preliminary broadband measurements of dielectric permittivity of

46th Lunar and Planetary Science Conference (2015)
D. Hickson3, R. Ghent1,4, and M. Daly3, 1Department of Earth Sciences, University of Toronto, 22 Russel St., Toronto, ON, Canada, M5S 3B1, Email: [email protected], 2Department of Physics, University of Toronto, 60
St. George St., Toronto, ON, Canada, M5S 1A7, 3Centre for Research in Earth and Space Science, York University,
4700 Keele St., Toronto, ON, Canada, M3J 1P3, 4Planetary Science Institute, 1700 East Fort Lowell, Tucson, AZ,
USA, 85719-2395.
Introduction and background: When considering radar observations of airless bodies containing regolith, the radar backscatter coefficient is dependent
upon the complex dielectric permittivity of the regolith
materials. In many current applications of imaging
radar data, uncertainty in the dielectric permittivity
precludes quantitative estimates of such important parameters as regolith thickness and depth to buried features (e.g., lava flows on the Aristarchus Plateau on the
Moon and the flows that surround the Quetzalpetlatl
Corona on Venus). For asteroids, radar is an important
tool for detecting and characterizing regoliths. Many
previous measurements of the real and/or complex
parts of the dielectric permittivity have been made,
particularly for the Moon (on both Apollo samples and
regolith analogues) [e.g. 1, 2]. However, no studies to
date have systematically explored the relationship between permittivity and the various mineralogical components such as presence of TiO2 and FeO. For lunar
materials, the presence of the mineral ilmenite (FeTiO3), which contains equal portions FeO and TiO2, is
thought to be the dominant factor controlling the loss
tangent (tanδ, the ratio of the imaginary and real components of the dielectric permittivity). Ilmenite, however, is not the only mineral to contain iron in the lunar
soil and our understanding of the effect of iron on the
loss tangent is insufficient. Beyond the Moon, little is
known about the effects on permittivity of carbonaceous materials. This is particularly relevant for missions to asteroids, such as the OSIRIS-REx mission to
(101955) Bennu, a carbonaceous asteroid whose regolith composition is largely unknown. Here we
present preliminary broadband (300 kHz to 14 GHz)
measurements on materials intended as planetary regolith analogs. Our ultimate goal is to establish a database of the effects of a wide range of mineralogical
components on dielectric permittivity, in support of the
OSIRIS-REx mission and ongoing Earth-based radar
investigation of the Moon. In addition to facilitating
quantitative interpretation of lunar radar data, our results will provide context into which samples returned
by OSIRIS-REx will fit, and will therefore inform future remote exploration of asteroids.
As shown by [2, 3] and others, the loss tangent is
particularly important when inverting for regolith
thickness from radar observations. In the simplest case
where scattering from roughness and rocks is ignored
and where the signal frequency is known, radar attenuation (and therefore the backscatter coefficient) is dependent on the parameter ɛ’’d/sqrt(ɛ’) where d is the
thickness of the regolith, ɛ’ is the real part of the complex permittivity (referred to as the dielectric constant),
and ɛ’’ is the complex part of the permittivity (referred
to as the loss factor)[2]. Often when considering the
dielectric loss of a material the loss tangent, tanδ, is
reported. The loss tangent is the ratio of ɛ’’ to ɛ’.
Since the dielectric properties of
powdered materials depend on several parameters (e.g.
grain size and size distribution, density, porosity, mineralogy, temperature, frequency, moisture) we intend to
systematically vary parameters with a focus on mineralogy. We begin by varying TiO2 content in a base
powder of alumina oxide under ambient conditions.
Future plans include measurements at different temperatures and under vacuum conditions. We will begin
to investigate mineralogy by varying FeO content as
well as a combination of both TiO2 and FeO. These
results will be compared with the mineral ilmenite
which will be crushed and sieved into a powder of
known grain size. Carbonaceous materials which are
relevant to carbonaceous asteroids such as Bennu are
also planed to be measured with this technique.
Method: We measure both parts of the dielectric
permittivity using an Agilent Technologies E5071C
ENA Series network analyzer capable of sweeping
through frequencies from 300 kHz to 14 GHz. The
network analyzer is connected to a coaxial transmission line (either 7 mm diameter and 10 cm length or 14
mm diameter and 15 cm length) which is filled with a
powdered material according to the methodology established in [4]. Signals generated from the network
analyzer are reflected from either end of the sample,
and transmitted in both directions through the sample;
the resulting reflection and transmission coefficients
result in S-parameters, from which we compute the
complex permittivity (dielectric constant and loss factor). We use the non-iterative method established by
46th Lunar and Planetary Science Conference (2015)
[5] and demonstrated for this method in [4]. During a
measurement the airline is kept upright to avoid powder displacement. The network analyzer is calibrated
using known calibration standards prior to each measurement.
Data: One example permittivity measurement is
shown for a powdered sample in Figures 1 and 2. Both
the dielectric constant (ɛ’) and the loss tangent are plotted for frequencies between 500 kHz and 14 GHz. Frequencies bellow 500 kHz were omitted due to large
error in that range. The sample consists of 76 µm alumina oxide powder with the addition of ~0.5 weight%
TiO2 powder. The sample was measured in the 7 mm
airline which has a volume of ~12.44 cm3. The density
of the powder in the airline was 0.41 g/cm3. Note that
dielectric properties of materials also depend on density, porosity and grain shape.
As can be seen in Figure 1 the dielectric constant
for the sample is stable with frequency while the loss
tangent (Figure 2) shows a slight frequency dependence. Relative error in the loss tangent measurements
is also much larger than for the dielectric constant with
the calculated loss tangent sometimes dipping below 0.
It is expected that refinement of the experimental
method, particularly better control on packing and
grain size, will yield more precise loss tangent results.
Figure 1: Plot of dielectric constant measured between 500 kHz and 14 GHz for one sample of 76 µm
alumina oxide powder + 0.5 weight% TiO2 powder
with a density of 0.41 g/cm3 in the 7 mm airline.
Dashed contours correspond the the uncertainty in the
measurement according to [4-6].
Figure 2: Plot of loss tangent measured between
500 kHz and 14 GHz for one sample of 76 µm alumina
oxide powder + 0.5 weight% TiO2 powder with a density of 0.41 g/cm3 in the 7 mm airline. Dashed contours
correspond the the uncertainty in the measurement
according to [4-6].
References: [1] Carrier W. D. et al. (1991) The
Lunar Sourcebook, CUP, 530-552. [2] Fa W. and
Wieczorek M. A. (2012) Icarus, 218, 771–787. [3]
Shkuratov Y. G. and Bondarenko N. V. (2001) Icarus,
149, 329–338. [4] Sotodeh S. (2014) York Univ. Masters Thesis. [5] Boughriet A-H et al. (1997) IEEE
Trans. On Microwave Theory and Techniques, 45,
52-57. [6] Baker-Jarvis J. et al. (1990) IEEE Trans. On
Microwave Theory and Techniques, 38, 1096-1103.