2028 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 51, NO.
8, AUGUST 2003
Two Advanced Solar Antenna “SOLANT” Designs
for Satellite and Terrestrial Communications
Stefano Vaccaro, Student Member, IEEE, Juan R. Mosig, Fellow, IEEE, and Peter de Maagt, Senior Member, IEEE
Abstract—A novel hybrid technology where amorphous silicon
(a-Si:H) solar cells are either integrated or physically combined
with printed slot antennas is presented. This basic idea is demon-
strated with the help of two innovative designs where the solar cells
are directly grown on a stainless steel ground-plane or glued onto
a standard copper layer printed on a dielectric substrate. To show
the compatibility of solar cells and antennas, the concept is used to
create a linearly polarized slot antenna, a circularly polarized slot
antenna and a slot array. The concept offers advantages in terms
of surface coverage, volume, weight and electric performance when
compared with a simple juxtaposition of antennas and solar cells.
Index Terms—Circularly polarized array, slot antennas,
SOLANT, solar antenna, solar cells.
I. INTRODUCTION
M OST satellite systems use separate solar cells and an-
tennas, which compete for the use of the limited surface
available. Therefore, combining both could save valuable “real
estate” and reduce the cost design of spacecraft missions, pro-
vided that the antennas and solar cells are compatible. In partic-
ular, solar panels constitute a significant part of communication
satellites, providing large flat surfaces on to which antennas can
be mounted or printed [1]. Printed antennas, commonly used
in microwave communications, are particularly well suited for
this integration, especially in configurations where their radi-
ating elements (patches or apertures) can be isolated from the
feed circuits.
After the work of Tanaka et al. [2], other publications, in-
cluding several by the authors of this paper, have explored the
possibility of placing solar cells directly on the patches [3]–[6].
Other combinations like putting solar cells behind reflectarray
antennas have also been studied [7], [8]. In the case of patch
antennas, the cells cannot be close to the patches’ radiating
edges due to destructive interactions. As technology has been
advancing rapidly, an idea emerged to integrate the antenna and
solar cell into (monolithic) building block units. These units
may then be replicated as desired to create a structure whose
transmitting aperture is also used as the light-collection area.
The available surface where the solar cells can be placed is
therefore limited. The techniques presented in this paper repre-
sent a major breakthrough in this respect. This has essentially
Manuscript received November 14, 2001; revised June 4, 2002.
This work was supported by the European Space Agency under Grant
AO/1-3308/97/NL/NB.
S. Vaccaro and J. R. Mosig are with the Laboratory of Electromagnetics
and Acoustics (LEMA), Ecole Polytechnique Fédérale de Lausanne (EPFL),
CH-1015 Lausanne, Switzerland.
P. de Maagt is with the European Space Research and Technology Centre
(ESTEC), European Space Agency, Noordwijk, The Netherlands
Digital Object Identifier 10.1109/TAP.2003.815424
0018-926X/03$17.00 © 2003 IEEE
been made possible thanks to two factors. Firstly, the introduc-
tion of slot antennas, the ground plane of which can be used
as “negative” collector cathode for the solar cells. Secondly,
the possibility of producing solar cells with arbitrary shapes
that don’t restrict the antenna layout. This has led to the novel
concept of a “fully-integrated” design where the two elements
are physically and functionally merged together into a single
device. This “fully-integrated” SOLANT design represents a
“best-of-both-worlds” solution, since a higher level of integra-
tion of both RF and dc functionality is the objective. Moreover,
the use of slots instead of patches, allows the solar cells to be
directly deposited on to the slot’s ground plane, resulting in an
efficient and lightweight device.
The concept can also have interesting spin-offs for terres-
trial applications, for example: roof tiles used for simultaneous
power generation and reception of satellite TV, isolated base sta-
tions for mobile telephone, alarm systems, sea buoys, container
tracking, etc.
II. SOLANT SATELLITE APPLICATIONS
Space-related applications of SOLANT antennas, in a first
approach, can be split into two categories of missions:
• Large spacecraft needed to comply with the increasing de-
mand for telecommunication, multimedia and increasing
scientific objectives. The need to implement several com-
plex instruments on one single spacecraft leads to an in-
crease in the number and size of antennas and solar gen-
erators.
