Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://selena.sai.msu.ru/Bus/Publications/EM&P2003(Bus-Dor-Mak).pdf Äàòà èçìåíåíèÿ: Tue Feb 17 19:58:42 2009 Äàòà èíäåêñèðîâàíèÿ: Thu Feb 27 20:40:24 2014 Êîäèðîâêà:

HYDRATED SILICATES ON EDGEWORTH-KUIPER OBJECTS ­ PROBABLE WAYS OF FORMATION
V. V. BUSAREV
Sternberg State Astronomical Institute, Moscow University, Russian Federation (RF) (E-mail: busarev@sai.msu.ru)

V. A. DOROFEEVA
Vernadsky Institute of Geochemistry, Russian Academy of Sciences (RAS), Moscow, RF

A. B. MAKALKIN
Institute of Earth Physics, RAS, Moscow, RF

Abstract. Visible-range absorption bands at 600­750 nm were recently detected on two EdgeworthKuiper Belt (EKB) objects (Boehnhardt et al., 2002). Most probably the spectral features may be attributed to hydrated silicates originated in the bodies. We consider possibilities for silicate dressing and silicate aqueous alteration within them. According to present models of the protoplanetary disk, the temperatures and pressures at the EKB distances (30­50 AU) at the time of formation of the EKB objects (106 to 108 yr) were very low (15­30 K and 10-9 ­10-10 bar). At these thermodynamic conditions all volatiles excluding hydrogen, helium and neon were in the solid state. An initial mass fraction of silicates (silicates/(ices + dust)) in EKB parent bodies may be estimated as 0.15­0.30. Decay of the short-lived 26 Al in the bodies at the early stage of their evolution and their mutual collisions (at velocities 1.5 km s-1 ) at the subsequent stage were probably two main sources of their heating, sufficient for melting of water ice. Because of the former process, large EKB bodies (R 100 km) could contain a large amount of liquid water in their interiors for the period of a few 106 yr. Freezing of the internal ocean might have begun at 5 â 106 yr after formation of the solar nebula (and CAIs). As a result, aqueous alteration of silicates in the bodies could occur. A probable mechanism of silicate dressing was sedimentation of silicates with refractory organics, resulting in accumulation of large silicate-rich cores. Crushing and removing icy covers under collisions and exposing EKB bodies' interiors with increased silicate content could facilitate detection of phyllosilicate spectral features.

1. Introduction Edgeworth-Kuiper Belt (EKB) objects orbit the Sun outwards of Neptune's orbit, 30 AU to 50 AU, and are possibly rather primitive solid bodies. According to presently accepted notions, the EKB objects formed in situ (Safronov, 1996; Farinella et al., 2000), though some part of their material could be brought by projectile bodies from the formation zones of giant planets, mainly of Neptune and Uranus. Contemporary models of the solar nebula (Makalkin and Dorofeeva, 1996; Mousis et al., 2000) yield very low temperatures and pressures of T = 15­ 30 K and P = 10-9 ­10-10 bar at the radial distance of 30­50 AU and the nebula
Earth, Moon and Planets 92: 345­357, 2003. © 2004 Kluwer Academic Publishers. Printed in the Netherlands.


346

V. V. BUSAREV ET AL.

age of about 106 ­107 yr when the EKB bodies of sub-planetary size were formed. At these T ­P conditions all volatiles excluding hydrogen, helium and neon were in the solid state (mostly ices and some organics), and the abundance of rocky (silicate) dust component was lower than that of ices in accordance with the solar ratios of corresponding elements. Nevertheless, the detected probable signs of hydrated silicates on some EKB objects (Boehnhardt et al., 2002) show that silicates in the bodies may be sufficiently abundant to be detected and that the silicates are probably aqueously altered (e.g., Vilas and Gaffey, 1989; Busarev and Taran, 2002). We have tried to indicate possible processes responsible for accumulation and aqueous alteration of silicates in EKB bodies. Obviously, a necessary condition for the last process should be a liquid state of water that requires a considerable elevation of temperature in the bodies' interior. Plausible factors for heating were decay of radionuclides (short-lived 26 Al and long-lived 40 K, 235 U, 238 U and 232Th) dispersed in silicate matter and mutual collisions between the bodies. As shown in model calculations, the long-lived radioisotopes were insufficient for total melting of ice fraction in icy satellites of giant planets of radii up to 800 km, although partial melting was possible (Consolmagno and Lewis, 1978; Prialnik and BarNun, 1990). A considerable role of collisional events in the EKB is probably confirmed by strong correlations between observed B­V and V­R colors of EKB bodies and their calculated mean random collision speeds (Stern, 2002). A basis for these calculations is the collisional resurfacing hypothesis. It suggests that the flux of cosmic rays darkening and reddening the upper layer of surface of icy airless bodies competes with impacts that excavate fresh material (more bright and blue or grey) from the interior to the surface (Luu and Jewitt, 1996). It is theoretically possible that phyllosilicates formed in the solar nebula at the earlier stage of its evolution, when T 400 K (Drouart et al., 1999), before accretion of planetesimals (e.g., Prinn and Fegley, 1989; Ganguly and Bose, 1995). In this case the mechanism of phyllosilicate formation was the interaction of silicate dust with water vapor, but the contribution of the process remains unclear.

