Improved astrophysical rate for the 18O(p,α)15N reaction by underground measurements

Contents lists available atScienceDirect

Physics

Letters

B

www.elsevier.com/locate/physletb

Improved

astrophysical

rate

for

the

18

O(p,

α

)

15

N

reaction

by underground

measurements

C.G. Bruno

a

,∗

,

M. Aliotta

a

,∗

,

P. Descouvemont

b

,

A. Best

c

,

T. Davinson

a

,

D. Bemmerer

d

,

A. Boeltzig

e

,

f

,

1

,

C. Broggini

g

,

A. Caciolli

h

,

F. Cavanna

i

,

T. Chillery

a

,

G.F. Ciani

e

,

P. Corvisiero

i

,

j

,

R. Depalo

h

,

A. Di Leva

c

,

Z. Elekes

k

,

F. Ferraro

i

,

j

,

A. Formicola

f

,

Zs. Fülöp

k

,

G. Gervino

l

,

A. Guglielmetti

m

,

C. Gustavino

n

,

Gy. Gyürky

k

,

G. Imbriani

c

,

M. Junker

f

,

M. Lugaro

o

,

P. Marigo

g

,

h

_,

_{R. Menegazzo}

h

_,

_{V. Mossa}

p

_,

_{F.R. Pantaleo}

p

_,

_{D. Piatti}

h

_,

_{P. Prati}

i

,

j

_,

K. Stöckel

d

,

q

,

O. Straniero

f

,

r

,

F. Strieder

s

,

T. Szücs

d

,

k

,

M.P. Takács

d

,

q

,

D. Trezzi

m a_{SUPA,SchoolofPhysicsandAstronomy,UniversityofEdinburgh,EH93FDEdinburgh,UnitedKingdom}

b_{PhysiqueNucléaireThéoriqueetPhysiqueMathématique,C.P.229,UniversitéLibredeBruxelles(ULB),B1050Brussels,Belgium} c_{UniversitàdiNapoli“FedericoII”,andINFN,SezionediNapoli,80126Napoli,Italy}

d_{Helmholtz-ZentrumDresden-Rossendorf,BautznerLandstr.400,01328Dresden,Germany} e_{GranSassoScienceInstitute,VialeF.Crispi7,67100L’Aquila,Italy}

f_{INFN,LaboratoriNazionalidelGranSasso(LNGS),67100Assergi,Italy} g_{INFN,SezionediPadova,ViaF.Marzolo8,35131Padova,Italy}

h_{UniversitàdegliStudidiPadovaandINFN,SezionediPadova,ViaF.Marzolo8,35131Padova,Italy} i_{INFN,SezionediGenova,ViaDodecaneso33,16146Genova,Italy}

j_{UniversitàdegliStudidiGenova,ViaDodecaneso33,16146Genova,Italy}

k_{InstituteforNuclearResearch(MTAATOMKI),POBox51,HU-4001Debrecen,Hungary} l_{UniversitàdegliStudidiTorinoandINFN,SezionediTorino,ViaP.Giuria1,10125Torino,Italy} m_{UniversitàdegliStudidiMilanoandINFN,SezionediMilano,ViaG.Celoria16,20133Milano,Italy} n_{INFN,SezionediRoma,PiazzaleA.Moro2,00185Roma,Italy}

o_{KonkolyObservatory,ResearchCentreforAstronomyandEarthSciences,HungarianAcademyofSciences,1121Budapest,Hungary} p_{UniversitàdegliStudidiBariandINFN,SezionediBari,70125Bari,Italy}

q_{TechnischeUniversitätDresden,InstitutfürKern- undTeilchenphysik,ZellescherWeg19,01069Dresden,Germany} r_{OsservatorioAstronomicodiCollurania,Teramo,andINFN,SezionediNapoli,80126Napoli,Italy}

s_{SouthDakotaSchoolofMines&Technology,501E.SaintJosephSt.,SD57701, USA}

a

r

t

i

c

l

e

i

n

f

o

a

b

s

t

r

a

c

t

Articlehistory:

Received23November2018

Receivedinrevisedform9January2019 Accepted11January2019

Availableonline17January2019 Editor:W.Haxton

Keywords:

Stellarhydrogenburning Hydrostaticstellarnucleosynthesis

The18O(p,

α

)15Nreactionaffects thesynthesisof15N,18Oand19Fisotopes, whoseabundancescanbe usedtoprobethenucleosynthesisandmixingprocessesoccurringdeepinsideasymptoticgiantbranch (AGB) stars. We performedalow-background direct measurement ofthe 18_O(p,