• Constellations of numerous smaller satellites (mini, micro,
nano-satellites) where the integration of solar cells with
antennas can lead to considerable reduction in spacecraft
size, mass and cost.
One can make a preliminary review of the missions that could
profit from integration of antennas with solar cells, where the
emphasis would be on:
• Missions where there is a natural coherence between sun
direction and antenna boresight. This option may be very
attractive for the Future Science Program; both in terms
of deep space missions to the outer planets (Mars and
beyond) and missions at the L2 Lagrangian point of the
earth-sun system. At this location, the sun, the earth, and
the moon are all located behind the payload which points
away from the Sun.
• Missions where there is an advantage to conform the an-
tenna to the spacecraft body (e.g., spinning satellites). An
obvious example of such a mission is a one-axis stabilised
satellite. In this case a large cylindrical part of the satellite
VACCARO et al.: TWO ADVANCED SOLAR ANTENNA “SOLANT” DESIGNS
(a)
(b)
Fig. 1. (a) Cross-section of an amorphous Silicon solar cell, showing the
different layers (thickness not to scale). (b) Picture of a a-Si:H SOLANT
module.
surface can be covered by integrated solar cells and radi-
ating elements. Each of them faces halftime the sun and
the earth. The subarrays of the integrated antennas with
solar cells should not be fed all at the same time, but an
electronic feeding network will have to allow for a con-
tinuous rotation of the beam that opposes to the satellite
spinning.
• Missions requiring very large antennas and/or solar arrays.
• Missions requiring very small spacecrafts.
For all of the above missions there is the option to save “real
estate” by combining two functions maintaining performance
equal or to increase capability (e.g., to offer communication in
a nonnominal situation). SOLANT could even be a mission en-
abling solution.
III. AMORPHOUS SILICON (A-SI:H) THIN FILM SOLAR CELL
TECHNOLOGY
To demonstrate the SOLANT principle, amorphous silicon
(a-Si:H) solar cells developed by IMT/University of Neuchâtel,
Switzerland have been selected [9]. These cells are thin, light,
flexible films that can be cut to fit arbitrary patch or array shapes
and, eventually, conformed to curved surfaces. Another funda-
mental advantage of these solar cells is the fact that they do not
involve a rigid semi-conductor wafer as substrate (germanium,
silicon, etc…).
They can also be grown on plastic or glass as well as on metals
such as stainless steel or aluminum.
Fig. 1(a) shows a lateral view of an amorphous silicon solar
cell on a plastic film (polyimide). The polyimide substrate
is covered by an aluminum layer (back contact) and by a
ZnO layer, which prevents aluminum diffusion. The layers
2029
mentioned above also form a mirror for the incident solar
light in order to increase absorption and consequently, produce
more current. The actual solar cell is made of three silicon
layers: a thin high conductivity phosphor doped n-layer (about
), an intrinsic (undoped) layer with a low defect
density and a very thin highly conductive p-layer.
Most photons are absorbed in the intrinsic layer, while the
doped layers are responsible for the build-up of an electric
field within the solar cell itself. The collector layer on top of
the cell must be both transparent and conductive and hence
it is made of transparent conductive oxides (TCOs)—either
indium tin oxide or zinc oxide. Since TCOs have a rather low
conductivity compared to metals, a finger pattern is deposited
on top [see Fig. 1(b)], made of a Cr adhesion layer covered
by a thicker Ag layer. Overall, the cell thickness is less than 5
but usually a 50- thick Polyimide substrate is used as a
support in standalone applications. Although this type of cell
is less efficient than GaAs solar cells which are currently used
in space applications, a-Si:H cells show a better watt/kilogram
ratio due to their light weight. Moreover, they are inexpensive
and have shown promising hardness against cosmic radiation
in initial tests [10].
