Micro-spectroscopic characterization of organic and hydrous components in weathered Antarctic micrometeorites

Abstract

Eight unmelted Antarctic micrometeorite (AMMs) recovered from Kuwagata Nunatak were studied on Al-foils by infrared (IR), Raman, and visible reflection micro-spectroscopy in combination with electron microscopy. Major element abundances of the AMMs studied were found to be similar to solar abundance, although all have the common characteristic of Mg-depletion. Absorption bands around 500 nm were detected for some of the AMMs by the visible micro-spectroscopic method, and these AMMs can be assigned to Fe-hydroxidelike materials. These results suggest that the studied AMMs experienced weathering in Antarctica. Four grains showed the presence of IR H₂O and CH bands similar to those of type 2 carbonaceous chondrites, while these were found to be absent in two grains, as in type 3 carbonaceous chondrites. D (disordered: 1360 cm^-1) and G (graphite: 1600 cm^-1) Raman band features of graphitic carbonaceous materials in these AMMs were not similar to those for type 3 but were rather close to those for type 2 and 1 carbonaceous chondrites, although some data showed a degree of deviation. The genetic classification of individual AMM grains can thus be studied by these methods, although the weathering effects and the atmospheric entry heating on organics and hydrous components need to be evaluated. These multiple micro-spectroscopic reflectance methods are useful for the characterization of precious small samples.

1. Introduction

Micrometeorites (size <1 mm) are the major fractions of extraterrestrial materials (Maurette, 2006). The flux of micrometeorites reaching the earth has been determined to be ∼40,000 tons per year by Love and Brownlee (1993), which is two- to threefold higher than that estimated by Grun et al. (1985) or Hughes (1978). Peucker-Ehrenbrink and Ravizza (2000) estimated extraterrestrial matter flux to be 30,000±15,000 tons per year based on Os isotopes. In contrast, the total mass flux of meteorites in the mass range from 10 g to 1 kg is between 2.9 and 7.3 tons per year (Bland, 2001). It can therefore be concluded that micrometeorites dominate the incoming mass flux of extraterrestrial materials to the earth.

The mineralogical and compositional features of unmelted micrometeorites are reported to be similar to those of hydrous carbonaceous chondrites (Kurat et al., 1994), although the hydrous phases in most micrometeorites are decomposed and dehydrated (Nakamura et al., 2001). In addition, electron energy loss spectroscopic (EELS) analyses have revealed that micrometeorites yield higher concentrations of carbonaceous material (average C content is ∼7 wt%; Engrand and Maurette, 1998) than carbonaceous chondrites (average C content is ∼2.5 wt%; Engrand and Maurette, 1998). Carbonaceous chondrites represent only small fraction (∼3.7% fall frequency) of all meteorite falls (Grady, 2000) and, consequently, their accretion rates are much lower than those of micrometeorites. Since about 60% of micrometeorites are classified into the unmelted type (Yada et al., 2004), which can contain about 7 wt% of carbonaceous materials, 30,000 tons per year micrometeorite flux would deliver about 1,260 tons of carbon (C) per year to the earth. On the other hand, only 3.7% (carbonaceous chondrite) of 7.3 tons (maximum) per year of meteorite falls can contain about 2.5 wt% C. This may correspond to about 7 kg C per year by meteorites. This is in the same order of magnitude as that estimated by Pasek and Lauretta (2008) (organic C flux from carbonaceous chondirtes is about 0.5 kg/year). Therefore, extraterrestrial carbonaceous materials are accreting to the earth mainly as micrometeorites (1.3 × 10⁶ kg (micrometeorites) vs. 7 or 0.5 kg (meteorites) C per year).

Micrometeorites recovered from Antarctica are called Antarctic micrometeorites (AMMs). Many AMMs are composed of fine-grained minerals and interstitial organic material. Currently, only a few studies have been performed to characterize organic materials in AMMs. Aminoisobutyric acid (AIB) and isovaline were identified in AMMs (Brinton et al., 1998). This organic analysis needed 30–35 AMM grains to detect these compounds, and organic natures for individual grains are not known. Micro two-step laser mass spectrometry (μL²MS) on individual AMM grains identified complex organic molecules, such as polycyclic aromatic hydrocarbons (PAHs) (Clemett et al., 1998). However, this μL²MS method elevates the temperature of AMM samples for ionization, and organic features in AMMs can be modified.

Antarctic weathering of AMMs has been suggested by major element abundance (Terada et al., 2001). Sephton et al. (2004) reported that the presence and amount of organic matter from carbonaceous chondrites in Antarctica are strongly associated with Antarctic weathering. Therefore, the effects of weathering should also be evaluated together with the characterization of organic components in AMMs.

