G i S l Wi d Al i Ab d Ch ll ith El t Mi b A l i f Al i Oli iGenesis Solar Wind Aluminum Abundance: Challenges with Electron Microprobe Analysis of Al in OlivineGenesis Solar Wind Aluminum Abundance: Challenges with Electron Microprobe Analysis of Al in Olivine Genesis Solar Wind Aluminum Abundance: Challenges with Electron Microprobe Analysis of Al in Olivine g p y1 2 1 1 1 3 1 1 1A E H f 1,2 J M P 1 D S B tt1 Y G 1 A J G J i 3 C M 1 d G R R 1 1C lif i I tit t f T h lA E Hofmann1,2 J M Paque1 D S Burnett1 Y Guan1 A J G Jurewicz3 C Ma1 and G R Rossman1 1California Institute of TechnologyA. E. Hofmann , J. M. Paque , D. S. Burnett , Y. Guan , A. J. G. Jurewicz , C. Ma and G. R. Rossman . California Institute of Technology, , q , , , , gy,

CA 9112 2 i C if i i f 488 O G /S 183 301 CA 9110Pasadena CA 91125 2Jet Propulsion Laboratory California Institute of Technology 488 Oak Grove Dr M/S 183 301 Pasadena CA 91107Pasadena, CA 91125, Jet Propulsion Laboratory, California Institute of Technology, 488 Oak Grove Dr M/S 183-301, Pasadena, CA 91107,Pasadena, CA 91125, Jet Propulsion Laboratory, California Institute of Technology, 488 Oak Grove Dr M/S 183 301, Pasadena, CA 91107, 3

p y gyh f @j l 3SESE A i St t U i it T AZ 85287amy e hofmann@jpl nasa gov 3SESE Arizona State University Tempe AZ 85287amy.e.hofmann@jpl.nasa.gov. SESE, Arizona State University, Tempe, AZ 85287. a y.e. o a @jp . asa.gov. S S , o a State U ve s ty, e pe, 85 87.

THE ALUMINUM PROBLEMTHE ALUMINUM PROBLEMTHE ALUMINUM PROBLEMTHE ALUMINUM PROBLEM Th G i i i d i d t t l d d ib d i [2] d ill t t dThe Genesis mission was designed to accurately done as described in [2] and illustratedThe Genesis mission was designed to accurately done as described in [2] and illustrated g y

d i l h i i f h l[ ]

i Fi 3 hi h h SIMS d thand precisely measure the composition of the solar in Fig 3 which shows a SIMS depthand precisely measure the composition of the solar in Fig. 3, which shows a SIMS depth and precisely measure the composition of the solar g , pfil f il ll i li i iwind (SW) The sample return capsule crashed in profile for Imilac pallasite olivine usingwind (SW) The sample return capsule crashed in profile for Imilac pallasite olivine usingwind (SW). The sample return capsule crashed in profile for Imilac pallasite olivine using ( ) p p

th d t f ti d t i ti l th C lt h C 7f G ith 3000the desert fragmenting and contaminating samples the Caltech Cameca 7f Geo with a 3000the desert, fragmenting and contaminating samples the Caltech Cameca 7f Geo with a 3000 , g g g p(Fi 1) MRP t l 27Al+ f 26M H+ If(Figure 1) MRP to resolve 27Al+ from 26MgH+ If(Figure 1). MRP to resolve Al from MgH . If (Figure 1). g

h li i l i k h ithe olivine Al content is known the imY t i i l f th t diffi lt t k i the olivine Al content is known, the im-Yet surprisingly one of the most difficult task in the olivine Al content is known, the imYet, surprisingly, one of the most difficult task in l t fl b l l t d i

, p g y,i SW f th G i l i li plant fluence can be calculated or vicemeasuring SW from the Genesis samples is cali plant fluence can be calculated, or vice measuring SW from the Genesis samples is cali- p ,

[2]g p

versa [2]brating the SIMS (secondary ion mass spectrome versa [2]. brating the SIMS (secondary ion mass spectrome- [ ]brating the SIMS (secondary ion mass spectromet ) t Control pieces of Si mounted beside thetry) measurements Control pieces of Si mounted beside thetry) measurements. Control pieces of Si mounted beside the y)

li i d i i l ti i thP i d h fil f h l i d l i olivine during implanting receive thePrecise depth profiles of the solar wind aluminum olivine during implanting receive the Precise depth profiles of the solar wind aluminum g p gfl d i

ec se dept p o es o t e so a w d a u usame fluence and serve as a primaryare available from backside depth profiling [1] of same fluence and serve as a primary

