An End-to-End Test of Neutron Stars as Particle Accelerators
a r X i v :a s t r o -p h /0511495v 1 16 N o v 2005
AN END-TO-END TEST OF NEUTRON STARS AS PARTICLE ACCELERATORS
Patrizia A.Caraveo
INAF IASF-Milano,Via Bassini,15;20133Milano;Italy
ABSTRACT
Combining resolved spectroscopy with deep imaging,XMM-Newton is providing new insights on the parti-cle acceleration processes long known to be at work in the magnetospheres of isolated neutron stars.Accord-ing to a standard theoretical interpretation,in neutron stars’magnetospheres particles are accelerated along the B ?eld lines and,depending on their charge,they can ei-ther move outward,to propagate in space,or be funnelled back,towards the star surface.While particles impinging on the neutron star surface should heat it at well de?ned spots,outgoing ones could radiate extended features in the neutron star surroundings.
By detecting hot spots,seen to come in and out of sight as the star rotates,as well as extended features trailing neu-tron stars as they move in the interstellar medium,XMM-Newton provides the ?rst end-to-end test to the particle acceleration process.
Key words:Neutron stars;pulsars;Geminga;ESA;X-rays.
1.INTRODUCTION
Isolated neutron stars (INSs)are natural particle accelera-tors.Their,presumably dipolar,rapidly rotating magnetic ?elds,naturally inclined with respect to the star rotation axis,induce electric ?elds ideally suited to accelerate par-ticles already present in the stars’magnetospheres or ex-tracted from the crusts.Following the seminal paper of Goldreich and Julian (1969)and Sturrock (1971),a lot has been done to work out the details of such an accel-eration,focusing on its most likely location(s)inside the INS magnetosphere and on its ef?ciency.Traditionally,two classes of models have been developed:on one side the polar cap ones(Ruderman &Sutherland,1975,Hard-ing &Daugherty,1998,Rudak &Dyck,1999),where the acceleration takes place near the star surface,just above the magnetic pole;on the other side,the outer gap ones(Romani,1996),where the acceleration is tak-ing place in the outer magnetosphere,not far from the light cylinder.Recently,the slot gap model,extending
from the polar cap to the light cylinder,has been added as a third alternative (Muslinov &Harding,2003,Dyck &Rudak,2003,Harding,2005).Notwhistanding im-portant differences between models,the interaction be-tween accelerated particles (typically electrons)and the star magnetic ?eld results in the production of high en-ergy gamma-rays which,in turn,are not able to escape the highly magnetic environment and are converted into electron positron pairs.This initiates a cascade rapidly ?lling the magnetosphere with energetic particles which,interacting with the magnetic ?eld,are responsible for the vast majority of the INSs’multiwavelength phenomenol-ogy.
2.ACCELERATION BY-PRODUCTS
INSs are mainly studied through their non thermal radio emission.Radio searches have been highly successful and the current radio catalogues list more than 1500pul-sars (Manchester et al.2005).
In spite of the sheer number of objects and their very di-verse phenomenology,INS radio emission accounts for a negligible fraction of the star rotational energy loss.A far more important fraction of the star energy reservoir goes into high-energy radiation,mainly in high-energy gamma-rays.While the number of objects shrinks to less than 1%of the radio ones (Thompson et al.,2001),in gamma rays the INSs’luminosity can reach a sizable fraction of the total rotational energy loss,with an in-crease in ef?ciency for the older objects.
The rich INSs’phenomenology encompasses now also X and optical emissions.While the numbers are slightly higher than the gamma-ray ones (a dozen in the optical (Caraveo,2000,Mignani et al.2004)and two scores in X-rays (Becker &Aschenbach,2002)the nature of the radiation is not as clear cut as in the radio or gamma-ray domains.In the optical,as well as in X-rays,aged neutron stars exhibit both thermal and non-thermal emis-sions.Indeed,when non-thermal emission somewhat weakens with age,the thermal one begins to emerge to tell the story of the cooling crust of the neutron star.