2. Some Estimates of Silicate Fraction in Sub-Planetary Bodies Composition of EKB bodies can be roughly estimated from the data on the most primitive objects in the Solar system ­ comets and interplanetary dust particles (IDPs) (e.g., Delsemme, 1988; Jassberger et al., 1988; Kissel and Krueger, 1987; Mumma et al., 1993; Pollack, 1994; Greenberg, 1998), using the solar system elemental abundances (Lodders and Fegley, 1998). According to the data, the bodies may consist of refractory dust and volatile ices with dust to ice mass ratio varying within 0.5­1.3. Dust contains inorganic (48­58 wt.%) and refractory organic fractions. Variations in the mass fraction of the former are caused mainly by the uncertainties of abundance ratios of Fe/Si and Mg/Si (from 0.34­0.5 to 0.9­1). Inorganic fraction or rock consists of silicates (mainly of magnesium and


HYDRATED SILICATES ON EDGEWORTH-KUIPER OBJECTS

347

iron silicates with mol relation FeO/(FeO + MgO) = 0.2­0.3), troilite (FeS) and metallic iron. Refractory organic fraction or CHON (52­42 wt.% in dust) is a complex insoluble polymer material with vaporization temperature 400­600 K. It includes aliphatic, cyclic and aromatic hydrocarbons (PAH), the last being probably the main component. The relative proportions of elements in this fraction are estimated as C:H:O:N = 1:1:0.5:0.12 (Jessberger et al., 1988). CHON contains about 50­70% of the total amount of C in comets. Ices include water ice (up to 80 wt.%), the volatile organics (10 wt.%) and gases (10 wt.%). Volatile organic compounds are methanol, formaldehyde and others with vaporization temperature near 300 K at normal pressure; gases (CO, CH3 OH, CH4 , H2 S, HCN and others) were incorporated with water ice in the gas-dust protoplanetary disk at T < 50 K (Fegley, 1999). Thus, the mass fraction of silicates (silicates/(ices + dust)) in parent EKB bodies may be estimated as 0.15­0.30. Such a low content of silicates in the bulk of EKB objects makes their easy detection by remote sensing methods questionable, especially in presence of the dark CHON-component in the material. Nevertheless, absorption bands at 600­750 nm were found recently in reflectance spectra of two EKB objects (Plutinos 2000 GN171 and 2000 EB173) (Boehnhardt et al., 2002). Taking into account the discovery of H2 O ice on EKB objects (e.g., Brown et al., 1999), one could consider the absorption bands as probable signs of hydrated silicates on the bodies. The spectral features are typical for Fe(2+)­Fe(3+) bearing phyllosilicates. Similar absorption bands were found in reflectance spectra of C­ PD­J­G-type asteroids (Vilas and Gaffey, 1989), hydrated M­S-type asteroids and carbonaceous chondrites (Busarev and Taran, 2002). A strong correlation between the spectral feature at 700 nm and the characteristic absorption band of OH groups at 3 µm was found for low-albedo asteroids (Vilas, 1994; Howell et al., 2001). We have predicted a possibility of silicate features' detection in reflectance spectra of EKB objects (Busarev, 2001). If the interpretation is correct, the detected spectral features point to aqueous alteration and dressing of silicate matter in EKB bodies during their evolution. Removing external ice covers and exposing EKB bodies' nuclei with elevated silicate content under subsequent collisions could facilitate detection of corresponding spectral features. We consider possible mechanisms supporting the processes.

3.

