Marco La Cognata INFN-LNS Catania

1. FLUORINE DESTRUCTION IN STARS: FIRST EXPERIMENTAL STUDY OF THE 19F(p,α0)16O REACTION AT ASTROPHYSICAL ENERGIES Marco La Cognata INFN-LNS Catania

2. Origin of the heavyelements Elements heavier than iron are produced by neutron capture. Slow neutron capture (s-process) is responsible of the nucleosynthesis along the stability valley AsymptoticGiantBranch (AGB) stars are the mains-process site. 13C(a,n)16O, active in the intershell, is the mainneutron source in low-mass AGB stars Relative abundances

3. Fluorine in astrophysics S-nuclei are produced and brought to the surface thanks to mixing phenomena, together with fluorine that is produced in the same region from the same n-source. 19F is a key isotope in astrophysics as it can be used to probe AGB star mixing phenomena and nucleosynthesis. But its production is still uncertain! In AGB stars, the largest observed fluorine overabundances cannot be explained with standard AGB models. A possible lack of proper accounting for the contribution from C-bearing molecules might provide an interpretation in Population I stars (Abia et al. 2010).

4. Fluorine in astrophysics S-nuclei are produced and brought to the surface thanks to mixing phenomena, together with fluorine that is produced in the same region from the same n-source. 19F is a key isotope in astrophysics as it can be used to probe AGB star mixing phenomena and nucleosynthesis. But its production is still uncertain! In the case of metal poor AGB stars our understanding is far from satisfactory (Lucatello et al. 2011, Abia et al 2011). We note that a significant fraction of the upper limits are located under the predicted lines (Lucatello et al. 2011)

5. The 19F(p,a0)16O reaction The S(E) factor shows several resonances around 1 MeV. Breuer (1959) claimed the occurrence of two resonances at around 400 keV Unpublished data (Lorentz-Wirzba 1978) suggests that no resonance occurs a0 is the dominant channel in the energy region of astrophysical interest Below Ecm = 460 keV, where data do not exist, a non resonant contribution is calculated for s-capture. The S-factor was adjusted as to the lower experimental points between 460 and 600 keV

6. direct break-up a s x c A C 2-body process Indirect Techniques: the THM Coulomb barrier exponentialdamping of the cross sectionatastrophysicalenergies + electron screening  low-energy, bare-nucleus cross sectionisexperimentallyavailableonlythroughextrapolationOR indirectmeasurements From A+a(xs)  c+C+s@ 10-60 MeV THM reaction A + x  c + C @ 5-20 keV By selecting the QF contribution Though EA >> VCoulitispossible to measureat the Gamowpeaksince: Ec.m.=EA-x-Qx-s Additional advantages: ● reduced systematic errors due to straggling, background… ● magnifying glass effect But... ● off-shell cross section deduced (x  virtual particle) ● no absolute units

7. direct break-up a s x c A C 2-body process Indirect Techniques: the THM Coulomb barrier exponentialdamping of the cross sectionatastrophysicalenergies + electron screening  low-energy, bare-nucleus cross sectionisexperimentallyavailableonlythroughextrapolationOR indirectmeasurements From A+a(xs)  c+C+s@ 10-60 MeV THM reaction A + x  c + C @ 5-20 keV By selecting the QF contribution Though EA >> VCoulitispossible to measureat the Gamowpeaksince: Ec.m.=EA-x-Qx-s Additional advantages: ● reduced systematic errors due to straggling, background… ● magnifying glass effect See C. Spitaleri talk for more details But... ● off-shell cross section deduced (x  virtual particle) ● no absolute units

8. The THM forresonancereactions In the “Trojan Horse Method” (THM) the astrophysically relevant reaction, say A(x,c)C, is studied through the 23directprocess A(a,cC)s: s a x c A F* C Upper vertex: direct abreakup Mi(E) is the amplitude of the transfer The process is a transfer to the continuum where x is participant, e.g. a proton or an alpha particle and s is the spectator Standard R-Matrix approach cannot be applied to extract the resonance parameters of the A(x,c)C reaction because x is virtual  Modified R-Matrix is introducedinstead (A. Mukhamedzhanov 2010) In the case of a resonant THM reaction the cross sectiontakes the form

9. Theexperiment • d: Trojanhorsenucleusp+n B=2.2 MeV |ps|=0 MeV/c n: spectator p: participant d(19F,16O)n Q3b=5.889 MeV 19F(p,) 16O Q2b=8.113 MeV (150 µg/cm2 ) Butane gas  52mbar

10. Theexperiment The experimentwasperformedat INFN-LNS Catania Beam: 19F@50 MeV Only in the coincidence 1-5 the a0channelwaspopulatedsignificantly  a1channel in the 1-4 n d p a 19F 20Ne* 16O

11. Selection of the reactionchannel • Kinematic locus of the reactions: • 2H(19F,116O)n • 19F(p,)16O • 2H(19F,016O)n • 19F(12C,15N)16O Three-body reaction in the exit channel -> coincidenceeventsgive rise to a kinematic locus. Identification of theparticles with Z=8 through the E-E tecnique.

