Printed antennas with variable conductive
ink layer thickness
J. Sidén, M.K. Fein, A. Koptyug and H.-E. Nilsson
Abstract: One of the complex tasks in mass production of RF electronics is printing the communication antenna using electrically conductive ink. For example, this is very common for radiofrequency identification (RFID) tags. Electrical properties of the ink are mostly determined by
conductive (e.g. silver) particles mixed into the ink solution and the way they ‘connect’ in
the cured ink. It is also desirable to minimise the amount of ink used per antenna, because
high-conducting metals like silver used in the ink are rather expensive. Metal-based inks have
limited conductivity, so the thicker the cured ink layer will be the better the antenna radiation efficiency can be achieved, but also the higher will be the costs. In the paper, the authors report on the
investigations of the possibility of minimising the amount of ink used per antenna. This can be
achieved by printing thicker ink layers, where antenna structures are known to have high current
density. Two common antenna structures and a dedicated antenna for passive RFID are used
in the investigation. The main result of the paper is that radiation efficiency depends primarily
on the total amount of ink used for printing the antenna, rather than on the variations of the
layer thickness within the antenna structure.
1
Introduction
Printed electronics is a relatively new area of research and
development targeting electronics manufacturing using
adapted commercial printing presses [1]. When the technology matures it will have a capacity to deliver very large
amounts of printed electronic units at a very favourable
price, as compared to the traditional silicon-based technology [2, 3]. Printed-radio-frequency-identification (RFID)
tags [4] and wireless sensor nodes [5] are assumed to be
among the key applications that could benefit from the
organic electronics developments.
While the semiconductor elements in RFID tags are
likely to stay silicon-based for quite some time, the communicating antennas are already printed utilising commercially
available techniques and electrically conductive ink.
Three-dimensional antenna structures such as horn antennas
can also be synthesised directly from CAD files using
rapidly deposited two-dimensional layers of polymers [6].
Commonly, metallic particles shaped as flakes are
blended into an ink, which is applied on a substrate with a
flexographic printer. In the dried ink, the metal particles
touch and overlap, creating continuous conductive traces.
There is a large potential of cheap bulk antenna fabrication
using printed conductor technology. However, ink-trace
conductors have at least two major drawbacks. First, their
conductivity is still not as good as that of corresponding
# The Institution of Engineering and Technology 2007
doi:10.1049/iet-map:20060021
Paper first received 8th February and in revised form 28th May 2006
J. Sidén and H.-E. Nilsson are with Mid-Sweden University, Electronics Design
Division, Sundsvall 85170, Sweden
M.K. Fein is with Precisia, LLC, 4600 Arrowhead Dr., Ann Arbor, MI 48108,
USA
A. Koptyug is with Sensible Solutions Sweden AB, Universitetsallén 32,
Sundsvall 85171, Sweden
E-mail: [email protected]
IET Microw. Antennas Propag., 2007, 1, (2), pp. 401– 407
solid metal traces. Despite the fact that there is less metal
in the ink-based conductor, as compared to the solid metal
one (such as etched copper), they are still rather expensive.
It happens because the best conductive inks use silver or
silver-coated particles blended into them, and the price
of silver and the manufacturing costs of silver powder
and silver flakes are rather high. The abovementioned
drawbacks result in a contradictory demand to the antenna
printing: thicker ink layers would be preferable for high
radiation efficiency and thinner ink would be desirable for
keeping production costs at a minimum.
Basic laws of conductivity for homogeneous media state
that ohmic losses decrease as conductor cross-section
increases [7]. Although printed ink traces are not homogenous, similar behaviour can also be applied in this
case. Total resistance of the printed electrical conductor
with a fixed trace width will have its resistance proportional
to the overall trace length and be inversely proportional to
the ink-layer thickness. Similarly, a printed antenna will
experience certain high-frequency resistivity influencing
its input impedance and radiation efficiency showing a
dependence on the ink-layer thickness and, possibly, layer
thickness variations throughout the structure. Radiation
efficiency of antennas is a crucial factor in passive RFID
tags because the RFID chip acquires all its power from
the electromagnetic wave of an interrogating RFID unit
[8]. Small radiation efficiency of the tag antenna immediately results in a low reading range of the RFID tag.
