Earth and Planetary Science Letters, 112 (1992) 91-99
Elsevier Science Publishers B.V., Amsterdam [UCl
The volatile element enrichment o ILF chondritic interplanetary dust particles -
Elmar K. Jessberger
and Kurt Traxel
a Max-Planck-histitut für Kernphysik, P. O. Box 10 39 80, 6900 Heidelberg, Gerrnany
6 Physikalisches Institut der Universitär, 6900 Heidelberg, Germar:y
Reccived April 3, 1992, revision accepted May 27, 1992
Interplanetary dust particles (IDPs) collected in the stratosphere provide new information on the chemical and isotopic composition of material in the vicinity of the Earth and thus possibly of asteroids and comets. The available data on major, minor and trace element abundances, mainly obtained with p1xE and SYXFA, suggest the presence of at least two groups of chondritic IDPs which differ significantly in their volatile element content. The ubiquitous enrichment of volatile lithophile. siderophile, and chalcophile elements is most plausibly explained by atmospheric contamination and thus the question arises of the factual relationship of all properties of interplanetary dust particles to early solar system processes.
1. Introduction In addition to meteorites and lunar material. interplanetary dust particles (IDPs) collected in the stratosphere by high-flying aircraft  constitute another source of extraterrestrial matter accessible for study in the laboratory. A major subclass of these particles differs from all known meteorite types in that the members of this subclass are highly porous . Such fragile material would never have reached the Earth's surface intact if not as tiny (5-50 Am diameter) objects which remained relatively cool during aerobraking in the atmosphere  . Certainly, immediately after their discovery these objects spurred much interest and experimental inquisitiveness, not least because their major source might be comets, which otherwise may only be accessible-and regrettably in the distant future-by costly space missions. When meteorite research began to intensify in tt, 111V
t q S .1I1<1VJ
basis for classification ~-f
Correspondence to : E.K. Jessberger, Max-Planck-Institut für Kernphysik, P.O . Box 10 39 80, 6900 Heidelberg, Germany.
meteorites was their bulk chemical composition  . The subsequently developed schemes provided important clues to their genesis and ensuing history, such as the profound distinction between more primitive and more evolved meteorites . In adition, minor and trace element contents also provided useful parameters for distinguishing the various groups and classes from one another  . In this spirit, survey programmes to analyze the contents of as many major, minor and trace elements as possible in a large number of individual stratospheric particles were initiated by van de Stap , Sutton and Flynn [5-7] and by ourselves [8-16] . The aim of these studies is to characterize the IDPs and ascertain whether a bulk chemical classification scheme emerges . We wish to establish a link with known extraterrestrial matter and ultimately to learn about the (most probably many different) origins and histories or the material and what they can tell us about the early evolution of our solar system . The techniques applied are PIXE  and SYXFA , which are highly sensitive and also non-destructive (SYXFA), or only marginally destructive . This allows subsequent destructive analyses with other methods such as sills [19-221.
0012-821X/92/$05 .00 © 1992 - Elsevier Science Publishers B.V . All rights reserved
Available major, minor and trace element data for stratospheric dust particles determined by P1XE and SYXFA. Elements are grouped according to their geomochemical behaviour and are ordered with falling condensation temperature . The numbers given are atomic abundances normalized to Fe and must be multiplied by the number (10 -') given below the element row heading. The reason for the Fe normalization is a technical one, in that Si, the most favoured normalization element, is in many IDPs not, or only poorly, determined . The uncertainties in the abundances are estimated to be about 15% . Ita'ics designate upper limits. For comparison the bulk abundances in major chondrite classes are also given .
.094 2.66 8.0 .100 3 .45 8.5 .113 4.41 9.1
1.11 1 .41 1 .03 3.9 1.19 1.47 .74 2.4 1.23 131 .60 2.1
6.33 2.44 1.65
5.31 2.44 1.0
5.34 2.26 1 .0 5.04 2.33 1 .0
.60 .46 37
130 .155 .74 .80 .080 .38 .64 .042 .22
1.40 S6 .71 .27 .40
decreasing condensation temperature References : (a) this work and [9-13], : (d) .