12. Selection of the reactionchannel • Kinematic locus of the reactions: • 2H(19F,116O)n • 19F(p,)16O • 2H(19F,016O)n • 19F(12C,15N)16O Full Q-valuespectrum, imposingthatZ=8 nuclei are detected in the telescope. Furthercuts are needed to remove background reactions Identification of theparticles with Z=8 through the E-E tecnique.

13. Selection of the reactionchannel • 0 channelStatisticallysignificantonlycoincidence PSD1-PSD5. (3) (4) (1) • Q-value of the three-body reactionislinked to the mass of the particles and not to anykinematicalvariable. (2) (3) (4) From energyconservation: Ebeam+Q=E1+E2+E3 Y=Ebeam-E2-E3 x=p32/2u • Events from d(19F,16O)n must gather along a straigth line (1) (2)

14. Selection of the reactionchannel • 0 channelStatisticallysignificantonlycoincidence PSD1-PSD5. Comparisionof the experimentalQ-valuewith the theoreticalone • Kinematic locus obtained by introducingall the mentionedcuts • Good agreement confirms the accuracy of the performed calibration

15. Selection of the QF reactionmechanism Experimentalstudy of the neutronmomentumdistribution inside deuteron. If the processisdirect, neutronshouldkeep the samemomentumas inside deuteron the momentumdistributionbears information about the kind of process Theoreticaldistribution: Hulthenfunction in momentumspace(redline): a=0.2317 fm-1 b=1.202 fm-1

16. If a levelispopulated-> increase in reactionyieldin correspondenceofitsexcitationenergy. (a) Manyhorizontal loci appear -> presence of sequentialmechanismcorresponding to the population of excitedstates of 17O. 2H+19F -> 17O* + 4He -> 16O+n+4He Vertical loci -> excitedstatesof20Ne -> sequential or QF (b) No levelfrom5He sequentialdecay are present in the experimentalplots.

17. Selection of the QF reactionmechanism Experimentalstudy of the neutronmomentumdistribution inside deuteron. If the processisdirect, neutronshouldkeep the samemomentumas inside deuteron the momentumdistributionbears information about the kind of process Theoreticaldistribution: Hulthenfunction in momentumspace(redline): a=0.2317 fm-1 b=1.202 fm-1 Onlyevents with0<pn<40 MeV/c will be selected.

18. Observed20Ne states Normalized coincidence yield obtained by dividing the selected coincidence yield by the product of the phase-space factor and of the p−n momentum distribution Three resonance groups corresponding to 20Ne states at: 12.957and 13.048 MeV; 13.222, 13.224, and 13.226 MeV; 13.529, 13.586, and 13.642 MeV. The normalized yield was fitted simultaneously with four Gaussian curves to separate the resonance contributions

19. Observed20Ne states Resonance at lowerenergy, nextto the regionofastrophysical interest, withimportantconsequences on the fluorine destruction

20. Observed20Ne states Resonancewithspin 4 and angular moment 4  suppressed in the directmeasurement due to the centrifugalbarrier

21. Observed20Ne states The dominantresonanceis the one with spin 3 therefore, as in the previous case, we do notexpect a contribution in the directmeasurement due to the centrifugalbarrier.

22. Observed20Ne states Theseresonanceshavealreadybeenmeasured and are thereforeusedfor the normalization

23. Observed20Ne states Theseresonanceshavealreadybeenmeasured and are thereforeusedfor the normalization

24. Fittingthe THM crosssection QF cross section of the 2H(19F,α016O)n reaction in arbitrary units. horizontal error bars  p−19F-relative-energy binning. The vertical error bars  statistical and angular-distribution integration uncertainties. Red band  the cross section calculated in the modified R-matrix approach, normalized to the peak at about 750 keV and convoluted with the experimental resolution.

25. The 19F(p,a)16O cross section R-matrix parameterization of the 19F(p,α0)16O astrophysical factor. Above 0.6 MeV, the reduced partial widths were obtained through an R-matrix fit of direct data Below 0.6 MeV, the resonance parameters were obtained from the modified R-matrix fit The non-resonant contribution is taken from NACRE (1999). Because of spin-parity, only the resonance at 12.957 MeV provide a significant contribution Gamow window: 27-94 keV this level lies right at edge of the Gamow window for extramixing in AGB stars

26. 19F(p,a)16O reaction reaction rate Reaction rate for the 19F(p,α0)16O. Upper and lower limits are also given, though they are barely visible because of the large rate range. Ratio of the reaction rate in panel (a) to the rate of the 19F(p,α0)16O reaction evaluated following the prescriptions in NACRE (1999). The red band arises from statistical and normalization errors.  A reaction rate enhancement up to a factor of 1.8 is obtained close to temperatures of interest for astrophysics (for instance, T9~0.04 in AGB stars)

27. Perspectives The contribution of the a0 channel only has been addressed since this is currently regarded as the dominant one at temperatures relevant for AGB stars  Spyrou et al. (2000) a1 channel is even more uncertain! An experimental campaign is scheduled to extract the cross section for the a1 channel and to improve the spectroscopy of the resonances discussed here.