_α

₎15_{N reaction}

cross-section atthe Laboratory for Underground Nuclear Astrophysics (LUNA) from center of mass energy

Ec.m.=340 keV down to Ec.m.=55 keV, the lowest energy measured to date corresponding to a

cross-section of less than 1 picobarn/sr. The strength of a key resonance at center of mass energy

Er=90 keVwasfoundtobeafactorof10higherthanpreviouslyreported.Amulti-channel R-matrix

analysisofourandotherdataavailableintheliteraturewasperformed.Overawidetemperaturerange,

T₌0.01–1.00 GK,ournewastrophysicalrateisbothmoreaccurateandprecisethanrecentevaluations. StrongerconstraintscannowbeplacedonthephysicalprocessescontrollingnucleosynthesisinAGBstars withinterestingconsequencesontheabundanceof18_{Ointhesestarsandinstardustgrains,specifically}

ontheproductionsitesofoxygen-richGroup IIgrains.

©2019TheAuthor(s).PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP3.

*

Correspondingauthors.

E-mailaddresses:[email protected](C.G. Bruno),[email protected](M. Aliotta).

1 _{Currentaddress:UniversityofNotreDame,DepartmentofPhysics,225NieuwlandScienceHall,NotreDame,IN46556,USA.}

https://doi.org/10.1016/j.physletb.2019.01.017

0370-2693/©2019TheAuthor(s).PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).Fundedby SCOAP3_.

1. Introduction

The18O(p,

α

)15Nreactioninfluencestheabundancesof15N,18O and19_{Fisotopes[1,2],criticaltoconstrainawidevarietyofstellar}

models. For example, the O isotopic ratios observed in asymp-totic giant branch (AGB) stars of different masses [3–5] can be usedto probethe nucleosynthesisandmixingprocessesin these stars.Furthermore,the18O/16Oabundanceratioiscriticalin clas-sifyingstardustoxideandsilicategrainsthatoriginallycondensed inAGB stars,supernovae, andnovae andcanbe found preserved inmeteorites [6,7]. Onestriking exampleis givenby oxygen-rich Group II grains whose experimentally measured 18O/16O ratios are significantly higher than predicted by models, and dilution withmatterofsolar compositioniscurrentlyassumedto explain thisdiscrepancy(e.g.,[8]).Alow-backgroundmeasurement ofthe

18_O(p,

_α

₎15_{N reaction atenergies of astrophysical interestcan}

re-sultin more accurate predictions forthe O isotopic composition andplacestrongerconstraintsonthestellarsitesfromwherethese grainsoriginate[8,9].

The18_O(p,

_α

₎15_{Nreaction( Q -value}

₌

₃

_.

_{98 MeV)hasbeen}

stud-iedusing bothdirect [10–12] and indirect[13,14] approaches.At temperaturesofastrophysicalinterest(T

=

0

.

01–1

.

00 GK),itsrate isdominatedbytheinterferenceofthree J π

=

1

/

2+resonancesat centerofmass energies Er

=

143, 610 and800 keV,respectively. Forthelattertwo resonances, resultsontheir energyandpartial widthsare largely inconsistent [14], andtensions havealso been reportedbetweenthecrosssectionofdifferentdatasetsatenergies

Ec.m.

≤

1 MeV[14]. While theexcitation function hasbeen mea-sureddirectlytoenergiesaslowasEc.m.

=

70 keV[11],significant

uncertainties remainthat affectthe cross-sectionextrapolationto lower energies. In addition, the energy and partial widths of a

Er

=

90 keVresonancewerequestionedinrecenttheoreticalwork [15].As aconsequence,thestellar reactionratestillcontains sig-nificantuncertainties.

This letter presentsthe results of a direct underground mea-surement ofthe 18_O(p,

_α

₎15_{N reaction cross-section from E}

c.m.

=

340 keVdown to Ec.m.

=

55 keV,the lowestenergy measuredto date. The primary aim of the presentstudy was to measure the non-resonant componentof the crosssection ofthe 18_O(p,

_α

₎15_N

reaction at proton beamenergies from Ep

=

360 to 60 keV, ex-tending therange ofdirect measurements to energies ofinterest forintermediate andlow-mass AGB stars.We alsoaimed to de-termine,withimprovedaccuracy,thestrengthofthreeresonances ofastrophysicalinterestat Er

=

90,200,320 keV,usingthe thick-target yieldapproach, inaddition tothe Er

=

143 keV resonance strengthalreadyreported[24]. Theexperimentwas performedat the underground LUNA-400 accelerator [16,17] of the Laboratori Nazionalidel GranSasso(LNGS), Italy, within aprogram of reac-tionstudiesofhydrogenburninginadvancedCNOcycles[18–23]. Thereducedbackgroundachievedunderground[24] allowedusto measurecross-sectionsaslowas1 picobarn/srwithunprecedented precision.

