Thesis for the degree of Doctor of Technology Sundsvall 2013
Antenna-Based Passive UHF RFID Sensor Tags
- Design and Application
Jinlan Gao
Supervisors: Professor Hans-Erik Nilsson Doctor Johan Sid ´en
Department of Electronics Design
Mid Sweden University, SE-851 70 Sundsvall, Sweden
ISSN 1652-893X
Mid Sweden University Doctoral Thesis 157
ISBN 978-91-87103-99-5
Akademisk avhandling som med tillst˚and av Mittuniversitetet i Sundsvall framl¨aggs till offentlig granskning f ¨or avl¨aggande av teknologie doktorsexamen, tisdagen den 18 juni 2013, klockan 13:15 i sal O102, Mittuniversitetet Sundsvall.
Seminariet kommer att h˚allas p˚a engelska.
Antenna-Based Passive UHF RFID Sensor Tags - Design and Application
Jinlan Gao
⃝Jinlan Gao, 2013c
Department of Electronics Design
Mid Sweden University, SE-851 70 Sundsvall, Sweden
Telephone: +46 (0)60 148835
Printed by Kopieringen Mittuniversitetet, Sundsvall, Sweden, 2013
Abstract
RFID, as a low cost technology with a long life time, provides great potential for transmitting sensor data in combination with the ordinary ID number. The sensor can, for example, be integrated either in the chip or in the antenna of an RFID tag.
This thesis focuses on the design of antenna-based UHF RFID sensor tags as wireless sensors at the lowest possible cost level compatible with standard communication systems in logistics. The applications of the sensor tags, in this work, mainly target remote humidity sensing.
Antenna-based sensory UHF RFID tags utilize the influence that the physical or chemical parameters to be sensed have on the electrical properties of a tag antenna.
The variations of the electrical properties of the tag antenna can be measured in many ways. In the thesis, a description is provided as to how these variations are normally measured by an RFID reader without any other assistant equipment.
Three structures of antenna-based RFID sensor tags are presented with detailed characterizations. The first one utilizes the sensitivity of the antenna to the surround- ing environment to construct RFID sensor tags, where a moisture absorbing layer providing wetness/humidity sensor functionality is placed on the RFID tag antenna to increase the humidity concentration surrounding the tag antenna and the thesis describes how to overcome certain limitations due to disturbances associated with background materials. The second structure directly integrates a small resistive sen- sor element into an RFID tag antenna and the sensor information can thus modulate the antenna performance by means of galvanic contact. The third structure embeds a small resistive sensor element into a loop which is positioned on top of the tag antenna and the sensor information can thus modulate the performance of the tag antenna by means of electromagnetic coupling. Both theoretical analysis and full wave simulations are presented for the latter two sensor tag structures in order to characterize the performance of the sensor tags.
An ultra-low cost printed humidity sensor with memory functionality is also designed and thoroughly characterized for integration into RFID tag antennas by means of galvanic contact or electromagnetic coupling. The sensor is a 1-bit write- once-read-many (WORM) memory printed using conductive ink. The WORM works as a pure resistive humidity sensor and can provide information about an historical event. The WORM sensor is presented by introducing its geometry, characterizing its behavior in humidity and explaining the principle of the humidity effect. The WORM sensors are also integrated into the RFID tags by means of both galvanic contact and electromagnetic coupling in order to experimentally verify the two con- cepts.
To lower the cost of the RFID tags, the antennas are normally printed, milled or etched on flexible substrates using low-cost high-speed manufacturing methods which in some cases cause a high degree of edge roughness. The edge roughness will affect the behavior of the antenna, however, the characteristics of edge roughness on
iii
iv
RFID antennas have previously not received any significant attention. Unforeseen antenna behavior can affect the antenna-based sensor tags, thus the influence of edge roughness is also investigated in the thesis.
Sammandrag
RFID, som en l˚agkostnadsteknik med en l˚ang livsl¨angd, tillhandah˚aller en stor potential f ¨or ¨overf ¨oring av sensordata i kombination med det ordinarie ID-numret.
Sensorn kan till exempel integreras i RFID-chippet eller i en RFID-taggs antenn.
Denna avhandling fokuserar p˚a design av antenn-baserade sensortaggar f ¨or RFID p˚a UHF-bandet som tr˚adl ¨osa sensorer till l¨agsta m ¨ojliga kostnad, kompatibla med standard-kommunikationssystem inom logistik. Applikationerna f ¨or sensortaggar, i det h¨ar arbetet, riktar sig fr¨amst mot fj¨arravl¨asning av fukt.
Antennbaserade sensoriska UHF RFID-taggar utnyttjar den p˚averkan som de fysikaliska eller kemiska parametrar som skall avk¨annas har p˚a de elektriska egen- skaperna hos en taggs antenn. Variationerna hos de elektriska egenskaperna hos taggens antenn kan m¨atas p˚a flera vis. I avhandlingen ges en beskrivning av hur dessa variationer normalt m¨ats med en vanlig RFID-l¨asare utan n˚agon annan assis- terande utrustning.
Tre stycken strukturer av antennbaserade RFID-sensortaggar presenteras med en detaljerad karakterisering. Den f ¨orsta utnyttjar antennens k¨anslighet till den omgivande milj ¨on f ¨or att konstruera RFID-sensortaggar. I detta koncept tj¨anar ett fuktabsorberande material till att tillhandah˚alla sensorfunktionalitet f ¨or v¨ata/fukt och placeras p˚a RFID-antennen f ¨or att ¨oka koncentrationen av luftfuktigheten kring taggens antenn och avhandlingen beskriver hur man kan ¨overvinna vissa begr¨ans- ningar p˚a grund av os¨akerheter i samband med bakgrundsmaterial. Den andra strukturen integrerar ett litet resistivt sensorelementet direkt i en RFID-taggs an- tenn och sensorinformationen kan d¨armed modulera antennens prestanda genom galvanisk kontakt. Den tredje strukturen b¨addar in ett litet resistivt sensorelement i en loopstruktur som ¨ar placerad strax ovanf ¨or taggens antenn och sensorinforma- tionen kan d¨armed modulera antennens prestanda genom elektromagnetisk kop- pling. B˚ade teoretiska analyser och fullv˚ags-simuleringar presenteras f ¨or de tv˚a sistn¨amnda sensortagg-strukturerna i syfte att karakterisera sensortaggarnas pre- standa.
En fuktsensor som kan tryckas till v¨aldigt l˚ag kostnad samt har minnesfunktion- alitet ¨ar ocks˚a designad och grundligt karakteriserad f ¨or integrering i RFID taggars antenner via galvanisk kontakt eller elektromagnetisk koppling. Sensorn ¨ar ett 1-bits write-once-read-many (WORM) -minne tryckt med ledande bl¨ack baserat. WOR- Men fungerar som en rent resistiv fuktsensor och kan tillhandah˚alla information om en historisk h¨andelse. WORM-sensorn presenteras genom att introducera dess geometri, karakterisera dess beteende i fukt och f ¨orklara principen av fukteffekten.
