Available online at www.sciencedirect.com

www.elsevier.com/locate/gca

Geochimica et Cosmochimica Acta 74 (2010) 4881–4892

Hydrous mineralogy of CM and CI chondrites frominfrared spectroscopy and their relationship with low albedo

asteroids

P. Beck a,*, E. Quirico a, G. Montes-Hernandez a,1,L. Bonal b, J. Bollard a, F.-R. Orthous-Daunay a, K.T. Howard c, B. Schmitt a,

O. Brissaud a, F. Deschamps d, B. Wunder e, S. Guillot d

a Universite de Grenoble, Laboratoire de Planetologie de Grenoble, Institut de Planetologie et d’Astrophysique de Grenoble, OSUG/CNRS,

122 rue de la piscine, 38000 Grenoble, Franceb University of Hawai’i at Manoa, Honolulu, HI 96822, USA

c Impacts and Astromaterials Research Centre (IARC), The Natural History Museum, Mineralogy Department, London SW7 5BD, UKd Universite de Grenoble, IsTerre, OSUG/CNRS, 1381 rue de la Piscine, Grenoble, France

e German Research Centre for Geosciences GFZ Potsdam, Telegrafenberg, 14473 Potsdam, Germany

Received 21 December 2009; accepted in revised form 12 May 2010; available online 25 May 2010

Abstract

IR spectroscopy is one of the few techniques that can directly probe water molecules in rocks. This method has been used tocharacterize the mineralogy of hydrated/hydrous carbonaceous chondrites, and to link known meteorite families with spectro-scopic observations of low albedo asteroids. In this paper, we present measurements of the infrared transmission spectra ofmatrix chunks from 3 CI and 9 CM chondrites. Spectra were measured at ambient conditions and then at different temperaturesalong a dehydration path toward high-T (�300 �C) under primary vacuum. At ambient conditions, the 3-lm spectral range isalways dominated by adsorbed atmospheric water molecules. Upon moderate (�100 �C) and high (�300 �C) heating under lowpressure (P < 10�4 mbar), adsorbed water and then phyllosilicates interlayer water are removed, revealing a residual absorptionband around 3 lm. This band is a characteristic IR feature of the phyllosilicate phases which dominate the mineralogical assem-blage of hydrated carbonaceous chondrites. Among the CM chondrites, the high-T spectra reveal a strong variability thatappears correlated with the alteration classification scheme of Rubin et al. (2007) and Howard et al. (2009a). The 3-lm bandcontinuously evolves from a broad feature peaking at 3550–3600 cm�1 for the weakly altered CMs (Murchison-type) to a sharpasymmetric peak at�3675 cm�1 for the more extensively altered samples (Cold Bokkeveld-type). We attribute this spectral evo-lution to variations in the chemistry of the phyllosilicate phases from Fe-rich to Mg-rich. On the other hand, the 10-lm spectralregion shows a single broad peak which does not compare with known terrestrial serpentine spectra, probably due to high struc-tural disorder of the chondrite phyllosilicate phases. The present work clearly shows that previously published reflectance spec-tra of chondrites are biased by the presence of adsorbed terrestrial water molecules. Laboratory data collected under dryconditions are needed to reinterpret the chondrite–asteroid connection from the comparison of their 3-lm absorption features.� 2010 Elsevier Ltd. All rights reserved.

0016-7037/$ - see front matter � 2010 Elsevier Ltd. All rights reserved.

doi:10.1016/j.gca.2010.05.020

* Corresponding author. Address: Laboratoire de Planetologie deGrenoble, Bat. Physique D, 122 rue de la piscine, BP53, 38041Grenoble Cedex 09, France. Tel.: +33 (0) 476635277; fax: +33 (0)476514146.

E-mail address: pierre.beck@obs.ujf-grenoble.fr (P. Beck).1 Present address: Universite de Grenoble, IsTerre, OSUG/

CNRS, 1381 rue de la Piscine, Grenoble, France.

1. INTRODUCTION

Carbonaceous chondrites record evidence for the actionsof water in the early Solar System. They can contain up to10 wt.% of H2O in the form of hydrated, hydroxylated oroxyhydroxylated mineral phases, that were produced

4882 P. Beck et al. / Geochimica et Cosmochimica Acta 74 (2010) 4881–4892

possibly during some nebular stage, but more likelythrough parent body processes (e.g., Brearley, 2006). Theextent of aqueous alteration varies across the different car-bonaceous chondrite groups, but is possibly also variableon a given parent body (McSween, 1979). Many workershave focussed on the numerous CM chondrites and triedto identify a mineralogical evolution among the samples.A variety of classification schemes have been discussedbased on different petrological, mineralogical and chemicalcriterion (McSween, 1979; Bunch and Chang, 1980; Burgesset al., 1991; Zolensky et al., 1993; Browning et al., 1996;Rubin et al., 2007; Howard et al., 2009a,b).

Infrared spectroscopy is one of the few techniques thatcan directly probe protons and then OH and H2O groups(Farmer, 1974). This technique is widely used in planetarysciences to identify ice and hydrated minerals at the surfaceof asteroids and planets from solar reflectance spectroscopy(Hiroi et al., 1996, 2001; Rivkin et al., 2000; Jouglet et al.,2007; Milliken and Rivkin, 2009). The diffuse reflectanceIR spectra of several carbonaceous chondrites were mea-sured by Miyamoto and Zolensky (1994) to quantify theamount of hydrous minerals they contain. The intensity ofthe so-called 3-lm band absorption feature is the combina-tion of the stretching m1 and antistretching m3 vibrationmodes of water. This intensity was found to slightly corre-late with the total hydrogen content of the samples. How-ever, the connection to the amount of phyllosilicatesappeared unclear (Miyamoto and Zolensky, 1994). A majorlimitation of this study lies in the fact that spectra were mea-sured at ambient conditions. Thus, the amount of adsorbedwater is high and potentially dominates the reflectance spec-tra of the samples (Pommerol et al., 2009). More recentlyOsawa et al. (2005) tentatively built a taxonomy of carbona-ceous chondrites from infrared transmission measurements.Spectra were obtained at ambient conditions and might alsobe perturbed by a significant amount of adsorbed water.