IV. SLOT BASED SOLAR ANTENNAS
As stated in the introduction, it is possible to carry the integra-
tion of antennas and solar cells a step further by resorting to slot
antennas instead of patches. Slot antennas provide large metallic
surfaces (the ground plane) on which solar cells can be laid or
directly grown. This yields a simple layered structure with high
area-coverage efficiency for the solar cells. In this case, it is pos-
sible to nearly completely cover the ground plane with the ex-
ception of a very small area close to the radiating edges of the
slot antenna, without degradation of the RF performance.
The choice of the slot antennas could introduce some draw-
backs that have to be considered, such as narrower bandwidth
and lower CP performance. However, these drawbacks can be
overcome through efficient design. From an overall system point
of view, slots remain an interesting alternative to patches, par-
ticularly when the effect of solar cells on the RF performance
must be minimized.
In the remainder of this section, we describe three prototypes,
which have been built and measured; a linear polarized slot an-
tenna; a circularly polarized slot antenna; a 6 1 elements cir-
cularly polarized slot array with beam tilt.
A. Linearly Polarized Fully-Integrated SOLANT [11]
The antenna described in this section, has been already pre-
sented in [11]. However, for the sake of completeness and in
order to give the reader a complete overview of all the different
variations of slot based SOLANT’s, some details given here.
This linearly polarized fully integrated unit (Fig. 2) has both
dc and RF functions intimately linked, sharing the same metallic
structure. Neither the antenna nor the solar cells can function
without this common layer.
The ground plane of the antenna is a metallic plate in which
the slots are laser-cut. A special “dog-bone” shape of the slot has
been used in order to obtain a better matching . After the
cutting, the solar cells are directly deposited on the ground plane
2030 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 51, NO. 8, AUGUST 2003

(a)

Fig. 3. Photo of the circularly polarized slot SOLANT.

(b)
Fig. 2. (a) Photo of the linearly polarized slot SOLANT. (b) Diagram of an
exploded view of the antenna.

which in turn is glued onto a 0.8 mm thick epoxy substrate with

a permittivity of . The ground plane therefore also acts
as a dc collector for the solar cells. Fig. 2(a) shows a photo of
the fully integrated SOLANT basic building block and Fig. 2(b) (a)
a diagram of an exploded view.
In order to simplify the chemical deposition of the cells, it
has been decided to use a stainless steel ground plane. Copper
cannot be used as substrate for the solar cells because its atoms
diffuse into the silicon layers, which has a negative impact on
the doping of the layers. This phenomenon does not occur with
a stainless steel substrate. Another characteristic that makes the
stainless steel attractive is the fact that, if properly machined,
it presents a very smooth surface, which is crucial for correct
solar cell deposition. From the antenna point of view, this type
of ground plane has the disadvantage of lower conductivity as
compared to conventionally used copper.
In order to determine the effect of the ground plane material
on the efficiency of the antenna, a comparative measurement
of gain between an antenna based on a copper and a stainless
steel ground plane has been performed [11]. The results of the
measurements indicate that the antenna based on a stainless steel
ground plane presents a lower efficiency than the copper based (b)
one, resulting in a degradation in gain between 0.65 and 0.85 Fig. 4. (a) Solar cell module with a cross slot, (b) Dimensions of the slot and
dB. the feeding line of the circularly polarized slot SOLANT.
A complete set of measurements of this antenna is given in
[11]. tenna, the solar cells have been specially tailored to fit the slot
shape [see Figs. 3 and 4(a)].
B. Circularly Polarized Slot SOLANT The circular polarization is obtained by exciting a cross slot
An alternative design described in this section utilizes a more with a line at 45 with respect to it. By adjusting the length
complex slot shape to obtain circular polarization. For this an- of each branch of the cross it was possible to generate two
VACCARO et al.: TWO ADVANCED SOLAR ANTENNA “SOLANT” DESIGNS 2031

Fig. 5. Measured S and axial ratio of the circularly polarized slot SOLANT with and without solar cells.