WORM-sensorerna integreras ocks˚a i RFID-taggar med hj¨alp av b˚ade galvanisk kon- takt och elektromagnetisk koppling f ¨or att experimentellt verifiera de tv˚a koncepten.
F ¨or att s¨anka kostnaden f ¨or RFID-taggar ¨ar antennerna normalt tryckta, fr¨asta eller etsade p˚a flexibla substrat med hj¨alp av billiga metoder f ¨or h ¨oghastighetstillverk- ning som i vissa fall kan orsaka en h ¨og grad av kantoj¨amnhet. Kantoj¨amnhet p˚averkar
v
vi
en antenns egenskaper, men kantoj¨amnhet p˚a RFID-antenner har tidigare inte f˚att n˚agon st ¨orre uppm¨arksamhet. Of ¨orutsett antennbeteende kan p˚averka antennbaser- ade sensortaggar varf ¨or kantoj¨amnhet unders ¨oks ocks˚a i denna avhandling.
Acknowledgements
First I would like to thank my supervisors Dr. Johan Sid´en and Prof. Hans- Erik Nilsson for giving me the opportunity and employing me as a Ph.D. student.
Thanks to Johan for all the guidance, suggestions and ideas on my research work during these years, and for always believing me and supporting me. Without his patience and encouragement, this thesis would not have been possible. Also thanks to Hans-Erik for the supervision and consistent support on my research work.
Thanks to Henrik Andersson and Anatoliy Manuilskiy for all their help in my research work.
Thanks to Prof. Youzhi Xu for introducing me to this opportunity to start my Ph.D. study at Mid Sweden University.
I would also like to thank all my colleagues at the Electronics Design Division, Mid Sweden University for your kindness and friendliness.
Also many thanks shall be given to all my Chinese friends whom I met in Sundsvall for bringing happiness to my life.
Financial support from Mid Sweden University, the Swedish KK Foundation, the Research Council Formas and the EU FP7 Project PriMeBits is also gratefully ac- knowledged.
Last but not least, thanks to my families for their endless love and support. Spe- cial thanks to my dear husband for standing behind me in all that I do.
Sundsvall, May 2013
Jinlan Gao
vii
viii
Table of Contents
Abstract iii
Sammandrag v
Acknowledgements vii
Table of Contents ix
List of Papers xiii
List of Figures xv
Terminology xvii
1 Introduction 1
1.1 Main contributions . . . 3
1.2 Outline . . . 5
2 Acquisition of the Sensor Information for an Antenna-Based Sensor Tag 7 2.1 Measurable output variables . . . 7
2.2 Measurement setup . . . 10
3 UHF RFID Sensor Tags With Covering of Sensing Material 13 3.1 Experimental results . . . 14
3.1.1 Power-up differences due to water drops . . . 15
3.1.2 Power-up differences due to humidity . . . 15
3.2 Discussion . . . 16
4 Printed Sensor (WORM) 19 4.1 Design of Printed Sensor . . . 19
4.2 Temperature Sintering . . . 20
4.3 Humidity Sintering . . . 21
4.4 Combination Effect of Temperature and Humidity . . . 23
4.5 Solvent effect on the resistance of printed patterns . . . 24
4.6 Applications . . . 29
5 UHF RFID Tags with Passive Sensors Directly Integrated into the Antennas 31 5.1 Analysis Model . . . 31
5.2 Antenna Designs . . . 36
5.3 Experimental Results . . . 40
5.4 Discussion . . . 43 ix
x TABLE OF CONTENTS
6 UHF RFID Tags with Passive Sensors Electromagnetically Coupled to the
Antennas 45
6.1 Analysis model . . . 45
6.1.1 For Shunt/Series Inductors Matched Antenna . . . 46
6.1.2 For Inductively Coupled Antenna . . . 52
6.2 Design Considerations . . . 56
6.2.1 Different Types of Antennas . . . 56
6.2.2 Different Types of Chips . . . 57
6.2.3 Various Distance between coupling loop and tag antenna . . . . 57
6.2.4 Assistance of the additional embedded component . . . 59
6.3 Experimental validation . . . 61
6.4 Discussion . . . 63
7 Investigation on the effect of the edge roughness on the antenna 65 7.1 Method . . . 65
7.2 Results . . . 67
7.2.1 Input return loss . . . 67
7.2.2 Bandwidth . . . 69
7.2.3 Ohmic losses . . . 70
7.3 Discussion . . . 70
8 Summary of publications 71 8.1 Paper I . . . 71
8.2 Paper II . . . 71
8.3 Paper III . . . 71
8.4 Paper IV . . . 72
8.5 Paper V . . . 72
8.6 Paper VI . . . 72
8.7 Paper VII . . . 73
8.8 Paper VIII . . . 73
8.9 Paper IX . . . 73
8.10 Paper X . . . 74
8.11 Author’s contributions . . . 74
9 Thesis Summary 77
References 79
PAPER I 85
PAPER II 91
PAPER III 97
PAPER IV 111
PAPER V 117
TABLE OF CONTENTS xi
PAPER VI 127
PAPER VII 137
PAPER VIII 145
PAPER IX 153
PAPER X 159
xii
List of Papers
This thesis is mainly based on the following papers, herein referred by their Ro- man numerals:
Paper I Microstrip Antennas for Remote Moisture Sensing Using Passive RFID Johan Sid´en, Jinlan Gao, and Bj ¨orn Neubauer,
Proceeding of Asia Pacific Microwave Conference (APMC 2009), pages 2375–
2378, December 2009.
Paper II Printed Temperature Sensors for Passive RFID Tags Jinlan Gao, Johan Sid´en, and Hans-Erik Nilsson,
Proceeding of the 27th Progress in Electromagnetics Research Symposium (PIERS 2010), pages 845–848, March 2010.
Paper III Printed Humidity Sensor With Memory Functionality for Passive RFID Tags
Jinlan Gao, Johan Sid´en, Hans-Erik Nilsson, and Mikael Gulliksson, IEEE Sensors Journal, volume 13, pages 1824–1834, 2013.
Paper IV Printed Electromagnetic Coupler with Embedded Moisture Sensor for Ordinary Passive RFID Tags
Jinlan Gao, Johan Sid´en, and Hans-Erik Nilsson,
IEEE Electron Device Letters, volume 32, pages 1767–1769, 2011.
Paper V An Analytical Model for Electromagnetically Coupled UHF RFID Sen- sor Tags
Jinlan Gao, Johan Sid´en, and Hans-Erik Nilsson,
Proceeding of the 7th Annual IEEE International Conference on RFID, pages 66–
73, April 2013.
Paper VI Characterization of UHF RFID Sensor Tags with Electromagnetically Coupled Passive Sensors
Jinlan Gao, Johan Sid´en, and Hans-Erik Nilsson,
Accepted to be published in Proceeding of European Conference on Smart Ob- jects, Systems and Technologies (Smart-SysTech 2013).