26

Al and Related Water Ice Melting, Aqueous Alteration and Sedimentation of Silicates in the EKB Bodies

Among other radionuclides 26 Al (half-life 7.2 â 105 yr) could play a key role in heating and initial thermal evolution of the main-belt asteroids and other sub-planetary bodies (up to hundreds-km-size) because it is widespread in the interstellar medium as a product of galactic supernovae and novae evolution. It was discovered in the galactic equatorial plane in the proportion of 26 Al/27 Al 10-5


348

V. V. BUSAREV ET AL.

(Mahoney et al., 1984) comparable to the same ratio (5 â 10-5 ) in the Ca­Al-rich inclusions (CAIs) (at the time of their origin) of the Allende meteorite (Wasserburg and Papanastassiou, 1982). Moreover, the detection of 26 Mg (the decay product of 26 Al) in a differentiated meteorite (Srinivasan et al., 1999) confirms the role of 26 Al for heating and differentiation of the parent bodies of the main-belt asteroids. But was the concentration of captured 26 Al sufficient for melting water ice in the EKB objects? If the time of EKB bodies' formation was substantially larger than the half-life of 26 Al, then independently of the isotope concentration it couldn't heat the EKB bodies with high efficiency, for instance, as it heated parent bodies of the main-belt asteroids. The formation time of hundreds-km-sized EKB bodies was from about one million years (Weidenscilling, 1997) to several tens of million years (Kenyon and Luu, 1998). It is assumed that the accretion of EKB objects was terminated by the formation of Neptune (Farinella et al., 2000) which began to disperse them via gravitational scattering. In this case an upper limit for accretion time of the EKB objects would be the formation time of Neptune, estimated as about a few 107 yr (Brunini and Fernandes, 1999; Bryden et al., 2000) to 108 yr (Pollack et al., 1996; Farinella et al., 2000). These timescales are at least one order of magnitude shorter than in previous models (Safronov, 1969; Wetherill and Stewart, 1989) due to incorporation of the stage of accelerated "run-away" accretion of giant planet embryos. Taking into account the model of cometary bodies formation by Weidenscilling (1997), accretion of bodies up to 100 km in radius at the EKB distances 35­50 AU within (1­1.5) â 106 yr seems to be possible, though this time is near the lower limit of accretion timescales. In this consideration we suppose that formation of planetesimals at the radial distances of the EKB could begin several 105 yr after the collapse of the protosolar cloud. Probably, this time was sufficient for formation of the protoplanetary disk and transport of the dust to the EKB distances. If the accretion of bodies of radius R = 100 km was complete no later than a few 26 Al half-life times, the decay of this isotope provides enough heat to melt the water ice in the interiors of these bodies. To check this conclusion we adopt the mass fraction of rock component of 30 wt.% (in accordance with data in the previous section). The rock component with chondritic (solar) abundances of refractory elements contains 1.3 wt.% of aluminum. We also adopt the 26 Al/27 Al ratio of 1 â 10-5 which is obtained from the "canonical" initial 26 Al/27 Al ratio of 5 â 10-5 and accretion time of a EKB body as a 1.6 Myr (after CAIs). This time possibly but not necessarily coincides with the age of the solar nebula (from the collapse stage). The above figures, giving the 26 Al abundance, should be added with the decay energy of 26 Al = 3 MeV per atom and its decay constant = 9.63 â 107 yr-1 to yield the heat production rate Q = 0.40 J kg-1 yr-1 . The time m required to heat


HYDRATED SILICATES ON EDGEWORTH-KUIPER OBJECTS

349

a large EKB body to the water-ice melting point and to melt the ice in its interiors can be estimated from the equation
m 0

Q exp(-t )d t =

Tm T0

cp dT + Lf mw ,

(1)

where T0 = 30 K is the adopted value for the initial temperature of the body, Tm = 273 K is the melting temperature of water ice (a good approximation to at P < 25 MPa, characteristic for interiors of the EKB body of radius R 300 km), Lf = 3.34 â 105 J kg-1 is the latent heat of fusion for H2 O, mw = 0.38 is H2 O mass fraction (as we take for calculation), cp is the thermal capacity at constant pressure per unit mass for the body's material. With some overestimation of cp at temperatures from 30 to 150 K we can take the temperature dependence of specific heat values for all main components similar to that for water ice: cpw = 7.67 T J kg-1 K-1 (Hobbs, 1974). In this approximation we obtain the following values of thermal capacities (all in J kg-1 K-1 ): cp r = 3.1 T for rocks (mainly silicates), cp CHON = 5.7 T for refractory organics, and cp vol = 10 T for volatile organics and gases (the approximation for gases is most crude, but this has little effect due to their low content). We use also mass fraction of CHON mCHON =0.22 and combined mass fraction of volatile organics and gases mvol+g = 0.10. With these values we obtain the thermal capacity for the mixture cp cp0 T , where cp0 = 6.1 J kg-1 K-2 . After substitution of this value in Equation (1) and integration we have the estimation for the time m : m = -
-1

ln{1 - [cp0 (Ti2 - T02 )/2 + Lf mw ]/Q} 1.9 â 106 yr.