2. Methodology

FulldetailsontheexperimentalsetuparereportedinRef. [24]. Briefly, a proton beam was accelerated onto solid Ta2O5 targets

enriched (98%) in 18O [26]. Targets were produced with thick-nesses corresponding to energy losses of 5 or 15 keV at proton

beam energy Ep

=

151 keV for cross-section measurements

re-spectively above and below Ep

=

103 keV. Alpha particles from the18_O(p,

_α

₎15_{Nreactionweredetected atbackwardanglesusing}

an array of eight silicon detectors: four placed at 135◦ with re-specttothebeamaxisandfourat102.5◦(ofthese,onlytwowere workingproperly during data taking) [24]. Protective aluminized

Mylar foils were mounted in front of each detector. Their thick-ness(5.5 μm)waschosensoastostopelasticallyscatteredprotons while atthe same time letting the alpha particles pass through withminimal energyloss (

≈

800 keV).Typical detectionenergies wereaboutEα

=

2

.

3 MeV,dependingontheprotonbeamenergy. For narrow and isolated resonances (as those investigated in this study), the resonance strength

ωγ

can be directly obtained fromthethick-targetyieldY as[27]:

ωγ

=

2

λ

2

eff

Y

W

η

(1)

where



eff is the effective stopping power in units of eV/(atom/

cm2_);

_η

_{isthedetectionefficiency;}

_λ

_{isthedeBrogliewavelength}

oftheprojectileattheresonantenergy;andW takesintoaccount theangulardistributionattheangleofthedetector.TheW factor

(atmost20%deviationfromunityinthisstudy)wascalculated us-ingan R-matrixapproach(seelater)basedon J π valuesreported inpreviousexperimentalstudies[11,13].

At eachbeamenergy,countsinthe alphapeak wereobtained by integration [24]. The natural background (

≈

0

.

04 counts/h/ detector) under the alpha peak of the 18_O(p,

_α

₎15_{N reaction}

(

∼

2 MeV)was negligible atall beamenergies asa resultof the ten-foldbackgroundreductionachievedundergroundatthese en-ergies[24].Theonlysourceofbackgroundwasfrombeam-induced reactions ontraceboron contaminantsinthetarget givingriseto abroadfeaturearound3 MeVthroughthe11_B(p,

_α

₎₂

_α

_reaction.Its

contribution to the 18O(p,

α

)15N alpha peak was estimated using a linear extrapolation and found to be less than 2% atall ener-giesinvestigated here.We conservativelyassignedan asymmetric uncertaintyof

−

3% tothenumberofcountsinthealphapeak. 3. Results

3.1. Narrowresonances

For the Er

=

200 and 320 keV (Fig. 1 top panel) resonances we first determined the non-resonant yield contribution from a second-order polynomial fit of the data points above and below the resonanceregion,andthenaddedthe polynomialfunction to afitofthethick-targetresonanceprofile[24]:

f

(

Ep

)

=

H



1

+

exp



ER−E δL

 

1

+

exp



E−ER−E δR



(2)

where H istheplateauheight; ER istheresonanceenergyinthe laboratory system;

δ

L and

δ

R describe,respectively, thesteepness

of rising andfalling edges of theprofile; and

E is the energy-equivalent target thickness(in thelaboratory). Thisprocedure al-lowed us to extract the net yield Y

(

=

H

)

used to calculate the resonancestrengthaccordingtoEq. (1).

The situationwas morecomplicatedforthe Er

=

90 keV reso-nance(Fig.1bottompanel)becausedatawereacquiredon5 keV-thicktargetsforbeamenergiesEp

>

103 keVandon15 keV-thick

targets for beam energies Ep

≤

103 keV to increase the

count-ing rate. Non-resonant yield data taken with both targets were first converted into cross-section data and these latter were fit with a single R-matrix calculation [28]. The fitted cross section was then convertedback into a yield curve in orderto establish the non-resonantyield contribution(dashedlineinFig. 1)to the thick-target resonanceyield.Afitto thethick-targetyieldplateau (solidlineinFig.1)wasfinallyperformedtoextracttheresonance strengthvalue.Notethatallfits tothick-targetyieldprofiles were

performed only on data acquired at 135◦ to minimize

system-atic uncertainties inefficiency dueto two non-workingdetectors at 102.5◦ (see[24] fordetails).