Already, it has been studied how the ink thickness affects
Q-values for low frequency RFID antennas (i.e. printed
inductive coils) [9 – 12] and how it affects the radiation
efficiency of RFID antennas working in the UHF band
[13 – 15]. It has also been shown how the total printed area
can be minimised by using gridded antenna patterns [16].
This paper reports on investigating possible benefits of a
variable ink-layer thickness approach. In this approach,
antenna pattern areas with highest current density are identified and printed with thicker ink layers. We compare the
401
results of computer modelling with experiment for two
common antenna structures. We also characterise the influence of the variable-layer-thickness tag-antenna printing on
the maximum reading range for a commercial RFID system.
Parameters for all printed antennas are also compared to the
ones of the solid copper-based antennas with the same
pattern. In the computer simulations, reference structures
of copper are assumed to be made of perfect electric
conductor (PEC).
Investigations were carried out for the structures working
at 868 MHz, one of the ISM bands for nonlicensed lowpower transmitting systems available within the EU [17].
The corresponding band allowed for such applications in
the USA is 902 – 928 MHz [18]. Although additional tests
must be conducted to specify antenna performance at
902 – 928 MHz, we expect that qualitative conclusions
could be directly extended from the mentioned EU band
to the US band.
2
Printed dipole and bow-tie antennas with
different sheet resistance of conductors
We have investigated the efficiency of different printing
approaches for two antenna structures, namely the halfwavelength dipole and the bow tie, as illustrated in
Figs. 1 and 2 and Figs. 3 and 4, respectively. The choice
of antennas represents one structure with relatively small
area and constant width throughout the antenna length, the
half-wavelength dipole. The other antenna, the bow tie represents a structure with relatively large area and increasing
width seen from the antenna centre.
Fig. 1 illustrates the current distribution in the dipole
antenna with the dimensions close to a half wavelength.
For the chosen frequency of 868 MHz the dipole antenna
dimensions were structure length ld ¼ 163 mm and trace
width w ¼ 2 mm. Fig. 3 illustrates the current distribution
in the bow-tie antenna, as presented when modelled using
electromagnetic simulation software package Ansoft
HFSS 9.2. The textured pattern is due to triangular
meshing, and the darker a triangle is the higher the
current density is in that area. The antenna dimensions for
the 868 MHz bow tie were structure length lb ¼ 124 mm
and flare angle a ¼ 308.
Test dipole and bow-tie antennas were manufactured
with ink layers of different thickness using a flexographic
printer, resulting in different sheet resistance of conducting
layers. The variable ink thickness approach is illustrated in
Figs. 2 and 4, where variations in ink-layer thickness were
Fig. 3 Ink pattern for half-wavelength dipole antenna structure
printed with thicker ink layer in the areas with highest current
density
achieved by using different anilox rollers and by adding
consecutive extra layers of paint at specific antenna areas.
Efficiency of the printed antennas was characterised by
measuring transmitted power in a set-up with a reference
antenna and an antenna under test placed at one metre distance in a wooden (balsa) fixture. The transmission parameter for transmission from the reference antenna to the
antenna under test is further denoted as S21 and is measured
using the Network Analyser E8364 (PNA Series) by Agilent
Technologies. Table 1 presents the comparison of the
simulated and measured transmission efficiency for the
structures printed with different layer thickness. Values
are presented as test antenna radiation efficiency and are
normalised to the power supplied to the reference, copper
antennas. The simulated radiation efficiency is acquired
by using the Ansoft HFSS EM modelling software
package, where the antennas are regarded as twodimensional structures with the sheet impedance corresponding to the one of the test structures. The relative
amount of used ink is calculated from the measurements
of thickness of printed traces and normalised to the thickest
trace used for each investigated antenna. In present prints
thickest trace tops were about 15– 20 mm, giving a sheet
resistance of about 50 mV/square.
