Adsorption of water to fractal dust grains during accretion has been proposed as a possible source of water for rocky planets. We have used computer simulations to study the feasibility of chemisorption onto forsterite dust grains by investigating the adsorption of dissociated water to stoichiometric and defective surfaces. Defects were modeled using steps, corner sites and vacancies on different forsterite surfaces. Our results show that water dissociation is expected on the stoichiometric (100) surface but not on the stoichiometric (010) surface. However, the energies released by dissociative adsorption at steps and corners indicate that the energetic barrier to chemisorption on the (010) surface would be favorable if these features were present. Steps and corners on all surfaces studied produced Mg sites that have low coordination and thus were highly reactive, favoring the dissociation of water. Terrace size between the steps was shown to have a limited effect on the final energies, although smaller terraces created more reactive Mg sites at corners. A simple Langmuir model was used with the energetic data from our simulations to examine the effectiveness of water adsorption at temperature and pressure conditions applicable to the accretion disk. The findings of this study suggest that water would be strongly chemisorbed at fractal forsterite surfaces even at low partial pressures suggesting that water could be retained during planetary accretion.

The presence of water in the Earth has long been an enigma. However, computer modelling techniques have shown that the adsorption of water onto the fractal surfaces of interplanetary dust particles, which are present in the planetary accretion disk, is sufficiently strong to provide a viable origin of terrestrial water.

INSIGHTS FROM ATOMISTIC AND ELECTRONIC-STRUCTURE CALCULATIONS. Krishna Muralidharan, Marilena Stimpfl, Nora H. de Leeuw, Pierre A. Deymier, Keith Runge and Michael J. Drake; Material Science and Engineering, University of Arizona, Tucson, AZ, 85721, USA, Lunar and Planetary Laboratory, University of Arizona, Tucson , AZ, 85721, USA, Department of Chemistry, University College, London , 20 Gordon Street, London WC1H 0AJ, UK

Muralidharan et al. [1] have shown that up to 8 Earth oceans of water could be associatively adsorbed on to grains in the accretion disk at 1 AU. Here we expand on that work using atomistic and electronic-level calculations to investigate dissociative adsorption, and use density functional theory to investigate the difference in D/H ratios between water on Earth and in the accretion disk. We find that dissociative adsorption leads to even stronger bonding than associative adsorption, demonstrating that adsorption must be a significant source of terrestrial planetary water. Using density functional theory we show that HDO may be preferentially retained relative to H2O in adsorption/desorption kinetics. Introduction: Comets, hydrous asteroids, phyllosilicates migrating from the asteroid belt, and hydrous minerals forming in the inner solar system have all been proposed as possible sources of inner solar system water, but there remain unresolved issues with these sources [2]. Recently, Drake [2] proposed that microscopic-level processes such as molecular adsorption of water in the accretion disk could in fact lead to the delivery of water to the inner solar system planets. Previously, adsorption on to grains had been dismissed due to the misconception that adsorption energies were too low to capture and retain gaseous water at high temperatures. We have shown [1,3] that the direct adsorption of water on to forsterite grains, the major silicate phase in protoplanetary disks, could be a significant source of terrestrial planetary water. Specifically, we [3,4] have carried out atomistic and electronic-structure calculations to map out the adsorption energy landscapes of water on forsterite surfaces and showed that water molecules could strongly chemisorb (via chemical bonds) on to forsterite (especially the {100} surfaces in a non-dissociative fashion. Next, we [1] used kinetic Monte Carlo (KMC) simulations in conjunction with the results of [3] to show that water could be retained even at conditions corresponding to the accretion disk (Fig. 1). We concluded that adsorption of gaseous water on to dust grains would start from the early stages of accretion. Many conservative assumptions were made in the above work. For example, only associative adsorption was considered. But forsterite surfaces are characterized by underbonded O and Mg atoms [4], leading to the surfaces being very reactive and consequently making dissociation of a water molecule into H and OH energetically favored. Here we calculate the dissociative adsorption energy landscapes of water on different forsterite surfaces and compare these results with associative adsorption [3]. If the resultant dissociative adsorption energy landscapes are similar to, or much lower than, the previous results, then adsorption is an even more significant source of terrestrial planetary water. This work also addresses why the D/H ratio of water on Earth is higher than in the accretion disk. Currently there is no quantitative theory. Using firstprinciples density functional theory (DFT), we calculate the activation barriers for adsorption/desorption of H2O and HDO molecules in order to see if differences could lead to preferential retention of HDO. If HDO is preferentially retained, as adsorption on to grains in the accretion disk would be a plausible alternative to the hypothesis that Earth’s water came from exogeneous sources such as comets or wet asteroids. Figure 1: Variation in the amount of adsorbed water expressed in terms of Earth-Oceans at different temperatures Models and Methods: Dissociative adsorption energy calculations: Using interatomic potential parameters derived for bulk Mg2SiO4 [4], the different surfaces ({100}, {010}, and {110}) were relaxed using the L-BFGS method [5] (which accurately locates the lowest surface energy configuration). Each surface is characterized by the presence of underbonded O and Mg atoms, with the number of underbonded atoms being a function of the surface geometry. Next, in order to calculate the adsorption energy, using an appropriate potential for water [4], a water molecule was dissociated on the relaxed surfaces such that an underbonded surface oxygen was protonated, while the OH group was bonded to a surface Mg atom. Then, the hydroxylated surface was relaxed using the L-BFGS energy-minimizer and the adsorption energy was calculated according to Eq. 1, [ ] ) ( 2O H s w s ads E E E E + − = + (1) where, Eads corresponds to the dissociative adsorption energy, Es+w is the total potential energy of the energy1882.pdf 40th Lunar and Planetary Science Conference (2009)