Fig. 6. Measured radiation patterns of the circularly polarized slot SOLANT at 3.84 GHz. Gain with cells
orthogonal resonant modes at two close frequencies [12]. At
some intermediate frequency, the modes are excited with equal
amplitude and phase shift of 90 . This type of circularly polar-
ized element, as depicted in Fig. 3, has a narrow bandwidth but
offers the advantage of requiring a very simple feeding circuit.
An open stub on the feeding line has been used to tune the
circularly polarized element’s frequency to the best matching
frequency. The exact dimensions of the cross-slot and feeding
line are given in Fig. 4(b). The antenna has been built on an
Ultralam 2000 substrate and a
= 3.7 dBi. Gain without cells = 3.8 dBi.
back reflector has been placed at 10 mm from the feeding
lines. The measured characteristics of the antenna, depicted
in Fig. 5, show an axial ratio (AR) under 3 dB between 3.82
GHz and 3.94 GHz (3%), that corresponds to a return loss
under (3.82–3.92 GHz). Furthermore, it is apparent
that these parameters are not affected by the presence of the
solar cells. The same can be said about the radiation pattern
and gain. In fact, the measurements depicted in Fig. 6, show
hardly any difference between the antenna with and without
the solar cells.
2032 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 51, NO. 8, AUGUST 2003

Fig. 7. SOLANT 6 2 1 slot array.

C. SOLANT 6 1 Beam Tilted Slot Array

The next step was to demonstrate that beam tilt can be ob-
tained with SOLANT antennas and that this is unaffected by
the solar cells. For the sake of simplicity, the previously de-
scribed circularly polarized slot cross design was used as array
element and a 6 1 linear array has been built to obtain a 15
beam tilt. The phases for each element were imposed by prop-
erly adjusting the length of the feeding network (phase shifts).
The spacing between the elements is 0.81 (64 mm) which
was given by the module size. To improve the gain, a back re-
flector was placed at 10 mm behind the antenna. Furthermore,
an MMIC amplifier has been integrated on the back of the an-
tenna, which is powered by solar cells that produce 0.821 Watt (a)
with 122 mA at 7 V.
As was explained in Section III, depending on the metal being
used, the solar cells can either be directly grown as a thin film
or glued directly onto the ground plane. For efficiency reasons
copper was selected as ground plane and the cells glued on to
it. The solar cells module, depicted in Fig. 4(a), consists of four
solar cells connected in series with a cross cut in the middle to
allow the RF radiation. Six solar cell modules have been used,
one per radiating element.
Fig. 8(a) and (b) present, respectively, the radiation patterns
in the XZ (six element direction) and YZ (one element direction)
planes for the antenna with and without solar cells at the fre-
quency of 3.8 GHz. As expected, the XZ pattern presents the (b)
principal lobe tilted at 15 while a secondary lobe appears at
2
Fig. 8. Measured radiation patterns at 3.80 GHz for the 6 1 slot array. (a)
60 . It can be seen that the presence of the solar cells does not
= =
XZ plane, (b) YZ plane. Gain without cells 11.00 dBi. Gain with cells 11.6
affect the radiating characteristics in this plane. dBi.
Due to the fact that in comparative measurements the an-
tennas need to be assembled and disassembled several times, it with and without solar cells are attributed to these imperfec-
is difficult to guarantee absolute repeatability between the radi- tions. Moreover, it is also important to point out the fact that
ation pattern measurements with and without solar cells. How- the patterns with and without solar cells are normalized to their
ever, the antenna replacement on the positioning system led to respective maximum value and not to the same reference. For
only minor changes in the radiation pattern measurements. this reason it is not possible to compare directly the levels for
The radiation pattern measurement in the YZ plane is most the two situations. Nevertheless, a comparison of the efficiency
sensitive to the antenna orientation, and consequently the slight between the antenna with and without solar cells can be made
differences that can be observed in Table I between the antenna with the gain values reported in Table I.
VACCARO et al.: TWO ADVANCED SOLAR ANTENNA “SOLANT” DESIGNS
TABLE I
GAIN VALUES OF THE 6 2 1 ARRAY WITH AND WITHOUT SOLAR CELLS AT DIFFERENT FREQUENCIES
Fig. 9. Return loss of the 6 2 1 slot array with and without solar cells.
