielectric function needs to be corrected for electron scattering As the dimen-sions of the thin film or particle near 10 nm, electrons behave more as waves and Drude’s model is only applicable after introducing some energy range modifica-tions that affect the dielectric function

Various models (Wood 1982 ; Kawabata and Kubo 19 66 ; Genzel et al 1975 ; Cocchini 1985 ; Bassani et al 1985) built on different assumptions have been devised

to capture the absorption behavior of nanoparticles and thin films at a range of quencies with different degrees of success Although many of these models predict the correct trends for the absorption spectra, they often disagree numerically with experimental data and diverge as the characteristic dimensions approach bulk length scales (Huang and Lue 1994) In his publication, Huang derived an expression for the real and imaginary parts of the dielectric function for small metallic particles using Lindhard’s equation Huang demonstrated with his model very good agree-ment with experimental data for structures with sizes confined to less than 10 nm The indication is that absorption spectra of metallic structures in the nanoscale shift towards the blue part of the spectrum and absorption peaks tend to broaden These findings can be accounted for and taken advantage of when designing optical com-ponents with dimensions confined to the nanometer range The optical properties of metal nanoparticles and more specifically noble metal nanoparticles will remain a continuous subject of research (Scarrafrdi et al., 2005, Scaffardi and Tocho, 2006) because of the potential applications in many fields such as spectrally selective coat-ings, nonlinear optics, and heterogeneous catalysis

Summary

This introductory chapter summarized the general design guidelines of a sensor regardless of the analyte the sensor is designed to measure, the environment the sen-sor is designed to operate in, or the application In real-life applications, all of these parameters significantly affect the design from material selection to the selection of the transduction scheme, the geometry of the device, and the package design

It is well established that nano- and microstructured sensing materials provide unique and novel functions unattainable using bulk materials More specifically, micro- and nanostructured transducers promise to be more sensitive, more selec-tive, and faster responding, but this comes at a cost At these scales, new design considerations need to be taken into account including difficulty of fabrication, integration with macroscale structures, sensitivity to environmental conditions, stability (thermal and chemical), and long-term reliability These issues are subjects

of numerous research efforts making great inroads towards bringing these materials

to mainstream everyday-life sensing devices The next chapters delve into the ous aspects and details of implementing thin films and nanostructures into sensors and the challenges and benefits of this endeavor

in a dielectric medium Surf Sci., 156:851–858

Coombes CJ (1972) Melting of small particles of lead and indium J Phys F: Met Phys ,

2 : 441 – 449

Dingreville R, Qu J, Cherkaoui M (2004) Effective elastic modulus of nano-particles, Proc 9th Int’l Symp on Adv Packaging Mat., pp 187–192

Feynman RP (1959) Plenty of room at the bottom, APS meeting

Genzel L , Martin TP , Kreibig U (1975) , Dielectric function and plasma resonances of small metal particles , Zeitschrift fur Physik B , 21 : 339 – 346

Huang WC , Lue JT (1994) Quantum size effect on the optical properties of small metallic cles Phys Rev B , 49 (24) : 279 – 285

Jin XC , Degertekin FL , Calmes S , Zhang XJ , Ladabaum I , Khuri-Yakub BT (1998) Micromachined capacitive transducer arrays for medical ultrasoundimaging , Proc Ultrasonics Symp ,

Pozhar LA (2000) Structure and dynamics of nanofluids: Theory and simulations to calculate viscosity Phys Rev E , 61 (2) : 1432 – 1446

Scaffardi LB , Tocho JO (2006) Size dependence of refractive index of gold nanoparticles Nanotechnology , 17 : 1309 – 1315

Scaffardi LB , Pellegri N , de Sanctis O , Tocho JO (2005) Sizing gold nanoparticles by optical extinction spectroscopy Nanotechnology , 16 : 158 – 163

Srikar VT , Spearing SM (2003) Materials selection in micromechanical design: An application of the Ashby approach J MEMS , 12 1 : 3 – 10

Truong V , Courteau P (1987) Optical properties of very fine Al particles: Quantum size effect