(Heidelberg) and SYXFA (Hamburg): (b)
(Brookhaven) ; (c) SYXFA (Brookhaven)
VOLATILE ELEMENT ENRICHMENT OF CHONDRITIC INTERPLANETARY DUST PARTICLES
2. Classification of interplanetary dust particles The presently available data base on stratospheric dust particles is given in Table 1 . The atomic abundances are normalized to Fe because this element is present in all particles (except W7027B14, where only traces of Fe could be found) and because it is always measured regardless of the filter used in front of the X-ray detector. The first observation is that the same data set is not available for all particles (the blank entries in Table 1), a fact that prevents rigorous cluster analysis and thus limits the sharpness of interpretation . The particles are divided into three major categories: The upper box in Table 1 contains those particles which--to varying degrees and with exceptions-grossly resemble chondrites . The second box contains the data for two Fe-S-Ni-rich spheres, and the third box those of non-ebondritic
particles which are either of terrestrial or extraterrestrial origin . The fourth box gives the elemental abundances in major chondrite classes for comparison  . In the this contribution we will concc,-itrate on the chondritic particles in the upper box. Chondritic particles are termed chondritic mainly on the basis of Si/Fe, Cr/Fe, Mn/Fe, Ni/Fe and S/Fe (Table 1). They are, however, non-chondritic in terms of Ca /Fe (a factor of 2.5 lower than Cl chondrites) and Co/Fe (factor of two higher), and there is an overall enrichment by a factor > 2 of the volatile elements K, Cu, Ga, Ge, and Br. A depletion of Ca has also been found in other IDPs  and in Antarctic micrometeorites . This is not accompanied by a comparable depletion of other refractory elements such as Al or Ti (Table 1). Some Ca can be present in IPDs in the form of carbonate  which could be lost by heating during entry into
Elemental composition of Groups 1 and 11 and ungrouped (Group 111) IDPs normalized to Fe and C1 chondrites . Given are the geometric means. In parentheses are the ranges of 1Q uncertainty. The Cl abundances (Si = 10"), according to which the elements 10-4 atm are from 1261 are ordered, are from 1251 and the 5017c condensation temperatures (°C) at Element
Condensation C1 Abundance Temp .
Group I Group II Group III C1 and Fe Normalized Abundances
12 38 62 119 522 1260 2250 2400 3770 5240
690 918 684 825 1037 660 1351 1549 1000 863
34 a3, 50) 2.9 (2 .1, 3.9) 2.2 (1.6. 3.0) 2.3 (1.6, 3.4) 2.8 (1 .9, 4s) 1.4 (1 .1 .1 .8) 1 .9 (1 .2, 2.9) 1.5 (1.3,1.71 2.2 (2.o, 2.5) 3.6 cis . 4.6)
14 (z . q) 5.6 (3.7, 8.4) 3.8 (v, s.3) 2.3 (1.5.3-M 2 .5 (2 .1, 3.0) 4.0 0.1.3.2) 1.8 (1 .4, 23) 1.2 (1.o,1 .4) 1.3 (o.s, 2.1) 1.2 (o.9,1 .4)
Il (3,31F) 1 .0 (0.2.5 .7) 0.8 (01, 2.6) 0.4 1.2 (o.2 . 6.s) 1.0 (o.1, 7.7)
1.0 (0.5, 1-9) 1.1 (0.8,1-5)
1.0 (0.4, 2.3) 0.6 (o.4, 1.o) -
0.4 (o.3, o.6) 1.4 (0 .8, 23) 0.8 (0 .6,1 .1) M 1.0 1 .2 (os, 1.7)
0.6 (0.4,1 .0)
1190 1277 1354 1518 1650 648 1336 1311
1.1 (0.8,1 .6) 1.1 (0.9,1 .4)
9550 13500 49300 61100 84900 515000
Ge Cu 2n Co Ti K Ce NIn Cr Ni Ca Al Fe
1 .3 (l .o, 1.7)
1.3 (0.7.2 .3) M 1 .0 1.3 (11, 2.0)
0 .3 (.OS, 1.5) -
E.K. JESSBERGER ET At..