Fig. 1. Thick-targetyieldprofileoftheEr=320 keV(top)and90 keV(bottom)

reso-nancesinthe18_O(p,α)15_{Nreaction.Theresonancestrengthwasobtainedfromafit} (redcurves)totheyieldprofile(Eq. (2))takingintoaccountnon-resonant contribu-tions(dashedlines)determinedeitherfromapolynomialfit(dashedline,toppanel) orfromanR-matrixcalculation(dashedlines,bottompanel).Errorbarsshow sta-tisticaluncertainties(seeTable2).Thedifferentialyieldshownontheordinateis definedhereastheyielddividedbythesolidangle.

Table 1

Resonance energiesand strengths ωγ for the three resonancesin 18_O(p,α)15_{Nmeasuredinthisworkandpreviously.Theuncertainty} ontheresonantenergiesofthisworkcorrespondstoourbeam en-ergyresolution[29].Statistical(st)andsystematic(sy)uncertainties arereportedforthiswork.Otherworksreporttotaluncertaintiesonly.

Er[keV] ωγ[meV] Ref.

90±3 (0.16±0.05)×10−3 _[11] 96.6±2.2 (0.18±0.03)×10−3 _[13] 90.3±0.3 (1.57±0.14st±0.12sy)×10−3 this work 143.9±0.9 170±20 [11] 142.8±0.1 167±12 [30] 143.2±0.3 164.2±0.9st±+₋1211..17 sy [24] 205±1 2.3±0.6 [11] 204.7±0.3 2.37±0.12st±0.18sy this work 316±1 57±10 [11] 317.0±0.3 85±9st±7sy this work

Resonancestrengthsobtainedinthepresentstudyforallthree resonances,aswellasforthe Er

=

143 keVresonancealready re-portedinRef. [24],areshowninTable1,withuncertaintiesgiven in Table 2. The

ωγ

values for the Er

=

200 and 320 keV reso-nancesare infair agreementwithprevious determinations while thestrength ofthe Er

=

90 keVresonance isan orderof

magni-Table 2

Errorbudgetforstatistical(tailasymmetry,background sub-traction,chargeintegration)andsystematic(stoppingpower, efficiency)uncertainties.

Source Rel. uncertainty Ref.

Tail asymmetry +2.0% [24]

Background subtraction −3.0% this work

Charge integration ±2.0% [24]

Stopping power ±4.0% [31]

Efficiency ±5.5% [24]

Table 3

DatasetsincludedinourglobalR-matrixfit.

Reaction Ec.m[keV] θlab[◦] Reference

18_O(p,α)15_{N 55–320 102.5, 135 this work}

18_O(p,α)15_{N 220–660 integrated [10]}

18_O(p,α)15_{N 70–886 90, 135 [11]}

18_O(p,α)15_{N 590–1670 several [12]}

18_O(p,p)18_{O 570–1340 90, 140 [36]}

tudehigherthan previouslyreportedin Lorenz-Wirzbaet al. [11] andLa Cognataet al.[13].WenotehoweverthatinherPhDthesis [32],Lorenz-Wirzbareports

ωγ

=

2

.

1

±

0

.

6 μeVinexcellent agree-mentwithourpresentvalue.Itisunclearwhydifferentvaluesfor thisstrength appearin Refs. [11,32] but thediscrepancyis likely due to an over-estimateof the poorly-known total width



tot of

thisEr

=

90 keV resonanceinRef. [11].Inourstudywehave suf-ficientdatatofittheyield(Fig.1),andwedonotneedtoassume a



totvaluetodeterminearesonancestrength.

3.2. R-matrixanalysis

Thedifferentialcross-sectionwascalculatedfromour differen-tialyielddatapointsusingthemedianenergyapproachdescribed in Ref. [33]. More details on this deconvolution analysis can be found in Ref. [34]. Electron screening corrections (at most 20% here) based on the adiabatic limit approximation [35] were ap-pliedtoeachcross-sectiondatapoint.