Fi 2are available from backside depth profiling [1] of s e ue ce d se ve s p yFigure 2are available from backside depth profiling [1] of

standard for analysis of Genesis Si samFigure 2 p p g

G i ili ll t C l l ti f standard for analysis of Genesis Si sam-Genesis silicon collectors Calculation of standard for analysis of Genesis Si samGenesis silicon collectors. Calculation of lh l i d Al fl ( / 2) plesthe solar wind Al fluence (atoms/cm2) ples. the solar wind Al fluence (atoms/cm ) p

Fi 1 the solar wind Al fluence (atoms/cm ) A l i b ( )

Figure 1 from these is based on a laboratory im Accurate electron microprobe (EMP)

gfrom these is based on a laboratory im- Accurate electron microprobe (EMP)from these is based on a laboratory im Accurate electron microprobe (EMP) yl t t d d (Fi 2) Th i l t Al anal ses of oli ine Al at the ≈100 ppmplant standard (Figure 2) The implant Al analyses of olivine Al at the ≈100 ppmplant standard (Figure 2). The implant Al analyses of olivine Al at the ≈100 ppm p ( g ) p

b i d 4 i h d iy pp

l l h ld b ibl I tibeam is rastered over 4 inches and is level should be possible In practicebeam is rastered over 4 inches and is level should be possible. In practice, beam is rastered over 4 inches and is p p ,i ifi blquite uniform therefore the Si pieces many significant problems were encounquite uniform therefore the Si pieces many significant problems were encoun-quite uniform, therefore the Si pieces

2many significant problems were encounq p

h th i l t fl (i / 2) tered hose mitigations are describedhave the same implant fluence (ions/cm2) tered whose mitigations are describedhave the same implant fluence (ions/cm ) tered, whose mitigations are described p ( )S C l li i

ghas San Carlos olivine hereas San Carlos olivine. here. as San Carlos olivine.

C tifi d i l i l t fl Our work is potentially valuable forCertified nominal implant fluences are Our work is potentially valuable forCertified nominal implant fluences are Our work is potentially valuable for pl k t b t ±20% i d d EMP t l t l i lonly known to about ±20% so independ EMP trace element analyses in generalonly known to about ±20%, so independ- EMP trace element analyses in general. y , p

lib i i i d hi by g

ent calibration is required This can beent calibration is required. This can beent calibration is required. This can be

RESULTSANALYTICAL METHODS RESULTSANALYTICAL METHODS RESULTSANALYTICAL METHODS RESULTSANALYTICAL METHODS RESULTS Al in San Carlos Olivine Al in Pallasite Oli ineX h t l t (XPS)El t i b l Al in San Carlos Olivine Al in Pallasite OlivineX-ray photoelectron (XPS)Electron microprobe analyses: Al in San Carlos Olivine Al in Pallasite Olivine X-ray photoelectron (XPS) Electron microprobe analyses: y p ( )p y

S t ti i t t spectrometrySystematic instrument errors We also explored the use of pallasite olivine as aProtocol used for analyses:spectrometry Systematic instrument errors We also explored the use of pallasite olivine as aProtocol used for analyses: Figure 7p yy We also explored the use of pallasite olivine as a y50 100

Figure 7

S ll d hi d S CT f Al i i f b i Genesis Al standard Imilac was implanted (Fig50x100µm rasterSpectra collected on sapphire and San Car-To test for Al contamination from carbon coating a Genesis Al standard. Imilac was implanted (Fig.50x100µm raster Spectra collected on sapphire and San Car-To test for Al contamination from carbon coating a Genesis Al standard. Imilac was implanted (Fig. µ100 A b tlos olivine400x400 µm area was SIMS cleaned on Si metal 1) Preliminary estimates of ppm Al are given in100 nA beam currentlos olivine 400x400 µm area was SIMS-cleaned on Si metal 1). Preliminary estimates of ppm Al are given in100 nA beam current