the Earth's atmosphere-and even more so by chemical reactions with, for example, H,SO4 there. On the other hand, hydrated phases in carbonaceous chondrites are also low in Ca and thus atmospheric Ca loss is not "mandatory" . The volatile element enrichment of IDPs will be discussed below. The chondritic particles in Table 1 are arranged in three groups . The five particles that occur in each of the two upper groups (Group I at the top and Group II underneath) have some similarities to each other, while the third group (Group III) simply contains thoie chondritic particles that do not fit into these two groups . Within a factor of two, both upper groups have nearly chondritic abundances of Al, Si, Cr, Mn, Ni and S (Table 2). Relative to bulk CI chondrites, they are depleted in Ca, somewhat enriched in K and about equally (x 2-2.5) enriched in Co, Cu, and Ge. However, they differ in their content of Cl, Ga, Se and Zn (Table 2). These differences provide the basis for the grouping. Figures 1 and 2 give an impression on the distinction between the groups . It is surprising that this first attempt to
10 K/Mn Fig . 1 . R-lationship of CI/N1n with K/Mn in stratospheric particles . Chondritic particles are as follows : triangles = Group I (average: shown by the black triangle), circles = Group 11 (average is shown by the filled-in circle) . Crosses are are non-chondritic particles . Mixing lines (dashed) are indicated with the CI /K ratio of the admixed material . Also given are the loci of bulk carbonaceous chondrites . The correlation coefficient of th ,-- straight line is (1 .98 . 0.1
Fig . 2 . Volatile siderophile and chalcophile element variations among stratospheric particles . The synu)ols are the same as in Fig . 1, except for the diamonds, which indicate non-grouped chondritic particles (average shown by the black diamond) . The correlation coefficients are given .
group IDPs using minor and trace element concentrations reveals differences primarily in some rare and volatile elements . "Therefore, in the following we will concentrate on these in some detail . Table 2 summarizes the average elemental abundances in the chondritic particles separately for Groups I and 11 and the ungrouped particles (Group 111) relative to Cl chondrites. The averages are the geometric means . The quoted uncertainties are lo, and are intended to signify the range of abundances . Because the spread among the ungrouped "Group III" IDPs is quite large, their average does not appear to be meaningful . Perhaps we should subclassify this group. However, this has not been done liere because such subclassification would lead to many very small classes. "The elements in Table 2 are ordered in increasing C1 chondritic abundance . Quite naturally the data then divide themselves into two blocks, i .e. below and aoove the heavy line. All
VOLATILE ELEMENT ENRICHMENT OF CHONDRITIC INTERPLANETARY DUST PARTICLES
elemental abundances below the line are indistinguishable from C1, with the exception of Ca, while all elements above the line are at least in Group I or in Group II, or in both, significantly enriched over C1 chondrites. The natural ordering by abundance of normal and enriched elements in Table 2 only to a marginal extent finds its counterpart in the sequence of condensation temperatures [261, a fact which suggests that the observed enhancements are only weakly dependent on the thermal behaviour of the elements. The same applies for the geomochemical properties, to which the enhancements appear not to be related-below and above the line in Table 2 are lithophile, siderophile and chalcophile elements . The association of C I abundance and enrichment of certain elements in chondritic IDPs may reflect the effects of contamination either in the atmosphere or during cleaning, handling and analyzing the particles. The ability to detect small deviations (by atom numbers) from some normal composition is small when that element is already very abundant in the original sample . 3. Volatile elements in interplanetary dust particles The enrichment of volatile elements in IDPs was already noted in the first report on a PIXE study of IDPs [81 and was subsequently corroborated by [4-161 . Van de Stap et al. [41 interpreted the enrichment to indicate that the source of these IDPs was enriched in volatiles after the more refractory elements were removed from the parent bodies of the chondritic meteorites; this would involve large-scale elemental fractionation of the early solar nebula in the region between 2 and 10 AU . With the exception of these IDPs, the material available for laboratory analysis which is richest in volatile elements is C1 chondrites. Cl chondrites constitute not only a certain. class of meteorites, but are distinguished from all others in that they contain the condensible ele ., .,.t15 tlll~. he pamn the solar 1.--111~:.11 JUl11W l. ;prmoorpolrmrtion wvm as u Photosphere [25,271 and are used as the "abundance standard" [251. Taking the solar photospheric abundances for the solar abundances and since the sun comprises 99.87% of the mass of the solar system, C1 chondrites represent the solar abun-
