MORPHOLOGICAL EFFECTS OF DUST GRAINS ON MID-IR EXTINCTION

SPECTRA

October 15, 2009CPS seminar

Akemi Tamanai (AIU Jena & TU Braunschweig)Harald Mutschke (AIU Jena)

Jürgen Blum (TU Braunschweig)Alexander Krivov (AIU Jena)

Chiyoe Koike (Osaka University)

Outline

Dust grains Experiment Results Astronomical application Database Summary

Presence of dust grains appear especially in N and Q bands via spectroscopy.

0

0.2

0.4

0.6

0.8

1

1.2

8 9 10 20 30 40

g-Her

Nor

mal

ized

Flu

x [J

y]

Wavelength [µm]

(Posch et al. 1999 & 2002)

AGB star: g-HerVega-type star: HD69830Comet: Hale-Bopp

(Lisse et al. 2007) (Crovisier et al. 1996)

Herbig Ae/Be star: HD100546

(Malfait et al. 1998)

What kinds of dust grains in these observed spectra?The chemical path way of dust grains

↓Which substance will be the first condensed species out of the gas phase?

At which temperature?At which abundance it occurs?

At what condition does the planet formation start actively?

Sorts of Dust Grains

Decisive role for planet formation, esp. Terrestrial planets

Enstatite (MgSiO3)

Amo-MgSiO3

Fayalite(Fe2SiO4)

Quartz(SiO2)

Corundum(Sapphire: Al2O3)

Corundum(Ruby: Al2O3)

Fe-rich olivine

Mg-rich olivine

The rocky planets form in high metallicity environments, where the dust grains are dominant (e.g. Shen et al. 2005)

Meteorite

Carbonaceous chondrite Allende (CV3.2)

NWA869 L4-6Chondrite

Meteorites hold a key to understand a formation of dust grains.

CAIs (calcium-aluminum inclusions) consist minerals which are predicted via a condensation model.

Minerals in CAIs are condensed from protoplanetary nebula.

Minerals in CAIs are• Spinel (MgAl2O4)• Corundum (Al2O3)• Rutile (TiO2)• Hibonite (CaAl12O19)• Olivine• Pyroxene more….

Factors to influence on MIR band profiles

CrystallinityPolymorph （多形）

同一物質が複数の異なった結晶構造を取る

e.g. SiO2α-quartz (trigonal: 三方晶系)α-tridymite(orthorhombic: 斜方晶系)α-cristobalite(tetragonal: 正方晶系)

Chemical compositonMg, Fe, Al, Si, Ti, Ca, S, …

e.g. OlivineSan CarlosMg1.9Fe0.1SiO4

Sri LanakaMg1.6Fe0.4SiO4

(Koike et al. 2003)

Morphology Size

Shape

Agglomeration

Experiment

Samples

Silicate Forsterite (Mg2SiO4) Fayalite (Fe2SiO4) Olivine (e.g. Mg1.9Fe0.1SiO4) Enstatite (MgSiO3) Diopside (CaMgSi2O6) Hypersthene ((Mg,Fe)SiO3)Wollastonite (CaSiO3) Kaolinite (Al2Si2O7▪2H2O) Talc (Mg3.33Fe0.1Si4O10 (OH)2)

Al2O3 Corundum (α-Al2O3) γ-Al2O3 χ−δ−κ-Al2O3

MgAl2O4 (Spinel)TiO2 (Rutile & Anatase)CaTiO3 (Pervoskite)Al2TiO5 (Tialite)SiO2

www.elbe.astro.uni-jena.de

<Medium effect>εmN2 1.0KBr 2.3CsI 3.0

The influence of its electromagnetic polarization.