In the study reported here, we have attempted to characterize organic features of individual AMMs using nondestructive infrared (IR) and Raman micro-spectroscopic methods. Raman spectroscopy has been conducted on carbonaceous chondrites (Raynal, 2003; Matrajt et al., 2004) and interplanetary dust particles (I_DPs) (Wopenka, 1988) with the aim of characterizing graphitic macromolecular organics. However, to date, no Raman data have been reported on AMMs. Although transmission infrared (IR) spectra on AMMs have been reported by Osawa et al. (2001), we test the IR transmission-reflection method (Okumura and Nakashima, 2004) in this study in order to be able to use the same sample mounting method on aluminum (Al)-foils for all of the measurements. For evaluating the Antarctic weathering of AMMs, we determined major element abundances using an electron probe micro-analyzer (EPMA), and colored weathering products were analyzed by visible micro-spectroscopy. We also compared the obtained organic features of AMMs with those of carbonaceous chondrites.

2. Materials and Methods

2.1 Sampling location

The AMMs used in this study were picked up from Kuwagata nunatak (72°06t’S, 35°15t’E) No. 11 point in 1998 (Yada and Kojima, 2000). The area is 2 km southwest of the most southern nunatak in the Meteorite Ice Field (Yada and Kojima, 2000). A detailed description of the separation procedure is provided by Yada and Kojima (2000).

2.2 Sample preparation

More than 50 grains were identified as micrometeorites from the above sample location using a scanning electron microscope (SEM) equipped with energy-dispersive X-ray analyzer (EDX) based on the compositional criteria (Nakamura et al., 2001). The micrometeorites can be classified into three major textural types based on the degree of melting: unmelted type (least-melted), scoriaceous type (partially-melted), and spherules (totally-melted) (Engrand and Maurette, 1998). Typical unmelted micrometeorites comprise mainly fine-grained materials with various amounts/sizes of vesicles (Engrand and Maurette, 1998). The classification of the Kuwagata point 11 samples are all listed in the AMM catalog of Gakushuin University (http://dust.cc.gakushuin.ac.jp/dust/index.html). The SEM image in this catalog for one of the AMM grains is shown in Fig. 1 (sample name: Y98K11KS020). In order to study organic and hydrous components in AMMs, our aim was to investigate the unmelted type micrometeorites from the Kuwagata point 11 samples. Only eight unmelted AMM grains were left available for this study, which have not been treated by other analyses (Table 1).

Fig. 1.

The secondary electron image of an Antarctic micrometeorite (AMM) sample No. Y98K11KS020 by scanning electron microscopy (SEM). The scale bar is 20 μm.

Table 1.

Description of Antarctic micrometeorite samples.

Each of eight AMM grains was gently pressed between two Al-foils in order to embed the grains in the Al-foils. Four AMM grains remained on the Al-foils, while the other four AMM grains were separated into two particles on two Al-foils. Successful IR and Raman spectra were obtained on only 11 particles among the above 12. FourAMMgrains and two separated particles from one grain were successfully measured by visible micro-spectroscopy.

For EPMA analysis, AMM particles on the Al-foils were fixed with petropoxi resin and polished with 1-μm Al₂O₃ compound to make flat surfaces. Seven particles were successfully measured by EPMA.

2.3 IR spectroscopy

IR micro-spectroscopy was conducted directly on the AMMs embedded in the Al-foils. We used here the transmission-reflection method on the Al-foils (Okumura and Nakashima, 2004) in which the IR light was first transmitted into the sample grain, which was approximately 12 μm thick, reflected onto the surface of the Al-foils, and again transmitted through the sample. The resulting spectra have absorption features of the sample with twice the thickness without too much reflection artifacts on the grain surface, except for the strong absorption bands, such as silicates. This method enables weak absorption peaks, such as aliphatic CH, to be better observed.

For IR measurements, a Fourier transform infrared micro-spectrometer (micro-FTIR), model JASCO FTIR 620 + IRT30 was used. A ceramic infrared light source, MCT (HgCdTe) detector and a ×16 Cassegrainian mirror were used in the micro-FTIR. The interferometer scan speed is 10 cm/s, yielding a wavenumber resolution of 4 cm^-1. One hundred scans were accumulated for each resultant spectrum. The background spectra were collected on each Al-foil at room temperature. Transmission-reflection spectra at room temperature in the range 4,000-750 cm^-1 were collected on these AMMs on Al-foils with a 400×400- μm² aperture to cover the whole grain of the AMMs. Small fragments of carbonaceous chondrites (Ivuna, Orgueil, Tagish Lake, Murchison, Murray, Allende, Moss) pressed on Al-foils were also measured by the transmission-reflection method with a 100×100-μm² aperture.

2.4 Raman spectroscopy

The Raman spectra of both the unmelted AMMs and the Moss CO3 carbonaceous chondrites were measured using a visible-Raman micro-spectrometer (Yamanoi et al., 2004). For measuring Raman spectra, the sample is excited by a 514.5-nm Ar laser. In our study, the laser’s intensity was set about 0.2 mW at the sample surface to avoid its overheating. A Raman spectrum of the sample was measured with a duplicate accumulation of signals at an exposure time of 100 s. The effective wavenumber region is from 100 to 2100 cm^-1. The sharp spike peaks often found in Raman spectra due to cosmic rays were then removed from the spectra. The wavenumber resolution is about 2 cm^-1. A ×20 long-working distance objective lens was used, giving a spot size of about 5 μmφ.

