Status and Science Results D. S Burnett and Genesis Science Team GPMC, August 2009

1. Status and Science Results D. S Burnett and Genesis Science Team GPMC, August 2009

2. Slides Background/Review Material 3-10 5 Year Publication Summary 11-13 Noble Gas elements and isotopes 14-16 Elemental Abundances; FIT Fractionation 17-27 Mg Isotopes in the Solar Wind 28-40 Contamination Issues Overview 41-45 Master Plan for Surface Cleaning 46-55 Status/Summary 56-60 Contents

3. What: Mission in a Nutshell • Placed a spacecraft outside the terrestrial magnetosphere • Exposed Materials • Solar wind ions (keV/amu) implant and stick • Exposed for 27 months • Fluences low, so materials must be ultrapure. • Returned materials to Earth for analysis in terrestrial laboratories.

4. Why: Genesis Science Objectives • Provide solar isotopic abundances to level of precision required for planetary science purposes. • Provide greatly improved knowledge of solar elemental abundances. • Provide a reservoir of solar matter to meet the needs of 21st century planetary science. • Provide elemental and isotopic data for the 3 different types (“regimes”) of solar wind.

5. Solar Wind Regimes • Three different kinds (“regimes”) of solar wind: • High speed (coronal hole) • Low speed (“interstream”) • Coronal Mass Ejections • Genesis separately sampled each of 3 solar wind regimes as well as bulk solar wind: • Allows correction for differences in composition between sun and solar wind • Agreement in derived solar composition from different regimes validates correction procedures

6. BMG Regime Arrays Al kidney Au kidney Concentrator Canister Cover & Array

7. Canister and Collector Materials pre launch target grids regimes Collector arrays Al kidney (PAC) Thermal closeout

8. Analysis Overview • Genesis sample analysis/testing is proceeding on a broad front in 28 laboratories worldwide. • Rates vary, but progress is being made. • The goal of Genesis is quantitative data; great emphasis on getting numbers right. • A major advantage of sample return missions is that important data can be verified, and in most cases, replicated with different techniques. • A major effort has been to make accurate, replicated measurements of the fluences of Mg and Ne. Most techniques can analyze one of these elements, which will then constitute primary quantitative reference fluences for other elements.

9. Analysis Overview, con’t Two distinct requirements: • Extract implanted solar wind from collector materials. • Analyzed extracted solar wind. Can mix and match approaches for extraction and analysis. Mass spectrometry is the most widely-used analysis technique.

10. Science Team Analysis Methods Secondary IonMass Spectrometry (SIMS) Solar wind extracted by ion beam sputtering Gas SourceMass Spectrometry Extraction by laser ablation or chemical etching (HNO3, Hg) Resonance Ionization Mass Spectrometry (RIMS) Extraction by ion beam sputtering Total Reflection X-ray Fluorescence in-situ analysis; unique in not requiring extraction. essentially non-destructive. Inductively-coupled Plasma Mass Spectrometry Extraction by differential chemical etching Accelerator Mass Spectrometry Extraction by differential chemical etching. Radiochemical Neutron Activation Analysis Extraction by differential chemical etching.

11. 5 Year Publication Summary. • Mostly refereed journals; no abstracts. • The 50-75 Genesis 2 page LPSC abstracts, not listed here, are major sources of information. Most can be regarded as progress reports. Those from the Curatorial Facility have special long term value.

12. Work published or submitted. A new time-of-flight instrument for quantitative surface analysis. Veryovkin et al. 16th Conf on Ion Beam Analysis, 2003 Ne in Genesis Bulk Metallic Glass solves long-standing lunar problem (ETH Zurich). Grimberg et al. Science 314, 1133, 2006. Proc. Symp. Comp. of Matter symposium, 3 papers. Space Sci. Rev. vol 170, 2007. Reisenfeld et al Solar Wind Conditions for Genesis samples, based on monitor data along with other spacecraft data for the same period. Wiens et al Genesis Solar Wind Overview Heber et al. Genesis Concentrator Performance Based on Ne Analysis of the Target Holder Gold Cross. Ne and Ar isotopic composition of different regimes (Wash U), Meshik et al., Science 318, 443, 2007 This paper produced additional favorable technical summaries: Science 318, 401, 2007 NatureNews 17 Oct 2007 nature.com/news/2007/071018/full/news.2007.175.html

