A short walk through the physics of neutron stars

Isaac Vidaña, INFN Catania

ASTRA: Advanced and open problems in low-energy nuclear and hadronic STRAngeness physics October 23rd-27th 2017, Trento (Italy) This short talk is just a brush-stroke on the physics of neutron stars. Three excellent monographs on this topic for interested readers are: Neutron stars are different things for different people

² For astronomers are very little stars “visible” as radio or sources of X- and γ-rays.

² For particle physicists are neutrino sources (when they born) and probably the only places in the Universe where deconfined quark matter may be abundant.

² For cosmologists are “almost” black holes.

² For nuclear physicists & the participants of this workshop are the biggest neutron-rich (hyper)nuclei of the Universe (A ~ 1056-1057, R ~ 10 km, M ~ 1-2 M ). ¤ But everybody agrees that …

Neutron stars are a type of stellar compact remnant that can result from the gravitational collapse of a massive

star (8 M¤< M < 25 M¤) during a Type II, Ib or Ic supernova event. 50 years of the discovery of the first radio

² radio pulsar at 81.5 MHz ² pulse period P=1.337 s

Most NS are observed as pulsars. In 1967 Jocelyn Bell & Anthony Hewish discover the first radio pulsar, soon identified as a rotating (1974 Nobel Prize for Hewish but not for Jocelyn) Nowadays more than 2000 pulsars are known (~ 1900 Radio PSRs (141 in binary systems), ~ 40 X-ray PSRs & ~ 60 γ-ray PSRs) Observables

§ Period (P, dP/dt) § Masses

§ Luminosity

§ Temperature http://www.phys.ncku.edu.tw/~astrolab/mirrors/apod_e/ap090709.html § Magnetic Field § Gravitational Waves ü NS-NS or BH-NS mergers ü NS oscillation modes: r-, f-, … modes The 1001 Astrophysical Faces of Neutron Stars Observation of Neutron Stars

X- and γ-ray telescopes Space telescopes

Chandra Fermi HST (Hubble) Atmospheric opacity opacity Atmospheric

Optical telescopes Radio telescopes

VLT (Atacama, Chile) Arecibo (Puerto Rico): d= 305 m Green Banks (USA): d= 100 m Nançay (France): d ~ 94 m GW: A New Way of Observing Neutron Stars GW170817: Binary LIGO/VIRGO GW detection with associated electromagnetic events observed by over 70 observatories

August 17th 2017 12:41:04 UTC + 1.7 seconds GW from a BNS merger detected GRB (GRB170817A) detected by FERMI γ-ray by Adv. LIGO & Adv. VIRGO Burst Monitor & INTEGRAL

+ 10h 52 min New bright source of optical light (SSS17a) detected in the galaxy NGC 4993 in the Hydra constellation + 11h 36 min Infrared emission observed + 15h Bright ultraviolet emission detected

+ 9 days X-ray emission detected + 16 days Radio emission detected PRL 119, 161101 (2017) Importance of the Event GW170817

Detecting GW from a NS merger allows us to learn more about the structure of these compact objects

This multi-messenger event provides confirmation that NS mergers can produce short GRBs

The observation of a allows us to show that NS mergers could be responsible for the production of most of the heavy elements, like gold and platinum, in the universe Multi-messenger Observations of the Event GW170817

Timeline of the discovery of GW170817, GRB170817A & the bright optical source SSS17a at NGC 4993 and the follow up observations

For details see: PRL 119, 161101 (2017)

Nature (2017)

ApJ L12 848 (2017)

ApJ L13 848 (2017)

… First Analysis & Implications of GW170817

The very first analysis of the event GW170817 seem to indicate:

² NS radii should be R < 13 km or even smaller than 12 km. Some analysis suggest R < 11 km Constraint on the EoS: those predicting large radii excluded ?

² Low value of the upper limit of tidal deformability indicates a soft EoS The Fingerprint of a Pulsar

Individual pulses are very different. But the average over 100 or more pulses is extremely stable and specific of each pulsar

² Top: 100 single pulses from the pulsar PSR B0950+08 (P=0.253 s) showing the pulse- to-pulse variability in shape and intensity

² Bottom: Average profiles of several pulsars

Hobbs et al., Pub Astr. Soc. Aust., 202, 28 (2011) Pulsar Rotational Period

The distribution of the Normal pulsars rotational period of pulsars shows two clear peaks that indicate the existence of two types of pulsars

§ normal pulsars with P ~ s Millisecond pulsars § millisecond pulsars with P ~ ms

