Notice: Wiley Online Library will be unavailable on Saturday 30th July 2016 from 08:00-11:00 BST / 03:00-06:00 EST / 15:00-18:00 SGT for essential maintenance. Apologies for the inconvenience.
 The intense dust devils and dust storms on Mars are believed to generate large electrostatic fields that significantly alter geophysical and geochemical processes on the planet. The existence of such fields must be related to a mechanism by which charged dust separates by polarity; it has been widely hypothesized that this separation originates from a particle-size dependence of the charge polarity, but this effect has never been demonstrated. To address this issue, we carry out experiments on the triboelectric charging of Martian regolith simulant (JSC-1 Mars), using a fluid flow apparatus wherein only particle-particle interactions occur, as is the case in Martian dust events. Our experiments show direct evidence that smaller particles tend to charge negatively and larger particles tend to charge positively, which provides a mechanism for the charge separation that creates electric fields in Martian dust events.
If you can't find a tool you're looking for, please click the link at the top of the page to "Go to old article view". Alternatively, view our Knowledge Base articles for additional help. Your feedback is important to us, so please let us know if you have comments or ideas for improvement.
 Electrostatic effects can significantly alter the geologic and atmospheric processes on Mars. Electrostatic charge of particles can enhance saltation [Kok and Renno, 2006, 2008], and can lead to the formation of geological features such as ‘razorbacks' that have been observed on Mars [Shinbrot et al., 2006]. Electron avalanches generated by the charged particles can activate chemical reactions such as the formation of hydrogen peroxide and the dissociation of methane, and thus alter the composition of the Martian atmosphere [Atreya et al., 2006; Farrell et al., 2006b; Kok and Renno, 2009]. Furthermore, in regard to robotic and human missions to Mars, charged dust can adhere to equipment and disrupt their operation (e.g., by covering solar panels) or cause permanent damage.
 Previous studies have examined the electrostatic charging behavior of Martian regolith simulant in order to elucidate the mechanism of electrostatic charging. The effective workfunction of regolith simulant was determined in studies carried out by contacting the simulant with various other materials [Gross et al., 2001; Sternovsky et al., 2002; Sharma et al., 2008]. Experiments also examined the electrostatic discharges that develop from charging the simulant by stirring the particles with a rod [Krauss et al., 2003; Fábian et al., 2001] and by mixing the stimulant with other materials [Krauss et al., 2006], and the electrostatic charging that develops during flow through a wind tunnel [Merrison et al., 2004]. However, to the best of our knowledge, no previous work has directly demonstrated a particle size dependent charging of Martian regolith, arising from only particle-particle contact, even though this is the key factor that generates electric fields in dust events. In this letter, we carry out experiments to probe this effect in Mars regolith simulant.
2. Experimental Procedure
 We have recently developed a methodology to investigate the triboelectric charging in granular materials due strictly to particle-particle interaction [Forward et al., 2009a, 2009b]; i.e., the particles being examined have no contact with any surface besides that of other particles, until after they are separated by charge polarity and collected. In our experimental apparatus, a bed of particles (∼350 mL) sits on a distribution plate that has a single hole at its center (hole diameter 200 μm). The bed is poured and sprayed with ionized air before each experiment to neutralize any initial charge on all particles. A stream of gas (nitrogen) is then passed through the hole in the distribution plate to fluidize the central region of the bed while the region near the container walls remains stagnant. This leads to a fountain-like flow, as shown in Figure 1a, in which the particles involved in the flow contact only other particles, and not any other surfaces such as the container wall. This particle flow apparatus is operated in a controlled low pressure (70 Torr) nitrogen environment to eliminate the variability in the ambient environment and reduce the effects of contaminants such as water vapor.
 After the particle flow apparatus is run for a sufficient time to reach steady-state charge behavior, the gas flow is stopped and the bed is removed from the vacuum chamber (the flow is run for 120 minutes which has been shown, using electrostatic probe measurements, to result in steady state charging [Forward et al., 2009b]). A non-contact method is then used to collect particles of a specific charge polarity. Particles are extracted from the bed using an electrically biased copper disk, 2.5 cm in diameter, suspended approximately 1 mm above the center of the bed (i.e., the region of the bed that was flowing). The disk is covered with parafilm that acts as a dielectric spacer to prevent the particles from losing their charge and falling back into the bed. A positive voltage (10 kV) is placed on the disk to extract negative particles from the bed, and a negative voltage (−10 kV) is placed on the disk to extract positive particles. Thus, the positively and negatively charged particles are extracted before touching any other surface (note that the particles touch the parafilm only after they are extracted). This charging and collection process is repeated five times for a single experimental trial in order to collect a total of ∼1.5 grams of particles of each polarity. After the particles are collected, they are rinsed with methanol to neutralize their charge and separate any charge-induced agglomerates. The size distributions of the positive and negative collections of particles are then determined with a Coulter LS 230 Particle Size Analyzer.
 The experiments were carried out with JSC-1 Mars regolith simulant (Planet LCC), which is volcanic material collected from the southern flank of the Mauna Kea volcano in Hawaii. The simulant is mainly composed of SiO2, AlO2 and Fe2O3 (43.5 wt%, 23.3 wt% and 15.6 wt%, respectively) with grain size, density, porosity, reflectance spectrum, mineralogy, chemical composition, and magnetic properties similar to the regolith found on the Martian surface [Allen et al., 1998; Fábian et al., 2001].