A theoretical assumption that conductor sheet resistance
is inversely proportional to ink-layer thickness was found
to be good with relatively thick ink layers and relatively
wide conductive traces. With the thin layers, measured conductivity was, however, commonly lower than predicted. In
the case of cured ink layers thicker than 10 mm measured
conductivity was very close to an expected 1 106 S/m. At
the same time, conductivity values down to 2 105 S/m
were measured for the layers of about 2 – 4 mm. Most probably, this is due to the fact that metal flakes responsible for
the conductivity of the ink are, in turn, several micrometres
in size. With the ink-layer thickness close to the average
metal particle size, the amount of continuous conductive
‘bridges’ with overlapping metal particles ‘connected in
parallel’ becomes significantly smaller. And one can also
expect that the ‘overlapping area’ between two adjacent
flakes in such a bridge, responsible for the resistance of
Fig. 1 Current distribution in the half-wavelength dipole
antenna
Fig. 2 Computer simulated current density distribution with
triangular meshing for a bow-tie antenna (darker areas correspond to higher current density)
402
Fig. 4 Ink pattern for bow-tie antenna structure printed with
thicker ink layer in the areas with highest current density
IET Microw. Antennas Propag., Vol. 1, No. 2, April 2007
Table 1: Radiation efficiency for printed antennas with different conductor sheet resistance
Dipole antenna
Sheet resistance
Simulated radiation
Total structure:
Total structure:
Total structure:
Inner one third:
Inner one third:
Inner one third:
2395 mV/sq
238 mV/sq
51 mV/sq
103 mV/sq
60 mV/sq
52 mV/sq
Outer two thirds:
Outer two thirds:
Outer two thirds:
6720 mV/sq
1012 mV/sq
213 mV/sq
38%
81%
95%
34%
73%
89%
29%
78%
93%
32%
57%
83%
21%
40%
100%
31%
48%
45%
1.4 mm3
2.7 mm3
6.6 mm3
2.1 mm3
3.2 mm3
3.0 mm3
Total structure:
Total structure:
Total structure:
Inner one third:
Inner one third:
Inner one third:
3242 mV/sq
178 mV/sq
54 mV/sq
86 mV/sq
41 mV/sq
35 mV/sq
Outer two thirds:
Outer two thirds:
Outer two thirds:
1097 mV/sq
1154 mV/sq
211 mV/sq
efficiency
Measured radiation
efficiency
Relative amount
of ink used
Total amount of ink
used (Cured Vol.)
Bow-tie antenna
Sheet resistance
Simulated radiation
58%
95%
98%
92%
92%
97%
54%
81%
93%
87%
76%
89%
18%
44%
100%
21%
33%
47%
7.4 mm3
18 mm3
41 mm3
8.6 mm3
14 mm3
19 mm3
efficiency
Measured radiation
efficiency
Relative amount
of ink used
Total amount of ink
used (Cured Vol.)
the flake-to-flake contact in such a bridge, also becomes
higher. It was also found that thin conductive ink layers
suffer from common problems known in flexographic printing as lining and the producing of a halo effect by squeezing
ink out at the edges of printed features. With the lining,
solid ink layers turn into a set of small narrow lines in the
direction of the print, creating significant anisotropy in conductivity (and sheet resistance). For thin layers anisotropy
of about 2mm was observed, i.e. the conductivity in the
print direction was twice higher than the one in the perpendicular direction. Thicker layers introduced a lower level of
anisotropy but still present at about 1.2 – 1.5. All manufactured antennas were printed in the direction of their length
and all sheet resistances were also measured in this
direction.
The first three columns in Table 1 refer to the antennas
printed with the uniform ink thickness and sheet resistance
all over the structure. Theoretically, using twice smaller
amount of ink should yield twice higher conductor sheet
resistance and vice versa. The last three columns in
Table 1 refer to the antennas printed with the ink thickness
lower at the periphery of the antenna and higher at the areas
with maximum current density. Figs. 5a and b gives an
additional and more visual view of the results presented in
Table 1.
As mentioned here, for relatively thin flexographically
printed layers, the proportionality of the trace conductivity
to the layer thickness (amount of ink used for the fixed
pattern) does not hold. For example, decreasing the
amount of ink used for printing the dipole by almost half,
from 40 to 21%, was causing tenfold change in the sheet
resistance instead of doubling it. Illustrations of
IET Microw. Antennas Propag., Vol. 1, No. 2, April 2007
corresponding cured ink traces are given in Figs. 6 – 9.