We analyzed five to ten points for each unmelted AMM sample. In total, 73 points of eight unmelted AMMs were measured for Raman spectra under a Laser Raman microscope. The Moss meteorite was measured by the same method as that used for the AMMs. The small fragment of Moss meteorite was pressed between two Al-foils, and two positions were measured for Raman spectra.

2.5 Visible spectroscopy

Measurements of visible reflectance spectra were performed by visible micro-spectroscopy (Yamanoi et al., 2009), which is composed of a visible-Raman microspectrometer (Yamanoi et al., 2004) with a dark field objective. The effective wavelength region is from 400 to 800 nm. The wavelength resolution is about 0.5 nm. A ×20 dark field objective lens was used, giving a spot size of about 50 μmφ.

A dark spectrum was measured by closing the CCD detector shutter; this dark spectrum was subtracted from all the spectra to remove CCD counts by a dark current. Dark field reflection intensity at 400–800 nm (reflection spectrum) of the samples on Al-foils was obtained with a duplicate accumulation of signals at an exposure time of 1 s with an unexposed interval of 1 s. Reflectance spectra of the sample were calculated by dividing the reflection spectra of the samples by those of the white reference (Yamanoi et al., 2009). Amorphous silicon dioxide powders (Kanto Chemical Co.) were used as the white reference material.

One to three positions of each of six AMM particles on Al-foils were measured by visible micro-spectroscopy. Three carbonaceous chondrites (Ivuna CI1, Murchison CM2, and Allende CV3) were also pressed on Al-foils, and three to seven points were measured by visible microspectroscopy.

2.6 Electron probe micro-analyzer

The elemental abundance of the polished surfaces of AMM particles were measured by EPMA (JEOL JXA-733 and -733 superprobe). Quantitative analyses with wavelength dispersive X-ray spectrometer were performed at 15 kV of accelerating voltage and a 9- or 10-nA electron beam current with a defocused beam of 20 μm in diameter.

3. Results

3.1 Weathering effects

3.1.1 EPMA analysis

Two to 19 points on each AMM were measured by EPMA and averaged. The elemental abundance of AMMs normalized to Si and solar abundance (Anders and Grevesse, 1989) is summarized in Fig. 2(a). The Al abundance is removed from the diagram because Al-foils and Al₂O₃ compounds were used for holding and polishing the samples, respectively. The Ti contents of AMMs are around the detection limit and will not be discussed here.

Fig. 2.

(a) Elemental abundance diagram (normalized to Si and CI chondrites; Anders and Grevesse, 1989) of Antarctic micrometeorites (AMMs) compared with average data on phyllosilicate and crystalline types AMMs (Kurat et al., 1994). Ti is around detection limit. (b) Average elemental abundance diagram (normalized to Si and CI chondrites) of CM2, CR2, CV3, CO3 (McSween and Richardson, 1977) and Tagish Lake (Greshake et al., 2005).

Y98K11KS010 was found to be phosphorus (P)-enriched compared with the solar abundance, Y98K11KS020-1 to be P- and iron (Fe)-enriched, while Y98K11KS093-1 was found to be potassium (K)-depleted, manganese (Mn)- enriched and Fe-depleted. Except for these anomalies, the chemical compositions of AMMs were found to be basically similar to one another. These characteristics of elemental compositions are similar to those for AMMs reported in Wright et al. (1997). Compared with the abundance patterns reported in Kurat et al. (1994), our data resemble those of the phyllosilicate micrometeorites (Fig. 2(a)). SEM observations of our AMM samples revealed fine-grained textures. This and the similarity in elemental abundance suggest that most of our AMM samples might be phyllosilicate-type micrometeorites. Phyllosilicates can be decomposed and dehydrated during atmospheric entry (Nakamura et al., 2001). However, they can also be formed by Antarctic weathering (Noguchi et al., 1999).

The abundances of major and minor elements of carbonaceous chondrites are summarized in Fig. 2(b). In these chondrites, magnesium (Mg) was found to be generally more depleted from the solar abundance than from the carbonaceous chondrites. K was found to be more enriched than the CM2 and CR2 carbonaceous chondrites, except for Y98K11KS093-1. Nickel (Ni) was significantly depleted from the solar abundance. Other elemental abundances were found to be similar to those found in either type 2 or type 3 chondrites (CM2, CR2, CV3, CO3 chondrite data from McSween and Richardson, 1977; Tagish Lake data from Greshake et al., 2005).

Atomic ratios of three major elements Si, Mg and Fe in AMMs are shown in Fig. 3(a). The AMM data are plotted in a relatively narrow compositional field that includes both low-calcium (Ca) pyroxene and olivine solid solutions (Fig. 3(a)). Compared with AMMs reported previously (Kurat et al., 1994; Genge et al., 1997; Gounelle et al., 2005) (Fig. 3(b)), our AMMs are apparently depleted in Mg (Fig. 3(a)). The same Mg depletion for the studied AMMs is also observed in Fig. 3(a) with lower Mg/(Si+Fe) ratios than those for the C1 (CI1), C2 (CM2, CR2 and Tagish Lake), C3 (CO3 and CV3) chondrites (C1, C2, C3 carbonaceous chondrites data from McSween and Richardson, 1977; Zolensky et al., 1993; Tagish Lake data from Greshake et al., 2005).