INS thermal emission,however,is not totally unrelated to the magnetospheric particle acceleration.Depending
on their electric charge,particles move in different direc-tions along the magnetic?eld lines.While those moving outward try to escape the INS magnetosphere,those mov-ing inward hit the star and heat its crust at well de?ned spots,that,under the assumption of a dipolar magnetic ?eld,should coincide with its polar caps(return currents, see e.g.Ruderman&Sutherland1975;Arons&Scharle-mann1979).Thus,thermal emission could be of use to trace non-thermal phenomena.
The escaping particles,on the other side,are part of the neutron stars’relativistic wind which is supposed to ac-count for the bulk of their observed rotational energy loss. Such relativistic wind can be traced through its interac-tion with the interstellar medium(ISM),both in the im-mediate surroundings of the stars,where the INS mag-netic?eld is still important,or farther away,where the wind radiation pressure is counterbalanced by the shocked ISM.An important player to determine the shape and the phenomenology of the resulting Pulsar Wind Nebula(PWN)is the actual neutron star speed.INSs are known to be high velocity objects and,plunging super-sonically through the ISM,they can give rise to a rich bow shocks phenomenology seen in the radio,optical and X-ray domains(e.g.Chattarjee&Cordes,2002).
Fig.1summarizes the neutron star acceleration tree: starting from the synoptic view(Harding,2005)of the mechanisms responsible the gamma ray emission(the as-pect most intimately related to the actual particle accel-eration)and following the destiny of the particle moving inward(left)and outward(right).Using past and present space observatories operating at X and gamma-ray do-main,we can construct such a tree with the aim to im-prove our understanding of the neutron star physics.
3.THE OBSERV ATIONAL PANORAMA
Our task is now to brie?y review the data related to the three steps highlighted in Fig.1in order to?nd INSs ob-served throughout the acceleration tree.
3.1.High-energy gamma-rays
Waiting for the next generation of gamma-ray instru-ments such as Agile(Tavani et al.,2003)and Glast (Michelson et al.,2003),we try to make the best use of the EGRET results.INSs(be they radio loud or radio quiet)are the only class of galactic sources?rmly iden-ti?ed as high-energy gamma-rays emitters.Indeed,pul-sars are especially appealing to gamma-ray astronomers since their timing signature allows to overcome the iden-ti?cation problem due to the relatively large gamma-ray error boxes.However,in spite of more than a decade of relentless efforts based on the EGRET data,only Crab, Vela,PSR1706-44,PSR1951+32,PSR1055-52and the radio quiet Geminga are con?rmed gamma-ray sources (Thompson et al.,2001).A few more pulsars have been proposed as sources of pulsed gamma-radiation,but the claims are still awaiting con?rmations.Several posi-tional coincidences between newly discovered pulsars and old Egret sources(Hartman et al.1999)have been reported,but,again,such suggestions cannot be con-?rmed without an operating g amma-ray telescope.Thus, for the moment being,our gamma-ray sample encom-passes a very young(and energetic)object such as the Crab,two slightly older pulsars(Vela,PSR1706-44)and three middle-aged INSs(PSR1055-57,Geminga and the fast spinning PSR1951+32).It is worth noting that the ef?ciency for conversion of rotational energy loss into gamma-rays changes as a function of pulsar age.It goes from the value of～0.009%,for the Crab,to several%, for Geminga and PSR1055-57.
3.2.Hot spots
The presence of hot spots on the surface of INSs has been long suspected on the basis of their overall X-ray spectral shape requiring more than a simple black-body to describe the https://www.wendangku.net/doc/c315967005.html,ually two black-body curves, characterized by different temperatures and emitting areas,are needed to?t the X-ray spectra for all but the very youngest INSs.A slightly colder black-body, covering the majority of the INS surface,provides the bulk of the X-ray luminosity while a hotter one,covering a smaller surface,is needed to obtain a satisfactory spectral?t.
Long XMM-Newton observations of Geminga, PSR0656+14and PSR1055-57,three middle-aged, rather similar INSs,have added an important piece of information.Taking advantage of their exceptional photon harvest,De Luca et al(2005a)were able to perform space resolved spectroscopy of the three INSs. For all objects they have shown that
a)the spectra are varying signi?cantly throughout the rotational phase
b)the hot blackbody contribution is the most dramati-cally variable spectral component.