The origin of water in the inner Solar System is not well understood. It is believed that temperatures were too high in the accretion disk in the region of the terrestrial planets for hydrous phases to be thermodynamically stable. Suggested sources of water include direct adsorption of hydrogen from the nebula into magma oceans after the terrestrial planets formed, and delivery of asteroidal or cometary material from beyond the zone of the terrestrial planets. We explore a new idea, direct adsorption of water onto grains prior to planetary accretion. This hypothesis is motivated by the observation that the accretion disk from which our planetary system formed was composed of solid grains bathed in a gas dominated by hydrogen, helium, and oxygen. Some of that hydrogen and oxygen combined to make water vapor. We examine quantitatively adsorption of water onto grains in the inner Solar System accretion disk by exploring the adsorption dynamics of water molecules onto forsterite surfaces via kinetic Monte Carlo simulations. We conclude that many Earth oceans of water could be adsorbed.

The origin of water in the terrestrial planets is controversial. Both comets and asteroids have isotopic properties inconsistent with Earth’s water. We are investigating if adsorption of water onto mineral surfaces in the accretion disk could be a viable source of water in the terrestrial planets. Gases coexist with dust in the accretion disk for long periods. The equilibrium H2O/H2 ratio in the accretion disk was 5 · 10 , which corresponds to a pH2O of 10 8 bar. Note that the equilibrium partial pressure is probably a lower limit. Astronomical observations show that dust clouds consist of Mg-rich olivine, pyroxenes, and other refractory minerals with radii 100 kJ/mol). In environments with low pH2O such highenergy sites are most likely the first sites to interact with water molecules. The calculations of (Stimpfl et al.; de Leeuw et al., 2006) using pure forsterite suggest that the inner disk could sustain adsorption of water, both associative and dissociative, onto perfect olivine surfaces (de Leeuw et al., 2000). To gain a better understanding of the mechanism of adsorption, we are performing minimum energy simulations with olivine with Fa up 14%. We are also conducting molecular dynamic simulations to gain an understanding of the kinetics of desorption. In this talk we will present the results from these latest simulations and quantitatively assess the contribution of this mechanism to the water budgets of the terrestrial planets.