Fig. 3.

(a) Ternary Si-Fe-Mg diagram showing the composition of several different points in 7 Antarctic micrometeorite grains (AMMs) investigated in this work. Olivine and pyroxene are represented by gray lines. Dotted and dashed circles are ranges for each group of carbonaceous chondrites (C1 (CI1), C2 (CM2, CR2 and Tagish Lake), C3 (CO3 and CV3). CI1, CM2, CR2, CO3 and CV3 data from McSween and Richardson (1977) and Zolensky et al. (1993). Tagish Lake data from Greshake et al. (2005)). (b) Ternary Si-Fe-Mg diagram of other AMM data from Kurat et al. (1994), Engrand et al. (1999), Genge et al. (1997) and Gounelle et al. (2005).

These elemental abundances of AMMs reveal Mgdepletion, especially for Y98K11KS004, Y98K11KS0201, and Y98K11KS037 (Fig. 2(a) and Fig. 3(a)).

3.1.2 Visible spectroscopy

The visible reflectance spectra of five AMMs are shown in Fig. 4. Carbonaceous materials generally absorb a shorter wavelength than the visible range (ultraviolet region) resulting in a yellowbrownish color (Stevenson, 1994). A weak absorption band around 500 nm was recognized for Y98K11KS004, Y98K11KS010, Y98K11KS020-2, and Y98K11KS037 (Fig. 4). This 500-nm band can be mainly due to electron pair transitions of Fe³⁺ in goethite or ferrihydrite structures (Scheinost et al., 1998) among ligand field and electron pair transitions of Fe³⁺ (Sherman and Waite, 1985). In fact, ferrihydrite has been reported in AMMs by Matrajt et al. (2001). However, Osawa et al. (2003) and Nakamura et al. (2001) did not detect goethite or other iron hydroxides by synchrotron X-ray diffraction. Therefore, the Fe³⁺- bearing-hydroxide-like materials may have an amorphous nature with relatively low crystalinity (Nagano et al., 2002).

Fig. 4.

Visible reflectance spectra of 5 Antarctic micrometeorite grains (AMMs). The Y98K11KS020 grain was separated into two particles on two Al foils (-1 and -2). Two locations on the Y98K11KS020-2 particle and three positions on the Y98K11KS010 grain were analyzed under the dark-field visible microspectrometer.

On the other hand, the 500-nm band of iron-hydroxidelike materials cannot be recognized in visible reflectance spectra of Ivuna (C1), Murchison (C2) and Allende (C3) carbonaceous chondrites (Table 3).

3.2 IR spectroscopy

IR spectra for the AMMs were measured here with the transmission-reflection method. All of the IR spectra of AMMs are shown in Fig. 5(a). Almost all of the spectra show the following absorption bands in common at 3,400, 2,960, 2,925, 2,850, and 1,630 cm^-1. However, the strong absorptions around 1,000 cm^-1 due to Si-O stretching of silicates show reflection artifacts. Osawa et al. (2001) already reported IR spectra of some AMMs by the transmission method on diamond plates. Their IR spectra are better resolved for silicate bands, which can also be saturated for thick samples. The present transmission-reflection method on Al-foils is useful for organic and hydrous IR features, but not for silicates.

Fig. 5.

(a) Infrared (IR) transmission-reflection spectra of 11 Antarctic micrometeorite particles with an aperture of 400×400 μm area under the IR micro-spectrometer. (b) IR transmission-reflection spectra of carbonaceous chondrites with an aperture of 100×100 μm area under the IR micro-spectrometer.

For most of the AMM particles, the features of absorption bands of molecular water are found at 3,400 cm^-1 and 1,630 cm^-1 (Aines and Rossman, 1984). The 3,400 cm^-1 band is observed for all the AMMs, except for Y98K11KS004 and Y98K11KS093-1, 2 (Fig. 5(a)). Three peaks are often found in the 2,800-3,000 cm^-1 region, corresponding to aliphatic hydrocarbon stretching features. The peaks at 2,925 and 2,850 cm^-1 correspond to the asymmetrical and symmetrical stretching vibrations of CH₂, respectively, while the peak at 2,960 cm^-1 corresponds to the asymmetrical stretching vibrations of CH₃ in aliphatic hydrocarbons (D’Hendecourt and Allamandola, 1986; Flynn et al., 2001). These CH peaks were recognized for Y98K11KS0010, Y98K11KS020-1, Y98K11KS037, and Y98K11KS077-1, 2 (Fig. 5(a)). The CH bending vibrations of these aliphatic hydrocarbons appear as very weak peaks around 1,465 cm^-1.

3.3 Raman spectroscopy

A representative Raman spectrum for each AMM particle is shown in Fig. 6(a). Twenty of 73 Raman spectra of the AMMs showed a peak around 700 cm^-1. This peak can be due to iron oxides, such as hematite (Amri et al., 2005) or ferrite (Graves et al., 1988). In fact, coarse-grained (5- 20 μm across) Fe oxides have been previously reported in AMMs (Noguchi and Nakamura, 2000).

