Laboratory Examination of the Electron Avalanche and Breakdown

46th Lunar and Planetary Science Conference (2015)
MARTIAN ATMOSPHERE. W. M. Farrell1. J. L. McLain2., M. R. Collier1., J. W. Keller1., T. L. Jackson1., and
G. T. Delory3; 1. NASA/Goddard Space Flight Center, Greenbelt MD ([email protected]); 2. University of
Maryland, College Park, MD; 3. University of California, Berkeley, CA
Abstract: Viking era laboratory experiments
show that mixing tribo-charged grains in a low pressure
CO2 gas can form a discharge that glows, indicating the
presence of an excited electron population that persists
over many seconds. Based on these early experiments
[1], it has been predicted that Martian dust devils and
storms may also contain a plasma and new plasma
chemical species as a result of dust grain tribo-charging
[2]. In this work, we examine the possible breakdown
in a Mars’s-like atmosphere under controlled circumstances. We conclude that in a Mars-like low pressure
CO2 atmosphere and expected E-fields, the electron
current remain in a dark ‘Townsend’ discharge where
the electron density is exponentially growing with applied E.
Laboratory Measurements: In order to
quantify the atmospheric currents generated under a
driving E-field in a low pressure CO2 gas, we performed a systematic laboratory study of the breakdown
process. In an analogy to thunderstorms, the
triboelectric process in dust devils tends to charge
smaller dust (< 20 microns) negative but leave larger
sand grains (~100 microns) and the surface positive.
Vertical winds in the storm then transport and separate
the charged grains, lofting the negatively charged light
dust to high altitudes relative to the larger sand grains
and the surface. In terrestrial dust devils, large E-fields
have been reported to develop by this electrical generation process [2].
In a simplified description, the charge concentration centers in the convective feature can be thought
of as a charged capacitor plate containing -Q at high
altitudes and +Q at the surface, and having a triboelectric generated E-field between the plates (i.e., behaves like a dipole and can be sensed externally from
the feature). To simulate this effect in the laboratory,
an electrostatic plate system has been placed in a CO2rich environment at low pressure to simulate the conditions on the surface of Mars.
Figure 1 shows inside the chamber containing
the parallel plates where the E-field is generated. The
chamber has the ability to maintain vacuum to 10-8
Torr but in our Mars applications that level of vacuum
is not required. All experiments are run for a CO2 gas
at 5 Torr. Ambient air is removed to a level of ~ 0.1
Torr and then ultra-high purity grade CO2 gas is leaked
into the chamber to obtain an ultimate pressure of 5
Figure 1. The two plates in the test chamber, with the photodiode board assembly located to the left.
Two circular parallel plates of 7 cm radius
form the capacitor that can be separated from ~0.1 cm
to ~7 cm via an computer-controlled manipulator, allowing plate separation to be set without having to
break vacuum. A UV photo-diode is used to stimulate a
low level of electron emission to initiate the electron
avalanche. A voltage drop across a resistor in series
with the plate capacitor provides an indication of the
plate-created atmospheric current.
Early Results. Figure 2 shows a plot of
measured (a) current and (b) equivalent electron conductivity as a function of driving E-field between the
plates. The value of E is V/d, where V is the voltage
applied to the plates from a high voltage power supply
and d is the plate separation. We make special note that
the effective conductivity shown in (b) is Je/E. At relatively small E values, the effective conductivity is the
ambient (bulk) isotropic conductivity, σ = Je/E. However, as E increases, the electron avalanche initiates an
exponential growth in electron density and electron
drift speeds increase, the avalanche current becomes
directional along E, and the conductivity is then that of
the σzz element in the conductivity tensor. This increase
in E field-aligned electron conductivity creates a subtle
but important effect in applications to grain charging:
Past studies have viewed the electron avalanche as an
effective increase in bulk conductivity, but it is actually
an electron beam flowing along the E direction.