(2)

Thus the water ice in the bodies can be melted in less than 2 million years after the body formation and, hence, at the age of the solar nebula of 3.5 million years. During this time only a surface layer of thickness R 10 km could remain solid, as follows from the simple estimation R , (3)

where is the thermal diffusivity related to the thermal conductivity k as = k/( cp ). The temperature dependence of for water ice is = 0 T -2 , where 0 9.1 â 10-2 m2 K2 s-1 (Kirk and Stevenson, 1987). However, the porosity of ices p = 0.5 decreases the thermal conductivity 5 to 50 times (Shoshany et al., 2002). The porosity would be at its maximum at the surface and reduces to the low values at the bottom of the layer. Thus the reasonable estimate for the thermal diffusivity of the layer is 10-6 m2 s-1 . The first outcome of radiogenic heating of the bodies (preceding the melting of water ice) should be evaporation of the most volatile species mentioned above as gases (CO, CH3 OH, CH4 , and so on). However their low integral fraction (5


350

V. V. BUSAREV ET AL.

wt.%) and probable moderate to high porosity of the early EKB objects would minimize the effect of their separation on the structure of the bodies. The consequences of water ice heating are much more important. First, huge amount of water ice evaporated at low pressures in the porous medium should recondense in the upper layers of the bodies, substantially reducing their porosity. As a result of insulation of the interiors from outer space, the pressure below the upper layer of thickness R would become higher than 1 bar and melting of water ice should occur when heated to T > Tm 270 K. Probable admixture of volatile organics might slightly decrease this temperature. Thus, as follows from Equations (1) and (2), internal water ocean in the young EKB bodies could form at their age m 1.9 Myr, that is after a + m 3.5 Myr after solar nebula CAIs and formation. Consider the evolution of the internal water ocean in a young EKB body of R = 100­300 km. The thermal convection in the ocean should be vigorous, if the Rayleigh number Ra is much higher than its critical value Racr 103 . We can estimate the value Ra = g d 3 T /( ), where 10-4 K-1 is the volumetric thermal expansion coefficient of the mixture, dominated by liquid water, g is the gravitational acceleration (g 4GR ), = 1.4 â 103 kg m-3 is the mean density ¯ ¯ of the body (calculated at the above fractions of components), d 0.8­0.9 R is the convective layer thickness, T is the temperature difference across the layer, 10-7 m2 s-1 and 10-5 m2 s-1 are the thermal diffusivity and kinematic viscosity of the water­solids mixture. At d = 70 km and the very low value of T = 1 K we nevertheless obtain a very high value Ra 1021 . The Nusselt number (Nu), which is the ratio of the total heat flow (including convective one) to the conductive flow is related to Ra by (Schubert et al., 1979) Nu 0.2Ra1/3 . With these data we can estimate the time scale for heat transport through the convective water ocean c by relation (3) where R is substituted for d 0.8 R and the molecular thermal diffusivity is substituted for the effective thermal diffusivity e which accounts for convection, with e = · Nu. At the above parameters we obtain c 103 yr. The time is very short relative to the time scale m 106 yr which is also the time scale for heat transport through the outer body's shell of thickness R 10 km and relative to the lifetime of the ocean till the onset of its freezing 0 (considered below). Owing to the rapid radial heat transport through the ocean its temperature is stabilized near the temperature of maximum water density 277 K (the adiabatic compression for hundreds-km-sized bodies is negligible) and probably never exceeds 280 K. After a lapse of time a continuing decrease of radiogenic heat production yields the freezing of the internal ocean beginning (as in a usual terrestrial ocean) from the upper layers. The lifetime of the water ocean till the beginning of its freezing in the early EKB body of radius R can be estimated by comparing the heat flux F1 from the ocean to the solid shell of thickness R above it and the heat flux F2 through the


HYDRATED SILICATES ON EDGEWORTH-KUIPER OBJECTS

351
26

shell. The flux F1 is generated in the interiors being heated by the quickly transferred to the lithosphere. Thus we can write F1 1 (R - ¯ 3 R )Q exp(-t )

Al decay and (4)