The topography maps were acquired using an FRT laser
optical profilometer and reveal the roughness of the conductor surface (i.e. significant variation in the layer thickness)
for the very thin traces and significant effects of squeezed
out ink is seen at trace edges. Table 1 clearly illustrates
that using printed structures with thick ink traces and low
sheet resistance results in better antenna parameters. Also
it should be noted that changing the layer thickness more
than two times, from 100 ink to 40% ink, when manufacturing the thick ink layer dipole antenna, measured radiation
efficiency only drops from 93 to 78% as compared to the
perfect conductor case. The same effect stays for the
bow-tie antenna.
A level of 100% ink is a normalised value for each
antenna structure’s thickest print, but the total pattern
surface area for the bow-tie antenna is significantly larger
than for the dipole one working at the same frequency. If
the cost saving is the dominating design issue, the obvious
choice should be the dipole antenna, simply because of its
smaller total area. However, ordinary thin halfwave dipole
antennas have certain limitations. For example, even a
small scratch on the paint could break a narrow conductor
trace and cause a complete loss of antenna functionality
[19]. This is a significant limitation for thin dipole antennas
used in harsh environments. Antenna structures for RFID
applications which demand robust structures should therefore preferably have reasonable conductor width.
As it also follows from Table 1, simulated values for radiation efficiency are relatively close to the measured ones,
confirming the adequacy of the antenna models used in
present research.
403
Radiation Efficiency, %
100
80
60
40
20
Simulated
Measured
0
0
20
40
60
80
Relative amount of ink used, %
a
100
Radiation Efficiency, %
100
80
Fig. 7 Height graph measured at intersect indicated in Fig. 6
60
40
20
Simulated
Measured
0
0
20
40
60
80
Relative amount of ink used, %
100
b
Relative RFID read range, %
100
80
60
Fig. 8 Topographic view of 3 mm section of 2 mm-wide dipole
antenna having average measured sheet resistance of 238 mV/
square
40
20
0
0
Simulated
Measured
20
40
60
80
Relative amount of ink used, %
100
c
Fig. 5
Radiation efficiency aganist relative amount of ink used
For a dipole and b bow-tie antenna, and c corresponding read range for
the RFID antenna
Fig. 6 Topographic view of V 3 mm section of 2 mm-wide dipole
antenna having average measured sheet resistance of 2395 mV/
square
404
Just like with a filter, an antenna’s Q-value also decreases
when introducing losses to the system. The VSWR ¼ 1.5
antenna bandwidth approximately doubles as it goes, from
about 100 MHz for the best dipole and bow-tie conductors,
to over 200 MHz for the antennas with sheet resistance of a
couple of ohms. Apart from the fact that maximum antenna
gain is reduced proportional to radiation efficiency, the
antennas’ radiation patterns remain the same for the highloss antennas as for the low-loss ones, with simulated
maximum directivity for the dipole and bow-tie antennas
of, respectively, 2.50 and 2.25 dB.
Fig. 9 Height graph measured at intersect indicated in Fig. 8
IET Microw. Antennas Propag., Vol. 1, No. 2, April 2007
3 Printed dipole and bow-tie antennas with
variable ink layer thickness
It is well known, that current density can vary significantly
through the antenna structure. For example, for half-lambda
antennas current density is highest in the vicinity of the feed
points and slowly decreases towards the structure periphery.
Figs. 1 and 3 illustrate this for the dipole and the bow-tie
structures correspondingly.
It is also known that power, P, loss is directly proportional to the current squared, P ¼ I 2 . R, which for a
constant resistance R implies higher losses at areas of
high current density. It was therefore suggested to increase
the conductor layer thickness only in most ‘vital’ parts of
the antenna structure, aiming for a result with increasing
radiation efficiency but without significant increase in manufacturing costs due to total amount of ink used. In the
discussed structures, only the innermost one third of the
overall structure length was printed with thicker ink layers.
Three different scenarios are investigated with different
mixes of thin and thick traces. Values are all presented
in the last three columns of Table 1 where the relative
amount of ink used is again normalised to the thickest
layers used in the preceding Section.