Fig. 6.

(a) Representative Raman spectra for different positions of about 1 μm spot size under the Raman micro-spectrometer in 8 Antarctic micrometeorite grains. (b) A representative Raman band fitting with 2 Lorentzian bands to obtain peak positions, peak heights (I_D, I_G), band areas (A_D, A_G) and full width at half maximum (FWHM) for D (disordered: 1360 cm^-1) and G (graphite: 1600 cm^-1) bands (sample: Y98K11KS010).

The peak position and the FWHM of the G band of the AMMs range from 1,582 to 1,595 cm^-1 and from 81 to 133 cm^-1, respectively (Fig. 7(a)). The peak position and the FWHM of the D band for the AMMs range from 1,357 to 1,373 cm^-1 and from 166 to 265 cm^-1, respectively (Fig. 7(b)). The I_D/I_G and A_D/A_(D+G) (%) range from 1.1 to 1.6 and from 65 to 78, respectively (Fig. 7(c)). These data are summarized in Table 2.

Fig. 7.

The correlation diagrams of Raman parameters for G and D bands of the Antarctic micrometeorites (AMMs). (a) FWHM vs. peak position of G band, (b) FWHM vs. peak position of D band, (c) the peak intensity ratio (I_D/I_G) vs. band area ratio (A_D/A_(D+G)).

Table 2.

Summarized Raman parameters such as peak position, FWHM, the peak intensity ratio (I_D/I_G), band area ratio (A_D/A_(D+G)) for G and D bands of the Antarctic micrometeorites (AMMs).

In this study, two major Raman bands around 1,360 and 1,600 cm^-1 were observed for all the AMMs together with Moss carbonaceous chondrite. These two bands have also been reported in carbonaceous chondrites (CI1, CM2, CR2, Tagish Lake, CV3: Cronin et al., 1988; Raynal, 2003; Matrajt et al., 2004). These Raman bands are assigned as D (disordered) and G (graphite) bands of aromatic carbon, respectively (Fig. 6(b)). The G band (1,600 cm^-1) is due to graphite-like sp₂ carbon bonds, while the D band (1,360 cm^-1) is from disorderd sp₃ carbons (Ferrari and Robertson, 2000). Although the D band originates from the diamond-like structure of C, diamond has been reported to be present only in trace amounts as nano-particles in meteorites and micrometeorites (Dai et al., 2002). Instead, the D band often indicates disordered natures of amorphous graphitic carbonaceous materials (Wopenka and Pasteris, 1993; Ferrari and Robertson, 2000), which are major C forms in cosmic materials (Ehrenfreund et al., 2002). Therefore, in our study, the D band corresponds primarily to “disordered” nature of graphite.

The Raman spectra exhibited significant fluorescence emissions as evidenced by the increasing background intensity toward the higher frequency (larger Raman shift). After a linear baseline correction from 1,000 to 2,000 cm^-1, the Raman spectra were fitted by two Lorentzian bands, with the aim of extracting individual band parameters (peak position, full width at half maximum (FWHM), peak intensity, and band area; Fig. 6(b)), as described in Raynal (2003). The obtained values of peak position and FWHM of the G band are plotted in Fig. 7(a), and the peak positions and FWHMs of the D bands are shown in Fig. 7(b). The peak intensity ratio of D band (I_D) to that of G band (I_G) (I_D/I_G) and the band area ratio of D band to the sum of D and G bands (A_D/A_(D+G)) (%) are presented in Fig. 7(c). These parameters indicate the degrees of structural organization (ordering) of terrestrial carbonaceous material (Wopenka and Pasteris, 1993). With increasing ordering, the FWHM of the G and D bands can be seen to become narrower and the peak position (frequency) of the G bands to become larger. These bands have already been reported for insoluble organic matter from carbonaceous chondrites (Busemann et al., 2007) and used as an indicator of thermal metamorphism of carbonaceous chondrites (Bonal et al., 2006).

4. Discussion

Within the framework of discussing the characteristics of organic components in AMMs, we will first evaluate their weathering after falling into Antarctic ice sheets, based on elemental composition and the visible micro-spectroscopy. Organic features in the IR and Raman spectra of AMMs will then be discussed and compared with those in carbonaceous chondrites.

4.1 Weathering in Antarctic ice sheets The EPMA analyses revealed that major element abundances of our AMMs are mostly similar to those for micrometeorites reported previously (Kurat et al., 1994; Genge et al., 1997; Gounelle et al., 2005)—with the exception of Mg abundance (Fig. 3(b)). Although relative to carbonaceous chondrites, AMMs are K-enriched and Mg- and Ni-depleted, the abundance of major and minor elements in AMMs grains is relatively similar to the average abundance patterns of carbonaceous chondrites (McSween and Richardson, 1977; Greshake et al., 2005) (Fig. 2(b)). Kurat et al. (1992a, b) suggested that the K enrichment of AMMs is likely to be of terrestrial atmospheric origin. The origin of the Ni depletion in AMMs is not yet understood due to conflicting explanations provided by earlier researchers (Presper et al., 1993; Brownlee et al., 1997), or it is due to atmospheric entry (Badjukov and Raitala, 2003).