Ab b k d ti t f Al 2µ

d t d ith C1). Preliminary estimates of ppm Al are given in

5 minute count time forAbove-background counting rates for Al 2pand recoated with C. Table 1 based on the nominal implant fluence5 minute count time for Above background counting rates for Al 2p 4

and recoated with C. Table 1 based on the nominal implant fluence.k d b k delectrons on sapphire were ~2x104 cpsEMP analyses around the Al k alpha peak showed a

Table 1 based on the nominal implant fluence. peak and backgroundselectrons on sapphire were ~2x10 cps EMP analyses around the Al k-alpha peak showed a These concentrations arepeak and backgrounds

Table 1:pp p

Oli i l i ti t 300y p p

decrease in co nting rate at the peak (Figure 4) These concentrations are p g2 i t li b k d

Table 1: Olivine aluminum counting rates were 300-decrease in counting rate at the peak (Figure 4). ese co ce t at o s a e

2 point linear background Aluminum in olivine ppmOlivine aluminum counting rates were 300700 500 700 d

decrease in counting rate at the peak (Figure 4). C d h k fi h b k precise to within a few per

2 point linear background Aluminum in olivine ppm I il 1 50700 cps on 500x700 µm areas due to sur-Counts are measured on the peak first then back- precise to within a few per-correction (Figure 7) Imilac 1 50 700 cps on 500x700 µm areas due to sur-Counts are measured on the peak first, then back- p ec se o w ew pe

h h il icorrection (Figure 7)

il 2 91face contaminationgrounds with 100 nA beam current 10 µm spot and cent thus the Imilac varia( g )

f 100 Al hi iImilac 2 91 face contamination grounds, with 100 nA beam current, 10 µm spot and cent, thus the Imilac varia-→for 100 ppm Al this gives

Al 2 k i ibl l bl f b kg , , µ p10 i t t ti

,i l f h

→for 100 ppm Al, this gives Imilac 3 66Al 2p peak visibly resolvable from back-10 minute count times. tions on a scale of hunpp , g

4% i ti t ti ti i tImilac 3 66 Al 2p peak visibly resolvable from back10 minute count times. tions on a scale of hun-4% one sigma counting statistics per point Eagle Station 33ground at all 12 analysis locations on oliTest done with repetitive 10 minute counts on same d d f i hi h

4% one sigma counting statistics per point Eagle Station 33 ground at all 12 analysis locations on oli-Test done with repetitive 10 minute counts on same dreds of microns are high S i t 40g y

ip

t (Fig re 5) dreds of microns are high- Springwater 40 vinespot (Figure 5). g

l i ifi S iTh S C l i h 60 70 Al d (SC4)p g

vine 1% f Al f SC3 ill b f

spot (Figure 5). h 10 d h i i i l d i ly significant SystematicThree San Carlos grains have 60-70 ppm Al and one (SC4)>1% of Al X-rays from SC3 will be fromThe 10 µm spot data show initial decrease in count ly significant. Systematic Three San Carlos grains have 60-70 ppm Al, and one (SC4)>1% of Al X-rays from SC3 will be from The 10 µm spot data show initial decrease in count y g y

i h b l i f 15 20%g pp , ( )

h 130 With ti i d l ti l diti 13 fcontamination (Figure 6)rate (believed due to contamination from electron errors in the absolute concentrations of 15 20%has ~130 ppm With optimized analytical conditions 13 ofcontamination (Figure 6) rate (believed due to contamination from electron errors in the absolute concentrations of 15-20% has 130 ppm. With optimized analytical conditions, 13 of ( g ) th SIMS i b t l th li

(b ) f ll d b t i th b hi h ill ll b li i d14 analyses gave the same counting rate within 1σ counting→use the SIMS ion beam to clean the oli-beam) followed by a steep increase as the carbon are present which will eventually be eliminated14 analyses gave the same counting rate within 1σ counting →use the SIMS ion beam to clean the olibeam) followed by a steep increase, as the carbon are present which will eventually be eliminated. y g g g

i ivine surfaces of contamination and to verifycoat was burned away by the electron beam Thep y

D i h i h i I il ldstatistics errors S C l li i SC3vine surfaces of contamination and to verify coat was burned away by the electron beam. The Despite the inhomogeneity Imilac could serve asstatistics errors. San Carlos olivine SC3 Fi 8y

th l liy y

count rate approaches that of uncoated siliconDespite the inhomogeneity, Imilac could serve as