Here, we report on mid-infrared measurements of thematrices of several CI and CM chondrites in a dry environ-ment allowing the removal of the spectral contribution of ad-sorbed terrestrial water. Transmission spectra were obtainedusing an environmental cell than can be heated up to 300 �Cand maintained under primary or secondary vacuum. Thisapproach enables to remove weakly bonded water moleculesand observe the mineralogical signatures of structural OHand H2O. An evolution, correlated with the alteration se-quence defined by Rubin et al. (2007) is observed amongthe spectra of CM chondrites. This spectral evolution mightcorrespond to the progressive replacement of the phyllosili-cates interlayer cations Fe2+ by Mg2+ as alteration progresseson the CM parent body. The present work also evidences thatunder ambient conditions the IR spectra of hydrated matricesare dominated by terrestrial water, hampering a reliable con-nection between laboratory and asteroid spectra.

2. SAMPLES AND ANALYTICAL METHODS

2.1. Samples

Twelve meteorites having experienced aqueous alter-ation on their parent bodies were considered in the present

work, among which 9 CM and 3 CI chondrites. The se-lected CM samples span almost the full range of aqueousalteration (from 2.6 to 2.0) as defined by Rubin et al.(2007). They are: Queen Alexandra Range 97990 (QUE97990, 2.6), Murchison (2.4), Murray (2.4/2.5), QueenAlexandra Range 99355 (QUE 99355, 2.3), Cold Bokkeveld(2.2), Mighei, Boriskino and Nawapali (not studied in Ru-bin et al., 2007). The CI chondrites Alais, Ivuna and Orgu-eil were studied as well. Samples were provided by theMuseum National d’Histoire Naturelle (Paris, France)and the Johnson Space Center – NASA (Houston, USA).

2.2. Infrared spectroscopy

Infrared spectra were obtained with a BRUKERHYPERION 3000 infrared microscope at the Laboratoirede Planetologie de Grenoble (LPG, France). The IR beamwas focused through a 15� objective and the typical size ofthe spot on the sample was �50 � 50 lm2. Spectra weremeasured at 4 cm�1 spectral resolution with a MCT detec-tor cooled with liquid nitrogen. Particular care was devotedto sample preparation, which is a critical issue in infraredmicro-spectroscopy. Samples must be thin (<100 lm) andtheir surface flat enough, in order to avoid absorption bandsaturation and scattering artefacts, respectively (Raynalet al., 2000). Matrix chunks were first finely powdered bycrushing them between two glass slides. Second, small ma-trix grains (<100 lm) were selected under a binocularmicroscope among chondrule fragments (only in CM chon-drites) and sulfate or carbonate grains (in CI and CM chon-drites). In a first serie of measurements, the matrix grainswere crushed between two sapphire windows. The windowhaving the largest amount of sample was kept for analysis.Because of the strong mid-IR absorption of sapphire, theseinfrared measurements covered only the 4000–1500 cm�1

range. In order to gain insights into the Si–O stretch bandat �1000 cm�1, some matrix grains were prepared onKBr windows with the protocol described by Raynalet al. (2000). Matrix grains were first pressed between twoglass slides, and transferred by pressing a KBr window ontoone of them. The sapphire or KBr windows were finally in-serted into an environmental cell, designed and built atLPG. This cell enables temperatures up to 300 �C to bereached under primary or secondary dynamic vacuum(from 10�4 down to 10�7 mbar). Optical access is permittedfrom both sides of the cell through ZnS or KBr windowsthus enabling measurements in transmission. Samples wereprogressively heated up to 300 �C with typical steps at 100,150, 200 �C, and spectra were recorded at each temperaturestep. Below, we qualify of “high-temperature” the spectraacquired at a temperature of 300 �C.

Synthetic Li–Mg–serpentine (chrysotile) was obtainedfollowing the method described in Wunder et al. (2009). Ini-tial solid starting-materials, c-Al2O3, Li2O, synthetic bru-cite, (Mg(OH)2), or synthetic talc, (Mg3Si4O10(OH)2) inthe presence of excess water were mixed in the stoichiome-tric composition of [(Mg44Li2Al2)Si32]O80(OH)64. Synthesiswas performed at 0.4 GPa, 400 �C for 31 days usingstandard cold-seal hydrothermal vessels.

IR spectroscopy of CI and CM chondrites matrices 4883

2.3. Spectra processing

The transmission spectra were converted into absor-bance by applying A = �log(T/T0), where T0 is the trans-mittance without sample and T the transmittance with thesample. In order to compare transmission spectra of sam-ples of different thicknesses, the spectra were normalized.In the case of measurements focusing on the 3 lm band,the maximum absorbance of this broad band was set to1. Prior to normalization, a spline baseline was calculatedand subtracted with control points at 1900, 2000, 2050,2100, 3750, 3800 and 3900 cm�1. The band full-width athalf maximum (FWHM) and barycenter were then calcu-lated. It should be emphasized that scattering effects couldnot be fully removed, leading to variable uncertainties inthe FWHM and barycenter spectral parameters. In the caseof acquisition with a KBr window, the spectra are normal-

Fig. 1. (A) Transmission spectra (in absorbance) of Orgueil at differentemperature steps. At ambient condition, spectra are dominated by the2300–2400 cm�1 spectral feature on all spectra is due to gaseous CO2. (Btemperatures. Spectra are shifted vertically for clarity.

ized to the absorbance of the Si–O stretching band around1000 cm�1.

3. RESULTS

3.1. Spectral evolution during dehydration in the 3 lm region

Typical IR spectra of the CI chondrite Orgueil and theCM Boriskino measured in the environmental cell areshown in Figs. 1 and 2. In the case of Orgueil, at room pres-sure and temperature, a broad feature is observed rangingfrom 3650 to 3000 cm�1 together with a sharp peak at3700 cm�1. This spectrum is consistent with publishedtransmission measurements (Zaikowski, 1979; Osawaet al., 2001, 2005). Upon heating to 300 �C, the broad fea-ture centered around 3350 cm�1 progressively disappears,until only a sharp asymmetric peak dominates the spec-

t temperatures. Spectra were successively measured at increasingsignature of molecular water (adsorbed and interlayer water). The) Spectral differences between the measurements of (A) at different

Fig. 2. (A) Transmission spectra (in absorbance) of Boriskino at different temperatures. Spectra were successively measured at increasingtemperature steps. At ambient condition, spectra are dominated by the signature of molecular water (adsorbed and interlayer water). (B)Spectral differences between the measurements of (A) at different temperatures. Spectra are shifted vertically for clarity.