Terada et al. (2001) found a correlation between the Mg/silicon (Si) ratio (SEM-EDS peak height ratio) at the surface of the unmelted AMM samples and the age of ice in which the AMMs are embedded. The Mg/Si ratio of samples which were taken from the same point as our AMMs (Kuwagata point No. 11) is reported to be 0.43±0.14 (Terada et al., 2001). Based on the Mg/Si ratio vs. age of ice sheets (Terada et al., 2001), this sample is supposed to have been buried in Antarctic ice for about 30,000 years, which is consistent with the ice core age (Terada et al., 2001). Therefore, the Mg depletion found in our AMMs can be explained by weathering in the Antarctic ice for a relatively long period and, therefore, can be taken as an indicator of Antarctic weathering (Table 3).

The visible reflectance spectra of AMMs showed a weak broad absorption band around 500 nm which can be due to Fe-hydroxide-like materials (Fig. 4). On the other hand, we did not find this 500-nm band in the visible spectra for carbonaceous chondrites (Ivuna C1, Murchison C2 and Allende C3). The presence of this 500-nm band can also be taken as an indicator of Antarctic weathering (Table 3).

4.2 IR spectroscopy

IR spectra of all the analyzed AMMs show a very broad band at 3,400 cm^-1 together with a smaller band around 1630 cm^-1 (Fig. 5(a)). These bands are considered to be due to hydrogen-bonded liquid-like molecular water (Nakamoto et al., 1955; Aines and Rossman, 1984), which may derive from the absorbed water in phyllosilicates. Nakamura et al. (2001), Noguchi et al. (2002), and Nakamura and Noguchi (2004) reported that the vast majority of phyllosilicate-rich AMMs contain saponite as the major phyllosilicate. This kind of phyllosilicate is generally known to be formed during aqueous alteration of carbonaceous chondrites (Jones and Brearley, 2006). However, Noguchi et al. (1999) also showed the possibility of minute phyllosilicates being generated by Antarctic weathering. The presence of the 3,400 cm^-1 band can therefore be an indicator of aqueous alteration with some possible contribution of Antarctic weathering (Table 3).

Table 3.

Summary of visible, IR and Raman micro-spectroscopic data together with EPMA data on the studied AMMs and carbonaceous chondrites. Classification of chondrite type based on IR OH, H₂O and CH bands is also given.

The IR spectra of some of the AMMs (Fig. 5(a)) display two peaks around 2,925 and 2,960 cm^-1, due to asymmetric CH₂ and CH₃ stretching vibrations, respectively. These two peaks indicate the presence of aliphatic hydrocarbons in some of the AMMs (Table 3).

4.3 Raman spectroscopy

Raman spectra of all the AMMs show two Raman bands, namely the D and G bands (Ferrari and Robertson, 2000) (Table 2). Both bands are broad, indicating the less ordered natures of the carbonaceous materials. Furthermore, significant fluorescence emission backgrounds in the Raman spectra (Fig. 6(b)) provide evidence for the presence of amorphous graphitic structures (Raynal, 2003).

The mean diameter of graphene-like sheets (L_a) (Pasteris and Wopenka, 1991) of the AMMs has been estimated to be 20∼30 Å based on the L_a vs. A_D/A_(D+G) relation (Wopenka and Pasteris, 1993). This L_a value is similar to that reported for Tagish Lake organics (about 20 Å ) (Nakamura et al., 2002). The graphene-like structures for the studied AMMs are possibly similar to those for aqueously altered carbonaceous chondrites.

4.4 Spectroscopic classification of carbonaceous chondrites

4.4.1 IR spectroscopy

The classification of carbonaceous chondrites by IR spectroscopy has been reported by Osawa et al. (2005) using the peak height ratios of hydrous components. We suggest here an alternative classification based on the criteria of carbonaceous chondrites in terms of the existence or absence of peaks of 3,685, 3,400, and 2,925 cm^-1 (Table 3).

Hydrous phyllosilicates, such as serpentine-group minerals, have generally a sharp peak around 3,685 cm^-1 due to OH species bound to the mineral lattice (Osawa et al., 2005). This phyllosilicate peak was observed for the CI1 and Tagish Lake carbonaceous chondrites, but not for the CM2 and CR2 carbonaceous ones (Fig. 5(b)). However, CM2 carbonaceous chondrites, such as Murchison, have been reported to contain phyllosilicates (Osawa et al., 2001). The IR spectroscopy appears to be unable to detect trace phyllosilicates. The presence of the 3,685 cm^-1 peak can indicate the presence of chondrite type 1 plus Tagish Lake type (Table 3).

A very broad band at 3,400 cm^-1 was observed for CI1, CM2, CR2, and Tagish Lake carbonaceous chondrites but not for the CO3 and CV3 ones (Fig. 5(b)). The presence of this 3,400 cm^-1 band can indicate the presence of chondrite types 1 or 2 (Table 3).