Toarriveatafinal reactionrate, weperformeda global multi-channel R-matrix [28] analysis on our differential cross-sections, at both 135◦ and 102.5◦, and the most recent 18O(p,

α

)15N and

18_O(p,p)18_{Odatasets (Table 3). The inclusion of elastic-scattering}

data in the fitting procedure provides strong constraints on the

R-matrixfitssincethepoleenergies andprotonwidthsare iden-tical.However,fortheadditionaldatasetsconsideredherewehad torelyondigitizedinformationfromthe exfor database[37] be-cause original data were not available. Similarly, error bars were either unavailableorunreliable (ifobtained fromthe digitization procedure, e.g. for errors smaller than the symbol size). There-fore, the statisticaluncertainty on all data points was arbitrarily setto

±

10%unlessthedigitizeduncertainty(whereavailable)was higher. Statistical uncertainties in our cross-section data points were calculatedasa combinationinquadratureofPoisson uncer-tainty, tail asymmetry and backgroundsubtraction (see Table 2). Furthermore,the 0.3 keV uncertainty in theenergy of each data point ofthiswork was converted intoan uncertainty inthe cor-responding cross-section and added in quadrature to the other statisticaluncertainties.

Our global R-matrix fitis shown inFig. 2 with ourdata and those from Ref. [11] (top panel) inthe form of the (differential)

Fig. 2. Global R-matrixfit(solidline)showntogetherwith 18_O(p,α)15_N differen-tialS-factordataat135◦fromourstudyandfromRef. [11] (toppanel);andwith 18_O(p,p)18_{Oelasticscatteringdataat140}◦_{from[36] (bottompanel). Thereddashed} linedintheinsetshowstheeffectofremovingthenewstateatEr=106 keV.

astrophysical S

(

E

)

factor,2 _{and with elastic scattering data from}

Ref. [36] (bottompanel).

SometensionsbetweenourdataandthosereportedinRef. [11] atresonant energies below Ec.m.

=

170 keV (Fig. 2top panel, in-set) maybe the resultof an incorrect(in the sense ofRef. [33])

deconvolution of the median energy near the strong Er

=

143

and 90 keV resonances. This deconvolution is less problematic at higher, non-resonant energies. The tensions observed at reso-nant energies could be resolved by increasing the uncertainties in data from [11], but this strategy would result in a lower re-ducedchi-square

χ

˜

2 _{valuewithoutanyaddedphysicsconstraints}

tothe data.Instead,we decided to discarddatafrom[11] below

Ec.m.

=

170 keV for both angles (approximately 10% of the data set).

R-matrix fits were carried out using both AZURE2 [38] and rmatrix2015 [39] independently.The R-matrixradius was set to

a

=

5fm



1

.

4fm

× (

181/3

+

11/3

)

, but we observed no signifi-cant differences using a

=

5

.

5 or 6 fm. We used a proton (al-pha) separation threshold value of Sp

=

7993

.

599

(

1

)

keV (Sα

=

4013

.

799

(

1

)

keV) [40]. Angular distribution values were taken

2 _{TheS(E)factorisdefinedasS(E)=σ(E)exp(2π η)E,whereσ(E)isthe} reac-tioncrosssection,ηistheSommerfeldparameter,andE istheinteractionenergy [27].

Table 4

R-matrixbest-fitparametersvalues(center-of-mass):inbold-face,valuesobtained fromfitstothick-targetyields(Table1)andincludedinthe calculation,butnot optimizedbythefit.Notetheinclusionofapreviouslyunobservedresonanceat

Er=106 keVwithtentativespin-parityassignmentsinbrackets,andthepresence

ofabackgroundpoleat7MeV.Uncertainties(fromMINUIT)arestatisticalonly.The lastcolumngivestheoff-diagonalinterferencesignbetweenresonances.

Jπ _E r[keV] p α Int. 1/2+ 142.8±0.3 164±12 meV 150±1 eV + 1/2+ 612.5±1.2 7.7±0.1 keV 163±1 keV − 1/2+ 799.8±0.3 24.4±0.3 keV 26.1±0.3 keV + 3/2− 597.6±0.3 36±2 eV 2.5±0.1 keV + 3/2+ 89.0±0.3 797±57 neV 121±5 eV + 5/2+ 204.7±0.3 791±56 μeV 12±1 eV + 5/2+ 317.2±0.3 28.2±2.0 meV 1.9±0.1 keV − (1/2−) 106±3 120±10 μeV 86±1.6 keV + 1/2− 7000 29±12 MeV 431±180 MeV +

from [11], except for the Er

=

90 keV resonance for which we usedmorerecentinformationin[13] instead.Inaddition,a back-ground pole [28] was also included. However, even by adding a background pole and by optimizing interference effects between resonances,we wereunable toreproducethebroadstructure ob-served around Ec.m.