Tamanai et al. 2009

KBr :(Potassium Bromide)Mixing ratio 1:500 (sample:KBr)d=13mm ; mass=0.2g

CsI :(Cesium Iodine)Mixing ratio 1:500(sample:CsI)d=13mm ; mass=0.22g

TiO2 (Rutile)Irregular shape w/ roundish edges

Irregular shape with not roundish edges

Long and thin Roundish

(TEM & SEM images: at Pathology w/ Dr. Nietzsche)

<Morphological effects>

0

0.2

0.4

0.6

0.8

1

10 15 20 25 30 35 40 45 50

Aldrich

Nor

mal

ized

Qex

t

Wavelength [µm]

TiO2 (Rutile)

Chitan No.101Tayca 150W

AJ

Peaks (microns)Aldrich 13.53 19.56 24.70 30.30 35.80AJ 15.33 23.51 27.95Tayca150 16.47 19.25 23.39 28.27CI101 13.85 19.27 25.15 35.87

Nor

mal

ized

Ext

inct

ion

<Aerosol vs. Pellet Measurements>

0

0.2

0.4

0.6

0.8

1

1.2AerosolCsI

10 15 20 25 30 35 40 45 50

(a) TiO2 (anatase)

0

0.2

0.4

0.6

0.8

1

1.2AerosolCsI

10 15 20 25 30 35 40 45 50

(b) TiO2 (anatase)

Nor

mal

ized

Ext

inct

ion

Wavelength [µm]

Caused by:

Using different dispersion methods.

Particles may transform during the grinding procedure.

Powdered sample structure deformation caused by the high pressurization required for the CsI pellet technique.

Particle orientation

Dust Grains in Debris Disks

Dust sources Replenished by larger objects (Weissman 1984)

Comets strew over Planetesimals collide with each other

Vega-type stars are surrounded by circumstellar debris disks.

Discoveries: Some planets Planetary companions Planet candidates (e.g. Mayor et al. 1995, Marcy & Butler 1998, Kürster et al. 2000, Lagrange et al. 2008)

Dust in Debris Disks

Large-sized dust grains 0.1 – 10 μm (Krivov et al. 2000)

Radiation pressure takes smaller grains away in gas-poor conditions

e.g. The disk surface area of β-Pic Size 1 to 20 µmDominant grain size 4 µm

(Artymowicz 1996)

Question?

As the grain size increases above 1μm, the spectral features of the infrared dust emission are broadened, and grow fainter.

The Vega-type stars have shown strong emission features in the mid-IR range.

Why?

Objective Extinction measurements of large-sized particles

(2 – 50 μm)

Mixture measurements --- How small-sized particles exert an influence on the extinction spectra when they are present in a same environment with larger ones (0.5 µm & 5 µm particles)

Fe-rich olivine measurements

Amorphous SiO2 (d=0.5 & ~5 µm)Cry-Mg2SiO4 (< 1µm) Cry-Mg2SiO4 (~ 16 µm)

Sample & Setup

Exp. 1 Large-sized particles Amorphous SiO2 (monosphere) Crystalline Mg2SiO4

Exp. 2 Mixture d=0.5 & d~5 µm amo-SiO2 (complete monosphere shape) d=0.5 & d~5 µm amo-SiO2 (5 µm particle are destroyed)

Exp. 3 Fe-rich Olivine Mg1.6Fe0.4SiO4 (Sri Lanka) vs. Mg1.9Fe0.1SiO4 (Peridot)

FTIR SpectrometerAerosol generator

Exp. 1 Large Particles (SiO2 monosphere)

0

0.2

0.4

0.6

0.8

1

1.2

7 8 9 10 20

d~5.0 µm

Nor

mal

ized

Ext

inct

ion

Wavelength [µm]

Amo-SiO2 (Aerosol)

d=0.5 µm8.93 µm

Peaks 8.93 µm 9.49 µm ∆λ=0.56 µm20.82 µm 21.21 µm ∆λ=0.39 µm

Log

9.49 µm

Crystalline forsterite (Mg2SiO4)

0

0.2

0.4

0.6

0.8

1

1.2

8 9 10 20 30 40

Large particlesSmall particles

Nor

mal

ized

Ext

inct

ion

Log wavelength [µm]

10.159.87

11.1511.08

16.3016.15

19.2419.03

27.2427.27

23.2223.25

33.6533.19

• Size Larger Longer wavelength• The extinction strength of the 23 µm peak is enhanced.