Aliphatic C-H bands were detected at 2,925 cm^-1 together with the smaller 2,960 cm^-1 band in CI1, CM2, CR2 and Tagish Lake carbonaceous chondrites but not in the CO3 and CV3 ones (Fig. 5(b)). The presence of the 2,925 cm^-1 band can correspond to chondrite types 1 or 2 (Table 3).

Therefore, the presence of the 3,685, 3,400, and 2,925 cm^-1 bands can correspond to type 1 chondrite, the presence of the 3,400 and 2,925 bands without the 3,685 cm^-1 band can correspond to type 2, and the absence of these bands can correspond to type 3. Although Tagish Lake is mineralogically grouped into type 2, it corresponds to type 1 based on the present criteria (Table 3).

4.4.2 Raman spectroscopy

The Raman parameters for primitive carbonaceous chondrites (Orgueil, Alais, Ivuna CI1; Murchison, Cold Bokkeveld, Murray CM2; Renazzo CR2; Tagish Lake C2, Allende and Axtell CV3) in the literature (Raynal, 2003; Matrajt et al., 2004) and Moss CO3 (this study) were also plotted in our figures (Fig. 8(a/2-c)) and summarized in Table 2.

Fig. 8.

The correlation diagrams of Raman parameters for G and D bands of carbonaceous chodrites. Except for CO3 measured in this study, CI1s, CM2s, CR2 and CV3s data are from Raynal (2003) and Tagish Lake data from Matrajt et al. (2004). (a) FWHM vs. peak position of G band, (b) FWHM vs. peak position of D band, (c) the peak intensity ratio (I_D/I_G) vs. band area ratio (A_D/A_(D+G)).

The peak position of the G band for the carbonaceous chondrites tends to become smaller—from about 1,605 cm^-1 for type 3 to about 1,585 cm^-1 for type 1 (Fig. 8(a))—indicating a more amorphous nature. The FWHM of the G band for the carbonaceous chondrites tends to become larger—from about 60 cm^-1 for type 3 to about 160 cm^-1 for type 1 (Fig. 8(a))—indicating a more amorphous nature (Ferrari and Robertson, 2000).

The peak position of the D band for the carbonaceous chondrites tends to become larger—from about 1,350 cm^-1 for type 3 to about 1,390 cm^-1 for type 1 (Fig. 8(b)). The FWHM of the D band for the carbonaceous chondrites tends to become larger—from about 70 cm^-1 for type 3 to about 380 cm^-1 for type 1 (Fig. 8(b))—indicating a more amorphous nature (Wopenka and Pasteris, 1993).

The band intensity ratio I_D/I_G for the type 3 carbonaceous chondrites is larger (1.8–1.2) than others reported. The one for type 2 has smaller values of between 1.1 and 0.7. This I_D/I_G ratio for type 1 shows similar ranges: 1.3–0.7 (Fig. 8(c)).

The band area ratio A_D/A_(D+G) (%) for the type 3 carbonaceous chondrites is smaller (45–60%) than the others. The one for type 2 has larger values of between 55 and 75%. This A_D/A_(D+G) (%) ratio for type 1 shows similar ranges of between 55 and 80% (Fig. 8(c)).

4.5 Spectroscopic classification of AMMs

4.5.1 IR spectroscopy

The classification for carbonaceous chondrites based on the presence of the 3,685, 3,400, and 2,925 cm^-1 bands can be applied to the AMMs studied here. The presence of the 3,685, 3,400, and 2,925 cm^-1 bands can correspond to type 1 chondrite, the presence of the 3,400 and 2,925 cm^-1 bands without the 3,685 cm^-1 band can correspond to type 2, and the absence of these bands can correspond to type 3 (Table 3).

Based on this classification criterion, Y98K11KS010, Y98K11KS020-1, Y98K11KS037, and Y98K11KS077- 1, 2 are classified as type 2. Y98K11KS023-1, 2 and Y98K11KS026 can be classified to type 2 by the presence of the H₂O band. However, they do not show the 2,925 cm^-1 band, and they can be designated as type 2t’. Y98K11KS004 and Y98K11KS093-1, 2 are classified as type 3 (Table 3).

4.5.2 Raman spectroscopy The peak position and the FWHM of the G band ranges of AMMs were found to mainly overlap those for CM2 carbonaceous chondrites (Fig. 9(a)) and those of the D band region for the AMMs to mainly overlap with the region for CI1s (Fig. 9(b)). The D peak position range of the AMMs can be seen to be similar to those for CM2 and CR2 chondrites (Fig. 9(b)). The I_D/I_G and A_D/A_(D+G) (%) distributions of the AMMs are close to those of the CI1 carbonaceous chondrites, and the A_D/A_(D+G) band area ratios (%) of the AMMs are in a similar range to those for CR2 and Tagish Lake carbonaceous chondrites (Fig. 9(c)). The results of Raman analysis suggest that, although some data deviates slightly, the aromatic carbonaceous material in the studied AMMs is different from type 3 carbonaceous chondrites and rather similar to types 1 and 2 carbonaceous chondrites. In fact, the PAHs in AMMs were reported to have some similarity to those in CM2 carbonaceous chondrites (Clemett et al., 1998).

Fig. 9.