4884 P. Beck et al. / Geochimica et Cosmochimica Acta 74 (2010) 4881–4892

trum. This remaining sharp feature at 3670 cm�1 is due tothe stretching mode of OH groups (Ryskin, 1977), whichprobes the local cationic environment of the hydroxylgroups in the octahedral layer of the phyllosilicate, withpossibly a contribution from tetrahedral cations. In the caseof Boroskino (CM), the room temperature spectrum dis-plays a broad asymmetric feature with a maxima between3350 and 3450 cm�1. Upon heating this feature is progres-sively reduced toward a broad asymmetric feature centeredaround 3650 cm�1.

The high-T (300 �C) spectrum reported here likelyprobes only the hydroxyl groups, as the water bendingmode m2 peaking at �1650 cm�1 disappears between 100and 300 �C (Fig. 3). Therefore, at room conditions, thebroad feature centered at 3350 cm�1 is due to the presenceof terrestrial molecular water. Hydroxyl groups are knownto remain stable up to at least 500 �C in the case of phyllo-silicates with a tetrahedral–octahedral (T:O) structure layer

(Beaufort et al., 1998) or tetrahedral–octahedral–tetrahe-dral (T:O:T) (Drits et al., 1998; Kloprogge et al., 2000),as are phyllosilicates in CI and CM chondrites.

In order to characterize the spectral signature of waterthat is expelled upon heating, we subtracted spectra mea-sured at different temperatures. We simply subtracted theabsorbance value measured at a high temperature by thevalue obtained at a lower temperature. This subtraction re-veals that a significant amount of water is released between25 and 100 �C in the case of Orgueil (CI, Fig. 1B) andBoriskino (CM, Fig. 2B). The spectral signature of waterexpelled between 25 and 100 �C and that of the water ex-pelled between 150 and 300 �C are significantly differentas is clearly seen for Boriskino (Fig. 2B). In the first case,the band appears symmetric and its maximum peaks at3350 cm�1 while in the second it is asymmetric and peaksaround 3520 cm�1. The low temperature water fraction(�100 �C) is most likely adsorbed water which is known

Fig. 3. Transmission spectra of Orgueil obtained during a dehy-dration experiment, showing both the 3 and 6 lm features. Thedisappearance of the 6-lm feature upon heating is indicative ofH2O loss. Residual peaks at 300 �C above 6 lm are likely due to thepresence of organics.

Fig. 4. High-T (300 �C), low pressure spectra of the CM and CIchondrites. Spectra are shifted vertically for clarity. There is a clearevolution from a Murchison-type to an Orgueil-type spectrum, thatappear to be related to the classification of Rubin et al. (2007). Thevertical black line corresponds to the absorption maximum of theOrgueil spectra. In the case of multiple measurements of a givensample, the different spectra are presented.

Fig. 5. IR spectra of various serpentine minerals (Salisbury et al.,1991) and Cold Bokkeveld (this study). Spectra are shiftedvertically for clarity. The vertical dashed lines correspond to theabsorption maximum of antigorite and chrysotile.

IR spectroscopy of CI and CM chondrites matrices 4885

to desorb at moderate temperature (<100 �C), while thehigh-temperature fraction is almost certainly interlayerwater within phyllosilicates.

3.2. High-T spectra of CI and CM chondrites in the 3 lm

region

The infrared spectra of CM and Orgueil (CI) chondritesmeasured at 300 �C and under primary vacuum are pre-sented in Fig. 4. Among CM chondrites, two end-memberspectral signatures were identified with type samples Mur-chison and Cold Bokkeveld.

The spectrum of Murchison shows a broad asymmetricfeature with a maximum around 3580 cm�1, while the spec-trum of Cold Bokkeveld reveals a sharp asymmetric featurewith a well-defined peak at 3675 cm�1 (Fig. 4). Among CMchondrites, there appears to be a continuous evolution be-tween those two end-members. But some variability withina sample is also encountered (Fig. 4). The spectra of CIchondrites look similar to some of the CM chondrites, inparticular Cold Bokkeveld and Nogoya. Although theroom temperature spectra might have a contribution fromferrihydrite, it should be negligible at high temperaturedue to decomposition of this metastable phase (Michelet al., 2007). The 3-lm band is thus likely to be controlledby phyllosilicates only. The strong absorption maximum at3675 cm�1 observed in Cold Bokkeveld like meteorites sug-gests the presence of Mg-rich serpentines that display sharpOH stretching absorptions in the near-IR spectral region(Ryskin, 1977; Salisbury et al., 1991). Indeed, the strongestabsorption of terrestrial lizardite, chrysotile and antigoriteare, respectively, found at 3686, 3691 and 3674 cm�1 (Salis-bury et al., 1991; Post and Borer, 2000). The absorptionmaximum for the Cold Bokkeveld-type spectra observedat 3675 cm�1 most closely corresponds to antigorite(Fig. 5). However, the 10-lm region reveals that the phyl-losilicate phases encountered in the CM and CI chondrites

are spectroscopically distinct from commonly encounteredterrestrial serpentines (see Section 3.3). Previous TEMand XRD studies show that there is no single poly-type

4886 P. Beck et al. / Geochimica et Cosmochimica Acta 74 (2010) 4881–4892

of serpentine in these samples, hence, we are likely measur-ing contributions from multiple or highly disordered ser-pentine phases.

The spectra presented in Fig. 4 have been ordered basedon naked-eye observation that reveals a continuous evolu-tion. Spectra obtained on different matrix chunks for a gi-ven sample show that some within sample variation isobserved as reported in Fig. 4 (in the case of Orgueil,Boriskino and Nogoya). However, the spectra show thatthis within sample variation is never as high as the inter-sample variations for the CI and CM we have studied sofar. Although we have extracted matrix chunks and thusthe petrological context is lost, we propose that this intra-sample variability is due to an heterogeneity in the extentof aqueous alteration, that lead to a spatial heterogeneityin the alteration products at the millimeter scale.