Analysing the data present in Table 1, we can conclude
that, for the dipole antenna, there is no significant benefit
in using variable-layer-thickness printing. Thus, antennas
printed with the same amount of ink distributed uniformly
in one case, and with thicker layer at the feed points and
thin at the periphery in the other, have similar radiation efficiency. It should be also noted that, even though results of
computer simulations agrees well with the measurements,
simulated efficiencies tend to be slightly higher than
measured. These higher theoretical values might partly be
due to lining and halo effects in the traces that are not considered in the simulations.
With the wider bow-tie structure there is a possible gain
in redistributing ink throughout the structure. A value of
21% of the relative amount of ink (Table 1, 4th column)
for example gives 87% radiation efficiency, which can be
compared to that of a uniformly printed bow-tie antenna
with 18% used ink (Table 1, 1st column) which measured
only 54% radiation efficiency. The more desirable and
higher radiation efficiencies, achieved when using uniformly 44% (Table 1, 2nd column) or unevenly distributed
47% (Table 1, 6th column) relative amount of ink however,
again gives more similar radiation efficiencies, 81 to 89%.
4
Printed RFID antenna
One of the major targets of present research was to assess
how the printed antennas would perform in an established
RFID system. One of the possible ways is to compare the
maximum reading distances for the commercial tags
having copper-trace antennas with that of printed-antennabased tags having the same antenna pattern and the same
RFID chip. We have used a commercial RFID system by
iPico Holdings.
A sketch of corresponding antenna structure is shown in
Fig. 10. As with the previously discussed printed dipole and
bow-tie antennas, structures with both uniformly distributed
ink and with a thicker layer at the inner one third of the
structure length were tested.
Input impedance of a passive RFID tag antenna is usually
far from the common 50 V, as it is intended to be loaded by
the RFID chip having input impedance far from those in
common RF systems [20]. For this reason, it is not possible
to measure S21 directly in the same set-up as described
IET Microw. Antennas Propag., Vol. 1, No. 2, April 2007
Fig. 10 Commercial RFID tag antenna pattern by iPico
Holdings
Our test antennas were printed in the direction of the antenna length
with a thicker layer of ink at the central third of the structure
(shown in bold)
earlier for the transmission between two resonant antennas.
Investigated RFID antennas are instead characterised by
measuring maximum RFID read range with the RFID chip
attached to the corresponding antenna.
In passive RFID systems, read range is a critical factor.
The only power is supplied to the tag by the transmitted
interrogating radio wave and maximum allowed power of
a transmitter is strictly regulated for each frequency band.
It is well known that the power available at a certain distance from a transmitting unit is inversely proportional to
the square of the distance. It follows from the Friis transmission formula [7]
Pr ¼
ðPt Gt Þ Gr l2
ð4pRÞ2
ð1Þ
where R is the distance between the tag and the interrogator,
Pt . Gt is the effective isotropic radiated power (EIRP) of the
interrogator unit, Gr is the gain of the tag (receive) antenna
and l is the wavelength. As it follows from (1), when the
distance between the interrogator unit and the tag is
doubled, power received by the tag falls four times. It
also follows from the equation that, to get the same power
with two antennas with different total radiation efficiency,
we need to move the less efficient antenna closer to the
interrogator by a factor of the square root of the ratio of
total radiation efficiency. If for example a printed antenna
has a total radiation efficiency of 64% as compared to the
reference one made of copper (100% radiation efficiency),
the operating RFID range
for a tag using the printed
pffiffiffiffiffiffiffiffiffi
antenna would be about 0:64 100% ¼ 80% of that of
the copper one. In this way, measured sheet resistances of
printed antennas can be used to characterise antenna radiation efficiency in computer simulations that, in turn, can
be converted to theoretical RFID read range, due to the
square root relation.
Table 2 presents a comparison of the simulated and
measured maximum reading distance values, together with
the simulated antenna radiation efficiency and the relative
amount of ink used for the ‘iPico antennas’ printed with
different conductor layer thickness. A visual view of this
is presented in Fig. 5c.