=

110 keV (Fig. 2), unless by assuming the existenceofahithertounobservednewresonanceatEr

=

106 keV. This newresonance isnot incompatible withprevious datafrom Lorenz-Wirzbaet al. [11] at 135◦ (Fig. 2). Because ofourlimited angular information (detectors were placed at two angles only), wewereonlyabletotentativelyassignaspin-parityof J π

=

1

/

2− tothisnewstatebasedonthelowest

χ

˜

2 _{valueobtainedfromthe} R-matrixfit(seebelow).

R-matrixresultswithAZURE2andrmatrix2015 werein excel-lentagreement.However,AZURE2offersthepossibilityofdefining scaling factors for each dataset that can be treated asadditional

degrees of freedom and optimized during the fitting procedure

[38].Thesescaling factorscan help attenuate systematictensions between datasets and effectively model their systematic uncer-tainty. We arbitrarily set all scaling factors to a value of

±

7%, corresponding to the total percent systematic uncertainty in our

data. Using AZURE2 with a proper normalization coefficient and

increaseduncertainties asdiscussedabove,we minimizedthe

χ

˜

2

usingtheMINUITpackageincludedinAZURE2.Best-fitparameters aregiveninTable4.

3.3. Reactionratecalculation

Recent18_O(p,

_α

₎15_{Nastrophysicalreactionratesavailableinthe}

literature [41,42] werecalculatedusing

RatesMC

[43],whichcan only simultaneously treat interferences betweenresonances with the same J π twoata time.However, sincethelow energycross sectionofthe18_O(p,

_α

₎15_{Nreactionisdeterminedbythepresence}

of three J π

=

1

/

2+ states,rates calculated with

RatesMC

have tocombinedifferentcalculationswithdifferentpairsofinterfering resonances, discarding the influence of one resonance in turn in eachcalculation.

A more accurate resultcan obtainedby fullytreating the in-terferenceeffectsofallstatessimultaneously.Tothisend,we nu-merically integratedthetotalcrosssectionobtainedwithAZURE2 using the parameters in Table 4(including the background pole) and adding the Er

=

20 keV resonance following Ref. [42]. Our recommended reaction rate, normalized to the Iliadis 2010 rate [41] (calculated with

RatesMC

), is shown in Fig. 3. Uncertain-ties werecalculated by varying theparameters in Table4 by

±

1

Fig. 3. Reactionratesforthe18_O(p,α)15_{Nreaction,calculatedbynumerical} integra-tionofthetotalcross-sectionusingtheparametersinTable4ofthiswork(black curve,dashed linesshowuncertainties),andIliadis et al.[41] (bluedashed and dottedline).Allcurvesarenormalized totheratereportedinIliadiset al.[41], cal-culatedwithRatesMC.Uncertaintiesforthislatterrateareshownasredlines.

sigmawithrespectto their central value. Tohighlightthe

differ-ence between the two approaches, we also show the rate that

wouldbe obtained by numerically integrating the cross sections calculatedwithAZURE2usingtheparametersinIliadiset al. [41]. Thisrate (blue dot-dashed line) is also normalized to the Iliadis 2010(

RatesMC

)rateinFig.3.

Thanksto theextensionofthe directdataenergyrangedown to Ec.m.

=

55 keVinthepresentstudy,andthesignificant reduc-tionoftheuncertaintyinthecross-sectionbelowEc.m.

=

340 keV, improvedconstraints arenow available forthe calculation ofthe interferenceeffects between the three J π

=

1

/

2+ resonances. In particular,attemperaturesT

<

0

.

1 GK,ourrateisuptoafactorof 2.5higherthantheIliadis2010rate(

RatesMC

),oraboutafactor of5 higher than the numericallyintegrated ratecalculated from theparametersinIliadis2010.Theuncertaintyinourreactionrate isalsoreducedbyup toanorderofmagnitudeoveran extended temperaturerange,T

=

0

.

01–1

.

0 GK.

4. Astrophysicalimplications

Thereported increase ofthe 18O(p,

α

)15Nreaction rateover a wide temperaturerange(T

=

0

.

01–1

.

0 GK) hasimplications in a numberofstellarsites.Inparticular,usingourrate,modelspredict

18_O/16_{O ratios for oxygen-rich Group II grainslower than}

previ-ously assumed[8], reinforcing the need fordilutionwith matter ofsolarcomposition to explain thediscrepancy.Fig. 4showsthe evolution of the 18O and 15N abundances during the Thermally PulsingAsymptoticGiantBranch(TP-AGB)phaseofastarwith ini-tial mass M

=

4

.