3.3. Si–O stretching vibration modes in the 10 lm region

The 10 lm region of infrared spectra is usually domi-nated by the Si–O stretching band, strongly controlled bythe crystallographic structure. Terrestrial serpentines typi-cally display multiple sharp diagnostic peaks used to distin-guish them. The spectra obtained on the CM chondritesMurchison, Cold Bokkeveld and Murray and the 3 CIchondrites are reported in Fig. 6, along with the spectraof several terrestrial phyllosilicates including Mg–serpen-tines, saponite and cronstedtite. The spectra of the 3 CIchondrites are very similar, with a single unstructuredbroad feature centered around 1020 cm�1. The apparent

Fig. 6. IR spectra of various serpentine minerals (dotted curves)(Salisbury et al., 1991), Cold Bokkeveld, Murchison, and MurrayCM and the three CI (Alais, Ivuna, Orgueil) in the 10 lm region.Spectra are shifted vertically for clarity. The vertical dashed linecorresponds to the absorption maximum of saponite.

secondary peak around 1100 cm�1 in the spectra of Alais(CI) and Orgueil (CI) can be attributed to the presence oftiny sulfate grains (Salisbury et al., 1991) not separated dur-ing sample preparation. The spectra of CM chondrites alsodisplay an unstructured band around 1000 cm�1 very simi-lar to that observed in CI spectra. This lack of spectralstructure in CM chondrites surprisingly does not supportthe presence of well-known serpentine phases identified bytransmission electron microscopic studies (e.g., Laurettaet al., 2000). We also compared spectra measured in ofCM and CI chondrites with spectra of synthetic Mg–ser-pentine (chrysotile) obtained following the method de-scribed in Wunder et al. (2009). The 10 lm region of thesynthetic serpentine still display three well-defined peaksand does not resembles CM or CI spectra (Fig. 6). It shouldbe noted that any linear mixture of the spectra of terrestrialphyllosilicates displayed in Fig. 6 fails to reproduce the CMand CI chondrite spectra in the 10 lm region. The phyllos-ilicate phases in CMs, and likely in CIs, appear spectro-scopically distinct from the terrestrial serpentines andsaponites when considering the 10 lm band as previouslynoted by Bunch and Chang (1980).

4. DISCUSSION

4.1. Origin of the spectral diversity among CI and CM

chondrites

The mineralogy of the phyllosilicates phases in CM andCI chondrites has mostly been studied using transmissionelectron microscopy (TEM), X-ray diffraction and electronmicroprobe. In CMs, the dominant phyllosilicates appearto be a serpentine related phase with a �7 A spacing(e.g., Tomeoka et al., 1989), in which the Fe–Mg abun-dance is variable within both the octahedral and tetrahedralsites. Transmission electron microscopy has the advantageto identify the phyllosilicate phases in a petrographical con-text. But this local mineralogy may not be representative ofthe bulk sample mineralogy. Because the samples are usu-ally very fine-grained (<100 nm), electron diffraction pat-terns are difficult to obtain. The identifications ofphyllosilicates are thus mainly based on the basal spacingdistance measured from high resolution TEM images onwell-crystallized grains. In addition to serpentine minerals,using this technique, a menagerie of hydrated and hydrox-ylated minerals has been described including, chlorite,smectite and brucite. Although there might be continuityin their chemistry, two distinct populations of serpentinesare observed in CMs (Tomeoka et al., 1989). An iron-richphase referred to as cronstedtite, which generally displaysplaty “long” crystals (>100 nm), as opposed to Mg-rich ser-pentines which usually present small grain sizes (20 nm). Inthe case of CIs, the hydrous mineralogy is dominated by amixture of serpentine and saponite (montmorillonite groupsmectite), that is found in association with ferrihydrite, ametastable disordered oxyhydroxide (Drits et al., 1993).The 3-lm region in the spectra we obtained on CM chon-drites reveals the presence of two end-members, a Murchi-son-type and a Cold Bokkeveld-type, the latter being quitesimilar to Orgueil. The Cold Bokkeveld-type spectrum

IR spectroscopy of CI and CM chondrites matrices 4887

resembles measurements performed on terrestrial serpen-tines in the fact that it presents a sharp asymmetric peakaround 3670 cm�1 (Salisbury et al., 1991; Post and Borer,2000). In the other hand, the Murchison-type spectra donot correspond to available spectra of terrestrial serpentineminerals in the 3-lm region. These observations suggesteither that the phyllosilicate phases either are not serpentineminerals or they are, but, with a distinct crystal chemistry.There is an abundant literature in the case of clay minerals(Madejova et al., 1994; Bishop et al., 2002a, 2002b), but lit-tle is known on the effect of the OH bounding cation on theIR spectra of serpentine. Mg-rich serpentines typically dis-play a sharp peak at about 3700 cm�1, but mid-IR spectraof Fe-rich serpentine minerals are poorly documented. Asan end-member, we measured IR spectra of terrestrial cron-stedtite, under vacuum conditions and moderate tempera-ture (�200 �C). The spectra reveal that H2O-freecronstedtite has a broad absorption band centered around3200 cm�1, but does not resemble our results for Murchi-son. However, it does not look like the Murchison spectra,a meteorite which hydrous mineralogy is expected to bedominated by Fe-rich serpentine (Howard et al., 2009a).Therefore, what is usually called “cronstedtite” in the car-bonaceous chondrites literature is strictly speaking not acronstedtite mineral, and the term Fe-rich serpentineshould be preferred. The IMA defines cronstedtite as Fe2+

2Fe3+(SiFe3+)O5(OH)4 hence when first TEM studies iden-tified Fe–Mg serpentine the term cronstedtite should neverhave been used.

Further informations are obtained by analyzing the 10-lm region of the IR spectra. The spectra we measured inthe Si–O stretching region are remarkably similar amongthe CM and CI chondrites, and remarkably different fromterrestrial serpentines. In the case of CI chondrites, the10-lm region spectrum appears consistent with saponitelike phyllosilicates, but not with serpentines, although sap-onite and serpentine are known to be present in approxi-mately equal proportion in CI chondrites (e.g., Tomeokaand Buseck, 1988). The same conclusion is obtained forCMs, whose mineralogy is dominated by serpentines(Papike, 1997). Two explanations can be proposed to re-solve this discrepancy. A first possibility is that TEM stud-ies that are made at the nanometer scale might not berepresentative of the global phyllosilicate mineralogy. How-ever, bulk meteorite X-ray diffraction studies also suggestthat serpentine minerals dominate the hydrous mineralogyof CI and CM chondrites matrices (Bunch and Chang,1980; Howard et al., 2009a) as well as electron probemicro-analysis (Zolensky et al., 1993; Browning et al.,1996). The second possibility concerns crystallographicorder. The lack of long-range ordering might explain thedifference between the IR spectra of CM and CI matricesand the considered terrestrial serpentines. Disordered oramorphous silicates are known to exhibit an unstructuredband peaking at �1000 cm�1, while crystalline silicatesalways exhibit a band with substructures due to crystallinesolid-state effects (e.g., Davydov splitting, field correlationeffect, LO–TO splitting; Poulet and Mathieu, 1970). Thebreakdown of long-range crystallographic ordering leadsto the optical excitation of numerous vibration modes.