0

500

1000

1500

2000

2500

3000

3500

8 9 10 20 30 40

Small particlesLarge particles

κ (c

m2 /g

)

Log wavelength [µm]

CsI 1:500

Exp. 2 Mixture

0

0.2

0.4

0.6

0.8

1

7.5 8 8.5 9 9.5 10 10.5 11

1:301:501:801:120

Qex

t(λ) /

Qex

t(Pea

k)

Wavelength [µm]

Peaks (d=0.5µm (8.93µm) & d=5.0µm (9.49µm))1:30 8.94µm 9.47µm1:50 8.95µm 9.45µm1:80 8.96µm 9.42µm1:120 8.97µm 9.47µm

(a) (b)

(c)

(d)

Nor

mal

ized

Ext

inct

ion

(monosphere vs. monosphere)

1:80 Mag. 2K 1:80 Mag. 10K

1:120 Mag. 1K

1:120 Mag. 7K

Exp. 2 Mixture ( monosphere vs. irregular shape)

0

0.2

0.4

0.6

0.8

1

7.5 8 8.5 9 9.5 10 10.5 11

1:30 (Sphere)1:50 (Sphere)1:30 (Irregular)1:50 (Irregular)

Qex

t(λ) /

Qex

t(Pea

k)

Wavelength [µm]

Peaks (d=0.5µm (8.93µm) & d=5.0µm (9.49µm)) Sphere Irregular ∆λ1:30 8.94µm 8.95µm 0.01µm 9.47µm 9.17µm 0.30µm1:50 8.95µm 8.96µm 0.01µm 9.45µm 9.20µm 0.25µm

Nor

mal

ized

Ext

inct

ion

(a) (b)

Survived 0.5 µm particles show a clear peak at 8.94 µm.

The peak at 9.5 µm of the 5 µm particles is strongly influenced by the morphological changes (size & shape).

0.5 µm monosphere : crashed 5 µm particles

Exp. 3 Metal ContentOlivine MgxFeySiO4Mg1.9Fe0.1SiO4 (Peridot)Mg1.6Fe0.4SiO4 (Sri Lanka) Peridot Sri Lanka

0

0.2

0.4

0.6

0.8

1

1.2

8 9 10 20 30 40

Peridot (Mg1.95

Fe0.16

Si0.91

O4)

Sri Lanka (Mg1.56

Fe0.40

Si0.91

O4)

Nor

mal

ized

Ext

inct

ion

Wavelength [µm]

9.899.94

11.0711.11

19.0419.04

16.2416.38

23.4623.91

33.5033.97

(Aerosol measurements)

More Fe

spectra shift to longer WL

(Koike et al. 2003)

PeridotSri Lanka

Astronomical ApplicationName Spec.

typeDistance (pc)

Age (Myr)

planets IRS spectrum

Remarks Reference

β Pic A5 V 19 12 + + - Chen `07

η Tel A0 V 48 12 - + Wide binary Chen `06

HD 113766 F3/5 V 130 10-16 - + Wide binary Lisse ̀ 08

HD 191089 F5 V 54 ? - + - Chen `08

BD+20307 G0 V 92 >300 - + Close binary Zuckerman `08

HD 69830 K0 V 13 2000 ? 3 (<0.7AU) + - Lisse ̀ 07

ς Lep A2 V 22 300 - + - Chen `06

HD 12039 G3/5 V 42 100 ? - ? - -

HD 86087 A0 V 98 50 - + - Chen `06

η Crv F2 V 18 600 ? - + - Chen `06

HD 181327 F5/6 V 51 ? - + - Chen `06

HD 72905 G1.5 V 14 400 - + - Beichman `06

HD 172555 A5 IV/V 29 12 - + Wide binary Lisse ̀ 09

HD 145263 F0 V 116 ? - - - Honda `04

HD 110058 A0 V 100 16 - + - Chen `06

β Leo A3 V 11 50 - - binary Chen `06

HD 23514 F6 V 120 100 - - ? - Rhee `08

HD 169666 F5 51 2200 - + - Abraham `08

HD 202406 F2 IV/V - - - - - Smith `08

HD 106797 A0 V 96 >10 - - - Fujiwara ̀ 09

Some main sequence stars known to have warm dust

Warm dust thermal emission: Infrared excess

About 30 main sequence stars known to have warm-dust emission (i.e. at 10-50 µm)

a few objects have strong emission at λ ~10 µm (requires T > 200K)