8 M_ andsolar-like metallicity Z

=

0

.

014, that undergoesamildHotBottomBurning(HBB)[44].Nucleosynthesis calculationsare carriedoutwithboththereaction ratepresented inthiswork(right)andinIliadis2010[41] (left).Asaresultofthe newreaction rate, the 18_{Omass fractionis reduced by an order}

ofmagnitude,andwithsignificantlyreduceduncertainties,atthe endoftheevolutionduetotheoperationofHBBcomparedtothe previouslyadopted rate. Thisreduction isobserved onlyforstars undergoingmildHBB (T



40–60 MK),correspondingtoanarrow stellarmassrangearoundM



4

.

5–5 M_,dependingonthestellar evolutioncodeused andthemetallicity. Ournewrateintroduces starsinthisnarrowmassrangeasapotentialnewsiteforthe pro-ductionofsomegrains,requiringnodilutionwithmaterialofSolar Systemcomposition.Inparticular,AGBstellarmodels[45] of5 M_

Fig. 4. Evolutionof15_Nand18_{OsurfaceabundancesduringtheentireTP-AGBphase} ofastarwithinitialmassof4.8 MandsolarmetallicityZ=0.014,computedwith theCOLIBRIcode[44].PredictionsobtainedwiththeratefromIliadis2010(left) andthe newonefrom LUNA(right)areshowntogetherwiththecorresponding uncertaintyranges.

andmetallicity Z

=

0

.

03 couldproducegrainswith18_O/16_Odown

to5

×

10−4and17O/16O

=

1

.

7

×

10−3.Thesevaluesareveryclose tothosereportedincorundumgrainT84,an outlyingoxygen-rich Group IIgrain,having18O/16O

=

4

.

5

×

10−4,17O/16O

=

1

.

5

×

10−3, withouttheneedtoassumedilutionwithmaterialofsolar compo-sition.Dataonthealuminum isotopicratiosofthistypeofgrains isneededto confirmthisnewpossibleproductionsite. Complete astrophysicalimplicationsofCNOratesrecentlymeasuredatLUNA [18,20,21,25], includingtheratepresented here, requiremore in-depth and comprehensive astrophysical analyses that go beyond the scopeof thisLetterandwill be thesubjectofa forthcoming publication.

5. Conclusions

We reported an improved 18_O(p,

_α

₎15_{N reaction measurement}

extending the reach of direct measurements down to Ec.m.

=

55 keV,the lowestenergytodate,withunprecedentedprecision. Resonance energies andstrengths forkeyastrophysical statesare ingoodagreementwithpreviouswork,exceptfortheEr

=

90 keV resonance whose strength value of 1

.

57

±

0

.

14stat

±

0

.

12syst μeV

isan orderofmagnitudehigherthanpreviously reported[11,13]. A multi-channel R-matrixanalysis ofour new cross-section data and other datasets available in the literature was performed to extrapolate thecross-section to energies ofastrophysical interest. The R-matrix fit presented in this work is the first attempt, to our knowledge, at fitting all available datasets, including elastic scattering,tominimizethestatisticalandsystematicuncertainties.

Thanks to the improved constraints offered by the new

under-groundmeasurement,ourrecommended18O(p,

α

)15Nreactionrate is moreprecise than the ratepresented inRef. [41] overa wide temperaturerange(T

=

0

.

01–1

.

00 GK)ofinterestinAGBstars.In particular, ourresultsreinforce theneed toassume dilutionwith matterofsolarcompositiontoreproduceobservedabundances of mostoxygen-richGroup IIgrainsoriginatinginintermediate-mass AGB stars[8]. Forsome outlyingoxygen-rich Group II grains, the

presentratesuggestsanewpotentialproductionsiterequiringno dilutiontomatchexperimental isotopic ratios.More dataon this typeofgrainsishighlydesirable.

Theauthors acknowledgeA.I. Karakasforproviding thestellar structureofthe5 M_ Monashmodel,andR.J.deBoerfora digiti-zationofdatafrom[11].ThisworkwassupportedinpartbyINFN, STFCUK(grantno.ST/L005824/1),NKFIH(grantno.K120666),HAS Lendület(grantno.LP2014-17),andDFG(grantno.BE4100/4-1). C.G.B. gratefullyacknowledges the support of a SUPAShort-Term Visit Award during his stay in Brussels. P.M. acknowledges the supportfromtheERCConsolidatorGrantfundingscheme(project STARKEY,G.A.n.615604).