Finally, we interpret from the shape of the 10-lm band thatmost of phyllosilicates in CI and CM are highly disordered,whatever their chemical composition (i.e., their Fe/Mgratio). Therefore, terrestrial Mg-rich serpentines (chrysotile,antigorite, lizardite) or almost pure Fe-bearing cronstedtite,cannot be considered as relevant mineralogical analogs ofCI and CM phyllosilicates.

4.2. The alteration sequence and the alteration scenario

A number of studies have attempted to characterize theextent of aqueous alteration as were recently reviewed byHoward et al. (2009a). In a pioneering work McSween(1979) remarked that the matrix abundances of CM chon-drites show variations that are correlated with a chemicalevolution: the meteorites that contain the most matrix werefound to be enriched in Mg compared to matrix-poor CMchondrites. From this observation, McSween (1979) sug-gested that the evolution of Mg/Fe might correspond toan alteration sequence, with a progressive increase of theMg/Fe ratio of the phyllosilicate phases as the alterationadvances.

Recognizing this chemical evolution of the phyllosilicatephases, Browning et al. (1996) tentatively defined a miner-alogical alteration index (MAI) from defocused electronmicroprobe analysis of CM matrices. The MAI is an indica-tor of the amount of tetrahedral sites occupied by Fe3+

within matrix serpentines, a quantity that is expected to de-crease as alteration progresses. Calculations of MAI are notstraightforward, and require some assumptions.

Mechanisms and processes have been proposed to ex-plain the variability of Mg/Fe in phyllosilicates of CMchondrites. Tomeoka et al. (1989) studied by TEM the pet-rographical relations between the alteration products. Theydeveloped an alteration model in which the different min-eral precursors are successively altered, starting by thetransformation of Fe/Ni metal into tochilinites [Fe(OH)2

(Fe,Ni)S]. Fe-rich matrix olivine subsequently reacts withthe fluid to produce Fe-rich serpentines that are usually re-ferred to as cronstedtite. In a following step, chondrulealteration begins which leaches Mg and Si into the fluid thatreacts with tochilinite to produce Mg-rich serpentine. Astochilinite hydrolysis continues, more Fe-rich serpentinesare produced together with sulfides.

One of the recognized properties of CM alteration is thefact that it is an isochemical process. With the exception ofa few fluid mobile elements, major elements concentrationsshow very minor variations among the group (see Rubinet al. (2007) compilation). Iron is no exception, which sug-gests a low mobility of the fluid (Howard et al., 2009a).

Our results on CM chondrites indicate a continuousspectral evolution of the 3-lm feature between the end-members Murchison and Cold Bokkeveld types. This evo-lution appears in agreement with the progressive alterationscheme of Rubin et al. (2007) (Fig. 4). Because serpentineminerals were shown to become enriched in Mg as alter-ation progresses (Zolensky et al., 1993; Howard et al.,2009a,b) we interpret the spectroscopic evolution to be aneffect of the bonding cation on the hydroxyl stretchingmode. IR spectra on synthetic serpentine minerals with

4888 P. Beck et al. / Geochimica et Cosmochimica Acta 74 (2010) 4881–4892

various Fe/Mg ratios are required to strictly confirm thisinterpretation.

Many workers have attempted to define a set of criteriato qualify and/or quantify aqueous alteration (McSween,1979; Bunch and Chang, 1980; Zolensky et al., 1993;Browning et al., 1996; Rubin et al., 2007; Howard et al.,2009a). This can be quantified on the basis of total phyllos-ilicate abundance for recovered meteorites but obviouslynot for asteroids. Here we propose to define a spectroscopiccriterion to qualify the degree of aqueous alteration thatcan be applied both to meteoritic sample measurements,and to infrared observations of asteroids. We focus onthe 3-lm region that displays significant variations in ourmeasurements on meteorites and that can be measured onasteroids from reflectance spectroscopy with minor contri-bution from thermal emission. Asteroid reflectance spectraare obtained by measuring the intensity of the sunlight re-flected by an object using Earth-based telescope or onboardIR spectrometer. These spectra typically involve multiplescattering within surface grains and cannot be related in adirect way to laboratory measurements of absorption spec-tra. However, in the case of absorption bands of moderateintensity (which is the case of the 3-lm band in phyllosili-cates) and multiple scattering, the so-called diffuse reflec-tance spectra is mostly controlled by the imaginary indexk, which is directly related to the absorption coefficient bya = 4pk/k (where a is the absorption coefficient, k the imag-inary index and k the wavelength). Therefore, band shape(including band maximum position) is expected to havesome first order resemblance when comparing a laboratoryabsorption spectra to a diffuse reflectance spectra. Undersome hypothesis, the conversion might even be performed(Hapke, 1993). The band shape can be described by threeparameters: the 3-lm band barycenter, the band maximumposition, and its full-width at half maximum (FWHM). Inorder to calculate these parameters, a spline baseline correc-tion (see Section 2.3) is performed on each spectrum beforenormalization. The barycenter is defined between 3400 and3700 cm�1 as::

B37003400 ¼

Pm¼3700m¼3400ðAðmÞ � CðmÞÞm

Pm¼3700m¼3400ðAðmÞ � CðmÞÞ

where A(m) is the absorbance at wavenumber m and C(m) isthe continuum intensity at the same wavenumber. Thisband parameter has been previously used to discriminatebetween different types of hydration on the Martian surface(Pommerol, 2009).

Fig. 7 presents the relationship between our three spec-troscopic criteria, and the total phyllosilicate content mea-sured by X-ray diffraction (Howard et al., 2009a). Theamount of phyllosilicates can be seen as a first order indica-tor of the extent of aqueous alteration, and has been shownto correlate with Rubin et al. (2007) classification (Howardet al., 2009b). Our three criteria appear to be related tosome degree with the extent of aqueous alteration(Fig. 7). The band maximum position and barycenter in-crease as alteration progresses, while the 3-lm band nar-rows and its FWHM becomes smaller. Although thecorrelations between the spectral parameters and the

amount of phyllosilicates are clear, some variability in thethree criteria is present within a given sample. It might bedue to sample heterogeneity, and possible optical artifactsdue to light scattering (Raynal et al., 2000).

