Astrophysical, observational and nuclear-physics aspects of r-process nucleosynthesis

T1/2P

nS

nsng

Sinaia, 2012

Karl-Ludwig Kratz

Part II

Summary “waiting-point” model

“weak” r-process

“main” r-process (early primary process; SN-II?)

superposition of nn-components

(later secondary process; explosive shell burning?)

seed Fe (implies secondary process)

Kratz et al., Ap.J. 662 (2007)

…largely site-independent!T9 and nn constant;instantaneous freezeout

To recapitulate…

Importance of nuclear-structure input !

Now…

from historical, site-independent “waiting-point” approaches

to more recent core-collapse SN nucleosynthesis calculations

Second part

r-Process calculations with MHD-SN models

2012 … new “hot r-process topic” magnetohydrodynamic SNe … but, nothing learnt about nuclear-physics input ???

“We investigate the effect of newly measured ß-decayhalf-lives on r-process nucleosynthesis. We adopt … a magnetohydrodynamic supernova explosion model… The (T1/2) effect slightly alleviates, but does not fullyexplain, the tendency of r-process models to underpro-duce isotopes with A = 110 – 120…”

“We examine magnetohydrodynamically driven SNe as sources of r-process elements in the early Galaxy… … the formation of bipolar jets could naturally providea site for the strong r-process…”

Ap.J. Letter 750 (2012)

Phys.Rev. C85 (2012)

In both cases FRDM masses have been used partly misleading conclusions

The neutrino-driven wind starts from the surface of the proto-neutron starwith a flux of neutrons and protons.

As the nucleons cool (≈10 ≥ T9 ≥ 6), they combine to α-particles + an excess of unbound neutrons.

Further cooling (6 ≥ T9 ≥ 3) leads to the formation of a few Fe-group "seed"nuclei in the so-called α-rich freezeout.

Still further cooling (3 ≥ T9 ≥ 1) leads to neutron captures on this seed compo-sition, making the heavy r-process nuclei. (Woosley & Janka, Nature, 2005)

The high-entropy / neutrino-driven wind model

Nevertheless, cc-SN “HEW” …still one of the presently favoured scenarios for a rapid neutron-capture nucleosynthesis process

• time evolution of temperature, matter density and neutron density

• extended freezeout phase

“best” nuclear-physics input (Mainz, LANL, Basel)

Three main parameters:

electron abundance Ye = Yp = 1 – Yn

radiation entropy S ~ T³/rexpansion speed vexp durations ta and tr

(Farouqi, PhD Mainz 2005)

The Basel – Mainz HEW model

• nuclear masses• β-decay properties• n-capture rates• fission properties

full dynamical network (extension of Basel model)

First results of dynamical r-process calculations

• Conditions for successful r-process „strength“ formula

• Neutron-rich r-process seed beyond N=50 (94Kr, 100Sr, 95Rb…) avoids first bottle-neck in classical model

• Initial r-process path at particle freezeout (T9 3) Yn/Yseed ≤ 150

• Total process duration up to Th, U τr ≤ 750 ms

instead of 3500 ms in classical model • Due to improved nuclear-physics input (e.g. N=82 shell quenching!) max. S 280

to form full 3rd peak and Th, U • Freezeout effects (late non-equilibrium phase) capture of “free” neutrons

recapture of bdn-neutrons

Baryonk

Baryonkmk

YSVk

YY

BSN

eExpSN

Seed

n

111

3

108

,parameters correlated

p

α

seed

Formation of r-process “seed”

Time evolution of temperature and densityof HEW bubble(Vexp=10,000 km/s)

extended “freeze-out” phase!

temperature

density

Recombination of protons and neutrons into α-particlesas functions of temperature and time

For T9 ≤ 7 α dominate;at T9 ≈ 5 p disappear, n survive, “seed“ nuclei emerge.

n

(Farouqi et al., 2009)

Distribution of seed nuclei

effect of different expansion velocities of the hot bubble, Vexp

distributions are robust !

effect of different electron abundances, Ye

distributions show differences,mainly for A > 100

Main seed abundances at A > 80 lie beyond N = 50 !