13. Published work Solar wind noble gases in targets from the Genesis mission PhD thesis, Ansgar Grimberg, ETH Zurich, 2007. Wieler et al., Consequences of the non-existence of the SEP component for noble gas geo- and cosmochemistry. Chemical Geology 244, 382, 2007. Grimberg et al. Solar wind He, Ne, and Ar isotopic and elemental composition. Data from the metallic glass flown on the NASA Genesis spacecraft.Geochimica, Cosmochemica Acta 72, 626, 2008. Mao et al. MegaSIMS: a SIMS/AMY hybrid for measurement of the Sun’s isotopic composition. Applied Surface Science 255, 1461,2008. Kitts et al. Application of grazing incidence XRF techniques to discover and quantify implanted solar wind. J. Applied Physics 105, 64908, 2009. Zimmerman et al. Laser ablation (193 nm), purification, and determination of very low concentrations of solar wind N in targets from the Genesis spacecraft. Geostandards and Geoanalysis 33, 183, 2009. Marty et al. N isotopes in the recent solar wind from the analysis of Genesis targets: Evidence for large scale isotopic heterogeneity in the nascent solar system. in press Geochimica and Cosmochimica Acta 2009. Heber et al. Noble gas composition of the solar wind as collected by the Genesis mission. submitted to Geochimica, Cosmochimica, 5/09.

14. Heber et al. Noble Gas Isotopic and Elemental Composition This is the #3 science objective of Genesis. Samples of Sandia diamond-like-carbon were laser ablated to release HeNeArKrXe for mass spectrometric analysis. Image shows pits from HeNeAr analyses. Those for KrXe were 3-7 times larger, given low fluences. Data in following slides refer to bulk solar wind.

15. HeNeAr isotopic compositions DOS refer to data from Heber et al. AloS and BMG are Genesis data from Meshik et al (2007) and Grimberg et al (2008). SWC refers to the Apollo foils (Geiss et al., 2004). Lunar ilmenite data from Benkert et al (1993) The good agreement between the DOS and lunar ilmenite data is somewhat surprising given evidence for significant erosion of lunar regolith grains (Grimberg et al., 2006). Except for He, Genesis data agree with the SWC within the larger errors of the latter. This is important since the SWC data refer to a few days in 1969-1972 and Genesis represents a 27 month average. Spacecraft data indicate isotopic variations on short time scales, but the SWC+Genesis data show mostly quick convergence to relatively well long term values. Although small, the He difference is real and warrants further thought.

16. Heber et al., Noble gas elemental compositions Labels on rows are as in previous slide. He is not retained in AloS and BMG. Large losses of He, Ne, and possibly Ar have occurred in lunar ilmenite. Although marginally significant, the differences between DOS and SWC are important with respect to issue of long term average. Relevance to model photospheric estimates are discussed below.

17. Science Issue: Do Sun and solar wind have same elemental composition? Slides 17-20 used in previous GPMC. • Spacecraft data have shown that high first ionization potential (FIP) elements are depleted in solar wind compared to solar surface (photosphere). • e.g. Fe/He is higher in SW than in photosphere. • Data for most easily-ionized elements (FIP < 9eV) appear unfractionated. • Most of elements in terrestrial planets have FIP<9eV • Genesis will provide a better test, but never will escape need to know a few photospheric elemental ratios accurately. • If fractionations due only to first ionization potentials, solar wind and photosphere isotope ratios expected to be same.

18. Fractionation Factor F = (X/Mg)SW / (X/Mg)photosphere

19. FIP Plot from spacecraft data

20. FIT (first ionization time) plot from spacecraft data FIT is an estimate of the time required to ionize a neutral atom upon transport from the lower temperature photosphere into the solar corona, from where the ion will be accelerated and incorporated into the solar wind. FIT is more physical than FIP, but is model-dependent. Ulysses data plots using FIT are cleaner than those with FIP with the 9eV fractionation cutoff (translated to about 10- 20 sec ionization time) showing clear depletions of high FIP/FIT elements, but no evidence of depletion of low FIP/FIT. Errors in SW abundances about 20%; errors in photosphere 10-20%. Note that Mg and Fe have same FIP but different FIT

21. Cr fluence • SIMS sensitivity for Cr is not high, but measurable. Figure shows profiles from Burnett et al (LPSC 2007). Crash-derived surface contamination (stainless steel?) is a big issue. • Fluence from 2007 data: 3.0 ± 0.5 x1010 atoms/cm2. • Wang et al (Met Soc 2009): • Aqua regia treatment (Humayun) lowered surface contamination significantly. • Fluence 3.3 ± 1.1 x1010 atoms/cm2 Agreement OK for now.