§ First millisecond pulsar discovered in 1982 (Arecibo)

§ Nowadays more than 200 millisecond pulsars are known

§ PSR J1748-2446ad discovered in 2005 is until know the fastest one with P=1.39 ms (716 Hz)

Minimum Rotational Period of a Neutron Star

Pulsar cannot spin arbitrarily fast. The absolute minimum rotational period is obtained when

Centrifugal Force = Gravitational Force

Keplerian Frequency

In Newtonian Gravity In

3 1/2 3/2 1/2 3/2 R " M sun % " R % ! M sun $ ! R $ Pmin = 2π ≈ 0.55$ ' $ ' ms Pmin = 0.96# & # & ms GM # M & #10km & " M % "10km % 2 ! 2π R6 $ log P = log#( ) B2 sin2 α&− log P  3 P log P = log P − log(2τ ) Pulsar distribution "# 6c I %& in the . P-P plane Normal pulsars

Pulsar equivalent of the Hertzprung-Russell diagram for ordinary stars

Millisecond pulsars Magnetic Field of a Pulsar

Type of Pulsar Surface magnetic field

Millisecond 108 – 109 G Normal 1012 G 1014 – 1015 G

Extremely high compared to … Magnetars Earth Magnet Sun spots 0.3− 0.5G 103 −104 G 105G

Magnetars € € € Largest continuous Largest magnetic field in lab. (USA) pulse in lab. (Russia)

Normal pulsars

4.5x105G 2.8x107G Millisecond pulsars

€ € Where the NS magnetic field comes from ?

A satisfactory answer does not exist yet. Several possibilities have been considered:

2 ² Conservation of the $ R ' φ = φ ⇒ B = B i magnetic flux during the i f f i& ) % R f ( gravitational collapse of the 2 iron core For a progenitor star with Bi ~ 10 G 6 12 & Ri ~ 10 km we have Bf ~ 10 G

² Electric currents flowing€ in the highly conductive NS interior

² Spontaneous transition to a ferromagnetic state due to the nuclear interaction Neutron Star Structure: General Relativity or Newtonian Gravity ?

Surface gravitational ~ 10−10 potential tell us how much compact an object is ~ 10−5 2GM c 2R € ~ 10−4 −10−3

è Relativistic effects are € ~ 0.2 − 0.4 very important in Neutron Stars and General Relativity € must be used to describe their€ 1 structure €

€ Neutron Stars Structure Equations

Ø The structure of a static (i.e., non-rotating) star with spherical symmetry in General Relativity is described by the Tolman-Oppenheimer-Volkoff (TOV) Equations

−1 dP m(r)ε(r)" P(r) %" 4πr3P(r)m(r)%" 2Gm(r)% = −G $1+ '$1+ '$1− ' dr r2 # c2ε(r)&# c2 &# c2r & dm = 4πr2ε(r) dr § boundary conditions

€ P(0) = Po , m(0) = 0 P(R) = 0, m(R) = M

€ Ø€ Rotation breaks spherical symmetry and makes the structure equations “slightly” more complicated Stability solutions of the TOV equations

² The solutions of the TOV eqs. represent static equilibrium configurations ² Stability is required with respect to small perturbations

dM dM G > 0, or G < 0 dρc dr The role of the Equation of State

The only ingredient needed Matter constituents to solve the TOV equations is the (poorly known) EoS (i.e., p(ε)) of dense matter Interactions EoS

“stiff” EoS TOV

“stiff” EoS

“soft” EoS

“soft” EoS Upper limit of the Maximum Mass

Mmax depends mainly on the behaviour of EoS, P(ε), at high densities. Any realistic EoS must satisfy two conditions: dP dP § Causality: ≤ c 2 § Stability: > 0 dρ dρ If the EoS is known up to ρ, these conditions imply:

1/2 " 5x1014 g / cm3 % € M max ≤ 3M $ € ' ¤ # ρ &

If rotation is taken into account Mmax can increase up to 20%: 1/2 " 5x1014 g / cm3 % M max ≤ 3.89M $ ' ¤ # ρ & How to Measure Neutron Star Masses

§ 5 Keplerian orbital parameters can normaly be determined in a binary system: ü orbital period: P

ü time of periastron passage: T0 ü eccentricity: e ü projected semi-major axis: a sin i ü angle of periastron: ω