Zhang Z , Li JC , Jiang Q (2000) Modelling for size-dependent and dimension-dependent melting

of nanocrystals J Phys D: Appl Phys , 33 : 2653 – 2656

Transduction Principles

Jeffrey Fortin

used in microsensors Each section provides an overview of the theory and then gives an example of a sensor that uses the transduction principle being described A classification of measurands is presented as well as the most common transduction techniques including piezoresistance, piezoelectricity, capacitive, resistive, tunneling, thermoelectricity, optical and radiation-based techniques, and electrochemical

Introduction

This chapter presents the most common fundamental transduction principles used

in microsensors Each section provides an overview of the theory and then gives an example of a sensor that uses the transduction principle being described

Wikipedia defines a transducer as follows

A transducer is a device, usually electrical, electronic, or electro-mechanical, that converts one type of energy to another for the purpose of measurement or information transfer Most transducers are either sensors or actuators In a broader sense, a transducer is sometimes defined as any device that senses or converts a signal from one form to another ( www.Wikipedia.com )

In a similar definition a transducer is defined as a device providing a usable output

in response to a specific measurand, where the measurand is defined to be the cal quantity, property, or condition that is to be measured (Norton 1982) It is further stated here that when one is designing a sensor or trying to choose the appropriate transduction technique there are a few questions one can ask, including: What is the measurand? What is the principle of transduction? What is the sensing element? What are the limits of the measurand to which the transducer will need to respond?

physi-J Fortin

GE Global Research Center 1 Research Circle, KW C314 , Niskayuna , NY 12309

e-mail: fortinje@research.ge.com

A Zribi and J Fortin (eds.), Functional Thin Films and Nanostructures for Sensors, 17

Integrated Analytical Systems,

DOI: 10.1007/978-0-387-68609-7_2, © Springer Science + Business Media, LLC 2009

White has presented a classification scheme for measurands or properties one may be interested in measuring The main categories and significant subcategories are given in Table 2.1 (White 1987)

Transducers are typically designed to sense a specific measurand and to ideally respond only to that particular measurand In reality a transducer will most likely respond to the measurand in question and will also respond to other energy sources that act on the sensor that are not of interest These are considered sources of noise, for example, when measuring strain with a piezoresistor the measurand of interest

is strain, however, the resistance will also change with temperature

Table 2.2 shows the most common transducer types used to quantify the major categories of measurand The sections below provide an overview of the common transduction mechanisms and principles such that one can begin to apply this knowledge to sensor design

Table 2.1 Classification of Measurands

Measurand Property of Interest

Acoustic Wave amplitude, phase, polarization Wave velocity Spectrum Biological Identity, concentration, state

Chemical Identity, concentration, state

Electrical Current, charge potential, potential difference Field (amplitude,

phase, polarization) Conductivity and permittivity Magnetic Field (amplitude, phase, polarization) Flux Permeability

Mechanical Position, velocity, acceleration Force Stress, strain Mass,

den-sity Flow Moment, torque Stiffness, compliance Viscoden-sity Crystallinity

Optical and Radiation Wave amplitude, phase, polarization Spectrum Velocity Energy Thermal Temperature Flux Specific heat Thermal conductivity

Table 2.2 Transduction Techniques for Common Measurands

Measurand

Most Common Transduction Techniques Utilized to Quantify the Measurand

Acoustic Piezoelectric Piezoresistive Capacitive Optical

Biological Piezoelectric Piezoresistive Electrical Optical

Chemical Piezoelectric Piezoresistive Electrochemical Electrical Optical Electrical Electrical Optical

Magnetic Piezoresistive Piezoelectric Electrical – capacitive, tunneling, Optical Mechanical Piezoelectric Piezoresistive Capacitive Optical

Optical and Radiation Thermoelectric (Seebeck) Photosensitivity (photovoltaic,

photoelec-tric, photoconductors, photodiodes, and phototransistors) Thermal Thermoelectric Photosensitivity Electric – resistive

Piezoresistivity

Piezoresistivity is, in its most basic form, the change in a material’s resistance resulting from a change in stress in the material The word piezo is derived from

the Greek word piezein , which means to press or squeeze Many materials exhibit

the piezoresistive effect and it is typically quantified by what is termed the gauge factor The gauge factor is the change in resistance per given strain per starting resistance and can be described via the following equation,

