Skip to main content

Full text of "New Manganese Silicide Mineral Phase in an Interplanetary Dust Particle"

See other formats


K.Nakamura -Messenger 1 ' 2 , L. P. Keller 1 , S. J. Clemett 1,3 , J. H. Jones', R. L. Palma 4 ' 5 , R. 0. Pepin 5 , W. Klock 6 , M.E. 
Zolensky and S. Messenger . Astromaterials Research and Exploration Science Directorate /NASA Johnson Space 
Center, Houston, TX 77058, USA, 2 ESCG/ Jacobs Engineering, TX 77058, USA, 3 ESCG/ ERC Inc., TX 77058. USA, 
Dept. of Physics and Astronomy, Minnesota State Univ., Mankato, MN 56001, USA, Dept. of Physics, Univ. of 
Minnesota, Minneapolis, MN 55455, USA, Rontgenanalytik Messtechnik GmbH, Taunusstein, D-65232, Germany. 

Introduction: Comet 26P/Grigg-Skjellerup was identified 
as a source of an Earth-crossing dust stream with low 
Earth-encounter velocities, with peak anticipated fluxes 
during April in 2003 and 2004 [1]. In response to this 
prediction, NASA performed dedicated stratospheric 
dust collections using high altitude aircraft to target 
potential interplanetary dust particles (IDPs) from this 
comet stream in April 2003. Several IDPs from this 
collection have shown unusually low noble gas 
abundances [2] consistent with the predicted short 
space exposure ages of Grigg-Skjellerup dust particles 
[1]. High abundances of large D enrichments [3] and 
presolar grains [4] in IDPs from this collection are also 
consistent with an origin from the comet Grigg- 

Here we report a new mineral from one of the cluster 
IDPs of the 'Grigg-Skjellerup" collection, L2055. Our 
report focuses on an unusual manganese-iron- 
chromium silicide phase that, to our knowledge, has not 
been observed previously in nature. This unique 
phase may also shed light on the genesis of the 
enigmatic low-Fe,Mn-enriched (LIME) olivine that has 
been previously reported in IDPs and meteorites [5]. 
Samples and Methods : Several IDPs from cluster#7 of 
L2055 were allocated for coordinated noble 
gas/mineralogy/isotopic studies between Univ. of 
Minnesota and NASA/JSC. IDP L2055 13 (4 urn in size, 
hereafter IDP 13) was embedded in low-viscosity epoxy 
and 70 nm-thick sections were obtained using ultrami- 
crotomy. Imaging and selected area electron diffraction 
(SAED) were performed using a JEOL 2000FX transmis- 
sion electron microscope (TEM) (200keV) and a JEOL 
2500SE field-emission scanning TEM (FE-STEM). The 
FE-STEM is equipped with a large area, thin window 
energy-dispersive X-ray detector (EDX) analysis sys- 
tem and Gatan Imaging Filter used to collect electron 
energy-loss spectra (EELS). Nanometer-scale composi- 
tional maps of the sample were acquired with a 2 or 4 
nm incident probe whose dwell time was minimized to 
avoid beam damage and element diffusion during map- 
ping. Image layers of each mapped region were ac- 
quired and combined in order to achieve sufficient 
counting statistics for major elements in each pixel to 
derive quantitative abundances. All the EDX data were 
reduced using the Cliff -Lorimer ratio method [6]. 

General mineralogy of IDP 13: Major components of 13 
include GEMS grains (glass with embedded metal and 
sulfides) mineral grains such as enstatite, forsterite and 
sulfides bound together by abundant carbonaceous 
material. Mineral grain sizes range from 20 to 200 nm. 
Individual enstatite and forsterite grains contained up 
to 5 wt% of MnO, having a typical composition of 
LIME olivine and LIME pyroxene [5]. Magnetite rims, 
which are an indicator of strong heating during atmo s- 
pheric entry, were not developed on the surfaces of 
mineral grains. Solar flare tracks were not detected in 
any enstatite grains ( -200 nm in size) in 13 consistent 
with a short space exposure time. Notably, IDP 13 also 
contained C-rich spherical hollow objects similar to N- 
rich organic globules reported from carbonaceous 
chondrites and IDPs [7,8]. 

Extremely Mn-rich crystalline grains : We found 
three unusual Mn-rich crystalline grains in IDP 13. 
Their grain sizes are 100, 200 and 250 nm in diameter. 
The major elements are Si and Mn, with minor Fe and Cr 
(Fig. lb), and O is below detection limits for both EDX 
and EELS analysis. Based on preliminary EDX data, the 
stoichiometry of the new phase is (Mn, Cr, Fe)Si al- 
though additional data are required to evaluate the ho- 
mogeneity of the phase and whether or not it is chemi- 
cally zoned. Mn-L,2,3 energy loss spectra show that the 
bulk of the Mn in the new phase is present primarily as 
Mn, not MnO, consistent with the absence of O by 
EELS and EDX (Fig. lc, [9]). Extensive SAED revealed 
that these Mn-rich grains are single crystals with a cu- 
bic symmetry. The crystal structure and <f-spacings are 
in excellent agreement with diffraction data for syn- 
thetic MnSi [10]. Synthetic manganese silicides 
(Mn Y Si v ) with composition x and y take quite a variety 
of forms and crystal structure changes depending on x 
and y, although only the simple MnSi assumes cubic 
symmetry (P2i3, a= 4.558A). 