Parameters HEW model Y(Z)

αn

seed

Ye=0.45

No neutrons no n-capture r-process!

Nucleosynthesis components:

S ≤ 100; Yn/Yseed < 1charged-particle process

100 < S < 150; 1 < Yn/Yseed< 15“weak” r-process

150 < S < 300; 15 < Yn/Yseed< 150“main” r-process

Synthesizing the A=130 and A=195 peaks

S=196

Process duration: 146 ms

S=261

Process duration: 327 ms

Pb

UTh

not yet enough

Abun

danc

e Y

full Nr, distribution superposition of “weighted” S-components

Ye S τ[ms]

0.49 290 4120.45 260 3270.41 235 213

Ye S τ[ms]

0.49 230 2340.45 195 1460.41 160 109

Reproduction of Nr,

Superposition of S-components with Ye=0.45; weighting according to Yseed

No exponential fit to Nr, !

Process duration [ms]Entropy S FRDM ETFSI-Q Remarks

150 54 57 A≈115 region180 209 116 top of A≈130 peak220 422 233 REE pygmy peak245 691 339 top of A≈195 peak260 1290 483 Th, U280 2280 710 fission recycling300 4310 1395 “ “

significant effect of “shell-quenching” below doubly-magic 132Sn

TTTT

TTTT

Width of A=130 and 195 peaks ca. 15 m.u.;width of REE pygmy peak ca. 50 m.u.;

Widths of all three peaks (A=130, REE and A=195) DS40 kB/Baryon

r-Process robustness due to neutron shell closures

Kratz, Farouqi, Ostrowski (2007)

Freezeout at A ≈ 130

Validity and duration of (n,g) ↔ (g,n) equilibriumSn(real) / Sn(n, g ) 1

freezeout

Validity and duration ofβ-flow equilibriumλeff(Z) x Y(Z) ≠ 0

(n, g )-equilibrium valid over 200 ms,independent of Z

freezeoutLocal equilibrium “blobs” withIncreasing Z, at different times

Definition of different freezeout phasesNeutron freezeout at 140 ms, T9 0.8, Yn/Yr = 1 130PdChemical freezeout at 200 ms, T9 0.7, Yn/Yr 10-2 130CdDynamical freezeout at 450 ms, T9 0.3, Yn/Yr 10-4 130In

For Ye≤0.470 full r-process, up to Th, U

For Ye0.490 still 3rd peak, but no Th, U

For Ye=0.498 still 2nd peak, but no REE

Superposition of HEW components 0.450 ≤ Ye ≤ 0.498

Farouqi et al. (2009)

“weighting” of r-ejecta according to mass predicted by HEW model: for Ye=0.400 ca. 5x10-4 M for Ye=0.498 ca. 10-6 M

„What helps…?“ low Ye, high S, high Vexp

r-process observables

Solar system isotopic Nr, “residuals”

T9=1.35; nn=1020 - 1028

r-Process observables today

Pb,Bi

Observational instrumentation

• meteoritic and overall solar-system abundances

• ground- and satellite-based telescopes like Imaging Spectrograph (STIS) at Hubble, HIRES at Keck, and SUBARU

• recent „Himmelsdurchmusterungen“ HERES and SEGUE

Elemental abundances in UMP halo starBD+17°3248

Detection of 33 n-capture elements, the most in any halo star

Observations I: Selected “r-enriched” UMP halo stars

Sneden, Cowan & GallinoARA&A, 2008

Same abundance pattern at the upper end and ??? at the lower end

Two options for HEW-fits to UMP halo-star abundances:• const. Ye = 0.45, optimize S-range;• optimize Ye with corresponding full S-range.