dance of the condensible elements. Anders and Grevesse [251 clearly demonstrate that the C1 chondrite abundances of the elements under discussion, Zn, Ga, Ge, Se and Br, which are not or only poorly analyzed in the solar photosphere, give a smooth function of the abundances of the odd-mass nuclides, supporting the contention that they indeed represent the solar system abundances. Minor deviations of Cl chondrites and solar photospheric abundances such as that of Mn (28 ± 12%) are irrelevant in the context of the accuracy of IDP data. Certainly, elemental fractionations occurred in the early solar nebula and especially later within planets and asteroids . But it appears to be asking a lot to claim from the data on < 10' g of rock that a large-scale fractionation had occurred which involved much more than 1024 g of material and by which a region of the solar nebula is enriched in certain elements by a factor of almost 2 orders of magnitude (Br). This region must be extended because the IDPs are derived from the many different sources supplying the zodiacal cloud with about 10 tons per second [28,291. Before such far-reaching conclusions are entertained, we should at least exclude the more earthly explanations. Among the lithophile elements, K and Cl show considerable variation from particle to particle (Table 1) and an overall enrichment (Table 2). Variations are the prerequisite in searching for correlations. Figure 1 demonstrates that the volatile elements K .and Cl, when normalized to another lithophile, but less volatile element Mn, do not vary randomly but arrange themselves in this double-logarithmic plot on a straight line: (ignoring those particles where for one or the other element only upper limits were obtained) . Also shown in this graph are the curved mixing lines which would result if some material (e.g. aerosol) with K/Cl ratios of 1, 5 or 10 were added to an arbitrary, but close to chondritic, starting composition . The best match to the data, however, remains the straight line. It is noteworthy that the non-chondritic particles W7027-B4 and -B9 also fall on the straight line. B4 is an elongate (9 x 10 Am) particle, whereas B9 is round (15 Am diameter) and, because it did not yield a clear EDX signal, B9 is possibly composed mainly of low-Z material [301; it does exhibit chondritic Ni/Fe, however (Table 1). The chemi-
EX . ]ESSBERGER ET AL .
Br/Fe x 10 4
Fig . 3 . Cu/Fe vs. Br/Fe in stratospheric particles. Symbols are the same as in Fig . 1 and 2 . The correlation coefficient excludes the highly deviant data .
cal composition (as far as it is known) of both particles is dominated by S (51-58 wt .%) and Ca (18-20 W,%) (Table 1) and thus both particles are possibly of terrestrial origin, or massively contaminated . The fact that their data fall on the same line as those from chondritic particles suggests that terrestrial and extraterrestrial particles have the same source of volatile elements . Most p'!.Ausibly. the volatiles were acquired during the residence time of the particles in the atmosphere . From the straight line in Fig. 1 we obtain the dependence CI/Cl u = (K/K )`.3. This simple relationship follows if it is assumed that the common denominator, here Mn and in Figs . 2 and 3 Fe, is not affected by atmospheric processes, which, however, is likely to be the case for the major elements Mn and Fe. The straight line suggests that not simple mixing, but a dynamic process was responsible for the enrichment of both K and Cl: the more efficiently Cl was taken u.,ErL , the .higher .~yv. . was NJ the Cl, //K ratin Figures 2 and 3 show variations in the siderophile and chalcophile elements which are enriched in chondritic 1DPs (Table 2). These scatter plots resemble that of the lithophile volatiles (Fig . 1). The data fields are, in general, far beyond the chondritic compositions . Indeed,