§ 3 unknown: M1, M2, i Kepler’s 3rd law

2 3 2 3 3 ! $ M 2 sini " 2π % a sin i G(M1 + M 2 ) 2π f M , M ,i ( ) = ( 1 2 ) ≡ 2 = $ ' 3 # & M M # P & G a " P % ( 1 + 2 ) mass function In few cases small deviations from Keplerian orbit due to GR effects can be detected

Measure of at least 2 post- High precision NS mass Keplerian parameters determination

−5/3 2/3 2/3 " Pb % 1 ω = 3T⊗ $ ' (M p + Mc ) # 2π & 1−ε Periastron precession 1/3 M M 2M 2/3 " Pb % c ( p + c ) γ = T $ ' ε Time dilation and grav. redshift ⊗ # 2π & 4/3 (M p + Mc ) Shapiro delay “range” r = T⊗Mc 2/3 −2/3 M M −1/3 " Pb % ( p + c ) s = sini = T⊗ $ ' x Shapiro delay “shape” # 2π & Mc " %−5/3 M M  192π 5/3 Pb p c Orbit decay due to GW emission Pb = − T $ ' f (ε) 5 ⊗ # 2π & 1/3 (M p + Mc ) An example: the mass of the Hulse-Taylor pulsar (PSR J1913+16) Recent measurements of high masses § PSR J164-2230 (Demorest et al. 2010)

ü binary system (P=8.68 d) ü low eccentricity (ε=1.3 x 10-6) ü companion mass: ~ 0.5M ¤ ü pulsar mass: M =1.928± 0.017M ¤

§ PSR J0348+0432 (Antoniadis et al. 2013)

ü binary system (P=2.46 h) ü very low eccentricity ü companion mass: 0.172 ± 0.003M ¤ ü pulsar mass: M = 2.01± 0.04M¤

Formation of Binary Systems

Figure by P.C.C. Freire Measured Neutron Star Masses (2017)

Observation of ~ 2 M neutron stars ¤

Any reliable EoS of dense matter should predict

M EoS > 2M max [ ] ¤ A natural question arises:

Demorest et al. Can a d.o.f. energetically favorable but that induces a strong softening of Antoniadis et al. the EoS (e.g., hyperons) be present in the interior of neutron stars in view of this new constraint ? updated from Lattimer 2013 Limits on the Neutron Star Radius

The radius of a neutron star with mass M cannot be arbitrarily small

General Relativity: 2GM a Neutron Star is not a R > c 2 Black Hole

Finite Pressure: 9 GM Neutron Star matter cannot R > € 4 c 2 be arbitrarily compressed

Causality: GM speed of sound must R > 2.9 € c 2 be smaller than c

€ How to measure Neutron Star Radii

Radii are very difficult to measure because NS:

² are very small (~ 10 km) ² are far from us (e.g., the closest NS, RX J1856.5-3754, is at ~ 400 ly)

A possible way to measure it is to use the thermal emission of low mass X-ray binaries: NS radius can be obtained from

² Flux measurement +Stefan-Boltzmann’s law ² Temperature (Black body fit+atmosphere model) ² Distance estimation (difficult) ² Gravitational redshift z (detection of absorption lines)

2 FD R∞ 2GM R∞ = 4 → RNS = = R∞ 1− 2 σ SBT 1+ z RNSc Recent Estimations of Neutron Star Radii The recent analysis of the thermal spectrum from 5 quiescent LMXB in globular clusters is still controversial

Steiner et al. (2013, 2014) Guillot et al. (2013, 2014)

+1.3 R = 9.1−1.5 km 2013 analysis R =12.0 ±1.4km R = 9.4 ±1.2km 2014 analysis Thermal Evolution of Neutron Stars Information, complemenary to that from mass & radius, can be also obtained from the measurement of the temperature (luminosity) of neutron stars

D. G. Yakovlev & C. J. Pethick, A&A 42, 169 (2004) Neutron Star Cooling in a Nutshell

Crust cools by Core cools by conduction neutrino emission Two cooling regimes Slow Fast Low NS mass High NS mass

Surface photon emission dominates at t > 106 yrs

slow cooling dE dT fast cooling th = C = −L − L + H dt v dt γ ν

ü Cv: specific heat

ü Lγ: photon luminosity ü Lν: neutrino luminosity ü H: “heating” Neutrino Emission

Anything beyond just neutrons & protons results in an enhancement of the neutrino emission Hyperonic DURCA processes possible

Additional as soon as hyperons appear Fast Cooling ê (nucleonic DURCA requires xp > 11-15 %)