Paper VII Electric and Electromagnetic Coupled Sensor Components for Passive RFID
Johan Sid´en, Jinlan Gao, Tomas Unander, Henrik Andersson, Peter Jons- son, Hans-Erik Nilsson, and Mikael Gulliksson,
Proceeding of IEEE International Conference on Microwaves, Communications, Antennas and Electronics Systems (COMCAS 2011), pages 1–5, November 2011.
xiii
xiv TABLE OF CONTENTS
Paper VIII Printable WORM and FRAM memories and their applications
Ari Alastalo, Tomi Mattila, Jaakko Lepp¨aniemi, Mika Suhonen, Terho Kololu- oma, Andreas Schaller, Henrik Andersson, Anatoliy Manuilskiy, Jinlan Gao, Hans-Erik Nilsson, Alexandru Rusu, Suat Ay ¨oz, Igor Stolichnov, Simo Siitonen, Mikael Gulliksson, Johan Sid´en, Tobias Lehnert, Jens Adam, Michael Veith, Alexey Merkulov, Yvonne Damaschek, J ¨urgen Steiger, Markus.
Cederberg and Miroslav Konecny,
Proc. of Large-area, Organic and Printed Electronics Convention, LOPE-C, pages 8 –12 , June 2010.
Paper IX Investigation of humidity sensor effect in silver nanoparticle ink sen- sors printed on paper
Henrik Andersson, Anatoliy Manuilskiy, Jinlan Gao, Cecilia Lidenmark, Johan Sid´en, Sven Forsberg, Tomas Unander, and Hans-Erik Nilsson, Manuscript submitted to IEEE Sensors Journal.
Paper X On the Influence of Edge Roughness in High-Speed RFID Antenna Manufacturing Processes
Jinlan Gao, Johan Sid´en, and Hans-Erik Nilsson, Accepted by PIERS 2013.
Related articles, but not included in the thesis:
Paper I The influence of paper coating content on room temperature sintering of silver nanoparticle ink
Henrik Andersson, Anatoliy Manuilskiy, Cecilia Lidenmark, Jinlan Gao, Tomas Ohlund, Sven Forsberg, Jonas Ortegren, Wolfgang Schmidt and Hans-Erik Nilsson,
Submitted to IOP Nanotechnology.
Paper II On the Impact of Edge Roughness to Narrowband and Wideband Flat Dipole Antennas
Johan Sid´en, Jinlan Gao, and Hans-Erik Nilsson,
In manuscript for the conference COMCAS 2013, submission during April 2013.
Paper III Home Care with NFC Sensors and a Smart Phone
Johan Sid´en, Vincent Skerved, Jinlan Gao, Stefan Forsstr ¨om, Hans-Erik Nilsson, Theo Kanter, and Mikael Gulliksson,
Proceedings of the 4th International Symposium on Applied Sciences in Biomed- ical and Communication Technologies (ISABEL 2011), pages 150:1–150:5, Oc- tober 2011..
List of Figures
1.1 RFID sensor tags . . . 2
2.1 Measurement setup . . . 11
3.1 RFID sensor tags with microstrip antennas . . . 14
3.2 Results of direct appliance of water drops on RFID sensor tags with microstrip antennas . . . 16
3.3 Results of placing RFID sensor tags with microstrip antennas in cli- mate chamber . . . 16
4.1 Printed WORM structures . . . 20
4.2 Temperature sintering of the WORM . . . 21
4.3 Sintering effect of water or humidity on the WORM . . . 22
4.4 Humidity sintering at various RH level . . . 23
4.5 Combined sintering effect of temperature and humidity . . . 24
4.6 Test pattern for investigating the solvent effect . . . 25
4.7 Results of testing solvent effect . . . 25
4.8 Test pattern for investigating the solvent spreading . . . 26
4.9 Results of testing solvent spreading . . . 27
4.10 Results of testing the patterns printed on pre-treated substrate . . . 28
5.1 Schematic diagram of the galvanically coupled sensor tags . . . 31
5.2 Galvanically coupled sensor tags and their electrical equivalents. . . . 32
5.3 Calculated power transfer coefficient vs. sensor resistance . . . 34
5.4 Calculated radiation efficiency vs. sensor resistance . . . 35
5.5 Calculated relative required minimum transmit power vs. sensor re- sistance . . . 36
5.6 Antenna designs . . . 37
5.7 Simulated power transfer coefficient vs. sensor resistance . . . 38
5.8 Simulated radiation efficiency vs. sensor resistance . . . 39
5.9 Simulated directivity vs. sensor resistance . . . 39
5.10 Simulated relative gain vs. sensor resistance . . . 40
5.11 Simulated relative required minimum transmit Power vs. sensor re- sistance . . . 40
5.12 Measured relative required minimum transmit power vs. sensor re- sistance . . . 42
5.13 Measured relative required minimum transmit power using WORM sensor in 80% RH . . . 43
6.1 EM coupled sensor tag . . . 45 xv
xvi LIST OF FIGURES
6.2 RFID tag antenna with shunt/series inductors matching network . . . 46
6.3 Equivalent circuit model of the RFID tag antenna with shunt/series inductors matching network . . . 46
6.4 Comparison between the equivalent circuit model and full wave sim- ulation for the RFID sensor tag shown in Fig. 6.2(b) . . . 51
6.5 Antenna impedance and radiation efficiency for various distance d . . 52
6.6 Antenna impedance vs. sensor capacitance CSor sensor inductance LS 52 6.7 RFID tag antenna with inductively coupled feed loop . . . 53
6.8 Equivalent circuit model of the RFID tag antenna with inductively coupled feed loop . . . 53
6.9 Antenna impedance and radiation efficiency for various distance d . . 56
6.10 Performance of the RFID sensor tags with different types of antennas . 57 6.11 Performance of the RFID sensor tags with different types of chips . . . 58
6.12 Performance of the RFID sensor tags with various chip impedances . . 58
6.13 Performance of the RFID sensor tags with various coupling coefficient 59 6.14 Relative required minimum transmit power vs. sensor impedance ZS, when the sensor impedance ZSin the model is represented by a resis- tance RSand a series capacitance CS, i.e. ZS = RS+ 1/(jωCS). . . 60
6.15 Relative required minimum transmit power vs. sensor resistance RS. The blue solid line is for when the sensor is pure resistive and the red dashed line is for when a series capacitor (C = 2.5 pF) is inserted together with the sensor. The results are calculated by setting ZS = RS+ 1/(jωC) in the model. . . 60
6.16 Simulated performance of RFID sensor tag with Rafsec tag . . . 62
6.17 Experimental results with Rafsec tag . . . 62
7.1 Samples of the conductor layer with edge roughness . . . 66
7.2 Edge roughness model . . . 66
7.3 The RFID antenna used in the experiments before receiving rough edges 67 7.4 Input return loss for different levels of edge roughness . . . 68
7.5 Bandwidth for different levels of edge roughness . . . 69
7.6 Radiation efficiency due to ohmic losses for the antennas with differ- ent levels of edge roughness . . . 70
Terminology
Abbreviations and Acronyms
ADC Analogue-Digital Converter
DI Deionized
GUI Graphical User Interface
HF High Frequency
IC Integrated Circuit
LF Low Frequency
RFID Radio Frequency Identification
RH Relative Humidity
UHF Ultra High Frequency
Mathematical Notations
The mathematical notations in the analysis on the RFID sensor tags are based on the following notations and the different superscriptions and subscriptions associated with them are used to distinguish the owners of the variables represented by them (detailed notations are introduced in the corresponding chapters of the thesis):
Ae the effective aperture of the antenna
Asc the radar scattering cross section of the antenna
C the capacitance
G the gain of the antenna
I the current
k the coupling coefficient between two inductors
L the self inductance
M the mutual inductance
P the power
Pbs the power backscattered by the antenna PC the sensitivity of the tag chip
Ploss the power dissipated as heat in the antenna Prad the power dissipated as radiation in the antenna
Pr the received power
Pt the transmitted power
Pt,readermin the required minimum transmit power from the reader
r the reading distance between the reader antenna and the tag an- tenna
xvii