Halo stars vs. HEW-model: Extremes “r-rich” and “r-poor”

Eu

Eu

Factor 25 difference for Sr - Zr region !

g g

gg

Elemental abundance ratios UMP halo stars

r-rich “Cayrel star” / r-rich “Sneden star” r-poor “Honda star” / r-rich “Sneden star”

normalized to Eu

r-poor “Honda star”

10 ≤ S ≤ 220incomplete main r-process

35% Nr,(Eu)

r-rich “Sneden star”

140 ≤ S ≤ 300

100% Nr,(Eu)

full main r-process

Extremes “r-rich” and “r-poor”: S-range optimized

Sr – Cd region underabundant by a mean factor ≈ 2 relative to SS-r

(assumption by Travaglio et al. that this pattern is unique for all UMP halo stars)

Sr – Cd region overabundant by a mean factor ≈ 8 relative to SS-r

“missing” part to SS-r = LEPP

Halo stars vs. HEW-model: Y/Eu and La/Eu

(I. Roederer et al., 2010; K. Farouqi et al., 2010)

Instead of restriction to a single Ye with different S-ranges,

probably more realistic, choice of different Ye’s with corresponding full S-ranges

39Y represents charged-particle component (historical “weak” n-capture

r-process)

57La represents “main” r-process

Clear correlation between “r-enrichment” and Ye

Caution!La always 100 % scaled solar; log(La/Eu) trend correlated withsub-solar Eu in “r-poor” stars

“… the Ge abundances…track their Fe abundances very well.An explosive process on iron peak nuclei (e.g. the a-rich freezeout inSNe), rather than neutron capture,appears to have been the dominantmechanism for this element…”

“Ge abundance seen completelyuncorrelated with Eu”

Ap.J., 627 (2005)

solar abundance ra

tio

Ge – Eu corre

lation

Zr – Eu co

rrelation

Correlation between the abundance ratios of [Zr/Fe] (obtained exclusivley with HST STIS) and [Eu/Fe]. The dashed line indicates a direct correlation betweenZr and Eu abundances.

Correlation between the abundance ratios [Ge/Fe] and [Eu/Fe]. The dashed line indicates a direct correlation between Ge and Eu abundances. As in the previous Figure, the arrow represents the derived upper limit forCS 22892-052. The solid green line at [Ge/Fe]= -0.79is a fit to the observed data. A typical error is indicated by the cross.

Relative abundances [Ge/H] displayed as a function [Fe/H] metallicity for our sample of 11 Galactic halo stars. The arrowrepresents the derived upper limit for CS 22892-052. The dashed line indicates the solar abundance ratio of these elements: [Ge/H] = [Fe/H], while the solid green line shows the derived correlation [Ge/H]=[Fe/H]= -0.79. A typical error is indicated by the cross.

Observations: Correlation Ge, Zr with r-process?

Can our HEW modelexplain observations ?

“Zr abundances also do not vary cleanly with Eu”

Ge okay! 100% CPRZr two components?

ELEMENT Y(Z) as fct. of S in %20 ≤ S ≤ 100 100 ≤ S ≤ 160 160 ≤ S ≤ 280

38Sr 98 2.2 0.2

39Y 98 2.2 0.2

40Zr 82 18 0.3

46Pd 3.6 82 14

47Ag 0.6 77 23

52Te 0.001 10 90

56Ba - - 100

HEW with Ye =0.45; vexp =7500

CP componentuncorrelated

weak-r componentweakly correlated

main-r componentstrongly correlated

n-rich α-freezeoutnormal α -rich freezeout

with Eu ?

Relative elemental abundances, Y(Z)

From Sr via Pd, Ag to Ba different nucleosynthesis modes

Halo stars vs. HEW model: Sr/Y/Zr as fct. of [Fe/H], [Eu/H] and [Sr/H]

Robust Sr/Y/Zr abundance ratios,independent of metallicity, r-enrichment, α-enrichment.

Same nucleosynthesis component:CP-process, NOT n-capture r-process

Correlation with main r-process (Eu) ?

–— HEW model (S ≥ 10); Farouqi (2009) – – average halo stars; Mashonkina (2009)

Zr in r-poor stars overabundant type “Honda star”Zr in halo & r-rich stars underabundant type “Sneden star”

Halo, r-rich and r-poor stars are clearly separated!