they suggest themselves as a natural-though fractionated--continuation of the trends displayed by the chondrite groups . And again the data are more (Fig . 2b and c) or less (Fig . 2a) aligned. Taking into the account the different data sources and the fact that the experiments were not specifically shaped for the measurement of volatiles, the alignments are convincing. Figure 3 demonstrates that Br is not a special case (perhaps only in terms of magnitude), which may be due to the fact that Br is also the rarest of the elements analyzed. Br enrichments have also been encountered in CM chondrites , but only by 40% and not relative to C1 chondrites but relative to elements of similar volatility. From the slopes in Figs. 2 and 3 we obtain the relationships of the "excess" volatiles as follows : Ga/Ga = (Se/Se,)' Zn/Zn,) = (Se/Se,)t .2' Ge/Ge,~ = (Se/Seta)", and Cu/Cu 11 = (Br/ Br(, )°Y (the subscript zero denotes the primary abundances). The exponents imply that Ga and Zn are similarly preferred over Se, Se is preferred over Ge, and Br over Cu. As with Cl and K, the addition of the volatiles most probably occurred in a dynarnic (i.e., chemical) process rather than by simple (mechanical) admixture. A possible source of the volatiles has been mentioned above-the terrestrial atmosphere . About 100 tons of extraterrestrial matter hit the atmosphere per day at speeds of greater than 11 .2 km/s. Most of this material > 50 Am is possibly heated above 500'C  and (partially) melts and thereby preferentially loses volatile elements, and certainly some other elements too. Molten Antarctic micrometeorites  and Deep Sea Spherules  are the end products of this process, although it should be noted that among Antarctic meteorites are many which do not show any influence of heat. The meteoritic infall generates the E-layer at a height of about 90 km, and the composition of this layer is highly variable, depending on the composition of the most recently combusted or molten meteoroid(s)  . The elements which are present at high (i .e., measurable) concentrations in the E-layer settle towards the Earth's surface, thereby being dihited with terrestrial atmospheric constituents . This makes them difficult to detect . Thus, to some yet unknown extent th .~ trace element composition of the atmosphere is governed by mete-
VOLATILE ELEMENT ENRICHMENT OF CHONDRITIC INTERPLANETARY DUST PARTICLES
oritic infall . The atmospheric trace element abundances certainly do not directly correspond to their meteoroidal progenitors because they are modified and fractionated (a) by the different temperature stabilities of the compounds in which they resided in meteoroids and (b) in the atmosphere by in-situ chemical reactions; different uptake efficiencies onto aerosols also plays a role in this modification and fractionation. For small interplanetary dust particles (< 50 Am) which remain relatively cool upon aerobreaking in the Earth's atmosphere , it takes typically months to years before they are removed from the stratosphere -or before they are eventually caught by the sampling device . During this time they are exposed to the atmosphere and can take up whatever is available there in suitable chemical form . The latter requirement probably gives rise to further elemental fractionations due to different trapping probabilities of elements and their compounds on different particle constituents of different grain sizes. Thus, the scenario envisaged is rather simple : Many larger micrometeorites (> 50 Am) are molten or partially molten during aerobraking in the Earth' atmosphere . The material lost is dispersed in the atmosphere and diluted with indigeneous material such as aerosols . Smaller mircrometeorites ( < 50 ,um) remain much cooler and during their residence time in the -tmosphere take up whatever there is in suitable form, which includes the losses from the larger micrometeorites. Thus, the size of the original extraterrestrial particles, their speeds, and their entry angles play decisive roles in material loss or gain during passage through and residence in the Earth's atmosphere . How do we test the model that atmospheric processes are indeed responsible for the enrichment of volatile elements of IDPs? Certainly, a plausible argument would be if one could demonstrate that the enrichments were surface correlated. The first study of this problem by Mackinnon and Mogk  using Auger spectrometry proved negative, which means that there was no detectable Br- or Zn-rich coating a few monolayers thick on their stratospheric particles. In a recent study, on the other hand, Rietmeijer  did detect a Br-bearing layer on an IDP. Another sign of surface correlation of a contaminant would