xviii LIST OF FIGURES
rmax the maximum reading distance between the reader antenna and the tag antenna
RadEff the antenna radiation efficiency
R the resistance
RA the resistance part of the tag antenna impedance RC the resistance part of the tag chip impedance RS the resistance part of the sensor impedance Rl the antenna loss resistance
Rr the antenna radiation resistance
U the voltage
XA the reactance part of the tag antenna impedance XC the reactance part of the tag chip impedance XS the reactance part of the sensor impedance
Z the impedance
ZA the impedance of the tag antenna ZC the impedance of the tag chip ZS the impedance of the sensor
λ the wavelength of electromagnetic waves at the targeted frequency Γ the voltage reflection coefficient
τ the power transfer coefficient
The mathematical notations in the investigation on the effect of the edge roughness:
p the period length of the triangular wave r the amplitude of the triangular wave
α the major angle formed by the triangular wave and the ideal edge of the antenna
β the minor angle formed by the triangular wave and the ideal edge of the antenna
Chapter 1
Introduction
For many reasons, Radio Frequency Identification (RFID) is one of the best can- didate technologies for remote wireless identification. Passive RFID tags have been successfully developed in industry as a replacement for or as a complement to the traditional barcodes. In order for RFID to fully replace barcodes there is an obvious requirement to produce the tag electronics at an extremely low cost and this is the reason why RFID has been the driving force for research and development projects targeting printed electronics. Reviews of the status of printed electronics are, for ex- ample, to be found in [1] and [2] in which the progress and the challenges for printing passive as well as active components and displays are discussed. Even though sig- nificant progress has been achieved in printed electronics during the last few years, there are still significant challenges to be overcome before enabling commercializa- tion of the fully printed RFID tags. This is particularly true for tags operating at UHF frequencies. Until the fully printed RFID has been achieved, printed electronics will have the potential to play an important role in hybrid solutions where printed elec- tronics are combined with traditional silicon-based electronics, such as a tag with a printed antenna and a silicon-based chip [3, 4].
RFID, as a low cost technology with a long life time, provides great potential for transmitting sensor data in combination with the ordinary ID number. To add a sensing functionality into RFID tags, there are several approaches. Chip-based sensor tag solution and antenna-based solution are two widely used approaches.
As shown in Fig. 1.1, chip-based solution is to design specific RFID chips with sen- sor modules [5–8] and the antenna-based solution is to add sensor functionality to RFID tag antennas [9–16]. An RFID sensor chip has a sensor module that transforms analog sensor information to digital numbers and communicates sensor data via an integrated finite state machine or microcontroller. RFID sensor chips are commonly powered by a battery or an energy harvesting module. An antenna-based RFID sen- sor tag integrates a sensor into the tag antenna in order to make the antenna’s elec- trical properties as a function of the sensor information and the RFID reader detects the sensor information by evaluating the tag’s communication performance. Both approaches have their advantages and drawbacks. For instance, the former offers a good compatibility with traditional sensors and the latter presents a longer lifetime and lower costs. The selection of the sensor tag type depends on the applications.
The antenna-based RFID sensor tags are usually application-oriented designs which are highly dependent on the specific applications and have not yet become available in the commercial market. While, there are several RFID sensor chips which have been introduced to the market, such as the HF RFID sensor chip MLX90129 from Melexis Microelectronic Systems [17] and the UHF RFID sensor chip SL900A
1
2 Introduction
(a) (b)
Figure 1.1: (a) Chip-based RFID sensor tag. (b) Antenna-based RFID sensor tag.
from IDS Microchip AG [18]. However, this is just the first step for sensor-enabled RFID tags to become widely used products. Enabling sensor functionality for RFID is still an attractive topic for research.
RFID tags with humidity sensors can provide a sensing solution for applications, for example, humidity surveillance within a construction structure where a high hu- midity condition can increase the risk of microbial growth in building materials [19].
Smart packaging of goods is another potential application. RFID sensor tags can add surveillance functions to ensure that, for example, food quality is preserved [20–23].
Smart packing solutions based on RFID technology with sensor functionality will provide transparency and traceability for packaging logistics, where the major ben- eficiaries are the stakeholders along the entire supply chain [21].
RFID sensor tags with humidity sensing functionality commonly work in the low frequency (LF) or high frequency (HF) band. Humidity sensing material is added to the coil antenna of the tag in order to modulate the resonant frequency of the tag as a function of the humidity level [23]. Such a humidity sensing tag has a low cost and a long lifetime, but its reading distance is limited at the level of 10 cm. To obtain humidity sensing over longer distances than a few centimeter, ultra high frequency (UHF) tags should be used. A few UHF RFID humidity sensing solutions have been proposed during recent years, however, they are mainly chip-based solutions and involve relatively high costs and short lifetimes since they require the assistance of extra chips and batteries.
In this thesis work we address the low-cost, long-lifetime, fully-passive antenna- based solutions for the integration of sensor functionality to standard UHF RFID technology. These solutions are potentially suitable for many different kinds of sen- sors and the work in this thesis mainly focuses on characterizing humidity sens- ing. Three structures of antenna-based passive UHF RFID sensor tags are presented in this thesis and they are hybrid solutions for RFID sensor systems utilizing both printed electronics and traditional silicon-based electronics.
Low-cost antennas naturally involve high-speed manufacturing processes which in turn can create high mechanical tolerances. For RFID antennas, mechanical man- ufacturing tolerances imply uncertainties in the antennas outer dimensions as well as in the surface- and line-roughness of the antennas geometry, which leads to un- certainties in antenna properties such as resonant frequency and input impedance.
Unforeseen antenna behavior can affect the measurements of the antenna-based sen-
1.1 Main contributions 3
sor tags, thus the tag antennas should be designed to be electromagnetically robust against edge roughness that might be introduced to the antennas during manufac- ture. The thesis’ last chapter shows an initial investigation on the effect of the edge roughness. It shows how rough patterns along the edges of an antenna structure can affect UHF RFID tags’ communication capabilities and the results indicate that an antenna with a wider line width is more robust and can stand a higher degree of edge roughness.