Halo stars vs. HEW model: Zr/Fe/Eu vs. [Eu/Fe], [Fe/H] and [Eu/H]

SS

Mashonkina (2009); Farouqi (2009)

Strong Zr – Eu correlationSS diagonal

HEW (av.)

r-poor

r-rich

Zr – Eu co

rrelation

Cowan et al. (2005)

similar metallicity different r- enrichment

ELEMENT Y(Z) as fct. of S in %20 ≤ S ≤ 100 100 ≤ S ≤ 160 160 ≤ S ≤ 280

38Sr 98 2.2 0.2

39Y 98 2.2 0.2

40Zr 82 18 0.3

46Pd 3.6 82 14

47Ag 0.6 77 23

52Te 0.001 10 90

56Ba - - 100

HEW with Ye =0.45; vexp =7500

CP componentuncorrelated

weak-r componentweakly correlated

main-r componentstrongly correlated

n-rich α-freezeoutnormal α -rich freezeout

with Eu ?

Relative elemental abundances, Y(Z)

From Sr via Pd, Ag to Ba different nucleosynthesis modes

Halo stars vs. HEW-model predictions: Pd

average Halo stars -1.5 < [Eu/H] < -0.6 log(Pd/Sr) - 0.95(0.09)

HEW model log(Pd/Sr) = - 0.81 (“r-rich”) - 1.16 (“r-poor”)

average Halo stars -1.5 < [Eu/H] < -0.6 log(Pd/Eu) +0.81(0.07)

HEW model log(Pd/Eu) = +1.25 (Pd CP+r) +0.89 (“r-rich”) +1.45 (”r-poor”)

r-poor stars ([Eu/H] < -3) indicate TWO nucleosynthesis components for 46Pd: Pd/Sr uncorrelated, Pd/Eu (weakly) correlated with “main” r-process

Pd relation to CP-element Sr

Pd relation to r-element Eu

Pd/Sr and Pd/Eu different from Sr/Y/Zr

Observations of Pd & Ag in giant and dwarf stars

C. Hansen et al.; A&A in print (2012)

indication of different production processes (Sr – charged-particle, Ag – weak-r, Eu – main r-process)

anticorrelation between Ag and Sr anticorrelation between Ag and Eu

correlation of Pd and Ag

Pd and Ag are produced in the same process (predominantly) weak r-process

All observations in agreement with our HEW predictions !

CS31082

CS22892

Observations: [Ag/Fe] vs. [Eu/Fe]

BD+17°3248

HD221170

HD186478

HD108577

HD88609

HD122563

“r-poor” “r-rich”

Selection of pure r-process stars; no measurable s-process “contamination”

SAGA data base: [X/Fe] vs. [Eu/Fe] (I)

Zn Sr

Zr

SS SS

SSSS

r-poor r-rich

Transition from CP-component to “weak” n-capture r-process

Ag

Th

SAGA data base: [X/Fe] vs. [Eu/Fe] (II)

m1

Ir

m1

Dy

m1

SS

…Th – U – Pb cosmochronology ?!

The SS A ≈ 130 r-peak

r-Process observables today

Observational instrumentation

• meteoritic and overall solar-system abundances

• ground- and satellite-based telescopes like Imaging Spectrograph (STIS) at Hubble, HIRES at Keck, and SUBARU

• recent „Himmelsdurchmusterungen“ HERES and SEGUE

From elemental abundances to isotopic yields …;More sensitive to nuclear physics input, here in the 120 < A < 140 regionXe-H shows strong depletion of lighter isotopes with s-only 128, 130Xe “zero”.

r-process observables

Xe in nanodiamond stardust (I)

(U. Ott)

Xe in nanodiamond stardust (II)

(U. Ott)

(U. Ott)

Xe in nanodiamond stardust (III)

The “neutron-burst“ model (I)

The “neutron-burst” modelis the favored nucleosynthesis scenario in the cosmochemistry

community;so far applied to isotopic abundances of Zr, Mo, Ru, Te, Xe, Ba, Pt

Historical basis: …neutron burst in shocked He-rich matter in exploding massive stars first ideas by Cameron Howard et al.; Meteoritics 27 (1992) Rauscher et al.; Ap.J. 576 (2002)