References

[1]M.Lugaro,etal.,Astrophys.J.615(2004)934. [2]S.Cristallo,etal.,Astron.Astrophys.540(2014)A46. [3]C.Abia,etal.,Astron.Astrophys.599(2017)A39. [4]O.Straniero,etal.,Astron.Astrophys.598(2017)A128. [5]K.Justtanont,etal.,Astron.Astrophys.578(2015)A115.

[6]E.Zinner,in:A.M.Davis(Ed.),TreatiseonGeochemistry,vol. 1,2nded., Else-vier,2014,pp. 181–213.

[7]L.P.Nittler,etal.,Astrophys.J.483(1997)475. [8]M.Lugaro,etal.,Nat.Astron.1(2017)0027. [9]S.Palmerini,etal.,Astrophys.J.764(2013)128. [10]H.-B.Mak,etal.,Nucl.Phys.A304(1978)210. [11]H.Lorenz-Wirzba,etal.,Nucl.Phys.A313(1979)346.

[12]N.S.Christensen,etal.,Nucl.Instrum.MethodsPhys.B51(1990)97. [13]M.LaCognata,etal.,Phys.Rev.Lett.101(2008)152501.

[14]M.LaCognata,C.Spitaleri,A.M.Mukhamedzhanov,Astrophys.J.723(2010) 1512.

[15]H.T.Fortune,Phys.Rev.C88(2013)015801. [16]H.Costantini,etal.,Rep.Prog.Phys.72(2009)086301.

[17]C.Broggini,D.Bemmerer,A.Caciolli,D.Trezzi,Prog.Part.Nucl.Phys.98(2018) 55.

[18]D.A.Scott,etal.,Phys.Rev.Lett.109(2012)202501. [19]A.DiLeva,etal.,Phys.Rev.C89(2014)015803. [20]F.Cavanna,etal.,Phys.Rev.Lett.115(2015)252501;

F.Cavanna,etal.,Phys.Rev.Lett.120(2018),239901(E). [21]C.G.Bruno,etal.,Phys.Rev.Lett.117(2016)142502. [22]R.Depalo,etal.,Phys.Rev.C94(2016)055804. [23]A.Boeltzig,etal.,Eur.Phys.J.A52(2016)75. [24]C.G.Bruno,etal.,Eur.Phys.J.A51(2015)94. [25]F.Ferraro,etal.,Phys.Rev.Lett.121(2018)172701. [26]A.Caciolli,etal.,Eur.Phys.J.A48(2012)144.

[27]C.Iliadis,NuclearPhysicsofStars,2ndedition,Wiley-VCH,2015. [28]P.Descouvemont,D.Baye,Rep.Prog.Phys.73(2010)036301.

[29]A.Formicola,etal.,Nucl.Instrum.MethodsPhys.A507(2003)609–616. [30]H.W.Becker,etal.,Z.Phys.A351(1995)453.

[31] J.Ziegler,SRIM2013.00,http://srim.org.

[32]H.-M.Lorenz-Wirzba,PhDthesis,WestfälischenWilhelms-Universitätzu Mün-ster,1978.

[33]C.R.Brune,D.B.Sayre,Nucl.Instrum.MethodsPhys.A698(2013)49–59. [34]C.G.Bruno,PhDthesis,TheUniversityofEdinburgh,2017.

[35]H.J.Assenbaum,K.Langanke,C.Rolfs,Z.Phys.A327(1987)461. [36]K.Yagi,J.Phys.Soc.Jpn.17 (4)(1962)604.

[37]N.Otuka,etal.,Nucl.DataSheets120(2014)272–276. [38]R.E.Azuma,etal.,Phys.Rev.C81(2010)045805.

[39]D.J.Mountford,etal.,Nucl.Instrum.MethodsA767(2014)359–363. [40]M.Wang,etal.,CPC36 (12)(2012)1603.

[41]C.Iliadis,etal.,Nucl.Phys.A841(2010)31–250. [42]A.L.Sallaska,etal.,Astrophys.J.Suppl.Ser.207 (18)(2013). [43]R.Longland,etal.,Nucl.Phys.A841(2010)1–30.

[44]P.Marigo,etal.,Mon.Not.R.Astron.Soc.434(2013)488–526. [45]A.I.Karakas,M.Lugaro,Astrophys.J.825(2016)1.