1.1 Main contributions
The main scientific contributions of this thesis are:
• An RFID sensor label utilizing moisture absorbing material to provide humidity sensor functionality to the tag antenna is presented, where the tags are based upon microstrip antennas in order to suppress the influence of background materials.
The sensitivity of an antenna to the surrounding environment can be utilized to construct RFID sensor tags. A moisture absorbing layer placed on an RFID tag antenna can increase the humidity concentration surrounding the tag antenna and then provide wetness/humidity sensor functionality to the tag antenna.
This thesis provides a solution to overcome certain limitations due to distur- bances associated with the background materials, in which microstrip anten- nas are used rather than dipole-based antennas in order to shield the influence of the background materials. Such an antenna-based sensor tag can be used for low-cost in-situ humidity surveillance.
• A concept of constructing RFID sensor tags is demonstrated, where a surface mount- able resistive sensor element is directly integrated into a tag antenna to construct RFID sensor tags.
The resistive sensor can be directly embedded into the tag antenna and the sensor resistance can thus modulate the antenna performance by means of gal- vanic contact. Integration of the sensor can, for example, be achieved by series connection or parallel connection with the RFID tag chip. A model-based theo- retical analysis is discussed for both two types of connections. A dipole-based antenna is designed to characterize the performance of RFID sensor tags with such integration structures. The concept is demonstrated by integrating the small printed resistive sensors into the printed tag antennas.
• A second method of integrating a surface mountable resistive sensor element into a tag antenna is also demonstrated.
The sensor element is embedded into a loop which is positioned on top of the tag antenna and the sensor information can thus modulate the performance of the tag antenna by means of electromagnetic (EM) coupling. The EM cou- pling mechanism is analyzed by showing how the antenna electrical properties
4 Introduction
change with sensor impedance. Two equivalent circuit models are respectively proposed for the EM coupled sensor tags with two different antenna structures and the models are verified by comparison with full-wave simulation results.
The proposed models can thus be a time-efficient approach to predict the per- formance of such EM coupled sensor tags through circuit-level calculations.
The proposed models are suitable for analysis of EM coupled sensor tags with not only the resistive sensors but also capacitive or inductive sensors. The key factors affecting the sensory performance and the methods for optimizing the sensory performance of the EM coupled RFID sensor tags are also discussed in the thesis. The concept is experimentally verified by attaching a printed loop with embedded resistive sensor on a commercial RFID tag.
• A printed resistive sensor element, used for demonstrating the concepts of galvanic contact integration and electromagnetically coupled integration of a sensor into a tag antenna, is thoroughly studied and characterized in the thesis.
The sensor is an inkjet printed structure with a very small geometry. The sen- sor changes its resistance by means of an irreversible sintering process, making it work as a write-once-read-many (WORM) memory. The WORM resistance can be changed by either high temperature or high humidity, which implies that the WORM can serve as a temperature sensor or a humidity sensor. The drawbacks and limitations of the WORM as a sensor are discussed. The heat sintering, humidity sintering and their combinational effect are characterized in the thesis. The humidity sensing mechanism of the printed WORM is inves- tigated through a series of experiments. The presented WORM sensor can be used to detect and record an event of excessive temperature or humidity and be readout at a later time to provide information regarding a historical event.
• An investigation on the influence of the edge roughness on an RFID antenna is per- formed.
Edge roughness is commonly associated with low-cost high-speed antenna manufacturing processes, e.g. printing, milling, and etching process. In some cases, these manufacturing processes can cause a high degree of edge rough- ness. The edge roughness will affect the behavior of an antenna, however, the characteristics of edge roughness on RFID antennas have previously not re- ceived any significant attention. Unforeseen antenna behavior can affect the measurements of the antenna-based sensor tags, thus the influence of the edge roughness is investigated in this thesis. The investigation is performed by analyzing the detuning of the antenna’s electrical properties caused by vari- ous degrees of edge roughness. This investigation utilizes a relatively simple structural model to simulate the edge roughness caused by the manufactur- ing process and shows a preliminary result, where future investigations could utilize other and more complicated structural models. However, the results of this investigation could advantageously constitute the basis for creating design guidelines for antennas that are robust against edge roughness.
1.2 Outline 5
• The presented three structures of the passive UHF RFID sensor tags provide both in- situ humidity sensing and historical humidity recording, at the lowest possible cost level and being compatible with standard communication systems in logistics.
1.2 Outline
Chapter 2: Describes the methods for the acquisition of the sensor information for an antenna-based sensor tag.
Chapter 3: Describes how a layer of moisture absorbing material can provide hu- midity sensing functionality to UHF RFID tag antennas and, for such a humidity sensor tag, how to overcome certain limitations due to distur- bances associated with background material.
Chapter 4: Describes the design and characterizations of a printed resistive humidity sensor (which is a write-once-read-many memory, WORM).
Chapter 5: Introduces the concept of UHF RFID sensor tags with sensors directly integrated into the antennas. The simulation results and the experimental results are presented.
Chapter 6: Introduces the concept of UHF RFID sensor tags with electromagneti- cally (EM) coupled sensors. Two analytical models are proposed for EM coupled sensor tags with two different types of antenna structures. The design considerations and potential optimization methods are discussed.
The experimental validation is also shown in this chapter.
Chapter 7: Investigates the effect of the edge roughness on the antenna performance.
Chapter 8: Summary of publications.
Chapter 9: Thesis summary.
Chapter 10: References.
6
Chapter 2
Acquisition of the Sensor Information for an
Antenna-Based Sensor Tag
An archetypal RFID system consists of an interrogator, more often known as a reader, and a transponder or tag. The reader normally has one or more antennas connected to a circuit board. The tag has an antenna and an integrated circuit (IC), often known as a silicon chip. The antennas play an important role in the commu- nication between the reader and the tag. There are many parameters that can be used to describe the electrical properties of an antenna. Input impedance, directiv- ity, gain and radiation efficiency are the most basic and most important parameters for a tag antenna. The degree of the impedance matching between the antenna and the load (which is the IC for the tag) determine the power transfer coefficient be- tween them. The directivity and gain describe the power magnification in different directions during antenna radiation. The ratio between the gain and the directivity is defined as the radiation efficiency. A radiation efficiency deviated from one indi- cates the existence of the ohmic losses in the antenna. These properties determine the performance of the tag antenna.
Antenna-based sensory UHF RFID tags utilize the influence of the sensed physi- cal parameters on the electrical properties of a tag antenna. The sensor information is added to the tag antenna by changing the antenna input impedance and also in- troducing ohmic losses to the antenna structure. Sometimes, the introduction of the sensor element might also cause variations of the antenna radiation pattern. The in- troduced ohmic losses cause the change in radiation efficiency of the antenna and the eventual variation of the antenna radiation pattern is expressed by the change in antenna directivity. The product of the directivity and radiation efficiency is defined as the gain of the antenna.Therefore, the sensor information is reflected in changes in the input impedance and the gain of the antenna.