¯ on the assumption of 26 Al homogeneous distribution, where =1.4 â 103 kg m-3 is the mean density of the body (calculated at the above fractions of components). The flux F2 can be written as F2 k T / R , (5)

where k is the thermal conductivity of shell, T 273 - 30 240 K. We assume k = 2 W m-1 K-1 for water ice, taking into account the empirical relation for crystalline ice k(T ) = 567/T Wm-1 K-1 and the compensating effect of increasing porosity from the base to the surface of the shell (Spohn and Schubert, 2003). The outer layer R 10 km is of primordial composition, identical to the bulk one; thus we obtain k 1.5 in the layer owing to the admixture of components other than water. The freezing of the water ocean begins when the incoming flow from interiors F1 becomes lower than the flux F2 coming from the shell. By equating two fluxes from (4) and (5) we obtain the estimate of the lifetime of the ocean of liquid water as 0 1.2 Myr for the bodies of radius R = 100­300 km respectively. This time is quite sufficient for silicates to form phyllosilicates by reaction with water. If the early EKB bodies, like comet nuclei, consisted of a conglomerate of ices and dust particles (the "dirty ice"), then sedimentation of solid particles (consisting of silicates and CHON) in the water ocean leads to formation of the core enriched in silicates (including phyllosilicates). However, convection hinders sedimentation and supports suspension. The criterion for sedimentation obtained by Solomatov and Stevenson (1993) includes the ratio of the settling velocity of particles in the non-convective medium up to the convective velocity uc (Rouse numbers S ): S= up ; uc up 4g r 2 (1 - )2 ; 1500 ¯ uc g d dF cp ¯
1 /2

,

(6)

where 0 10-6 m2 s-1 is the kinematic viscosity of the liquid water with admixture (< 10 wt.%) of volatile organics, 3­100 is the viscosity of the convecting liquid-solid mixture, (2.2 - 1) â 103 = 1.2 â 103 kg m-3 is the density difference between settling particles and fluid. (The particles contain silicate rocks and refractory organics (CHON) in proportion 0.57/0.43 by mass.) Parameter r is the radius of the settling particle, = 0.52 is the mass fraction of the particles. The initial thickness of the internal ocean is taken as d 0.9R . For the case of the early EKB bodies we assume that the thermal flux F is equal to F1 , where F1 is defined by Equation (4). The thermal capacity cp for the mixture of water and solids is taken equal to 2.4 â 103 J kg-1 K-1 . Values for parameters and g are


352

V. V. BUSAREV ET AL.

shown above. For the case of a low mass fraction of the settling particles ( 1) Solomatov and Stevenson (1993) obtained the simple relation which we extend to the case of any < 1 and obtain the following criterion for suspension (the case that sedimentation is inhibited): S< r ( · Re) d
1 /2

1- ,

(7)

where 0.01 is the efficiency factor equal to the fraction of the maximum available power of the convection that is spent on the gravitational work against sedimentation, Re = uc d/ is the Reynolds number. For the early EKB bodies we find Re 1012 that is much higher than the critical value Recr 30­100 for transition from laminar to turbulent convection. Thus the convection in the bodies is turbulent. With the parameter set for these bodies we obtain from (7) that sedimentation is inhibited for the particles smaller than 10 cm, but takes place at r > 10 cm. It is unclear whether the particles of rock composition can grow up to decimeter size even in 106 years, but the presence of the organic compounds in the solid particles can greatly increase their sticking probability, playing the role of a sticking agent in formation of rock-CHON aggregates. The growth rate of the particle mass m and radius r at collisions with smaller ¯ particles is described by the simple accretion equation dm/d t = r 2 ur ,where ur is the mean relative velocity of particles, is the sticking probability. For the parameters of the bodies under consideration, including velocities from (6), one can show that the time scale for the particle growth to decimeter size is rather low (10 yr), if = 1. The very low sticking probability 10-8 is sufficient to reach r 10 cm in 1 Myr. The settling time for the particles of r 10 cm is much lower than 1 yr. Thus the formation of the core enriched in silicates and refractory organics could happen in the early EKB bodies during the lifetime of the water ocean of a few 106 years. With the assumed mass fractions of the components we obtain the radius of the rock-CHON core rc 0.7 R . Because of their fluffy fractal structure some IDPs probably originated in the protoplanetary cloud (e.g., Rietmeijer and Nuth, 2001), organic coatings on silicate particles could not be an obstacle for penetration of liquid water to silicate particles and for the process of silicate hydration. The volatile organics dissolve in water, forming mineral acids. As is known from terrestrial conditions, an acidic medium accelerates transformation of inorganic compounds, in particular silicate hydration (Veselovskij, 1955). Three major hydrous minerals that are predicted to form in the sufficiently large planetesimals of the solar nebula are serpentine containing 13.0 wt.% H2 O, talc (4.8 wt.% H2 O), and brucite (8.3% H2 O) (Fegley, 2000). Hydrous phases similar to these minerals are found in CI and CM2 carbonaceous chondrites. Thus the silicates in the EKB bodies could contain about 10% of water. From our estimates it follows that the processes of aqueous alteration of silicates and formation of the core were probably completed before the onset of the internal