be a negative correlation of particle size and enrichment factors. With the present data set, however, such a correlation is not obvious when we inspect the published dimensions of the particles  and, for example, the Br overabundances (Table 1) . Rietmeijer , however, did indeed find a clear correlation between Br content and atmospheric residence times, corroborating our notion from 1991 that the Earth's atmosphere is probably one source of volatile elements in IDPs [161 . An inverse hint would be the preferential loss of volatile elements from the larger (> 50 Am) micrometeorites, which reach higher entry temperatures than their small (< 50 Am) couterparts . Flynn et al .  report that three 50-160 Am micrometeu :ites are depleted in Zn and Ge, and that two out of these three are normal in Cu and Ga ; the third is depleted in Cu and Ga as well as in Zn and Ge . In other words, two particles may have supplied the atmosphere with Zn and Ge, but not with Cu and Ga, while the third particle may have contributed to the atmosphere's inventory of all four elements . In addition, in the search of surface correlation Flynn et al . [7j found no variation in the Fe/Zn ratio across one of the micrometeorites devoid of Zn and Ge and with chondritic Cu and Ga . It is not vet clear if this absence is due to Zn loss from the volume of the particle during passage through the atmosphere or if it reflects a pre-atmospheric . intrinsic Zn deficiency . If, however, the loss is volume-c , . . lated, the volatile element gain of smaller Mrs may also be . Certainly the size and, most importantly, the porosity, mineralogy and chemistry, together with the entry speed and angle, influence the magnitude of the losses from an individual particle . The opposite situation may also prevail : these very same parameters and, in addition, the chemical form of the volatile elements in the atmosphere as well as the reactions between, for example, aerosols and incoming IDPs, probably also govern-to an unknown extent-the uptake of elements from the atmosphere . In this respect it is noteworthy that the two analyzed FSN _e,. 11daLr ,.1, `v~x . llu~ ee Alement ., .., .. . ., ... enrirhments_ possibly because of their compactness Fe-S-Ni spheres react differently to the atmospheric environment than porous chondritic matter . At present, no detailed physical or chemical model for the atmospheric contamination can be
E .K. JESSBERGER ET AL.
given because of the lack of relevant data. Simple surface contamination or mixing seems to be ruled out (cf. Fig . 1 and [6,35]) . One must consider that the particles spent an extended time in the chemically agressive aerosol environment of the atmosphere and the different constituents of the IDPs may have reacted differently to it. Alone from bulk chemical data-as reported here-as well as from highly sophisticated analytical TEM  with its high spatial resolution, but unsufficient sensitivity, the problem is difficult to solve. T. Stephan [pers. commun .], who employed TOFSIMS with high spatial resolution and sufficient sensitivity, reported a highly imhomogeneous distribution of excessive halogens in one IDP section. Studies of this type will ultimately reveal the origin and thus the significance of the volatile element enrichment of IDPs. Another argument pointing towards alteration of extraterrestrial material in the atmosphere is provided by the fact that Antarctic micrometeorites larger than 0-^t t 50 Am are commonly totally enveloped by faicrometre-thick rims [G. Kurat and M. Maurette, pers. commun .]. These rims probably consist of magnetite and they are unrelated to any chemical or mineralogical feat;:re of the micrometeorites themselves. Seeing as it appears that the Fe was diffusing into the material ^-Ijacent to the rims from the outside [G. Kurat, pers. commun .], these rims are probably also the result of-albeit unspecified-atmospheric processes . One could even speculate that the pyroxene whiskers found by Bradley et al.  are not Early Solar System condensates but instead condensates in the atmosphere at certain very restricted physicochemical conditions. Study of stratospheric particles may open up new opportunities for learning more about upper atmosphere chemistry and processes . Further, if dedicated experiments were to prove the atmospheric origin of at least part of the volatile enrichment of IDPs and of other suspected atmospheric features such as the rims around micrometeorites, the results of IDP research would have to be interpreted extremely cautiously . It must be positively established that any new feature encountered in stratospheric particles is truly of extraterrestrial origin and not related to terrestrial or to ruminated meteoritic matter in the Earth's atmosphere .