22. FIT plot from Genesis data We can combine the noble gas and Cr fluences from previous slides with the Mg and Fe fluences from the May 2009 GPMC and produce a preliminary FIT plot using only Genesis data. As for Ulysses data, Mg is used as normalizing element. Note that most literature FIP/FIT plots use O as the normalizing element. Errors in either the Genesis Mg fluence or the photospheric Mg abundance only affect the value of F for other elements. The pattern of points on the plot is unaffected. Ne and Ar are not plotted, as there are no spectroscopoic data for their photospheric abundances. Kr and Xe are special cases. There are no spectroscopic photospheric abundances; however, CI abundance curves are sufficiently smooth in the Xe, and especially Kr, mass regions that interpolation gives relatively precise abundances subject to two assumptions: the overall validity of CI abundances. the assumption that Kr and Xe, as volatile elements, are not depleted or enhanced relative to neighboring nonvolatile elements. Genesis can eventually test both of these assumptions. Light element photospheric abundances are too variable to interpolate abundances for Ne and Ar.

23. 1.E+01 Genesis F(Mg) 1.E+00 ?? FIT, s 1.E-01 0.01 0.1 1 10 100 1000 FIT plot from Genesis data The dashed lines are obviously not a unique description of the data, but as more and better Genesis data are obtained, the true systematics of the data will be revealed. As noted earlier by Vogel et al (LPSC 09), the solar wind Xe/Kr is significantly higher than the interpolated solar ratio. This is an issue deserving of further study. Mg Cr Fe Xe H He Kr

24. 1.E+01 Genesis F(Mg) 1.E+00 ?? FIT, s 1.E-01 0.01 0.1 1 10 100 1000 Solar Ne and Ar abundances from Genesis At this point, this is just an illustration of the approach rather than a serious attempt to derive solar abundances. The plot shows the location of FIT for Ar and Ne. Taking the dashed line seriously, values of F for Ar and Ne can be interpolated. Using the measured solar wind Ar/Mg and Ne/Mg the photospheric ratios can be derived. The derived abundances can converted to abundancs relative to H and compared with the literature. Ne Ar

25. Comparison of Genesis solar Ne and Ar with literature. Looks respectable relative to other estimates, especially Grevesse. GFIT Previous page GPMC 11/07: extrapolation of Genesis He/Ne vs He/H for regime data of Heber and monitor H Grevesse: abundances originally in Anders and Grevesse (1989) unchanged in subsequent compilations. Lodders 03 Ap J. 591, 1220, 2003 Lodders 08 Ap J 674, 607, 2008 PJ Palme and Jones Treatise of Geochemistry B(07) Bochsler A&A 471, 315, 2007.

26. Fe K Cr CaMg P S O N H Al C He Xe Na Kr Genesis + SOHO MTOF + Ulysses FIT plot

27. Discussion: Genesis + SOHO MTOF + Ulysses FIT plot Previous slide compares Genesis bulk data with those for many elements from Giammonco et al (Ap J 681, 1703 2008) for slow solar wind. Bulk is mostly slow. Spacecraft He, CNO, and Fe are from Ulysses slow (via Reisenfeld). Given errors in spacecraft data, flat plus dropoff at high FIT curve would still be OK. However, taking data at face value, monotonically-decreasing trend is suggested. Key test is solar wind Al fluence. Giammonco et al calculate that Al has a very low FIT and does appear enriched. Genesis Al fluence should be measurable and has gone to top of element priority list.

28. Photosphere/SW Isotope Fractionation? • More an issue to Genesis than elemental fractionation since isotopes are our highest priority objective. • Spacecraft data not precise enough for planetary science? • FIP/FIT are atomic properties, wouldn’t expect isotopic fractionation. • “Coulomb Drag” effects associated with acceleration of solar wind from corona would be mass dependent; specific model by Bochsler (2000); relatively large fractionations predicted. • Regimes are different kinds of solar wind, so if isotopes fractionated, expect to see differences. • Test with He, Ne and Ar extracted by laser ablation at Zurich (Sandia) and Wash U St. Louis(AloS) from 5 mm-sized fragments. • no cleaning required.