 We first present results of our control test where the particle extraction is performed before the fluid flow apparatus is operated. As shown in Figure 1c, very few particles are collected on the disk, because the particles are essentially uncharged at this point (the ionized air treatment neutralized the particle charges). In comparison, many particles are collected on the disk when the procedure is carried out after the fluid flow apparatus is run, as shown in Figure 1d.
 The particle size distributions for the samples of positive and negative particles are shown in Figure 2a. These results represent the average of three experimental trials, and the error bars represent the standard error of the results from the three trials. In Figure 2b, the particle size distribution for the original sample is compared with the total particle size distributions of the collected samples, obtained by averaging the size distributions of the positive and negative samples.
 It is clear from Figure 2a that the size of negatively-charged particles tends to be smaller than the positively-charged particles. The mean diameters of the negatively- and positively-charged particles are estimated to be 296 ± 21 μm and 400 ± 31 μm, respectively.
 The comparison in Figure 2b shows that the particle size distribution of the collected samples does not exactly match that of the original sample. In particular, the collected samples are depleted of very large and very small particles. This result is understandable, in that the very large particles are more difficult to extract due to gravity (their mass is larger), and the smaller particles are more difficult to extract due to their greater interparticle cohesive forces [Kok and Renno, 2006].
 It remains unclear why triboelectric charging would occur in a single component system such as the Mars dust. In particular, it is counter-intuitive that charging would occur at all, since charge transfer should be driven by differences in material properties (e.g., work function). However, the results found in this study suggest that particle size differences provide a driving force for charge transfer. This idea is also supported by previous reports of agitated samples of Mars regolith simulant producing more electrical discharges when a sample contains a broad particle size distribution [Krauss et al., 2003].
 Our results also show that bipolar charging in the Mars regolith simulant occurs such that smaller particles tend to charge negatively and larger particles tend to charge positively. We believe that this size dependence is universal (i.e., independent of material properties) in single component granular systems, due to its being observed in a wide range of systems described below. We have recently demonstrated this result in both soda lime glass systems and polyethylene systems [Forward et al., 2009a], and previous experiments by Castle, Inculet and co-workers have led to the same conclusion in various polymer systems [Ali et al., 1998; Zhao et al., 2002, 2003; Inculet et al., 2006]. Furthermore, this size dependence of particle charges is consistent with the direction of electric fields observed in terrestrial dust devils [Freier, 1960; Crozier, 1964; Stow, 1969; Ette, 1971; Farrell et al., 2004] and in volcanic plumes [Miura et al., 2002]. However, the reason why triboelectric charging would depend on the particle size is unclear because the particles are macroscopic (10–1000 μm) and thus their sizes would seem to be irrelevant in regard to charge transfer (i.e., the chemical properties are not a function of physical size as may be the case for nanoscale materials).
 We have recently proposed a mechanism to explain a universal dependence of charging and polarity on particle size [Lacks et al., 2008]. The key idea is that the electrons in insulators are not at equilibrium, and surface contact can allow electrons trapped in high energy states on one surface to relax to lower energy states on another surface [Lowell and Truscott, 1986]. We have shown that this simple effect alone causes smaller particles to charge negatively and larger particles to charge positively, respectively, as illustrated in Figure 3. In this schematic, red (darker) dots represent electrons in high energy states, green (lighter) dots represent electrons that have relaxed to low energy states, and the surface area of the large particle is exactly twice that of the small particle. Initially, the two particles have an equal surface density of high energy electrons, given as the number of electrons trapped in high energy states per surface area (Figure 3a). The first collision between the particles (Figures 3a–3d) does not lead to net electron transfer, but does lead to an asymmetry in the surface density of high energy electrons (Figure 3d); this asymmetry is produced even though there was no asymmetry prior to the collision and the electron transfer during the collision was symmetric. After this first collision, a second collision can occur in one of 4 ways (Figures 3e–3h), depending on whether a high energy electron on either particle is near the point of collision; the probability of each type of collision is shown in Figures 3e–3h. Net electron transfer occurs in collisions iii and iv, since in these cases only one of the two particles has a high energy electron near the point of collision. Since collision iv is more likely to occur than collision iii, the smaller particle is more likely to gain an electron than the larger particle, and thus is more likely to charge negatively. Note that the key to this mechanism is the non-equilibrium dynamics, in which the high energy electrons can transfer to low energy states on another particle, but the electrons in low energy states do not transfer (they are already in a stable state). This mechanism is described in more detail and more rigorously elsewhere [Lacks and Levandovsky, 2007; Duff and Lacks, 2008; Lacks et al., 2008], and compared quantitatively with experiment [Kok and Lacks, 2009].
 We present the first results that flowing JSC-1 Mars regolith simulant undergoes triboelectric charging such that the smaller particles tend to charge negatively and larger particles tend to charge positively. This particle-size dependence of charge polarity leads to the charge separation that explains the existence of electric fields in Martian dust events. These electric fields, in turn, are believed to significantly alter geophysical and geochemical processes on the planet.
 We thank the American Chemical Society Petroleum Research Fund and the National Science Foundation (grant DMR-0705191) for support of this research, and Aaron Jennings for help with the particle sizing.