Acknowledgements EKJ thanks the Naturhistorisches Museum at Vienna for its hospitality during a sabbatical visit and is especially grateful to Gero Kurat for much instructive discussion . Gero Kurat, Michele Maurette and an anonymous reviewer made valuable comments on the manuscript . We also thank Thomas Stephan, who made unpublished results available . Frans J.M. Rietmeijer provided some recent data. This work was in part supported by the Deutsche Forschungsgemeinschaft . References 1 D.E. Brownlee, Cosmic dust, Annu . Rev, Earth Planet Sci. 13,147-173,1985. 2 G.J . Flynn, Proc. 19th Lunar Planet . Sci. Conf. 673-682, 1989 . 3 B. Mason, Cosmochemistry . Part 1 . Meteorites, in : Data of Geochemsitry, M. Fleischer, ed ., Geol. Suiv. Prof. Pap. 440-B-1, 1979 (6th ed .) . 4 C.C.A .H . van de Stap, R.D . Vis and H. Verheul, Interplanetary dust : Arguments in favour of a late stage nebular origin, Lunar Planet . Sci. XVII, 1013, 1986 . 5 S.R . Sutton and G.J . Flynn, Stratospheric particles: Synchrotron X-ray fluorescence determination of trace element contents, Proc . 18th Lunar Planet . Sci. Conf. 607614, 1988. 6 G .J . Flynn and S.R . Sutton, Synchrotron X-ray fluorescence analyses of stratospheric cosmic dust: New results for chondritic and low-nickel particles, Proc . 20th Lunar Planet . Sci. Conf . 335-342, 1990 . 7 G.J . Flynn and S.R . Sutton, Average minor and trace element contents in seventeen "chondritic" IDPs suggest a volatile enrichment, Meteorites 26, 1991 . 8 E.K. Jessberger, R. Wallenwein, H. Blank and K. Traxel, Pixe analysis of interplanetary dust particles, Meteoritics 20, 673, 1985 . 9 E.K . Jessberger and R. Wallenwein, PIXE characterization of stratospheric micrometeorites, Adv. Space Res. 6, 5-8, 1986 . 10 R. Wallenwein, H. Blank, E.K. Jessberger and K. Traxel, Proton microprobe analysis of interplanetary dust particles, Anal . Chim . Acta 195, 317-322, 1987. 11 R. Wallenwein, Ch . Antz, E.K . Jessberger and K. Traxel, Proton microprobe analysis of interplanetary dust particles, in: Proc . 10th Regional IAU Meet . (Prague), Z. Ceplecha and P. Pecina, eds., Vol. 2, pp . 245-248, 1987 . 12 Ch . Antz, M. Bavdaz, E.K . Jessberger, A. Kn6chel and R. Wallenwein, Chemical analysis of interplanetary dust particles with synchrotron radiation, in : Proc . 10th Regional IAU Meet . (Prague), Z. Ceplecha and P. Pecina, eds., Vol. 2, pp . 249-252, 1987 . 13 R. Wallenwein, Ch. Antz, E.K. Jessberger, A. Buttkewitz, A. Knöchel, K. Traxel and M. Bavdaz, Multielement anal-
VOLATILE ELEMENT ENRICHMENT OF CHONDRITIC INTERPLANETARY DUST PARTICLES yses of interplanetary dust particles with PIXES and SYXFA, Lunar Planet. Sci. XX, 1171-1172, 1989 . 14 E.K . Jessberger, Halley's grains: interplanetary and interstellar dust, Meteoritics 24, 281, 1989 . 15 E.K . Jessberger, Asteroids, meteorites, interplanetary and Cometaryy dust, in : Proc . Enrico Fermi Int . Sch. Phys ., E.
Bussoletti and G. Strazzulla, eds., pp. 107-127, Soc. Ital. Fis., 1991, 16 E.K. Jessberger, J. Bohsung, S. Chakaveh and K. Traxel, New PIXE analyses of interplanetary dust particles, Meteoritics 26, 352, 1991 . 17 F. Bosch, W. Herth, B. Martin, R. Nobiling, B. Povh, H.D . Reiss and K. Traxel, The Heidelberg proton microprobe, Nucl, Sci. Appl . 1, 33-35, 1980 .