HYDRATED SILICATES ON EDGEWORTH-KUIPER OBJECTS

353

ocean freezing, which happened for the bodies of R = 100­300 km respectively at the time a + m + 0 1.6 + 1.9 + (1-2) = 4.5­5.5 Myr after formation of the solar nebula (and CAIs). This time is shorter than the formation period of giant planets (107 yr for Jupiter and Saturn and at least a few 107 yr for Uranus and Neptune). Hence the EKB bodies could become layered and have silicate-rich cores before the onset of their heavy bombardment by the bodies dispersed from the region of giant planet formation. The cores very probably contained phyllosilicates which could be exposed during the heavy-bombardment stage after many cratering and destructive events. Even at the zeroth sticking probability of the silicate-CHON particles the core could form in the early EKOs during the freezing of the water ocean by the following mechanism. At the top of the ocean a thin nonconvective layer exists because of the negative thermal expansion coefficient of water between 0 and 4 C. By equating the thermal fluxes in the layer and in the upper ice shell with the help of Equation (5) we estimate the layer's thickness to be less than 1 km. According to our estimates the downward velocity of the upper boundary of the water ocean is less than the sedimentation velocity (Equation (6)) for particles of radius a few microns. So during the freezing of the ocean the particles concentrate in its lower, liquid fraction and form the core to the end of the ocean freezing. We find the duration of this process to be of the order of 10 Myr. We considered a short accretion time scale of the EKB bodies which is slightly higher than the lower limit of the possible range of the accretion times. This means that we deal with a "border" case. So we can't fully rely on 26 Al as the main and only heat source for thermal evolution and hydrosilicate formation in the EKB bodies. Nevertheless, we suggest a possibility of radiogenic heat accumulation sufficient for origin of an aqueous media, aqueous alteration and sedimentation of silicates in the bodies of radius R = 100­300 km within the first 5­10 Myr of the solar nebula evolution Myr. The effect of long-lived radionuclides at the stage of formation and early evolution of EKB objects is negligible relative to 26 Al (e.g., Choi et al., 2002). At timescales, comparable to the Solar System age, the long-lived radionuclides 40 K, 235 U, 238U and 232Th may have a dominant role in heating of the EKB objects, but the power of this heat source is sufficient to evaporate (partially or totally) only the ices more volatile than water ice (De Sanctis et al., 2001).

4. The Role of Collisions in Evolution and Silicate Aqueous Alteration in the EKB Objects The relative velocities of the EKB bodies during the stage of their accretion should be lower than 50 m s-1 , as follows from the numerical models (Stern and Colwell, 1997a). This result is consistent with the theory (Safronov, 1969) at the value of 2 Safronov number 3, where 2 = e / 2 , e is the escape velocity from the largest body in the inverse power-law mass distribution, is the mean square-root


354

V. V. BUSAREV ET AL.

velocity of the EKB bodies relative to their mean circular Keplerian motion (it is assumed that the radii of the largest bodies in the distribution are about 100 km). Simple estimates show that for the initial stage the mean temperature increase of a EKB body in the distribution at such moderate-velocity collisions is less than 1 K, which is negligibly small for any cosmochemical applications. The subsequent process was dispersion of EKB, when the growth of Uranus and Neptune gravitationally perturbed the orbits of the remnant bodies in their zones and ejected them to the EKB region. Their high-velocity collisions with the EKB bodies yielded not only erosion and fragmentation of the bodies, but also increased eccentricities of them up to 0.3 between 30 and 50 AU (Stern and Colwell, 1997b). According to results of these authors, no less than 99% of the original mass of the population was lost from the EKB during this stage of 109 yr (the mass decreased from 10­35 M to 0.1­0.3 M ). The relative velocities of collisions at this stage, according to calculated eccentricities, could be higher than 2 km s-1 . A simple estimation of the mean increase of the body's temperature T = T2 - T1 can be made for collisions of large bodies of comparable masses. The energy balance can be expressed as (m1 + m2 )
T2 T1

cp (T )d T

1 kh µ 2 , 2

(8)