ê Processes

Process R (Schaab, Shaffner-Bielich & Balberg 1998)

only N Λ → p + l +νl 0.0394 − Σ → n + l +νl 0.0125 − 0.2055 Σ → Λ + l +νl N+Y − 0 0.6052 Σ → Σ + l +νl − Ξ → Λ + l +νl 0.0175 − 0 Ξ → Σ + l +νl 0.0282 0 + Ξ → Σ + l +νl 0.0564 − 0 Ξ → Ξ + l +νl 0.2218 + partner reactions generating neutrinos, Hyperonic MURCA, … R: relative emissitivy w.r.t. nucleonic DURCA (−Δ/k T ) Pairing Gap suppression of C & ε by B v ~ e

1 3 1 1 § S0, SD1 ΣN & S0 ΛN gap § S0 ΛΛ gap

(Balberg & Barnea 1998) (Wang & Shen 2010) 1 § S0 ΣΣ gap

NSC97e

(Zhou, Schulze, Pan & Draayer 2005) (I.V. & Tolós 2004) Anatomy of a Neutron Star

Equilibrium composition determined by

ü Charge neutrality

∑qiρi = 0 i

ü Equilibrium with respect to weak interacting processes € b → b + l + ν ∂ε 1 2 l µ = b µ − q µ − µ , µ = i i n i ( e ν e ) i b2 + l → b1 + ν l ∂ρi

€ € The Hyperon Puzzle

Hyperons are expected to appear in the core of neutron stars at ρ ~ (2-3)ρ0 when µN is large enough to make the conversion of N into Y energetically favorable.

But

The relieve of Fermi pressure due to its appearance EoS softer reduction of the mass to values incompatible with observation “stiff” EoS

ü can YN & YY interactions solve it ? “soft” EoS ü or perhaps hyperonic three-body forces ? ü what about quark matter ? Possible Solutions to the Hyperon Puzzle

YN & YY Hyperonic TBF Quark Matter

• YY vector meson repulsion Natural solution based on the known importance of Phase transition to φ meson coupled only to 3N forces in nuclear deconfined QM at hyperons yielding strong physics densities lower that repulsion at high ρ hyperon threshold

• Chiral forces (see Weise’s talk)

YN from χEFT predicts Λ To yield M > 2 M s.p. potential more max ¤ repulsive than those from QM should be meson exchange § significantly repulsive to guarantee a stiff EoS

§ attractive enough to avoid reconfinement The r-mode instability

Ω : Absolute Upper Limit Instabilities prevent NS Kepler of Rot. Freq. to reach ΩKepler

r-mode instability : toroidal mode of oscillation ü restoring force: Coriolis

ü emission of GW (CFS mechanism) r-mode unstable due to GW emission • GW makes the mode unstable Kepler • Viscosity stabilizes the mode Ω / c

Ω −iω(Ω)−t/τ (Ω,T ) A ∝ A0e shear viscosity viscosity shear bulk viscosity viscosity bulk

r-mode damped by damped r-mode 1 1 1 r-mode damped by damped r-mode = − + τ (Ω,T) τ GW (Ω) τ Viscosity (Ω,T) Hyperon Bulk Viscosity ξY (Lindblom et al. 2002, Haensel et al 2002, van Dalen et al. 2002, Chatterjee et al. 2008, Gusakov et al. 2008, Shina et al. 2009, Jha et al. 2010,…)

(Haensel, Levenfish & Yakovlev 2002) Sources of ξY:

non-leptonic N + N ↔ N +Y weak reactions N +Y ↔Y +Y

Y → B + l +νl Direct & Modified B'+Y → B'+ B + l +ν URCA l N +Y ↔ N +Y strong reactions N +Ξ↔Y +Y Y +Y ↔Y +Y

Reaction Rates & ξY reduced by hyperon superfluidity but (again) hyperon pairing gaps are poorly known Critical Angular Velocity of Neutron Stars

−iω(Ω)t−t/τ (Ω) § r-mode amplitude: A ∝ Aoe (I.V. & C. Albertus in preparation) 1 1 1 1 = − + + τ(Ω,T) τGW (Ω) τξ (Ω,T) τη (T) 1 = 0 r-mode instability region è τ (Ωc,T) Ω < Ω stable c € Ω > Ωc unstable

€ As expected: smaller r-mode instability region BHF: NN (Av18)+NY (NSC89) due to hyperons (M=1.27M ) ¤ The final message of this talk

Neutron stars are excellent observatories to test fundamental properties of matter under extreme conditions and offer an interesting interplay between nuclear processes and astrophysical observables For your time & attention