The values of the maximum gain in the direction of the beam
range between 9.7 and 11.6 dBi, depending on frequency. The
anomalous result at 3.91 GHz is most likely due to measure-
ment instability. Overall, a check over the useful frequency band
shows that the variation of the antenna gain due to the solar cells
is less than 1 dB, which is within the accuracy of the measure-
ment setup.
Finally, Fig. 9 represents the return loss for the antenna with
and without solar cells. The antenna presents return losses lower
than between 3.7 and 3.9 GHz, with the minimum at
3.84 GHz. It should be noted once more that the array was built
with the basic element as described in Section IV-B. As such,
the array performance is not optimum. However, the mechanism
of beam tilt is demonstrated and the effect and the effect of the
solar cells on radiation performance is shown to be minimal.
V. CONCLUSION
A novel concept of integration between antennas and solar
cells has been presented and demonstrated. Thanks to the use of
the versatile amorphous silicon solar cells, it has been possible
to develop slot antenna solutions that make more efficient use of
the space available for solar cells. The amorphous silicon solar
cells can be adapted to each antenna shape, either by cutting
them or by directly growing them on the antenna surface.
The presented designs represent different levels of integra-
tion, potentially useful in diverse space-related and terrestrial
applications. The fully-integrated slot design can simultane-
2033
ously comply with all of the standard RF requirements (gain,
polarization, radiation pattern) as well as the standard dc
requirements (power, voltage, current) in many situations.
The results obtained in this paper, utilizing standard technolo-
gies, show that it should be possible to realize large arrays of
SOLANT antennas with circularly polarized operation, beam
tilt capability and sizeable dc power production.
ACKNOWLEDGMENT
The authors particularly wish to thank Dr. P. Torres,
Dr. M. Götz, J. Kündig, and Prof. A. Shah of IMT, Uni-
versity of Neuchâtel, Switzerland, Prof. A. Skrivervik and
Mr. J.-F. Zürcher of LEMA-EPFL, Switzerland. Finally, the
authors wish to acknowledge the support of the Swiss Space
Office at Bern, Switzerland.
REFERENCES
[1] B. Pattan, Satellite-Based Cellular Communications. New York: Mc-
Graw-Hill, 1997.
[2] M. Tanaka, Y. Suzuki, K. Araki, and R. Suzuki, “Microstrip antennas
with solar cells for microsatellites,” Electron. Lett., vol. 31, pp. 263–266,
1996.
[3] S. Vaccaro, P. Torres, J. R. Mosig, A. Shah, J.-F. Zürcher, A. K.
Skrivervik, F. Gardiol, P. de Maagt, and L. Gerlach, “Integrated solar
panel antennas,” Electron. Lett., vol. 36, no. 5, pp. 390–391, Mar. 2000.
[4] S. Vaccaro, P. Torres, J. R. Mosig, A. Shah, A. K. Skrivervik, J.-F.
Zürcher, P. de Maagt, and L. Gerlach. Integration of antennas and solar
cells for satellite communications. presented at Proc. Millenium Conf.
Antennas and Propagation, Symp.. [CD-ROM] Session 4A7, paper No
0954
2034 IEEE TRANSACTIONS ON ANTENNAS AND PROPAGATION, VOL. 51, NO. 8, AUGUST 2003
[5] , “Combination of antennas and solar cells for satellite communi-
cations,” MOTL, vol. 29, no. 1, pp. 11–16, Apr. 2001.
[6] J. Huang, “Mars rover antenna for solar array integration, JPL TMO
progress report 42-136,”, JPL report is publicly available, 1999.
[7] , “Capabilities of printed reflectarray antennas,” in Proc. IEEE
Phased Array Systems and Technology Symp., Boston, MA, Oct. 1996,
pp. 131–134.
[8] M. Zawadzki and J. Huang, “Integrated RF antenna and solar array for
spacecraft application,” in Proc. IEEE Phased Array Systems and Tech-
nology Conf., Dana Point, CA, May 2000, pp. 239–242.