Best for mineralogical analysis

Spitzer IRS Spectra – what can be infered from them?

Small grains dominating (surprising because of blowout limits, but biased by selection)

Arrows: am. Silicate band, cryst. Olivine bands, pyroxenes/SiO2

Crystalline olivine bands vs. 9.8 µm amorphous silicate band Short-wavelength wing (9.3µm) indicates presence of pyroxenes or SiO2 Grain sizes: large grains (> 1 µm) should broaden 10 µm band to longer

wavelengths lack of small grains should flatten the band strongly

Large grains

Example of a large-grain emission spectrum

Moor et al. 2009

Lisse et al. 2009 (accepted)

HD 172555 (A5 star) 40 % silica and tektite, 35% olivines and pyroxenes, SiO gas, hypervelocity collision !

Detailed mineralogical analyses (Lisse et al.)Spitzer HD 113766 Disk Spectral Model

HD 113766 (F3/5 binary) 29 % olivines, 14 % pyroxenes,

12 % amorphous silicate 21 % FeS, 12 % ice Warm asteroidal dust (T=450K) at

1.8 AU Icy dust at 4-9 and 30-80 AU

Some of the identified minerals do not have distinct bands, they just contribute to continuum emissionand they can hardly be distinguished.

Most of the comparative spectra have been obtained by the embedded dust samples, which make a different band profile.

The grain shape influences the band profiles.

Are such detailed fits possible at all?

Some concerns:

Dominating minerals first, consider grain shape effects (also composition and size effects) on sharp bands

Use spectra of well-defined particulates (known shape and size) dispersed in N2-gas – aerosol spectra

Our approach:

First steps: Some own fitting tests for HD 69830 – IDL programme IRSA by Lutz Bornschein

Three-component model:• Fo60+ olivine (synthetic, Koike)• Astronom. Silicate (amorphous, D&L93)• „carbon“ cold dust

Intention: simplification, search for real „hard facts“ (constraints on dust mixture)

Used dust spectra:- olivines and pyroxenes from aerosol

measurements- Koike et al. (2003) olivines- amorphous silicates- „carbon“ for grey emission (may also

mimic colder dust component)

100 % Mg

60% Mg+ Large forsterite

Some own fitting tests for HD 69830

K0 V13 pc

2000 Myr ?3 planets

• Beichman et al. (2005): Spitzer IRS• T(dust) ~ 400K• M(dust) ~ 1000 zodi

Best fit for HD69830

[µm]

Database

http://elbe.astro.uni-jena.de

Peak Search: to search for your desired spectra via inputting a peak position.

Basic information: Chemical formula, density, particle size, product info., etc…

Plot: Aerosol and CsI pellet measurements. Numerical data sets are downloadable.

Images: obtained by scanning and transmission electron microscopes (SEM & TEM)

Created by David Schmitz

Summary<Large-sized particles vs. Small-sized particles (Forsterite)> Peak positions are not strongly influenced by the particle size, but

the bandwidth beyond the 11 µm peak is significantly broadened.

<Mixture of Small- and Large-sized particles (SiO2)> Even mass of the 0.5 µm monosphere particles is 120x less than

that of the 5 µm ones, the peak of the 0.5 µm particles clearly appears on the extinction spectrum.

<Metal Content in Olivine> Peaks of Fe-rich olivine take place at slightly longer WL in aerosol

measurements as well.

<HD69830> Emission spectrum of HD69830 fits well with Fe-rich olivine

spectrum obtained by aerosol measurement.