2.1 Measurable output variables
In a passive UHF RFID system, the reader transmits interrogation signals to the tags and the tags reply to the reader by means of modulating the backscattered sig- nals. The communication channel carrying information from the reader to the tag is regarded as forward link and that carrying information from the tag to the reader is regards as reverse link. In forward link, the reader transmits a power of Pt,reader
7
8 Acquisition of the Sensor Information for an Antenna-Based Sensor Tag
to its transmitting antenna (which has a gain of Greader) for power radiation and this results in a power density of Pt,readerGreader/(4πr2)at the position of the tag antenna, where r is the reading distance between the reader antenna and the tag an- tenna. The power received by the tag antenna (Pr,tag) is the product of the incident power density and the effective aperture of the tag antenna (Ae,tag), as presented in Eq. 2.1. A certain percentage of the received power, determined by the power trans- fer coefficient (τ ), is transferred to the silicon chip of the tag, as described in Eq. 2.2.
The power backscattered by the tag antenna (Pbs,tag) is determined by the incident power density and the radar scattering cross section of the tag antenna (Asc,tag), as presented in Eq. 2.3. In a similar manner to that for the forward link, the backscat- tered power received by the reader antenna (Pr,reader) is the product of the backscat- tered power density and the effective aperture of the reader antenna (Ae,reader), as presented in Eq. 2.4.
Pr,tag= Pt,readerGreader
1
4πr2 · Ae,tag (2.1)
Pr,chip= Pr,tag· τ (2.2)
Pbs,tag= Pt,readerGreader
1
4πr2 · Asc,tag (2.3)
Pr,reader= Pbs,tag 1
4πr2· Ae,reader (2.4)
The effective aperture of the tag antenna and the reader antenna can be calcu- lated by using the standard formulas, Eq. 2.5 and Eq. 2.6, where λ is the wavelength of electromagnetic waves at the targeted frequency and Gtag is the gain of the tag antenna. The power transfer coefficient indicates the efficiency of the antenna due to impedance mismatch and, for the tag antenna, is calculated by Eq. 2.7 where ZC = RC + jXC and ZA = RA+ jXA are respectively the chip impedance and antenna impedance. The radar scattering cross section of the tag antenna is deter- mined by the impedance and the gain of the antenna, as presented in Eq. 2.8.
Ae,tag= λ2
4πGtag (2.5)
Ae,reader= λ2
4πGreader (2.6)
τ = 1− |Γ|2= 1−
ZC− ZA∗ ZC+ ZA
2= 4RCRA
|ZC+ ZA|2 (2.7)
Asc,tag = λ2
4πG2tag 4RA2
|ZC+ ZA|2 = λ2
4πG2tagRA RC
τ (2.8)
2.1 Measurable output variables 9
To power up the chip, Pr,chipshould not be less than the chip sensitivity PCwhich is the minimum power required to activate the tag IC.
Pr,chip≥ PC (2.9)
Since RFID systems are commonly forward link limited [24] and the chip sensi- tivity PC is constant, the maximum reading distance of the RFID tag and the mini- mum reader transmit power required for IC power-up can be derived from forward power transmission equations (Eq. 2.1, 2.2 and 2.9). With successful forward power transmission, the backscattered power is received by the reader and the power level relates to the radar scattering cross section of the tag antenna.
Based on the analysis above, we can conclude that
(a) the maximum reading distance (rmax) is proportional to the square root of the power transfer coefficient and the gain of tag antenna when the transmit power from the reader is a constant, as presented in Eq. 2.10;
rmax ∝ √
τ· Gtag (2.10)
(b) the required minimum transmit power (Pt,readermin) is inversely proportional to the power transfer coefficient and the gain of tag antenna when the reading distance is fixed as a constant, as presented in Eq. 2.11;
Pt,readermin ∝ 1 τ· Gtag
(2.11)
(c) the backscattered power received by the reader (Pr,reader) is proportional to the radar scattering cross-section (Asc,tag) of the tag antenna when the reader transmit power and the reading distance are constants, as presented in Eq. 2.12.
Pr,reader ∝ Asc,tag (2.12)
As mentioned at the beginning of this chapter, the sensor information is reflected in changes in the input impedance and the gain of the antenna. At the same time, changes in the impedance and the gain of the antenna can be obtained by measuring the change in one of the three variables, rmax, Pt,readerminand Pr,reader. The changes in these three variables can thus be related to the sensing information.
It should, however, be mentioned that Eq. 2.1-2.4 are only really accurate for free- space propagation and anechoic environments. A more exact model must account for additional factors such as multipath and small- and large-scale fading [25, 26].
Therefore, the above equations present a simplistic approximation to an ideal situa- tion and are thus optimistic.
10 Acquisition of the Sensor Information for an Antenna-Based Sensor Tag
2.2 Measurement setup
To facilitate the acquisition of the sensor data, setting the reader transmit power as a constant is easily accomplished by setting up the parameters for the reader.
However, for long term measurements of a sensor tag, maintaining the reading dis- tance at a fixed constant value is not a good option. Thus, to remove the requirement of fixed reading distances, a twin tag setup is necessary. The twin tag setup was firstly presented in [10]. In such a twin tag setup, one of the tags is a sensor tag and the other is a normal tag, which is used as a reference. The reference tag has the same kind of antenna and chip as those in the sensor tag, but without the sensor element. Two identical RFID tags are applied within the same sensor label and they are always interrogated at the same reading distance, thus allowing for a differential power readout.
According to Eq. 2.11, the ratio between the required minimum transmit power for a sensor tag and for a normal tag is determined by the inverse ratio of the power transfer coefficient and the gain of the tag antennas when the two tags are interro- gated at the same reading distance. That is, the relative required minimum transmit power for a sensor tag compared to a normal tag is determined by Eq. 2.13 and Equa- tion 2.14 produces results in decibel units. The relative required minimum transmit power in decibel is also referred to as the transmitted power difference. Similarly, according to Eq. 2.12, the relative value of the reader received power for a sensor tag compared to a normal tag can be expressed by the ratio between the radar scattering cross section for a sensor tag and for a normal tag when the two tags are interrogated at the same reading distance. Since the radar scattering cross section is also a func- tion of the antenna impedance and gain, the relative reader received power can be finally expressed through the ratios of the antenna impedance and gain, as presented in Eq. 2.15 and Eq. 2.16. The detuning of the antenna electrical properties caused by the introduction of a sensor into the tag antenna can thus be evaluated in terms of the relative required minimum transmit power or the relative reader received power without the restriction of fixed reading distance.
Pt,readerrelative
min= Pt,readersensor
min
Pt,readernormal
min
=
( τ· Gtag
τ0· Gtag0
)−1
(2.13)
Pt,readerrelativemin[dB] = Pt,readersensormin[dBm]− Pt,readernormalmin[dBm] =−10 log10
( τ· Gtag
τ0· Gtag0
)
(2.14)
Pr,readerrelative= Pr,readersensor Pr,readernormal = Asc
Asc0
= G2tag· RA· τ G2tag0· RA0· τ0
(2.15)
Pr,readerrelative[dB] = Pr,readersensor [dBm]−Pr,readernormal[dBm] = 10 log10
( G2tag· RA· τ G2tag0· RA0· τ0
) (2.16)
2.2 Measurement setup 11
In Eq. 2.13-2.16, the variables with subscript “0” represent the parameters of the normal tag antenna and the corresponding variables without subscript “0” repre- sent the parameters of the sensor tag antenna. This is always true throughout the remainder of this thesis.