In order to visualize the full dataset in a 2D space, theband barycenter is plotted against the FWHM in Fig. 8.This figure reveals a common evolution of these twocriteria: the FWHM decreases as the barycenter shifts tohigher frequencies. The least altered sample according tothe Rubin et al. (2007) or Howard et al. (2009a) classifica-tions are on the lower-right end of the diagram (QUE97990), while the heavily altered sample Nogoya is locatedon the upper left side. According to this diagram, the CMchondrites aqueous alteration sequence is (for increasingalteration): Cold Bokkeveld (2.2), Niger, Nogoya (2.2),Boriskino, Nawapali, Murray (2.4/2.5), Murchison (2.5),QUE 97990 (2.6) and QUE 99355 (2.3). With the excep-tion of QUE 99355, there is a general agreement betweenour relative alteration degree and the classification ofRubin et al. (2007). In our classification of Fig. 8, QUE99355 appears to be less altered than expected from Rubinet al. (2007) classification. However, naked-eye spectralcomparison in Fig. 4 appears to show that QUE 99355is intermediate between Nogoya and Murray, and thenin agreement with Rubin et al. (2007). A possible explana-tion of the discrepancy when using our two spectral crite-rions is that they might be sensitive to small interferenceartifact presents in the IR spectra of QUE 99355(Fig. 4) that are due to multiple light reflection on betweenthe samples flat surfaces.

The diagram presented in Fig. 8 might still be quite help-ful in trying to classify the extent of aqueous alterationfrom IR spectroscopy. Note that the technique is easy toimplement and measurements are very fast and relativelycheap.

4.3. The CI–CM connection

Based on IR spectroscopy, two important observationscan be made: the measurements show diversity within agiven sample and there appears to be a continuous evolu-tion from the heavily altered CI Orgueil toward the moreweakly altered CM chondrites (Murchison-like) (Fig. 4).The intra-sample heterogeneity was observed in most sam-ples for which measurements were duplicated, including theCI Orgueil and the CM Cold Bokkeveld. This heterogeneitymight be explained by the mm-size chemical heterogeneitiesprior to alteration, or by a spatial heterogeneity of alter-ation for a given sample, but also by a combination ofthe two. Alteration has been observed to be heterogeneousin CMs, and the presence of alteration veins has been usedto argue for a parent body process. In the case of ColdBokkeveld, the rock was described as a breccia thatcontains clasts of various alteration degrees (Metzleret al., 1992). Our measurements were performed on randommatrix chunks, and a similar study in a petrological context(i.e., in situ) might help in understanding the heterogeneityof the hydrous mineralogy. It is interesting to note that byXRD Cold Bokkeveld is the only CM yet analyzed to showsignificant inter sample heterogeneity.

Fig. 7. 3-lm Band maximum, band barycenter and FWHM (full-width half maximum) as a function of the total phyllosilicates content asmeasured by X-ray (Howard et al., 2009a,b). The black curves are quides for the eye.

IR spectroscopy of CI and CM chondrites matrices 4889

The continuous evolution between CM and CIpreviously raised the question of the genetic link betweenthe two chondrite groups. From mass balance calculation

Zanda et al. (2006, 2009) suggested that the oxygen isotopesystematic and volatile content variability can be explainedby a mixture between three components and they suggest a

Fig. 8. 3-lm Band barycenter as a function of its FWHM for the different CM and CI chondrites samples. The black curve is a second orderpolynomial regression to the dataset.

Fig. 9. IR transmission spectra (in absorbance) of Murchison andCold Bokkeveld (CB) at ambient and dry (T = 300 �C, primaryvacuum) conditions. The reflectance spectrum of Pallas (Joneset al., 1990) is also presented for comparison.

4890 P. Beck et al. / Geochimica et Cosmochimica Acta 74 (2010) 4881–4892

chemical continuity among carbonaceous chondrites. Onthe contrary, H isotopic compositions of CM and CI chon-drites matrix measured by Eiler and Kitchen (2004), tend toindicate a different protolith or fluid composition for CMand CI meteorites. Although the fluids might have been dif-ferent in term of hydrogen isotopes, the alteration productsappear to be quite similar between highly altered CM andCI which means that the mineralogical processes were likelysimilar.

4.4. The CM chondrite/asteroid connection

C-type asteroids are the most abundant asteroids in themain belt and display spectral features that suggest a rela-tionship with carbonaceous chondrites (Jones et al.,1990). In particular, C-type asteroids have a low albedoand some of them exhibit an IR absorption around 3-lm(Jones et al., 1990; Lebofsky et al., 1990). These asteroidshave been tentatively connected to specific meteorite groupsby comparing observed solar reflectance spectra to labora-tory measurements.

The spectra obtained at low pressure and moderate tem-perature in the present work reveal that terrestrial adsorbedwater significantly perturbs the IR signature at 3 lm. In thecase of Murchison (CM), the spectrum at ambient condi-tions is dominated by a broad peak centered at�3300 cm�1 (2.95 lm), while the spectra recorded in “dry”

atmospheric conditions show an asymmetric peak centeredat �3670 cm�1 (2.72 lm) with a broad long wavelengthwing (Fig. 9). It is extremely likely that the signature of ad-sorbed water is strongly present in all laboratory reflectancespectra measured at ambient conditions (Pommerol et al.,2009) and thus most previously published measurementsof terrestrial minerals and chondrite infrared spectra arelikely biased for asteroid comparison. At the surface ofasteroids, only minute amounts of adsorbed water can beexpected. In order to obtain valid analogs of infrared reflec-

tance spectra for altered parent bodies (e.g., CM, CI), mete-orite reflectance spectra acquired in dry vacuum conditionsare required.

Pallas is a large low albedo asteroid on which a 3-lm fea-ture was observed (Jones et al., 1990). We will not discusshere the nature of Pallas surface material, but rather useits reflectance spectrum as an illustration of the 3-lm band

IR spectroscopy of CI and CM chondrites matrices 4891

problem. Fig. 9 presents the spectrum of Pallas (2) togetherwith our measurements of Murchison and Cold Bokkeveldboth in ambient and dry conditions. The comparison be-tween the ambient meteorite spectra and the Pallas observa-tion displays a clear incompatibility in band position andshape, despite the limited spectral coverage of the asteroidspectrum. However, when adsorbed water molecules are re-moved, the absorption maximum is shifted toward lowerwavelengths and the two CMs spectra appear to be possiblecandidates for interpreting Pallas surface material reflec-tance spectra. However, asteroids observations in the 2.6–2.8 lm range are requisite for a detailed comparison of bandposition and shape with chondrite spectra.