As it is the case for the dipole and bow-tie antennas (see
Section 2), simulations agree relatively well with measured
values and, for half of the cases, the measured RFID read
range actually exceeds the simulated ones. One unexpected
exception, however, appeared for the antenna with average
sheet resistance of 1130 mV/square, where measured RFID
read range was significantly lower than expected. This is
derived to the problem that printing the antennas with variable conductor thickness becomes problematic with very
thin layers at the structure periphery, where lining and
halo effects become significant. The severity of this is illustrated through the topography map in Figs. 11 and 12
405
Table 2: Maximum reading range for a commercial RFID tag with antenna printed with different ink layer thickness
Sheet resistance
Simulated radiation
Total structure:
Total structure:
Total structure:
Inner one third:
Inner one third:
Inner one third:
1130 mV/sq
245 mV/sq
50 mV/sq
84 mV/sq
88 mV/sq
43 mV/sq
Outer two thirds:
Outer two thirds:
Outer two thirds:
2900 mV/sq
261 mV/sq
199 mV/sq
17%
42%
72%
11%
45%
53%
41%
65%
85%
33%
67%
73%
6%
56%
97%
28%
72%
75%
16%
29%
100%
21%
28%
31%
0.7 mm3
1.3 mm3
4.3 mm3
0.9 mm3
1.2 mm3
1.3 mm3
efficiency
Simulated RFID
read range
Measured RFID
read range
Relative amount
of ink used
Total amount of ink
used (Cured Vol.)
Also, from Table 2, it is concluded that there is again no
obvious benefit gained from redistributing the ink throughout the structure as compared to uniformly distributing the
same amount of ink.
The radiation pattern for the investigated RFID tag
antenna is very similar to the doughnut shape of the ordinary half-wavelength dipole commonly found in the literature [7], with a simulated maximum directivity of
2.17 dB. No change in radiation pattern properties was
observed for antennas with thinner ink traces.
5
Fig. 11 Results from optical laser profilometer showing 3.0 mm
of a 1.0 mm wide ink trace for the printed conductor section indicated in Fig. 4, showing a pronounced lining effect
Fig. 12
Height graph at the intersection indicated in Fig. 11
showing the surface roughness for about lt ¼ 3.0 mm of the
wt ¼ 1.0 mm width periphery trace section of the antenna
depicted in Fig. 10. The depletion of the conductive layer
in the ‘valleys’ going in the direction of printing is so
significant that complete breaks in the conductor layer
occurred in some produced samples that were discarded.
406
Conclusions
A new approach of distributing ink over a printed antenna
structure has been studied. Applied ink thickness is determined in accordance to the distribution of the antennas’
current density. The main application is for mass-produced
RFID-antennas, where saving fractions of a gram of ink
per antenna can save thousands of dollars per day in a
production line.
Investigation, however, shows that it is mainly the total
amount of ink used that determines the printed antennas’
radiation efficiency and not how it is distributed throughout
the structures.
We have found that ink layers having sheet resistances
of less than 50 mV/square gives close to 100% radiation
efficiency, while, if we can settle with about 80% radiation
efficiency, the ink layer thickness could be halved. It was
also found that with the flexographic printing technique
and ink used, lines of width 1 mm or less are not preferable
as they can become relatively rough. For very low ink layer
thickness, flexographic lining and halo effects can also
appear as severe problems, creating anisotropy in conductivity and can also lead to nonconductive traces for
narrow traces printed perpendicular to the direction of the
print.
A good observation is that computer simulations of
antennas printed with limited conductivity show radiation
efficiency that agrees well with measured antenna
samples. The same holds for printed RFID tag antennas,
where, unless for very low radiation efficiencies, calculated
relative maximum read range also agrees well with
measured values.
IET Microw. Antennas Propag., Vol. 1, No. 2, April 2007
6
Acknowledgments
Financial support from the forest industrial research
program FSCN at the Mid-Sweden University, the
Knowledge Foundation and our industrial partners are
gratefully acknowledged.
Support in high-frequency conductivity measurement of
printed ink traces by Jerzy Krupka at Warsaw Technical
University, is also gratefully acknowledged. We also
acknowledge support with the laser optical profilometer
by Christina Westerlind at SCA Research.
7
References
1 Molesa, S.E., Volkman, S.K., Redinger, D.R., Vornbrock, Ad.F., and
Subramanian, V.: ‘A high-performance all-inkjetted organic transistor
technology’. Electron Devices Meeting, 2004, IEDM Tech. Digest.