18 Ch. Antz, Kombinierte chemische Multielementanalyse interplanetarer Staubteilchen mit Hilfe der Protonenmikrosonde und der Synchrotron-Röntgen-Fluoreszenz,
Thesis, Univ. Heidelberg, 1988. 19 E. Zinner, K.D . McKeegan and R.M. Walker, Laboratory measurements of D/H ratios in interplanetary dust, Nature 305, 119-121, 1983 . 20 F.J . Stadermann, Rare earth and trace element abun-
dances in individual 1DPs, Lunar Planet . Sci. Conf. XXII, 1311-1312,1991 . 21 E. Zinner, Cometary and interstellar dust grains : Analysis by ion microprobe mass spectrometry and other techniques, Space Sci. Rev . 56, 147-156, 1991 . 22 F.J . Stadermann, Messung von Isotopen- und Elementhäufigkeiten in einzelnen Interplanetaren Staubteilchen mittel Sekundärionen-Massenspektrometrie, Ph .D . Thesis, Univ . Heidelberg, 1991 . 23 M. Maurette, C. Olinger, M. Christophe Michel-Levy, GKurat, M. Pourchet, F. Brandstätter and M. Bourot-Denise, A collection of diverse micrometeorites recovered from 100 tonnes of Antarctic blue ice, Nature 351, 44--47,
1991 . 24 J .P . Bradley, S.A . Sandford and R.M . Walker, lnterplanu tary dust particles, in : Meteorities and the Early Solar System, J.F . Kerridge and M.S . Matthews, eds., pp. 861895, Univ . Arizona Press, Tucson, 1955 .
25 E. Anders and N. Grevesse, Abundances of the elements : Meteoritic and solar, Geochim. Cosmochim. Acta 53, 197214, 1989. 26 J.T. Wasson, Meteorites : their Record of Early Solar-Sys-
tem History, 267 pp ., Freeman, New York, 1985 . 27 E. Biemont, M. Baudoux et al ., The solar abundance of iron : a "final" word, Astron. Astrophys. 249, 539-544, 1991 . 28 Ch . Leinert and E. Griin, Interplanetary dust, in : Physics
and Chemistry in Space (Space and Solar Physics, Vol . 20, Physics of the Inner Heliosphere 1), R. Schwerin and E. Marsch, eds., pp. 207-275, Springer, Heidelberg, 1990. 29 F.L. Whipple, On maintaining the meteoritic complex, in : The Zodiacal Light and the Interplanetary Medium, J.L .
Weinberg, ed ., pp . 409-426, NASA, 1967 (SP-150). 30 Cosmic Dust Catalogue 83, NASA Johnson Space Cent ., Houston, 1983 . 31 G.W . Kallemeyn and J.T. Wasson, The compositional clas-
sification of chondrites-1 . The carbonaceous chondrite groups, Geochim. Cosmochim. Acta 45, 1217-1230, 1981 . 32 A. Steinweg, D. Krankowsky, P. Lämmerzahl and B. Anweiler, Metal ion layers in the auroral lower E-region measured by mass spectrometers, J. Atmos. Terr. Phys. . in
press, 1991 . 33 G.J . Flynn, Atmospheric entry heating: a criterion to distinguish between asteroidal and Cometary sources of interplanetary dust, Icarus 77, 287-310, 1988 . 34 O.B . Toon and N.H . Farlow, Particles above the tropopause, Annu Rev . Earth Planet . Sci. 9, 19-58, 1981 . 35 I.D .R . Mackinnon and D.W . Mogk, Surface sulfur mea-
surements on stratospheric particles, Geophys. Res. Lett . 12, 93-96, 1985 . 36 F.J .M . Rietmeiier, Bromine in interplanetary dust paricles [IDPs] : Evidence for stratospheric contamination, Meteoritics 27, submitted, 1992 . 37 J.P. Bradley, D.E . Brownlee and D.R . Veblen, Pyroxene whiskers and platelets in interplanetary dust : evidence of vapour phase growth . Nature 301, 473-477, 1983 .