5. CONCLUSION

We have measured the near-infrared spectra of CM andCI carbonaceous chondrites, under ambient and dry condi-tions. Adsorbed water strongly dominates the 3-lm regionin the spectra measured under ambient conditions. Whenadsorbed water is removed by moderate heating under vac-uum, the OH absorption band is clearly observed and weshow that it displays a continuous spectral evolution be-tween samples. We qualify this 3-lm band spectral evolu-tion by mean of spectroscopic criteria (band barycenter,position of the maximum and FWHM), which appear tobe correlated to the extent of aqueous alteration as definedby Rubin et al. (2007) and Howard et al. (2009a) from inde-pendent methods. We propose that our spectroscopic crite-ria reflect the evolution of the chemistry of thephyllosilicates phase from magnesian to ferrous as alter-ation processes. In the future, synthetic serpentines withchemistry analogous to that of chondrites are needed tocharacterize the effects of the tetrahedral and octahedral cat-ions on the infrared spectra, and possibly to use the near-IRspectra of chondrites to quantify their chemical changes in-duced by alteration.

Previously published infrared spectra of carbonaceouschondrites were measured under ambient conditions. Simi-larly to our transmission spectra under similar conditions,those reflectance measurements are very likely perturbedby adsorbed water, especially in the 3 lm spectral range.Therefore reflectance spectra of carbonaceous chondritesin “dry conditions” are strongly needed in order to have va-lid analog measurements for direct comparison with pub-lished asteroid reflectance spectra and future spectralobservations of 21 Lutetia by the VIRTIS instrumentaboard the Rosetta spacecraft.

ACKNOWLEDGMENTS

This study was funded by the French Centre National d’EtudesSpatiales (CNES), and the pole TUNES from University JosephFourier – Grenoble I (France). JSC – NASA (Houston, USA)and Museum National d’Histoire Naturelle (Paris, France) arewarmly acknowledged for having provided valuable chondrite sam-ples. F.-R. Orthous Daunay was supported by a Ph.D. grant fromMinistere de l’Enseignement Superieur et de la Recherche. K.T.H.was supported by S.T.F.C. through the UK Cosmochemistry Anal-ysis Network (U.K., CAN). G.K. Benedix is acknowledged forthoughtful comments on an early version of the manuscript.

Insightful comments from two reviewers were appreciated as wellas the editorial handling by Alexander N.Krot.

REFERENCES

Beaufort D., Cassagnabere A., Petit S., Lanson B., Berger G.,Lacharpagne J. C. and Johansen H. (1998) Kaolinite-to-dickitereaction in sandstone reservoirs. Clay Miner. 33, 297–316.

Bishop J., Madejova J., Komadel P. and Froschl H. (2002a)The influence of structural Fe, Al and Mg on the infrared OHbands in spectra of dioctahedral smectites. Clay Miner. 37,

607–616.

Bishop J., Murad E. and Dyar M. D. (2002b) The influence ofoctahedral and tetrahedral cation substitution on the structureof smectites and serpentines as observed through infraredspectroscopy. Clay Miner. 37, 617–628.

Brearley A. J. (2006) The action of water. In Meteorites and the

Early Solar System II (eds. R. P. Bonzel, D. S. Lauretta and H.Y. McSween Jr.). The University of Arizona Science Series,

Tuczon.

Browning L. B., McSween H. Y. and Zolensky M. E. (1996)Correlated alteration effects in CM carbonaceous chondrites.Geochim. Cosmochim. Acta 60, 2621–2633.

Bunch T. E. and Chang S. (1980) Carbonaceous chondrites. II:Carbonaceous chondrite phyllosilicates and light element geo-chemistry as indicators of parent body processes and surfaceconditions. Geochim. Cosmochim. Acta 44, 1543–1577.

Burgess R., Wright I. P. and Pillinger C. T. (1991) Determinationof sulfur-bearing components in C1 and C2 carbonaceouschondrites by stepped combustion. Meteoritics 26, 55–64.

Drits V. A., Lindgreen H., Salyn A. L., Ylagan R. and McCarty D.K. (1998) Semiquantitative determination of trans-vacant andcis-vacant 2: 1 layers in illites and illite-smectites by thermalanalysis and X-ray diffraction. Am. Miner. 83, 1188–1198.

Drits V. A., Sakharov B. A., Salyn A. L. and Manceau A. (1993)Structural model for ferrihydrite. Clay Miner. 28, 185–207.

Eiler J. M. and Kitchen N. (2004) Hydrogen isotope evidence forthe origin and evolution of the carbonaceous chondrites.Geochim. Cosmochim. Acta 68, 1395–1411.

Farmer V. C. (1974) The Infrared Spectra of Minerals. London, p.539.

Hapke B. (1993) Theory of Reflectance and Emittance Spectroscopy.Cambridge University Press, Cambridge.

Hiroi T., Zolensky M. E., Pieters C. M. and Lipschutz M. E. (1996)Thermal metamorphism of the C, G, B, and F asteroids seenfrom the 0.7 micron, 3 micron and UV absorption strengths incomparison with carbonaceous chondrites. Meteorit. Planet.

Sci. 31, 321–327.

Hiroi T., Zolensky M. E. and Pieters C. M. (2001) The Tagish lakemeteorite: a possible sample from a D-type asteroid. Science

293, 2234–2236.

Howard K. T., Benedix G. K., Bland P. A. and Cressey G. (2009a)Modal mineralogy of CM2 chondrites by X-ray diffraction(PSD-XRD). Part 1: Total phyllosilicate abundance and thedegree of aqueous alteration. Geochim. Cosmochim. Acta 73,

4576–4589.

Howard K. T., Benedix G. K., Bland P. A. and Cressey G. (2009b)PSD-XRD modal mineralogy of CM chondrites spanning thecomplete petrographic range. In: 72nd Annual Meeting of the

Meteoritical-Society, Nancy, France.Jones T. D., Lebofsky L. A., Lewis J. S. and Marley M. S. (1990)

The composition and origin of the C, P, and D asteroids-wateras a tracer of thermal evolution in the outer belt. Icarus 88, 172–

192.