Several steps:1) start with SOLAR isotope distribution2) exposure to weak neutron fluence (t = 0.02 mbarn-1)

mimics weak s-processing during pre-SN phase3) s-ashes heated suddenly to T9 = 1.0 by SN shock4) expansion and cooling on 10 s hydrodynamical timescale

resulting n-density (from (a,n)-reactions) during burst ≈ 1017 n/cm3

for ≈ 1 s final n-exposure 0.077 mbarn-1

shift of post-processed isotope pattern towards higher A

strong time and temperature dependence

Howard‘s Xe-H abundances N* are pure, unmixed yields;in addition, arbitrary mix with 5.5 parts Nּס => N-H

N-H = (N* + 5.5Nּס)/5.5N 0.182 + 1 = ּס N*/Nּס

N-H = 1 s-only 128,130Xe totally eliminated

…however,“there is little physical reasonto admix N* with solar abun-dances,other than the tradition ofinterpreting isotopic anomaliesas a binary admixture with SSabundances.“

arbitrary Nּס-mixing somewhat too strong!

The “neutron-burst“ model (II)

Two SN models a) site-independent “waiting-point“ approach b) high-entropy-wind ejecta of core-collapse SN-II

full parameter space in a) nn-components b) S-ranges

No reproduction of Xe-H !

special variants of r-process?

• “hot“ • “cold“ r-process• “strong”

Xe in SN-II

The A 130 SS r-abundance peak

HEW reproduction of the solar-system A 130 r-abundance peak

Nuclear-physics input: ETFSI-Q; updated QRPA(GT+ff); experimental data

Ye = 0.45

“cold” r-processVexp = 20000; S(top) 140peak top at A 126

“hybrid” r-processVexp = 7500; S(top) 195peak top at 128 < A < 130

“hot” r-processVexp = 1000; S(top) 350peak top at A 136as, e.g. in Arcones & Martinez-Pinedo, Phys. Rev. C83 (2011)

Successive “cuts“ of low-S components

exclusion of “weak“ r-process

“strong“ r-process !

Xe-H – HEW “old“

“old“ nuclear physics input

e.g. 129Ag T1/2(g.s.) = 46 ms; 130Pd T1/2 = 98 ms; P1-2n = 96 %; 4 %

Experiment: T1/2 = 68 ms; Pn = 3.4 %

Surprising b-decay properties of 131Cd

QRPA predictions: T1/2(GT) = 943 ms;

Pn(GT) = 99 %

T1/2(GT+ff) = 59 ms;

Pn(GT+ff) = 14 %

Remember: …just ONE neutron outside N=82 magic shell

Nuclear-structure requests: higher Qβ, lower main-GT, low-lying ff-strength “new” nuclear-physics input for A ≈ 130 peak region

…but, still „cuts“ of low-S region necessary

again, exclusion of “weak“ r-process “strong“ r-process !

129Xe, 131Xe, 134Xe okay;132Xe still factor ≈ 2 too high

Xe-H – HEW “new“

“new“ nuclear physics inputoptimized to measured T1/2, Pn and decayschemes of 130Cd and 131Cd

e.g. 129Ag stellar-T1/2 = 82 ms; 130Pd T1/2(new/old) ≈ 0.2; P1-2n(new/old) ≈ 0.8; 4.5• reduction of 129Xe• increase of 136Xe

128 130 132 134 136Mass number A

0.0

0.5

1.0

1.5

2.0

i Xe/

136 X

e

Xe-HHEW total SHEW S>190HEW S>200HEW S>210

128 130 132 134 136Mass number A

0.0

0.5

1.0

1.5

2.0

i Xe/

136 X

e

128 130 132 134 136Mass number A

0.0

0.5

1.0

1.5

2.0

i Xe/

136 X

e

128 130 132 134 136Mass number A

0.0

0.5

1.0

1.5

2.0

i Xe/

136 X

e

128 130 132 134 136Mass number A

0.0

0.5

1.0

1.5

2.0

i Xe/

136 X

e

…same HEW r-process conditions as Xe-H !

towards Rauscher & Nʘ L'09

towards HEW

Pt-H in nanodiamonds

… at the end, an easy case for HEW SN-II “cold“ r-process !recent AMS measurement by Wallner et al.