MnSi-LIME olivine shell: One of the MnSi grains has a 
core -mantle structure, having multiple concentric layers 
apparent in the bright field image (Fig. la). A series of 
spectral maps (Fig. Id) and a dark field image ([012], 
Fig.le) clearly show that only the core area (indicated 
as position 1 and 2 in Fig. la) is MnSi. The MnSi core is 
surrounded by LIME olivine with MnO=4.5 wt % 
(layer# 3) and MnO= 2.0 wt% (layer 4). A high resolu- 
tion image (Fig. If) shows that LIME olivine is epitaxial 

New MnSi phase in an IDP: K. Nakamura-Messenger, et al. 

to the MnSi with the MnSi (200) parallel to the olivine 
c*. LIME olivine was first reported from chondritic 
porous IDPs and some ordinary chondrites [5], and has 
been commonly observed in IDPs since then. Recently, 
LIME olivine was also found in comet Wild -2 dust 
samples returned by NASA Stardust mission [11], indi- 
cating that LIME olivine is a very common mineral 
component of comets. LIME olivine has been pro- 
posed to form from condensation in the protosolar 
nebula [5]. 

Petrogenesis: MnSi is an exotic phase and is not a 
predicted nebular condensate. The origin of our MnSi 
phase is uncertain but it must have formed under rather 
reducing conditions. If the reaction MnO + SiCh --> 
MnSi + 3/2 2 pertains, then at 1 500 K the implied fo 2 of 
formation is IW-5.2. (Thermo data for MnSi from [12].) 
For comparison, a gas of solar composition at this tem- 
perature would have an fi)2 of about IW-5.5. Therefore, 
our MnSi phase need not be extrasolar. 

The relationship between our MnSi phase and 
LIME olivines is tantalizing, but also unclear. Simple 
oxidation of the MnSi phase should produce rhodonite 
[Mn pyroxene], not olivine. Also, Mg addition would 
be required to produce LIME olivine from our MnSi 
phase. These observations imply a complex process 
that is not currently understood. We speculate that the 
MnSi phase acted as a nucleus or substrate for forster- 
ite condensation and that some oxidiation of MnSi oc- 
curred during this process. This suggests that redox 
conditions changed somewhat between the time that 
MnSi formed and the later condensation of LIME oli- 
vine. Clearly, more work is necessary to explore MnSi 

and LIME olivine petrogensis. We are planning to 
measure Si isotope of this phase using the JSC 
NanoSIMS 50L, which may help to constrain the origin 
of this unusual phase. 

References: [1] Messenger S. (2002) MAPS 37, 1491- 
1506. [2] Palma R. L., et al. (2005) MAPS (Supp.), 40, 
Abst#5012 [3]Nittler L. R, et al. (2006) LPS XXXVII, 
Abst#2301 [4] Nguyen A. N., et al. (2007) LPS XXXVIII, 
Abst#2332 [5] Klock, W., et al. (1989) Nature, 339, 126- 
128 [6]Cliff, G, and Lorimer, G.W. (1975) J. Micr., 103, 
203-207.[7] Nakamura-Messenger, K., et al. (2006) 
Science, 314, 1439-1442 [8] Messenger, S., et al. (2008) 
in this volume. [9] Bradley, J.P. (1994) GCA, 58, 2123- 
2134 [10] JCPDS (1998) [11] Zolensky, M.E. et al. (2006) 
Science, 314, 1735-1739. [12] Barin I. et al.. (1977) 
Thermodynamic Properties of Inorganic Substances. 

Figure 1 (a): A Bright field TEM micrograph of MnSi- 
LIME olivine shell. Orange numbers indicate different 
layers from the core -mantle to the crust of the shell, (b): 
EDX spectra from Mn, Fe, Cr-silicide core (top), and 
LIME olivines (layer#3 in the middle, layer#4 in the 
bottom), (c): EELS spectra of Mn-L2,3 from one of the 
MnSi grains, (d): Si-K, 0-K, Mg-K, Mn-K, and Cr-K 
spectral mapping of the grain of Fig. 1(a). (e): a dark field 
image of boxed area of Fig. 1(a). Red lines indicate the 
hexagonal crystal boundaries between MnSi and LIME 
olivine. The SAED pattern of zone [012] is associated 
with this image, (f) A high resolution TEM image of the 
epitaxial boundary between MnSi and LIME olivine. 






' 3; outer mantle: LIME olivine 






4* P 



d: outer marine: LIMEclMne 
1 <wno;-zww 



[ a Vi! 

Mn-L 2 3 R 9 1c