Both the differential output power transmitted by the reader and the differential backscattered power received by the reader can be obtained using a twin-tag setup and they are related to the sensing information. The work in this thesis was carried out mainly by measuring the differential output power. Fig. 2.1 illustrates this twin tag method for the case used in this thesis work. The sensor tag and the reference tag are placed in a label and an RFID reader antenna is placed in front of the tags.
The distance between the two tags is relatively small in comparison to the distance between the RFID reader and the sensor label. The distance from the reader to each tag can thus be treated as the same value. A GUI PC software was developed to control the reader to sweep the output power and record the minimum power levels required to read each tag within a label. The power sweep can be set to automatically run several times in order to create an average value for each experiment since a variation of±0.5 dB for output power is relatively common.
Figure 2.1: Illustration of the twin tag concept.
12
Chapter 3
UHF RFID Sensor Tags With Covering of Sensing Material
It is well known that the performance of low cost tags, constructed with simple one-layer antennas, is very sensitive to the surrounding environment and especially to nearby metallic surfaces and water [27–29]. The water content nearby an RFID antenna will directly cause ohmic losses in the antennas near-field and also change its resonance frequency. It has previously been characterized as to how this property can be used to measure the wetness in soil and snow by connecting a transmission line to a buried monopole antenna [30].
If the RFID tag is covered by or totally embedded in a moisture absorbing ma- terial, the tag performance will relate to the water concentration in the moisture ab- sorbing material. Considering the previously mentioned differential readout method using a twin-tag setup, one of the RFID tags is covered by or totally embedded in a moisture absorbing material while the other tag remains naked. In a humid envi- ronment, the humidity concentration will thus be higher in the moisture absorbing material than in the vicinity of an naked tag [31]. The performance of the embedded tag will be worse than that of the naked tag. The difference in performance of the two tags will thus be related to the level of relative humidity.
[10] demonstrates a humidity RFID sensor solution with an antenna covered by moisture absorbing paper. Paper material is known to withdraw water and in a hu- mid environment, the water concentration in a paper material is a function of the relative humidity in the surrounding air [32]. As water will increase both the real and imaginary parts of the papers dielectric constant, the tag antenna will operate with lower efficiency due to the changes in input impedance and ohmic losses. That is, the humidity will eventually cause a degradation of the tag performance. An RFID reader positioned in front of the twin-tag label must thus emit a stronger in- terrogating signal in order to power up the embedded tag than that required by the naked tag. By comparing the minimum power levels required to power up each tag, it is therefore possible to determine the humidity level at the tag’s location. In the field the procedure requires a lookup table where moisture levels previously have been characterized versus differences in power up levels.
As mentioned, the basic concept regarding the RFID sensor tags utilizing mois- ture absorbing material has been characterized previously by embedding ordinary commercially available RFID tags in a paper material. Readout problems can how- ever occur when these one-layer antennas are placed on a background material that also contains a moisture absorbing material, such as wood, or containing nails or screws or something else that might disturb the measurement.
13
14 UHF RFID Sensor Tags With Covering of Sensing Material
Therefore, in this chapter, we present how UHF RFID tags with microstrip anten- nas could be used for measuring levels of relative humidity at hidden locations and thus shield the background materials. The microstrip antennas’ low influence from background materials in combination with making them more narrow-banded than commercial tags ensures that the differential readout becomes more reliable.
3.1 Experimental results
It is well known that the performance of low cost tags, constructed with simple one-layer antennas, is very sensitive to the surrounding environment and especially to nearby ]. Water content nearby an RFID antenna will directly cause ohmic losses in the ange its resonance frequency.
ized how this property can be used to measure humidity at a hidden location with aid of pairs of ordinary passive RFID tags by covering or totally embedding one of the two tags with a moisture absorbing material while the other tag is left untouched [4]. Paper material is known to withdraw water and in a humid environment the humidity concentration will thus be higher in the moisture absorbing material than in the vicinity of an As water will increase both the real and imaginary parts of the paper’s dielectric constant, the tag antenna will operate with lower efficiency due to ohmic losses and change in input impedance. If the tags are passive, an RFID reader positioned at the same distance from both tags in the label
characterized versus differences in power up levels.
Fig. 2. Drawing of the pair of microstrip antennas where one antenna patch is covered by a moisture absorbing material.
Figure 3.1: Drawing of the pair of microstrip antennas where one antenna patch is covered by a moisture absorbing material and the other one is left open.
A relatively simple microstrip antenna structure is mirrored on a PCB where one of the antennas is covered by a paper based moisture absorbing material and char- acterized for the impact of moisture through remote measurements in a homemade climate chamber as well as by the direct appliance of water drops. The microstrip antenna in Fig. 3.1 is fabricated from 1.53 mm thick Rogers RO3003 and has dimen- sions l = 75 mm, w = 65 mm, lp= 47mm, wp = 15mm, lsw = 10mm, lf = 10mm, wf = 4.5mm and d = 25 mm. In order to maintain it at this small size but still operating at 867 MHz it has an lsw = 10mm long ”shortcircuiting wall” between the patch and the ground plane indicated as an elongated rectangle in the rightmost part of the patches in Fig. 3.1. The feed lines are similarly connected to the ground plane at the top and bottom of the drawing and the Alien Gen-2 Monza RFID chips used are indicated as diamond shapes between the feed lines and antenna patches. This unit with the relatively small antennas gives a read range in the open air of about 2 meters. After assembling the antennas and the RFID chips to a unit, which we refer to as a twin-tag label, it was painted with transparent lacquer in order to not allow unwanted moisture to enter the PCB material.
3.1 Experimental results 15
The upper antenna in Fig. 3.1 was covered with, respectively, one and three 500 µm thick blotting papers known to have a high ability to withdraw moisture.
The reason for attempting two different total thicknesses is to evaluate how much a thicker moisture absorbing layer affects the antenna degradation. To further evalu- ate how to increase the impact of moisture, some absorbing papers were also doped with NaCl in order to increase the ohmic losses when moisture enters the paper.
The experimental tags were placed in a 1 x 1 x 0.1 meter large climate chamber and a SAMSys RFID reader was placed outside with its reader antenna 0.80 meters from the tags inside the chamber. For each experiment the output power of the SAMSys reader was swept from 12.0 dBm to 28.0 dBm with a 5 dBi reader antenna.
The minimum output powers required to read the open tag and the embedded tag were recorded and compared in order to obtain the differential power value.
3.1.1 Power-up differences due to water drops
The first characterizing experiments were carried out by applying 20 µg water drops directly onto the blotting paper. Fig. 3.2 shows the resulting graphs with re- gards to how much more power was required to read the embedded and wetted tag as compared to the open one.