Jouglet D., Poulet F., Milliken R. E., Mustard J. F., Bibring J. P.,Langevin Y., Gondet B. and Gomez C. (2007) Hydration state

4892 P. Beck et al. / Geochimica et Cosmochimica Acta 74 (2010) 4881–4892

of the Martian surface as seen by Mars Express OMEGA: 1.Analysis of the 3 mu m hydration feature. J. Geophys. Res.

(Planets) 112, E08S06, doi:10.1029/2006JE002846.Kloprogge J. T., Frost R. L. and Hickey L. (2000) Infrared

emission spectroscopic study of the dehydroxylation of somehectorites. Thermochim. Acta 345, 145–156.

Lauretta D. S., Hua X. and Buseck P. R. (2000) Mineralogy of fine-grained rims in the ALH 81002 CM chondrite. Geochim.

Cosmochim. Acta 64, 3263–3273.

Lebofsky L. A., Jones T. D., Owensby P. D., Feierberg M. A. andConsolmagno G. J. (1990) The nature of low-albedo asteroidsfrom 3-lm multi-color photometry. Icarus 83, 16–26.

Madejova J., Komadel P. and Cicel B. (1994) Infrared study ofoctaedral site populations in smectites. Clay Miner. 29, 319–

326.

McSween H. Y. (1979) Alteration in CM carbonaceous chondritesinferred from modal and chemical variations in matrix.Geochim. Cosmochim. Acta 43, 1761–1770.

Metzler K., Bischoff A. and Stoeffler D. (1992) Accretionary dustmantles in CM chondrites – evidence for solar nebula processes.Geochim. Cosmochim. Acta 56, 2873–2897.

Michel F. M., Ehm L., Liu G., Han W. Q., Antao S. M., Chupas P.J., Lee P. L., Knorr K., Eulert H., Kim J., Grey C. P., CelestianA. J., Gillow J., Schoonen M. A. A., Strongin D. R. and PariseJ. B. (2007) Similarities in 2-and 6-line ferrihydrite based onpair distribution function analysis of X-ray total scattering.Chem. Mater. 19, 1489–1496.

Milliken R. E. and Rivkin A. S. (2009) Brucite and carbonateassemblages from altered olivine-rich materials on Ceres. Nat.

Geosci. 2, 258–261.

Miyamoto M. and Zolensky M. E. (1994) Infrared diffusereflectance spectra of carbonaceous chondrites: amount ofhydrous minerals. Meteoritics 29, 849–853.

Osawa T., Kagi H. and Nagao K. (2001) Mid-infrared transmissionspectra of individual Antartic micrometeorites and carbona-ceous chondrites. Antarct. Meteorit. Res. 14, 71–88.

Osawa T., Kagi H., Nakamura T. and Noguchi T. (2005) Infraredspectroscopic taxonomy for carbonaceous chondrites fromspeciation of hydrous components. Meteorit. Planet. Sci. 40,

71–86.

Papike J. J. (1997) Planetary Materials. Mineralogical Society ofAmerica, Washington.

Pommerol A., Schmitt B., Beck P. and Brissaud O. (2009) Watersorption on Martian regolith analogs: thermodynamics andnear-infrared reflectance spectroscopy. Icarus 204, 114–136.

Pommerol A. (2009) Hydratation minerale a la surface de Mars:

caracterisation spectroscopique, thermodynamique et application

a la teledetection. Ph.D. thesis.

Post J. L. and Borer L. (2000) High-resolution infrared spectra,physical properties, and micromorphology of serpentines. Appl.

Clay Sci. 16, 73–85.

Poulet H. and Mathieu J. P. (1970) Spectres de vibration et symetrie

des cristaux. Gordon & Breach, New York.Raynal P. I., Quirico E., Borg J., Deboffle D., Dumas P.,

d’Hendecourt L., Bibring J. P. and Langevin Y. (2000)Synchrotron infrared microscopy of micron-sized extraterres-trial grains. In: International Conference on Mars Exploration

Program and Sample Return Missions. Pergamon, ElsevierScience Ltd., Paris, France.

Rivkin A. S., Howell E. S., Lebofsky L. A., Clark B. E. and BrittD. T. (2000) The nature of M-class asteroids from 3-mu mobservations. Icarus 145, 351–368.

Rubin A. E., Trigo-Rodrıguez J. M., Huber H. and Wasson J. T.(2007) Progressive aqueous alteration of CM carbonaceouschondrites. Geochim. Cosmochim. Acta 71, 2361–2382.

Ryskin A. S. (1977) The vibrations of protons in minerals:hydroxyl, water and ammonium. In: Infrared Spectra of

Minerals. Mineralogical Society of Great Britain & Ireland.Salisbury J. W., Walter L. S., Vergo N. and D’Aria D. (1991)

Infrared (2.1–25 lm) Spectra of Minerals. Johns HopkinsUniversity Press, Baltimore.

Tomeoka K. and Buseck P. R. (1988) Matrix mineralogu of theOrgueil CI carbonaceous chondrite. Geochim. Cosmochim. Acta

52, 1627–1640.

Tomeoka K., McSween, Jr., H. Y. and Buseck P. R. (1989)Mineralogical alteration of CM carbonaceous chondrites: areview. Antarct. Meteorit. Res. 2, 221.

Wunder B., Deschamps F., Watenphul A., Guillot S., Meixner A.,Romer R. L. and Wirth R. (2009) The effect of chrysotilenanotubes on the serpentine–fluid Li isotopic fractionation.Contrib. Mineral. Petro. 159, 781–790.

Zaikowski A. (1979) Infrared-spectra of the Orgueil (C-1) chon-drite and serpentine minerals. Geochim. Cosmochim. Acta 43,

943–945.

Zanda B., Hewins R. H., Bourot-Denise M., Bland P. A. andAlbarede F. (2006) Formation of solar nebula reservoirs bymixing chondritic components. Earth Planet. Sci. Lett. 248,

650–660.

Zanda B., Le Guillou C. and Hewins R. H. (2009) The relationshipbetween chondrules and matrix in chondrites. In: 72nd Annual

Meeting of the Meteoritical-Society, Nancy, France.Zolensky M., Barrett R. and Browning L. (1993) Mineralogy and

composition of matrix and chondrule rims in carbonaceouschondrites. Geochim. Cosmochim. Acta 57, 3123–3148.

Associate editor: Alexander N. Krot