where µ = m1 m2 /(m1 + m2 ) is the reduced mass, kh is the fraction of impact energy converting to heat. The energy loss to fragmentation of the bodies and scattering of fragments is lower than the loss to the body's heating, as follows from many experimental and theoretic data. For large bodies we adopt kh = 0.7­0.8. The heat capacity cp for the assumed mixture of components of the early EKB bodies is approximated by the above finction: cp 6.1T J kg-1 K-1 . For collisions with > 2km s-1 we obtain from (8) the temperature T2 500 K, which is destructive for the hydrosilicates. However, bodies subjected to the high-velocity collisions, were mostly swept out of the EKB, and the remaining ones probably very rarely, if ever, experienced such impacts. After accomplishing this destructive stage the EKB was close to the modern low-mass state with rather rare collisions onto 100-km-sized bodies. The collision lifetimes for disruption large objects in the present-day EKB are much longer than the age of the Solar System (Durda and Stern, 2000). The collision velocities in the EKB, as follows from the observed orbit eccentricities, are 1.5 km s-1 . The mean temperature rise T at collisions with velocities =1.5 km s-1 we estimate from Equation (8) at 240 K, appropriate for hydrosilicate formation. The thermal consequences of mutual collisions of the EKB bodies include evaporation of volatiles, melting of water ice, impact dressing of silicates and creation of heat centers under the cratered areas which are "buried" for a long time in bodies' interiors and preserve favorable conditions for silicate hydration. However, kinetic restrictions on the processes in the subsurface layers of EKB bodies are to be studied. Collisions at such velocities lead also to erosion of EKB objects (Durda and Stern,


HYDRATED SILICATES ON EDGEWORTH-KUIPER OBJECTS

355

2000). Removal of icy covers of the bodies, excavation of phyllosilicates formed in their interiors and/or exposition of the interiors with higher silicate content were probable consequences of the events. The processed areas of the bodies would cover dozens of percent of their surfaces. In the latest thermal models of comets, icy satellites of giant planets and EKB objects (e.g., Prialnik and Bar-Nun, 1990; De Sanctis et al., 2001; Choi et al., 2002) radiogenic heating is considered as the main factor of their evolution. In our opinion the effect of collisions of EKB objects on their thermal evolution was probably no less important than the decay of 26 Al. A combined effect of collisions, radiogenic heating, and (to less degree) of insolation could considerably increase the internal temperature of EKB objects (or some of them) from initial 15­30 K up to at least 210­240 K when a process of diagenesis (low-temperature aqueous alteration of silicates) (e.g., Rienieijer, 1985) could start. But as shown above, the temperature might have been substantially higher.

5. Conclusions As follows from the above consideration, the interiors and/or undersurface layers of the EKB objects are the most proper places for formation of phyllosilicates. Two main mechanisms of heating of the bodies, partial evaporation and elimination of volatiles, melting of water ice and aqueous alteration of silicates probably existed. The first mechanism is the decay of radionuclides (mainly short-lived isotope 26 Al) in the rocky fraction of the EKB bodies during first 5 Myr after formation of the solar nebula (and CAIs). The melting of water ice and origin of internal water ocean in the sufficiently large bodies (200 km) probably led to vast aqueous alteration of silicates. It could also yield the sedimentation of the silicate-organic fraction of matter and accumulation of silicate-rich cores. The second mechanism is the impact heating at the mutual collisions of the bodies with velocities 1.5 km s-1 at much later stage of their evolution. There is a theoretical possibility of phyllosilicate formation in the solar nebula at the earlier stage of its evolution, before accretion of planetesimals, at interaction of the silicate dust grains with water vapor (e.g., Prinn and Fegley, 1989; Ganguly and Bose, 1995). However, if the temperature in the Kuiper-belt region never is higher than 150 K, H2 O was probably always in the form of ice. In this case the formation of phyllosilicates in situ in the gaseous nebular environment through the gas-solid reaction was impossible, and phyllosilicate formation could happen only in large bodies which could contain liquid water. The authors thank the anonymous reviewer for useful comments. A.B.M. acknowledges support from the Russian Ministry of Science, Industry, and Technology via Contract 40.022.1.1.1108.