29. Importance of Solar Wind Mg Isotopes • Mg isotopes provide a good test for isotopic differences between photosphere and solar wind if an assumption is made: • With exception of evaporation effects in CAIs, no systematic mass dependent fractionations for non-volatile elements greater than about 1 per mil / amu. • Thus, assume comparison of Earth and Solar Wind is comparison between Photosphere and Solar Wind. • For noble gases or other volatiles, have no apriori knowledge of photospheric isotopic compositions. • Validity can be tested with other non-volatile elements, e.g. Fe • Assume variations will be mass dependent fractionations.

30. Bochsler (2000) Coulomb Drag Model All solar wind samples are depleted in He/H elemental relative to helioseismology solar ratio by about a factor of 2.5. Model assumes that all He/H fractionation all due to Coulomb Drag. With this assumption, fractionations for other isotope ratios can be calculated. 26Mg/24Mg depletion of 20 per mil in solar wind relative to Sun predicted.

31. Kallenbach et al. 2001 SOHO/Celias/MTOF Data show 26Mg depletion consistent with Bochsler.

32. Genesis Mg isotopes by SIMS • Genesis planning assumed that ICPMS measurement required for good test, but with -10 to -20 permil per amu, ion probe looked feasible. • ASU 6f: Jurewicz, Hervig, Rieck • Genesis bulk solar wind diamond-like-carbon (DLC) and Si samples. • Focus on DLC here

33. SIMS mass fractionation correction • “Srni” implant: separate 24Mg, 26Mg implants: • 26/24 =0.10 • High 24 fluence, 3e15/cm2 • Permitted ICPMS calibration (Wadhwa) • (26/24)implant = 0.09989 • Good agreement with implanter integrated current ratio. • - Kroko relative fluences appear good.

34. Adequate Sensitivity: DLC Sample 60065 Sufficient mass resolv. power to rule out contributions from C2+ Solar wind = cons. v ions; Implant = cons E ions. In both cases, depth profiles differ for isotopes; calculate ratios of integrated profiles. Effect of choice of end points for integration not important. 24 25 & 26 Per profile counting statistics errors in 25/24 or 26/24 about 6 per mil one sigma. Analysis times short enough that many profiles feasible.

35. Dead time corrections For SIMS, high 24Mg fluences in Srni implant standard creates dead time problems. Issue somewhat unique to Genesis. • Depth profiling requires rastered beam. • Measured average counting rates much lower than instantaneous. • Semiconductor industry rarely interested in precise isotopic ratios, so little work on accurate correction procedures. • Our deadtime corrections should be good, but they are approximate, not exact.

36. Dead time: 6% at peak; 4% on integral.40 permil correction on M/24 ratios

37. 25Mg only: all SW ratios agree within 2  Error of mean justified.

38. Good agreement of F26. Don’t agree with average F25 For mass dependent fractionation, F26 = F25 expected. Not observed; very surprising result. Confirmation required.

39. Compare (26Mg / 25Mg) • Very insensitive to dead time errors 11 profile average: Solar Wind (26Mg / 25Mg) depleted by 2.2 ± 0.8 permil (2 sigma) Much smaller than F derived from 26Mg / 24Mg or 25Mg / 24Mg. The 0.8 permil is what formal uncorrelated error propagation gives. But this seems wrong because it appears to beat counting statistics. Within counting statistics, 26/25 could be terrestrial.

40. Conclusions • SW data relative to 24Mg not consistent with mass fractionated terrestrial Mg. • Problem quite possibly with us, not the Sun. • 26Mg / 24Mg depletion significant unless we have underestimated dead time corrections by 50%, which seems large. • 24Mg-rich compositions, in same sense and magnitude as Bochsler (2000) model and Kallenbach et al SOHO data. • Good news is that sensitivity and precision of solar wind Mg isotopic data are good, for both DLC and Si. • Problems with standard, not with Genesis sample. • Good news, as this can be fixed. • With lower fluence, Srni III 26Mg and 25Mg implant, can reduce deadtime corrections by large factor. • dead time errors should cancel with 26/25 =1 in implant. • Can also improve method of deadtime correction. • ICPMS of Srni III is feasible.

41. Cleaning up to Recover Science after Crash Three necessary steps to recovery of science objectives: 1. Recover Collector Materials intact. Done Expected 250 samples, have ~ 15,000 > 3mm; 1700 > 1 cm . Priority given to allocation, but major progress made on catalogs, see JSC GPMC contribution or Genesis JSC web page. • Remove surface contamination. Required for essentially all analyses, especially from here on out. • Learn to allocate and analyze smaller samples than planned. Items 2 and 3 worked simultaneously. This and the following 2 slides are background from previous GPMC,

42. Basic Approach: • Contamination levels are highly variable. • Cleanliness requirements vary for different analytical techniques. • No one-size-fits-all solution. • Basic Curatorial cleaning services: UPW, uv-ozone. • Rest is responsibility of PIs • but Curatorial Facility supports with characterization • Particle counting (JSC) • Ellipsometry (JSC) • XPS (EAG commercial lab) • Lab TRXRF (new; see following pages) • Ellipsometry doesn’t work for some materials and doesn’t give quantitative information.