IEEE Int., pp. 1072–1074
2 Berggren, M., Kugler, T., Remonen, T., Nilsson, D., Miaoxiang, C.,
and Norberg, P.: ‘Paper electronics and electronic paper’. IEEE
Conf. on Polymers Adhesives Microelectron. Photonics, Polytronic
2001, pp. 300–303
3 Huebler, A., Hahn, U., Beier, W., Lasch, N., and Fischer, T.: ‘High
volume printing technologies for the production of polymer
electronic structures’. IEEE Conf. on Polymers Adhesives
Microelectron. Photonics, Polytronic 2002, 2002, pp. 172–176
4 Subramanian, V., Frechet, J.M., Chang, P.C., Huang, D.C., Lee, J.B.,
Molesa, S.E., Murphy, A.R., Redinger, D.R., and Volkman, S.K.:
‘Progress toward development of all-printed RFID tags: materials,
processes, and devices’, Proc. IEEE, 2005, 93, (7), pp. 1330–1338
5 Beale, D.A.R., and Hume, A.L.: ‘Capability and potential of
disposable organic sensor networks’. IEEE Aerospace Conf., 2004,
vol. 3, p. 2064
6 Huang, Y., Gong, X., Hajela, S., and Chappell, W.J.: ‘Layer-by-layer
stereolithography of three-dimensional antennas’. IEEE Int. Symp.
Antennas Propag., 2005, vol. 1A, pp. 276–279
IET Microw. Antennas Propag., Vol. 1, No. 2, April 2007
7 Cheng, D.K.: ‘Field and wave electromagnetics’ (Addison-Wesley,
UK, 1989, 2nd edn.), pp. 204, 640
8 Finkenzeller, K.: ‘RFID handbook’ (John Wiley and Sons Ltd, UK,
2003), pp. 133– 140
9 Redinger, D., and Molesa, S.: ‘An ink-jet-deposited passive
component process for RFID’, IEEE Trans. Electron Devices, 2004,
51, (12), pp. 1978–1983
10 Sangoi, R., Smith, C.G., Seymour, M.D., Venkataraman, J.N., Clark,
D.M., Kleper, M.L., and Kahn, B.E.: ‘Printing radio frequency
identification (RFID) tag antennas using inks containing silver
dispersions’, J. Dispersion Sci. Technol., 2004, 25, (4), pp. 513–521
11 Chen, S.C.Q., and Thomas, V.: ‘Optimization of inductive RFID
technology’. IEEE Int. Symp. Electron. Environment, 2001,
pp. 82–87
12 Cichos, S., Haberl, J., and Reichl, H.: ‘Performance analysis of
polymer based antenna-coils for RFID’. IEEE Conf. on Polymers
Adhesives Microelectron. Photonics, Polytronic 2001, pp. 120–124
13 Sidén, J., Fein, M.K., Koptioug, A., and Nilsson, H.-E.: ‘On the
efficiency of printed RFID antennas’, IEEE Trans. Electronic
Devices, submitted for publication
14 Bechevet, D., Vuong, T.-P., and Tedjini, S.: ‘Design and
measurements of antennas for RFID, made by conductive ink on
plastics’. IEEE Int. Symp. Antennas Propag., 2005, vol. 2B,
pp. 345– 348
15 Nikitin, P.V., Lam, S., and Rao, K.V.S.: ‘Low cost silver ink RFID tag
antennas’. IEEE Int. Symp. Antennas Propag., 2005, vol. 2B,
pp. 353– 356
16 Sidén, J., Olsson, T., Koptioug, A., and Nilsson, H.-E.: ‘Reduced
amount of conductive ink with gridded printed antennas’. IEEE
Conf. on Polymers Adhesives Microelectron. Photonics, Polytronic
2005, pp. 86– 89
17 European standard: ‘ETSI 302 208: ‘European Telecommunications
Standards Institute’, 2005
18 FCC§18.101– §18.311: ‘Federal Communications Commission’, 1998
19 Olsson, T., and Koptioug, A.: ‘Statistical analysis of antenna
robustness’, IEEE Trans. Antennas Propag., 2005, 53, (1), pp. 566–570
20 De Vita, G., and Iannaccone, G.: ‘Design criteria for the RF section of
UHF and microwave passive RFID transponders’, IEEE Trans.
Microw. Theory Tech., 2005, 53, (9), pp. 2978–2990
407