The gauge factor for silicon decreases with increasing impurity concentrations and this can be predicted by model Controlled doping is typically accomplished via ion implantation of the specific doping ion into the silicon to define the piezoresistor

An alternative technique is to deposit a film containing the doping ion over a terned Si surface and then use a temperature treatment to drive the dopant into the depth of the Si

The piezoresistive gauge factor decreases as temperature increases, and this can also be predicted The coefficients increase linearly with the inverse of temperature One can find a deep description of the mathematics in Sze (1994)

When deciding if piezoresistance can be used as a transduction measurand for a particular measurand one only has to determine if the sensor can be designed such that the measurand can produce a stress on a portion of the device where a piezore-sistor can be located Examples of measurands that are quantified via piezoresist-ance include: pressure, vibration, acceleration, and magnetic field Once it has been decided that the measurand could be quantified via piezoresistance one must deter-mine if it is the best approach that will meet all the specifications of the application,

as described in Chapter 1

A great example of a transducer that uses the piezoresistive effect is a based MEMS pressure sensor Silicon-based pressure sensors have been around since the late 1950s and they are a very mature technology GE Sensing offers a multitude of Si-based pressure sensors for a variety of applications including blood pressure sensing, tire pressure sensing, industrial process measurement, and so on

silicon-An overview of one pressure sensor is presented here that is designed to measure tire pressure ( www.GESensing.com )

This sensor is approximately 1 by 1 mm and is an absolute pressure sensor, meaning the reference cavity is a vacuum and, once calibrated, the sensor gives the absolute pressure inside the tire As with most pressure sensors, the electronic readout technique utilizes a Wheatstone bridge

The sensor element is a thin silicon membrane with embedded piezoresistors The piezoresistors are formed via an implantation step and the proper doping level

is chosen to provide the highest gauge factor The piezoresistors are positioned in the areas of the membrane that see the highest strain due to the force of the pressure

Many membranes are formed on a silicon wafer using typical microfabrication processes and then this wafer is wafer-bonded in a vacuum environment to a bottom silicon wafer with cavities lined up with the membrane This forms a drumlike structure with a vacuum cavity lying below a thin membrane As the external pres-sure fluctuates, the membrane moves to balance the force between the external

Piezoelectricity

Overview of Theory Piezoelectricity, as is piezoresistivity, is an electrical effect

caused by a change in the strain of a material In the cause of piezoelectricity, when

a piezoelectric material is stressed (compressive or tensile) a charge is induced across the material’s faces in response to the magnitude and direction of the strain

Fig 2.1 Side view of GE Sensing silicon pressure sensor (Courtesy of GE Sensing.)

Reference Cavity Diaphragm

Piezo - Resistor Wire - Bond Pad

Fig 2.2 Side view representing diaphragm deflection of GE Sensing silicon pressure sensor (Courtesy of GE Sensing.)

P2 Reference CavityPressure

P1 AppliedPressure

A piezoelectric transducer therefore converts a change in a measurand into a change

and quartz Typical thin film-based piezoelectric materialso include ZnO, AlN, and

for sensing in applications such as accelerometers, vibrometers, ultrasound, and high dynamic range ac pressure sensing

In microsensors, piezoelectric materials can be either directly deposited onto the device or they may be integrated into the device, for example, lamination of a piezo-electric polymer film Because PZT has an order of magnitude higher piezoelectric effect than ZnO and AlN many techniques have been developed to integrate PZT with

a microdevice, including sol–gel and sputtering ZnO and AlN films have also been integrated into microdevices for the purpose of transduction (Royer et al 1983 ; Ried

et al 1993 ; Ko et al 2003 ) The definition of methods and measurement of tric crystal units is reported in detail in Halfner (1969)

An example microdevice utilizing a piezoelectric transduction technique was recently presented at the 2006 Transducers Conference at Hilton Head (Horowitz

et al 2006 ) The device presented was a micromachined piezoelectric microphone for aeroacoustics applications Although there had been previous research done on MEMS-based microphones most of them have been developed for audio applica-tions The microphone reported at 2006 Hilton Head was designed for high sound pressure applications in excess of 160 dB with a bandwidth of >50 kHz