The correlation diagrams of Raman parameters for G and D bands of the Antarctic micrometeorites (AMMs) and carbonaceous chodrites. (a) FWHM vs. peak position of G band, (b) FWHM vs. peak position of D band, (c) the peak intensity ratio (I_D/I_G) vs. band area ratio (A_D/A_(D+G)).

4.6 The relation of AMMs with carbonaceous chondrites

The weathering effects on the studied AMMs can be indicated by the Mg depletion, as evidenced from the EPMA analyses, and the presence of a 500-nm band due to Fe-hydroxide-like materials, based on visible microspectroscopy results (Table 3). The IR and Raman data for AMMs and carbonaceous chondrites are also summarized in Table 3.

All of the AMMs appear to show varying degrees of weathering. The weathering effects on organic materials during residence in Antarctic ice (Tyra et al., 2007) are not yet well understood. Sephton et al. (2004) suggested that Antarctic weathering may have similar effects as the parent body aqueous alteration by removing ether bonds from organic material and alkyl side chains from its constituent units (Sephton et al., 2004).

The IR H₂O and CH bands of the AMMs showed no clear relation to these weathering effects (Table 3). It should be noted that, despite the Mg depletion and the presence of Fe-hydroxide-like materials, Y98K11KS004 is similar to type 3 carbonaceous chondites without the presence of the hydrous phases (Table 3). The Raman parameters of the AMMs are in a relatively similar range regardless of the weathering degree. However, G band peak positions for some AMM grains deviate from the clustered region to the smaller values (left hand side in Fig. 7(a)). This slight disordering of graphic materials for some AMMs may possibly be due to Antarctic weathering.

Other effects, such as heating effects during atmospheric entry on the AMMs, should also be considered. Matrajt et al. (2006) reported that organics in unmelted micrometeorites can be lost either by thermal degradation or evaporation upon atmospheric entry. They also pointed out the possible survival of some organic molecules in micrometeorites after atmospheric entry. The G band FWHM values for one AMM grain (Y98K11KS004) deviated somewhat from the clustered region to smaller values (downward in Fig. 7(a)). This slight ordering may be due to heating during atmospheric entry. However, the heating effects may be variable and could not be evaluated in the study reported here.

Although Y98K11KS004 have clear weathering evidences, it does not show hydrous IR features. Hydrous phyllosilicate formation during Antarctic weathering might not be significant compared with Mg-depletion and ironhydroxide- like material formation. Raman parameters for this AMM particle are similar to type 2 or 1 carbonaceous chondrites despite the heating during atmospheric entry.

Therefore, we conclude that the IR and Raman features of the AMMs may retain at least some original characters before Antarctic weathering. Our multi-method approach can be useful for the genetic classification of AMMs in terms of any discussion on its parent body. Tomeoka et al. (2003) suggested that hydrous asteroids will produce much more abundant fragments than anhydrous asteroids during the asteroidal collision. This may be due to the wet and porous natures of these hydrous asteroids. Abundant dust particles are formed before arriving on the earth’s surface as carbonaceous chondrites and micrometeorites (Tomeoka et al., 2003). Reflectance spectroscopy measurements (Jones et al., 1990) indicate that 30–40% of known asteroids are hydrated CI-CM-like asteroids, which are possible parent bodies of CI-CM chondrites (Tomeoka et al., 2003). Since the micrometeorites are defined only by size—i.e., smaller than 1 mm, AMMs may also have an origin in similar parent bodies as carbonaceous chondrites.

For the further understanding of original organic and hydrous components in AMMs without the weathering effects, fresh AMMs should be measured by the present multimethods.

5. Conclusions

We investigated eight unmelted Antarctic micrometeorite grains (AMMs) by pressing these on Al-foils and measuring by IR, Raman, and visible reflection micro-spectroscopy in combination with EPMA analyses. These data were compared with those for carbonaceous chondrites. The following results were obtained.

• 1) The elemental compositions of AMMs show Mg depletion (Table 3), suggesting their weathering in Antarctic ice sheets.

• 2) The visible reflectance spectra of some of the AMMs show a weak absorption band around 500 nm due to the presence of Fe-hydroxide-like materials. On the other hand, this band cannot be recognized in visible reflectance spectra of carbonaceous chondrites (Table 3).This difference may also indicate the Antarctic weathering.

• 3) The OH, H₂O, and CH bands in the transmissionreflection IR spectra can be used to classify the AMMs into carbonaceous chondrite types (Table 3). Using these criteria, four AMM grains were classified into type 2, while two grains were classified into type 3.

• 4) The Raman features of D (disordered: 1360 cm^-1) and G (graphite: 1600 cm^-1) bands of graphitic carbonaceous materials in the AMMs are not similar to those for type 3 but rather close to those for type 2 and 1 carbonaceous chondrites, although some data shows a certain degree of deviation.

• 5) The weathering effects in Antarctic ice sheets and heating effects during atmospheric entry on AMMs should be evaluated for the presence of organic and hydrous components.

• 6) The genetic classification of individual AMM grains can be conducted using the multiple microspectroscopic reflectance methods on the same sample configuration pressed on Al-foils. These methods will also be useful for characterizating any precious small samples.