Summary

Still today• there is no selfconsistent hydro-model for SNe, that provides the necessary

astrophysical conditions for a full r-process

• parameterized dynamical studies (like our HEW approach) are still useful to explain r-process observables;

• astronomical observations & HEW calculations indicate that SS-r and UMP halo-star abundance distributions are superpositions of 3 nucleosynthesis components: charged-particle, weak-r and main-r

• the yields of the CP-component (up to Zr) are largely uncorrelated with the “main” r-process;

• the yields of the weak-r component (Mo to Cd) are partly correlated with the “main” r-process; elements ≥ Te belong to the “main” r-process

• HEW can explain the peculiar isotopic anomalies in SiC-X grains and nanodiamond stardust

Therefore,

K. Farouqi (Mainz)B. Pfeiffer (Mainz & Gießen)O. Hallmann (Mainz)U. Ott (Mainz)P. Möller (Los Alamos)F.-K. Thielemann (Basel)W.B. Walters (Maryland)

L.I. Mashonkina (RAS, Moscow)C. Sneden & I. Roederer (Austin, Texas)A. Frebel (MIT, Cambridge)N. Christlieb (LSW, Heidelberg)C.J. Hansen (LSW, Heidelberg)

Main collaborators:

Discussion

???

Neutron star mergers

Pn effects at the A≈130 peak

Significant differences

(1) smoothing of odd-even Y(A) staggering

(2) importance of individual waiting-point nuclides, e.g. 127Rh, 130Pd, 133Ag, 136Cd

(3) shift left wing of peak

in clear contrast to Arcones & Martinez-PinedoPhys. Rev. C83 (2011)

Halo stars vs. HEW model: “LEPP” elements

HEW (10 < S < 280)WP (Fe seed; 1020 < nn < 1028)weak-r (Si seed; nn 1018)

LEPP-abundances vs. CP- enrichment (Zr)

HEW reproduces high-Z LEPP observations (Sr – Sn); underestimates low-Z LEPP observations (Cu – Ge)

additional nucleosynthesis processes ?(e.g. Fröhlich et al. np-process; Pignatari et al. rs-process; El Eid et al. early s-process)

(Farouqi, Mashonkina et al. 2008)

Th/U – a robust and reliable r-process chronometer?

Neutron-density range [n/cm3]

r-C

hron

omet

er a

ge [G

yr]

norm.13.2 ± 2.8

Th/U robust – yesreliable – no

(not always)

Correlate Th, U with other elements, e.g. 3rd peak, Pb !

How about Pb as r-chronometer ?

In principle, direct correlation of Pb to Th, U ≈ 85% abundance from α-decay from actinide region

Sensitivity to age:Pb

0 Gyr Y(Pb) = 0.375 (Mainz) // Sneden et al. (2008) Y(Pb) = 0.622 ?5 Gyr 0.433

13 Gyr 0.451 Th,U/Pb

13 Gyr: Th/Pb ≈ 0.048 U/Pb ≈ 0.007

New HEW results:1) agreement Th/U and Th/Pb with NLTE analysis of Frebel-star (HE 1523-0901; [Fe/H] ≈ -3)2) consistent picture for Th/Ba – Th/U for 4 “r-rich” halo-stars with Ye ≈ 0.45; and for the “actinide-boost” Cayrel-star with Ye ≈ 0.44.

Effect of n-capture rates on A=130 and A=195 peaks

blue curve: <σv>green curve: 1/100 x <σv> red curve: 100 x <σv>

at A=130 peak fast freezeout no significant effect

at A=195 peak slow freezeout significant effect

T-dependent NON-SMOKER rates; ETFSI-Q + AME2003