MECHANISM TO EXPLAIN WATER IN THE INNER SOLAR SYSTEM. M. Stimpfl , N.H. de Leeuw , P. Deymier, M. J. Drake and A.M. Walker 4 1 Lunar and Planetary Laboratory, University of Arizona, Tucson , AZ, 85721, USA, mstimpfl@lpl.arizona.edu; 2 Department of Chemistry, University College London, 20 Gordon Street, London WC1H 0AJ, UK; 3 Material Science and Engineering, University of Arizona, Tucson, AZ, 85721, USA; 4 Department of Earth Sciences, University of Cambridge, Downing Street, Cambridge CB2 1TS, UK.

The origin of water in the inner solar system is not yet well understood. Because of the coexistence of water and small solid particulates in the accretion disk from which our planetary system formed, we propose that adsorption of water onto the surfaces/pores of forsterite could play and important role in the delivery of water to the rocky planets. In this work we employ energy minimization techniques to understand the surface energy structure and the distribution of surface adsorption sites for the {0 1 0} and {1 0 0} planes in forsterite. This study indicates that most of the surface area is not involved in the initial adsorption of water molecules and that the stronger adsorption sites coincide with the most underbonded surface Mg atoms.

In situ Raman spectroscopy at high pressure was utilized to follow the phase transition of a synthetic sample of Li-aegerine pyroxene (LiFeSi2O6) from its low-pressure (C2/c) phase to its high-pressure (P21/c) phase. The phase change occurred between 0.7 and 1 GPa and was accompanied by a change in coordination of the Li atom from 4 to 5, which was confirmed by single-crystal X-ray diffraction. This is the first report of the Raman spectrum of Li-aegerine in the P21/c phase. As was previously observed with other pyroxenes, additional changes in the Raman spectra were observed at pressures higher than the phase transition, including the splitting of the peak near 700 cm−1, which has traditionally been utilized to indicate the phase transition. Comparisons with the Raman spectra of spodumene in both symmetries are utilized for a discussion of modes. Copyright © 2005 John Wiley & Sons, Ltd.

Two main scenarios can account for the delivery of water to the inner planets: either the water originated outside of the inner solar system and was later delivered to the terrestrial planets or the source of water was local. A source for this ‘endogenous’ water could be represented by dust grains present in the accretion disk onto which water had been previously adsorbed [1]. Monte Carlo simulation of adsorption at nebular T, P and fH2O onto spherical grains showed that this mechanism can store up to 3 times the Earth’s oceans [2]. This model, however, did not take into account the specific surface interactions between water gas and the crystalline surface nor did it investigate rigorously the role of porosity. To fill this gap, we are performing molecular dynamics simulations of the system water olivine using the open source code LAMMPS [3]. The bulk olivine was modelled using periodic boundary conditions (PBC) and Buckingham potentials for the short-range interactions with a cut off distance of 10 Ǻ for both short and long-range interactions. Long-range Coulombic interactions were calculated using the Ewald method. To allow for partially covalent bonds in the silica group and in water molecules we allowed for angle dependent forces by introducing angle interactions among triplet of atoms. The “virtual crystal” was then cleaved by removing the PBC in the positive z direction, thus creating a free surface. After the top layered relaxed we inserted water molecules and studied their trajectories. This simulation will allow construction of surface site adsorption probabilities at P, T condition of the nebula.

The equilibrium intracrystalline distribution of Mn and Mg between the M1 and M2 sites of a Mn-rich/Fe-free orthopyroxene (donpeacorite) was investigated by means of annealing experiments at temperatures between 980 and 800 degreesC and single-crystal X-ray diffraction. The data show that Mn, as does Fe2+ in Fe-Mg orthopyroxene, preferentially orders at the M2 site. However, comparison of the distribution coefficient k(D(Mn-Mg)) determined in this study with k(D*) measured for Fe-Mg orthopyroxene shows that Mn has a much stronger preference for the M2 site relative to Fe2+. This result implies that the practice to partition Fe2+ + Mn = Fe* as one species, typically implemented to determine the quenched-site occupancies in Fe-rich/Mn-poor orthopyroxene, should be abandoned and that Mn should be considered totally ordered at M2. The partitioning method, i.e., Fe vs. Fe*, has implications for the determination of cooling rates from the observed ordering state of orthopyroxene, particularly for Fe-poor compositions (Fs < 0.16).