The microphone was fabricated by combining a sol–gel PZT (lead zirconate–titanate) deposition process on a silicon-on-insulator wafer The PZT was deposited onto a 1.80 mm diameter 3 m m thick Si diaphragm The PZT was processed and lithographically defined to an annular ring near the diaphragm edge to maximize the sensitivity The PZT layer was 270 nm thick and was placed between two thin metal electrodes A diffusion barrier separated the PZT from the silicon diaphragm

As acoustic energy impinged upon the diaphragm it moved As it moved, the

piezo-electric material experienced stress in the z -axis and the charge at its surfaces

changed in response to this stress and this charge was measured

Fig 2.3 ( a ) Piezoelectric material with no externally applied strain has no net charge on its surfaces ( b ) The same material under strain produces charges at surface that can be measured

Test results for the microphone showed a sensitivity of 0.75 uV/Pa with a linear dynamic range from 47.8 to 169 dB and a resonant frequency of 50.8 kHz

Electrical—Resistance, Capacitance, Impedance, Tunneling

There are many ways to convert a change in an external measurand into a change

in an electrical signal directly, particularly in a MEMS device, where there are moving parts that can be accessed electrically or integrated thin films that often have electrical properties based on the environment to which they are exposed

Resistance: One key transduction technique that converts the measurand to a

change in resistance is piezoresistance, which, because of its widespread use in microdevices was covered separately above Other resistance-based transduction techniques also rely on the measurand interacting with a film or bulk structure and hence changing an electrical property For example, certain polymers have a moisture-

or gas-sensitive resistivity Another example is in the measurement of temperature, where because resistance of a material is a function of temperature it can be directly used to measure temperature; this type of device is referred to as a thermistor

An example of a device that uses a materials change in resistance due to sure to the measurand can be seen in the work of Valentini et al (2004 ) This device uses carbon nanotubes (CNTs) as the functional transducer material Carbon nano-tubes present extremely high surface-to-volume ratios and have recently seen significant attention for their gas adsorption properties (Treacy 1996) Valentini et al used an interdigital electrode structure made from platinum deposited and patterned on top of a silicon nitride film The CNTs were then grown from a catalyst between the Pt electrodes to heights of approximately 200 nm The results showed that the

Capacitance: In this transduction technique the measurand interacts with the

device to change the capacitance value of a capacitor This change can be induced

by changing the effective distance between the two plates or electrodes of the capacitor or by changing the dielectric constant of the insulator material Examples

of both are given below

The capacitive transduction technique that can be used to measure pressure is capacitance In a typical silicon-based capacitive pressure sensor the design is similar

to a piezoresistive pressure sensor as described above Instead of implanting zoresistors into the diaphragm, the diaphragm itself is used as one plate of a capaci-tor Alternatively a metal layer can be placed on or embedded in the diaphragm The second plate or electrode is located at the bottom of the gap (see Fig 2.5 ) In the sensor shown in Fig 2.4 the substrate is a degenerately doped silicon wafer and the membrane has been wafer-bonded under vacuum to a patterned oxide layer Contact

pie-is made to the substrate through an opening in the oxide and directly to the also

degenerately doped silicon membrane A capacitor with an oxide dielectric as a

As the external pressure changes the diaphragm moves and the distance between the diaphragm and the lower electrode changes, thus changing the capacitance of the device One advantage of a capacitive-based pressure sensor over a piezoresis-tive approach is the lower power consumption of the sensor itself Another advan-tage is the capacitive approach tends to have higher sensitivity if properly designed (Eaton and Smith 1997 )

The G-Cap Moisture Sensor offered by GE Sensing is an example of a device that utilizes the change in capacitance of a thin film in response to the measur-and In this case the functional material is thin polymeric film sandwiched between two thin patterned electrodes The polymeric film was developed to allow for measurement of a wide range of relative humidity from 0 to 100% and

it can survive total immersion in water without loss of accuracy The typical capacitance of the sensor is in the range of 140–190 pF and it changes linearly with %RH The capacitance is measured between 1 kHz and 1 MHz (Fig 2.5 ) The sensitivity to changes of temperature can be calibrated and is typically less than 0.05% RH/°F

Fig 2.4 Capacitive-based pressure sensor

Pressure

Ground

Si Wafer

Reference Capacitor

0 10 20 30 40 50 60 70 80 90 100

G-CAP2, calibrated, 10 kHz, 1 VAC, 25⬚C (77⬚F)