43. Brown Stain (slide unchanged from previous GPMC) Non-crash issue Polymerized organic contamination film (‘brown stain”) Thicknesses measured by ellipsometry (JSC,) XPS (EAG, JPL), and FIB/TEM (LLNL). Up to about 75 A thick, but Highly variable; some samples appear essentially free of stain. If less than 100A: negligible SW attenuation ( C. Olinger, LANL calculations). Brown stain must be removed for most, but not all, analyses: uv-ozone (demonstrated by Open U) most successful to date. JSC unit is operational and demonstrated to remove C effectively For some applications, greater amount of removal may be required. good correlation between XPS and ellipsometery on same Si samples. We have learned to work around Brown Stain. Important Boundary Condition: Because amounts of contamination highly variable, cherry-picking good (low brown stain) samples is an acceptable contamination control.

44. Particulate Contamination Overview. Crash-related issue: Particulate contamination on all samples. • Variety of wet cleaning techniques work to varying degrees: • Any solvent (e.g. ultra pure water) will take off 1/2-2/3 of particles and almost all big (>5 micron) ones. • For most samples, JSC Megasonic ultra-pure-H2O (UPW) in routine use for materials for which this possible. • Probably not applicable for AloS samples and must be done with care for AuoS • Particulate contamination is themajor obstacle to completion of the Genesis science objectives. • Our success so far has been with techniques such as SIMS or RIMS that can analyze areas of 50-200 micron size, can dodge micron-size particles and can recognize, and afford to lose, a particle-contaminated profile. • None of these benefits are available for large area analysis ( > 1 cm size ), for which in some cases a single contaminant particle can ruin the analysis. • Some of the science objectives require analysis of large areas. • Some success with acid cleaning, but not good enough. Systematic approach needed.

45. Master Plan for Sample Cleaning Plan was presented in May 2009 GPMC. Following 9 pages, repeated, describe plan. Implementation is in progress.

46. UPW PI Curatorial Facility UV - O3 Lab TRXRF Wet cleaning SRTRXRF TOF-SIMS Lab TRXRF SEM Clean? yes no Master Plan for Sample Cleaning

47. Master Cleaning Plan Previous GPMC show that we know a lot about particulate surface contamination. However, we don’t know enough to successfully clean samples for large area analysis. We need approach(s) capable of efficient before-after measurements on samples subject to various wet cleaning techniques. Analysis must be non-destructive; need to be quantitative but high accuracy not required. Efficiency and access important because a lot of trial and error will be required in wet cleaning tests. We need to do a large number of analyses. XPS used previously does not have adequate sensitivity. Synchrotron radiation TRXRF (SRTRXRF) (APS) and TOFSIMS (Manchester) have adequate sensitivity. Genesis time for SRTRXRF only a few days per year. Need to emphasize solar wind analysis. Access to Manchester TOFSIMS has been good, but not possible to process a large number of samples solely for cleaning studies.

48. Laboratory TRXRF • Use tube X-rays rather than synchrotron radiation; otherwise technique is same as described earlier. • XPS detection limits are ~ 1014 atoms/cm2 • SRTRXRF has ~ 1010 atoms/cm2 atoms/cm2 detection limits. • Lab TRXRF ~ 1011- 1012 atoms/cm2 detection limits achievable because of good signal/background and use of 104 - 105 sec counting times. • Minimum sample handling; samples analyzed, as received, in air. • Samples handled in laminar flow benches. • TRXRF demonstrated to work on all collector materials (except diamond-like-C for unknown reasons). • Not sensitive for elements lighter than Si; works best for 1st row transition elements (Ca-Ge), but samples clean of all these elements is way beyond where we are now. • Two laboratories identified with adequate capabilities: Loyola Chicago (M. Schmeilng) and NIU (L. Lurio). Sample spectra from Genesis flight sapphire in following slides.

49. Loyola. UPW-washed sapphire 60679 Spectrum slightly below critical angle to sample surface contamination

50. NIU UPW-rinsed sapphire 21022 Two different angles near the critical angle.