[9] P. Pernet, M. Goetz, X. Niquille, D. Fischer, and A. Shah, “Front contact
and series connection problems of a-SI:H solar cells on polymer film
substrates,” in Proc. 2nd World Conf. Photovoltaic Energy Conversion,
vol. I, Vienna, Switzerland, July 1998, pp. 976–979.
[10] J. Kuendig, M. Goetz, J. Meier, P. Torres, L. Feitknecht, P. Pernet, X.
Niquille, A. Shah, L. Gerlach, and E. Fernandez, “Effect of proton irra-
diation on the characteristics of different types of thin-film silicon solar
cells,” presented at the Proc. 16th EU Photovoltaic Solar Energy Con-
ference, Glasgow, U.K., May 2000.
[11] S. Vaccaro, P. Torres, J. R. Mosig, A. Shah, J.-F. Zürcher, A. K.
Skrivervik, P. de Maagt, and L. Gerlach, “Stainless steel slot antenna
with integrated solar cells,” Electron. Lett., vol. 36, no. 25, pp.
2059–2060, Dec. 2000.
[12] T. Vlasits, E. Korolkiewicz, A. Sambell, and B. Robinson, “Performance
of a cross-aperture coupled single feed circularly polarized patch an-
tenna,” Electron. Lett., vol. 32, no. 7, pp. 612–613, Mar. 1996.
Stefano Vaccaro (S’00) was born in Locarno, Switzerland. He received the
M.Sc. degree in electrical engineering from the Ecole Polytechnique Fédérale de
Lausanne (EPFL), Switzerland, in 1998, where he is working toward the Ph.D.
in the Laboratory of Electromagnetics and Acoustics.
After a short industrial experience, he joined the Laboratory of Electromag-
netics and Acoustics at EPFL in 1999. His current research interests are in the
development of microstrip antennas and in their integration with solar cells.
Juan R. Mosig (S’76–M’87–SM’94–F’99) was born in Cadiz, Spain. He
received the electrical engineer degree from the Universidad Politecnica
de Madrid, Spain, in 1973 and the Ph.D. degree from the Laboratory of
Electromagnetics and Acoustics, Ecole Polytechnique Fédérale de Lausanne
(EPFL), Switzerland, in 1976.
In 1976, he joined the Laboratory of Electromagnetics and Acoustics, Ecole
Polytechnique Fédérale de Lausanne (EPFL), Switzerland. Since 1991, he has
been a Professor at EPFL and, since 2000, the Head of the Laboratory of Elec-
tromagnetics and Acoustics. In 1984, he was a Visiting Research Associate at
Rochester Institute of Technology, Rochester, NY. He has also held scientific
appointments at universities of Rennes (France), Nice (France), Technical Uni-
versity of Denmark and University of Colorado at Boulder. He is the author of
four chapters in books on microstrip antennas and circuits. He is Coorganizer
and Lecturer of yearly short intensive courses in numerical electromagnetics
(Europe and USA). His research interests include electromagnetic theory, nu-
merical methods and microstrip antennas.
Dr. Mosig is a Member of the Swiss Federal Commission for Space Appli-
cations and is responsible for several research projects for the European Space
Agency.
Peter de Maagt (S’88–M’88–SM’02) was born in Pauluspolder, The Nether-
lands, in 1964. He received the M.Sc. and Ph.D. degrees from Eindhoven Uni-
versity of Technology, Eindhoven, The Netherlands, in 1988 and 1992, respec-
tively, both in electrical engineering.
Currently, he is with the European Space Research and Technology Centre
(ESTEC), European Space Agency, Noordwijk, The Netherlands. His research
interests are in the area of millimeter and submillimeter-wave reflector and
planar integrated antennas, quasioptics, photonic bandgap antennas, and mil-
limeter- and submillimeter-wave components.
Dr. de Maagt was a corecipient of the H. A. Wheeler Award of the IEEE
Antennas and Propagation Society for the Best Applications Paper of 2001 and
a recipient of a 2002 ESA Award for Innovation.