356

V. V. BUSAREV ET AL.

References
Boehnhardt, H., Delsanti, A., Hainaut, O. et al.: 2002, Proceedings of Asteroids, Comets, Meteors (ACM 2002), ESA-SP-500, Berlin, Germany, pp. 47­50. Brown, R. H., Cruikshank, D. P., and Pendleton, Y.: 1999, Ap. J. 519, L101­L104. Brunini, A. and FernÀndez, J. A.: 1999, Planet. Space Sci. 46, 997­1001. Bryden, G., Lin, D. N. C., and Ida, S.: 2000, Astrophys. J. 544, 481­495. Busarev, V. V.: 2001, AAS 198th Meeting, BAAS 33(2), 892. Busarev, V. V. and Taran, M. N.: 2002, Proceedings of Asteroids, Comets, Meteors (ACM 2002), ESA-SP-500, Berlin, Germany, pp. 933­936. Choi, Y.-J., Cohen, M., Merk, R., and Prialnik, D.: 2002, Icarus 160, 300­312. Consolmagno, G. J. and Lewis, M. D.: 1978, Icarus 34, 280­293. Delsemme, A. H.: 1988, Phil. Trans. R. Soc. 325, 509­523 De Sanctis, M. C., Capria, M. T., and Coradini, A.: 2001, Astron. J. 121, 2792­2799. Drouart, A. B., Dubrulle, D., Gautier, D. et al: 1999, Icarus 140, 129­155. Durda, D. D. and Stern, S. A.: 2000, Icarus 145, 220­229. Farinella, P., Davis, D. R., and Stern, S. A.: 2000, in V. Mannings, A. P. Boss, and S. S. Russell (eds.), Protostars and Planets IV, University of Arizona Press, Tucson, pp. 1255­1282. Fegley, B. Jr.: 1999, Space Sci. Rev. 90, 239­252. Fegley, B. Jr.: 2000, Space Sci. Rev. 92, 177­200. Ganguly, J. and Bose, K.: 1995, Lunar Plan. Sci. XXVI, 441­442. Greenberg, J. M.: 1998, Astron. Astrophys. 330, 375­380. Howell, E. S. et al.: 2001, in Asteroids 2001: From Piazzi to the Third Millenium, Osservatorio di Palermo, Sicily, p. 62. Jessberger, E. K., Christoforidis, A., and Kissel, J.: 1988, Nature 332, 691­695. Kenyon, S. J. and Luu, J. X.: 1998, Astron. J. 115, 2136­2160. Kerridge, J. F.: 1999, Space Sci. Rev. 90, 275­288. Kirk, R. L. and Stevenson, D. J.: 1987, Icarus 69, 91­134. Kissel, J. and Krueger F. R.: 1987, Nature 326, 755­760. Lodders, K. and Fegley, B.: 1998, The Planetary Scientist's Companion, Oxford University Press. New York, Oxford. Mahoney, W. A. et al.: 1984, Astrophys. J. 286, 578­585. Makalkin, A. B. and Dorofeeva, V. A.: 1996, Solar Sys. Res. 30, 440­455. Mousis, O., Gautier, D., Bockelee-Morvan, D. et al.: 2000, Icarus 148, 513­525. Mumma, M. J., Weissman, P. R., and Stern, S. A.: 1993, in E. H. Levy and J. I. Lunine (eds.), Protostars & Planets III, University of Arizona Press, Tucson, pp. 1177­1252. Pollack, J. B., Hollenbach, D., Beckwith, S. et al.: 1994, Astrophys. J. 421, 615­639. Pollack, J. B., Hubickyi, O., Bodenheimer, P., et al.: 1996, Icarus 124, 62­85. Prialnik, D. and Bar-Nun, A.: 1990, Astrophys. J. 355, 281­286. Prinn, R. G. and Fegley, B., Jr.: 1989, in S. K. Atreya et al. (eds.), Origin and Evolution of Planetary and Satellite Atmospheres, University of Arizona, Tucson, pp. 78­136. Rietmeijer, F. J. M.: 1985, Nature 313, 293­294. Rietmeijer F. J. M. and Nuth, J. A.: 2001, Lunar Plan. Sci. XXXII, 1219 (abstract) [CD-ROM]. Safronov, V. S.: 1969, Evolution of the Protoplanetary Cloud and the Formation of the Earth and Planets, Nauka Press, Moscow, USSR [in Russian] [NASA TTF-667 (Engl. transl.), 1972]. Safronov, V. S.: 1996, Solar Sys. Res. 30, 251­257. Schubert, G., Cassen, P., and Young, R. E.: 1979, Icarus 38, 192­211. Shoshany, Y., Prialnik, D., and Podolack, M.: 2002, Icarus 157, 219­227. Solomatov, V. S. and Stevenson, D.: 1993, J. Geophys. Res. 98, 5375­5390. Spohn, T. and Schubert, G.: 2003, Icarus 161, 456­467. Srinivasan, G., Goswami, J. N., and Bhandari, N.: 1999, Science 284, 1348­1350.


HYDRATED SILICATES ON EDGEWORTH-KUIPER OBJECTS

357

Stern, S. A. and Colwell, J. E.: 1997a, Asron. J. 114, 841­849. Stern, S. A. and Colwell, J. E.: 1997b, Astrophys. J. 490, 879­882. Veselovskij, V. S.: 1955, Chemical Nature of Combustible Minerals, Mining Institute, Academy of Sciences Press, Moscow, USSR [in Russian]. Vilas, F. and Gaffey, M. J.: 1989, Science 246, 790­792. Wasserburg, G. J. and Papanastassiou, D. A.: 1982, in C. A. Barnes, D. D. Clayton and D. N. Schramm (eds.), Essays in Nuclear Astrophysics, Cambridge University Press, p. 77. Weidenschilling, S. J.: 1997, Icarus 127, 290­306.