It can be seen how the tag covered with one paper affects and degrades the RFID antenna more for a specific amount of water than occurs when the tag is covered by three papers. It can also be noted that the salt-doped paper has a significantly larger impact on the required power-up levels and that the only one layer of salt-doped paper has the largest impact, giving the clearest readout.
It is shown how pre-doping the covering moisture absorbing material with NaCl significantly increased the impact to the embedded antenna. In relation to measuring high humidity values, a high doping level even produced too large an impact since it proved impossible to read the embedded tag and lighter doping was required. For detection of wetness, one embedding paper has a stronger influence on the antenna for a specific amount of water than embedding the antenna in a thicker layer consist- ing of three stacked papers. The best results regarding wetness detection was thus achieved by only one layer of salt-doped paper, where even two water drops of a total of 40 µg created a difference in power-up levels of 5 dB.
3.1.2 Power-up differences due to humidity
Placing the tag in a chamber with controlled humidity gave similar results to those for the water drops. The results are shown in Fig. 3.3. Salt-doped paper was also attempted at this point. At this time the graphs proved not to differ as much regarding the different number of embedding papers as was the case when a certain number of water drops were applied. There is already a slight power difference at 60% RH while the higher values, that are most often of interest, require up to approximately 11 dB more power to read the embedded RFID antenna as compared to the open one. It is interesting to note that above 80% RH, a value when mold and putrefaction are likely to start appearing, there is already a power difference of about 7 dB, which it should be easy to distinguish by using a not perfectly aligned
16 UHF RFID Sensor Tags With Covering of Sensing Material
Figure 3.2: Results of direct appliance of 20 µg water drops on RFID sensor tags with mi- crostrip antennas.
that the combination of only one paper and having that one doped with salt has the largest impact, giving the clearest
Placing the tag in a chamber with controlled humidity gave similar results as for the water drops as is shown in Fig. 4.
Salt-doped paper was tried also here but the heavily doped had too large impact on the antenna as the embedded tag could not be read even at 80%
RH. A less doped paper was therefore used with the results
Limitations of the proposed measurement concept includes
Fig. 4. Results from experiments in climate chamber. There is a slight power difference already at 60% RH while the higher
Figure 3.3: Results of placing RFID sensor tags with microstrip antennas in climate chamber.
hand-held reader.
3.2 Discussion
It has been shown how pairs of microstrip antennas equipped with passive RFID chips can be utilized for remote measurements of humidity or wetness with an or- dinary RFID reader. This was performed by equipping one of the antennas with a paper that degrades the antenna efficiency in proportion to the level of relative hu- midity or an amount of water, thus forcing it to require a higher level of power in order to operate. The microstrip antennas work better for these kinds of measure- ments as compared to the previously proposed one-layer antennas since they tend
3.2 Discussion 17
to be less sensitive to it background material. Evaluated microstrip antennas were also designed to have a narrower bandwidth than the common commercial one- layer RFID antennas, making them more sensitive to dielectric changes in a material embedding the antenna.
It was also shown how pre-doping the covering moisture absorbing material with NaCl significantly increased the impact to the embedded antenna. In relation to mea- suring high humidity values, a high doping level even produced too large impact since the embedded tag was no longer readable and lighter doping was required.
For the detection of wetness, it was further experienced that one embedding paper had a stronger influence on the antenna for a specific amount of water than embed- ding the antenna in a thicker layer consisting of three stacked papers. The best results for wetness detection were thus achieved by using only one layer of salt-doped pa- per, where even just two water drops totaling 40 µg created a difference in power-up levels of 5 dB.
The method of covering the tag with a layer of special material could also, poten- tially, be utilized to measure other physical quantities as long as there is a material to be placed over the tag antenna that changes its electrical properties in proportion to the physical quantity of interest.
18
Chapter 4
Printed Sensor (WORM)
In the previous chapter, the moisture absorbing material provided humidity sens- ing functionality to RFID tag antennas that in turn provided information regarding the current humidity status through an RFID readout method. This chapter presents an ultra-low cost humidity sensor element with memory functionality. It is a 1-bit write-once-read-many (WORM) memory printed using conductive ink based upon nanometer-sized silver particles. The printed WORM is suitable for direct integra- tion into a printed RFID antenna or other printed circuits. As a sensor component the WORM exhibits two states, “On” and “Off”, corresponding to “WORM bit = 1”
and “WORM bit = 0”. The printed WORM memory is defined by means of its resis- tance and in which a logical zero equals a high resistance and a logical one equals a lower resistance. The setting for this bit is accomplished by means of sintering.
4.1 Design of Printed Sensor
The presented WORM is relatively simple in its geometry since it consists solely of a short segment of one or several printed narrow lines in between two pads as shown in Fig. 4.1. As is the case in normal metallic wire, the resistance for the WORM line is decided by the length, width and thickness of the printed line in combination with its resistivity. The WORM line is printed by means of separated dots. The rea- son for this special design is to obtain a high initial resistance. The separated ink dots become slightly connected due to the spreading of the ink dots when they are printed and the poor connection between them offers the printed line a relatively high initial resistivity. In addition, for a fixed length of printed line, the design of the separated ink dots reduces the volume of the ink used for printing the line and thus decreases the across section of the printed line which will lead to a higher resis- tance. Two half-round pads on both sides of the lines are printed in order to further increase the resistivity of the printed line and also to ease the contact with other printed structures. The reactance of the WORM impedance is extremely close to zero at the frequencies of interest and is the reason why the WORM characterization has been solely treated as resistive. The resistance of the WORM can also be adjusted by printing one or more lines between the two contact pads. Fig. 4.1 shows the de- signs and printed patterns of both a one-line WORM and a five-line WORM. All the WORMs in this thesis were produced by using the silver ink DGP-40LT-15C from Advanced Nano Products (ANP) Co. Ltd and printed using the inkjet Dimatix Ma- terial Printer DMP-2800 with a 10 pL printer head on nano-porous photo paper. The resistances of the WORMs were measured with an Agilent 34405A Digit Multimeter.
19
20 Printed Sensor (WORM)
(a) (b)
(c) (d)
Figure 4.1: The printed WORM structure with (a)-(b) one sensor line and (c)-(d) five parallel sensor lines.
4.2 Temperature Sintering
In nano-particle inks, the metal particles are stabilized by one or more disper- sants or by capping agents. The capping agents create an electrostatic and/or steric barrier which prevents the aggregation of the particles. In a printed trace, the nano particles are embedded in an insulating polymer matrix which impedes electrical conductance [33]. Sintering is required in most cases to reach high conductance. Sin- tering is a process used in order to enhance junctions among particles and to reduce resistivity by means of thermal heating in an oven or by applying a current flow, intense pulsed light, microwave or a laser [34–47]. High temperatures can cause this reduction of resistivity by melting the encapsulating polymers of the silver nano particles and also by evaporating the solvent and binders in the ink, which, in turn, cause a neck formation and particle growth. Fig. 4.2 shows the effect of thermal sintering at different temperatures, taking the five-line WORM as an example. The WORM could be sintered by a temperature higher than 30◦C. The sintering velocity is proportional to the temperature.