46th Lunar and Planetary Science Conference (2015)
As evident in Figure 2, there are three separate current regimes:
1) For E-fields below about 25 kV/m, the current varies linearly with E-field, J = σE, behaving like
a nominal atmosphere of conductivity near ~ 10-12 S/m.
Fig 2. The a) current and b) effective electron conductivity as
a function of E-field for plate separations (from left to right
in the figure) of 60, 40, 20, and 8 mm.
2) Between 25 and 100 kV/m and for currents
below ~1 µA, the gas is undergoing an electron avalanche process, where the electron density is exponentially increasing with driving E-field. This portion of
the curve is commonly called the ‘Townsend’ discharge, with the atmospheric conductivity increasing
exponentially with E. This ‘dark’ discharge, having no
obvious illumination, occurs preceding a spark discharge [3]. In the electron avalanche, the electron density increases as n = no exp(α(Ε)d), where α(Ε) = αο
exp(-Eo/E). The quantities αο and Eo for a low pressure
CO2 gas can be derived [2] or found in texts on the
subject (e.g., Table 6.1 in [4]) The electron conductivity in Figure 2b is no longer directly proportional to E
in the Townsend discharge regime, but instead displays
the obvious exponential increase with E (i.e., the effective conductivity along E is now σ ~ no exp(α(Ε)d) eµ,
where µ is the electron mobility). The ions, on the other hand, are not as easily accelerated along E, with
their relative drift speeds in proportion to their mass.
3) At a threshold electric field (which differed
for each value of plate separation distance, d), the currents make an abrupt increase by over a factor of 100,
to initiate an observable spark discharge in the chamber. These discharges typically are found with currents
exceeding ~ 0.1 mA, and the increase or jump in cur-
rent at the threshold E is nearly a factor of 100 (note
the clear ‘gap’ in measurement values from 10-5 to 10-3
Ampere levels, from ‘Townsend’ to ‘spark’ discharges,
in Figure 2a).
We should note that the term ‘spark’ discharge is a misnomer: Even at currents at 10’s of µA in
between the plates, the electron density is still very
low, near 1 part in 1-10 billion of the neutral gas density. Hence the gas is still very weakly ionized. As such,
this is not a filamentary discharge in the same sense of
terrestrial lightning, where the gas is very hot and
~100% ionized.
Conclusion: Typical dust devil tribo-electric
currents found at Earth and expected for Mars are illustrated by the dotted line in Figure 2a. We find that for
typical dust devil charging currents, we should not expect the dust devil feature to initiate a spark discharge.
However, enhanced electrification and electron avalanche should be expected to be driven in the system.
Specifically, the dust devil is expected to be in the dark
discharge ‘Townsend’ regime, with electron densities
exponentially growing with increasing E.
We also find that the current densities generated in the Townsend discharge are not capable of
short-circuiting a tribo-charging dust devil (having
charging currents at Jc ~10 µA/m2). For the most part,
the currents in the electron avalanche will remain below Jc. However, as the E-field increases, the elecrtron
current, Je will become comparable to Jc and in doing
so will limit the E-field growth, dE/dt, (but not shut
itself off or quench).
References. [1] Eden HF and B Vonnegut,
(1973). Electrical breakdown cause by dust motion in
low pressure atmospheres: Consideration for Mars,
Science, 180, 962. Mills, AA, (1977). Dust cloud and
frictional generation of glow discharges on Mars, Nature, 268, 614. [2] Delory, GT et al. (2006), Oxidant
Enhancement in Martian Dust Devils and Storms:
Storm Electric Fields and Electron Dissociative Attachment, Astrobiology, 6, 451. Jackson, T. L., and W.
M. Farrell (2006), Electrostatic fields in dust devils:
An analog to Mars, IEEE Trans. Geosci. Remote Sens.,
44(10), 2942– 2949, doi:10.1109/TGRS.2006.875785.
[3] Llwellyn-Jones, F. (1966), Ionization and breakdown in gases, Chapman and Hall, London. [4] Brown,
S. C. (1966), Introduction to electrical discharges in
gases, Wiley and Sons, NY.