W h SIMS l d 11San Carlos olivine SC3 Figure 8the cleanlinesscount rate approaches that of uncoated silicon

p g y,SIMS d d f G i EMP lWe have SIMS cleaned 11 Figure 8 the cleanliness cou t ate app oac es t at o u coated s co

(d h d li ) i hi h a SIMS standard for Genesis as EMP analysesWe have SIMS cleaned 11 (dashed line) supporting this theory a SIMS standard for Genesis, as EMP analyses areas on SC3 in preparation(dashed line), supporting this theory. , y

l t l d SIMS i t iblareas on SC3 in preparation San Carlos olivine Fi 6 very close to an analyzed SIMS point are possible

p pf ddi i l l

San Carlos olivine Figure 6 very close to an analyzed SIMS point are possible. for additional analyses SIMS l dFigure 6 y y p p

Th E l St ti d S i t li i Al ifor additional analyses SIMS cleaned Figure 4 The Eagle Station and Springwater olivine Al is

y(Fi 8) Th l ill areasFigure 4 The Eagle Station and Springwater olivine Al is (Figure 8) The sample will areas g g p g

i i tl l f EMP l i H(Figure 8). The sample will

inconveniently low for EMP analysis Howeverbe carbon coated at the same inconveniently low for EMP analysis. However, be carbon coated at the same y y ,th S C l i if Oi h d d d the San Carlos grains appear more uniform Oncetime as the standard and ana- the San Carlos grains appear more uniform. Once time as the standard and ana- g pp

lib t d i l t i il bl SIMS l fl d i th t l a calibrated implant is available SIMS analyses oflyzed using the protocol a calibrated implant is available, SIMS analyses of lyzed using the protocol p , yAl l ti l ffi i tl t Al ilisted above Al can relatively efficiently map out Al zoninglisted above. Al can relatively efficiently map out Al zoning. s ed bove.

Af h l iy y p g

Th E l St ti d S i t l t kAfter the electron micro- The Eagle Station and Springwater analyses tookAfter the electron micro- The Eagle Station and Springwater analyses took b l l t

g p g yb t 1 h i l di j l t lprobe analyses are complete about 1 hour including major element analysesprobe analyses are complete about 1 hour, including major element analyses.

the same spots will be analyzed in the SIMSg j y

Thi ld b d h ffi i t Ab tthe same spots will be analyzed in the SIMS. This could be made much more efficient Aboutt e sa e spots w be a a y ed t e S S.Wi h hi d l i h i l l h l i d This could be made much more efficient. About With this data along with implant results the solar wind

25% f thi ti t i Al fWith this data, along with implant results, the solar wind

25% of this time was spent removing Al surface, g p ,

l i fl fil ill b l l t d 25% of this time was spent removing Al surface aluminum fluence profiles will be recalculatedt i ti

aluminum fluence profiles will be recalculated.

contaminationcontamination. Miti tiMitigationMitigation g

S lid b l Fi 5 f lSolid symbols on Figure 5 are from anal-

CONCLUSIONSSolid symbols on Figure 5 are from anal

i 50 100 i ( ill i CONCLUSIONSyses using a 50x100µ spot size (still using CONCLUSIONSyses using a 50x100µ spot size (still using CONCLUSIONS 100 nA beam current and 10 minute count100 nA beam current and 10 minute count ti )time).

Caution must be taken when performing trace element analyses with the electron microprobe Unantime).

Caution must be taken when performing trace element analyses with the electron microprobe. Unan-Depositional decreases in count rate over Caution must be taken when performing trace element analyses with the electron microprobe. UnanDepositional decreases in count rate over ticipated analytical artifacts such as those we encountered during long analyses need to be corrected

ptime for oli ine are m ch smaller than for ticipated analytical artifacts, such as those we encountered during long analyses, need to be corrected.time for olivine are much smaller than for ticipated analytical artifacts, such as those we encountered during long analyses, need to be corrected. time for olivine are much smaller than for ili i bsilicon; correction may not be necessary

References: [1] Heber V S et al (2014) 45th LPSC Abstract #1203 [2] Burnett D S et al (2015) Geostandards and Geoanalytical Res 39silicon; correction may not be necessary.

References: [1] Heber V. S. et al. (2014) 45th LPSC, Abstract #1203. [2]. Burnett, D. S. et al. (2015) Geostandards and Geoanalytical Res. 39, [ ] ( ) , [ ] , ( ) y ,265 276 265-276.