%RH

Impedance: This is very similar in nature to the resistance or capacitance

tech-nique In impedance-based transduction the AC impedance of a component of the device is measured and monitored Often there is a material that is used to measure

an environmental parameter, such as humidity or local magnetic field

One impedance-based transduction method that has received a lot of tion in recent years is giant magneto impedance (GMI) This effect refers to a material’s very large change in resistance due to an applied magnetic field This effect has been seen in amorphous wire, ribbons, and thin films An analogous effect is called giant stress impedance (GSI), where a change in stress causes a large change in impedance The GMI effect is seen in many high-permeability materials such as amorphous soft magnetic wires with Fe-based or Co-based compositions (Zribi et al 2005 ; Han et al 2005 ; Garcia et al 2005 ) This effect depends on the material’s permeability which is a function of many factors including domain configurations, material geometry, anisotropy, and excitation frequency The GMI effect can be expressed as an impedance ratio given by Equation (2.2),

D Z Z/ =[ ( )Z H −Z H( max)] / (Z Hmax), (2.2)

fashion where the magnetic field is replaced with stress

Han et al (2005) provide a nice overview of sensors utilizing the GMI and GSI effect The applications of these sensors include magnetic field measurement, a position measurement for the location of a catheter in the human body, nondestruc-tive testing, and electronic surveillance Zribi et al (2005) also present an oil-free stress impedance pressure sensor for harsh environments

Tunneling: Another interesting technique employed in micro- and

nanosys-tems for transduction of a measurand value into an electrical signal is electrical tunneling Typically what is done in this case is a mechanical component becomes one electrode in a two-electrode circuit This mechanical component will have a tip or surface placed within a few to tens of nanometers from the second stationary electrode The mechanical component with the tip will move

in response to the measurand As it moves, the distance between the tip and the stationary electrode will change and hence the tunneling current will change This technique can be used to measure very tiny changes in the measurand if properly designed

A number of publications have addressed the design, fabrication, and

MEMS-based high-precision, wide-bandwidth micromachined tunneling ometer with a resolution of 20 ng/sqrt Hz and 5 Hz–1.5 kHz bandwidth The design consists of a cantilever tip substrate, a proof mass, substrate, and a cap substrate all wafer-bonded together The accelerometer is operated at a pressure of 10 mTorr to

acceler-reduce thermomechanical noise and increase Q to above 100 A feedback controller

is used to maintain the tunneling gap at 10 Å

Thermoelectricity

Overview of Theory

Thermoelectricity: This transduction technique converts the value of the measurand

into a voltage (or electromotive force) generated by the potential difference between the junctions of two selected dissimilar materials due to the Seebeck effect The Seebeck effect is well known and is the technology employed in thermocouples, where, as the temperature of the junction changes, the voltage across the junction changes Thus, a thermocouple works by measuring the difference in potential caused by the dissimilar wires Several thermocouples in series are called a thermo-pile This technique is also used in silicon-based devices to measure temperature using a noncontact approach The Seebeck effect can be described by Equation (2.3) referring to Fig 2.6 :

V =(S B−S A) / (T2−T1), (2.3)

A good example of a microsensor utilizing the Seebeck effect for transduction

is the GE silicon-based IR thermopile ( www.GESensing.com ) The IR thermopile sensor consists of a number (about 40) of thermocouple pairs connected in series and covered with a high emissivity coating The hot junctions are thermally isolated from the cold junctions and are exposed to the incident IR radiation The cold junc-tions are attached to a heat sink GE’s device is built on a silicon wafer and the thermocouple radiation detection junctions are placed on a thin, low thermal mass diaphragm, and the reference junctions are placed off the membrane on the thick silicon wafer (Figs 2.7 and 2.8 ) A thermistor is placed in the finished package as

The IR thermopile device allows for measurement of temperature without direct contact with fast, millisecond response times due to the low thermal mass of the diaphragm Applications include tympanic thermometers for body temperature measurement, food temperature measurement in microwave ovens, measuring temperatures inside vehicles, and many others

Fig 2.6 Thermocouple consisting of wire of material A connected to wire of material B

V

Wire A

Wire B

T1 T2