This is shown in Fig.2where the emitting radii,com-puted on the basis of the phase-resolved spectral?ts,are shown as a function of the pulsar rotational phase.While for PSR B0656+14the modulation in the emitting radius wrt.the average value is<10%,similar to the value found for the cool blackbody component,in the case of PSR B1055-52we see a100%modulation,since the hot blackbody component is not seen in4out of10phase intervals.A similar,100%modulation is observed also for Geminga,although in this case the hot blackbody component is seen to disappear in just one phase interval, and the pro?le of its phase evolution is markedly broader. It is natural to interpret such marked variations as an effect of the star rotation,which alternatively brings into view or hides one or more hot spots on the star surface. As outlined above,such hot spots arise when charged particles,accelerated in the magnetosphere,fall back to the polar caps along magnetic?eld lines.Straight estimates of neutron star polar cap sizes,based on a simple“centered”dipole magnetic?eld geometry(polar
Figure1.Neutron star acceleration tree.Top panel:schematic view of a pulsar magnetosphere showing the gamma-ray emitting regions,according to the various classes of models(from Harding,2005).Bottom left:a similar view of a pulsar magnetosphere showing the hot spots on the pulsar polar caps.Bottom left:hydrodynamic simulation of a bow shock(BS) generated by the interaction of the isotropic relativist wind of a neutron star(marked with a cross),moving horizontally from right to left,with the ISM(Gaensler et al,2004).A indicates the pulsar wind cavity,where the electrons propagate freely,B is used for the shocked pulsar wind material,while C represents the shocked ISM.The termination shock,TS,is where the energy density of the pulsar wind is balanced by the external pressure,while CD is the contact discontinuity
bounding the shocked pulsar wind material.
cap radius R P C=R c),where R is the neutron star radius,?is the angular frequency and c is the speed of light),predict very similar radii for the three neutron stars,characterized by similar periods(233m for PSR B0656+14,326m for PSR B1055-52and297m for Geminga,assuming a standard neutron star radius of10 km).The observed radii are instead markedly different, with values ranging from～60m for Geminga to～2 km for PSR B0656+14(see De Luca et al,2005a for a detailed discussion).
While waiting to enlarge the sample of deeply scrutinized X-ray pulsars,it does not come as a surprise that two of the three objects showing direct evidence for the pres-ence of rotating hot spots are highly ef?cient gamma-ray sources.
3.3.Pulsar Wind Nebulae
When the particle wind from a fast moving INS inter-acts with the surrounding ISM,it gives rise to com-plex structures,globally named“Pulsar Wind Nebulae”(PWNe)where～10?5?10?3of the NS˙E rot is con-verted into electromagnetic radiation(for recent reviews see Gaensler et al2004,and Gaensler2005).The study of PWNe may therefore give insights into aspects of the neutron star physics which would be otherwise very dif-?cult to access,such as the geometry and energetics of the particle wind and,ultimately,the con?guration of the INS magnetosphere and the mechanisms of particle ac-celeration.Moreover,PWNe may probe the surrounding medium,allowing one to measure its density and its ion-isation state.
A basic classi?cation of PWNe rests on the nature of the external pressure con?ning the neutron star wind(e.g. Pellizzoni et al.2005).For young NSs(
Static PWNe(Slane,2005,for a review)usually show complex morphologies.Striking features such as tori and/or jets(as in the Crab and Vela cases),typically seen in X-rays,re?ect anisotropies of the particle wind emitted by the energetic,central INS and provide important con-strains on the geometry of the system.A remarkable axial symmetry,observed in several cases,is assumed to trace the rotational axis of the central INS.For the Crab and Vela PWNe,such an axis of symmetry was found to be coincident with the accurately measured direction of the INS proper motion(Caraveo&Mignani1999,Caraveo et al.2001).This provided evidence for an alignement be-tween the rotational axis and the proper motion of the two neutron stars,with possible important implications for the understanding of supernova explosion mechanisms(Lai et al.2001).The alignement between spin axis and space velocity,directly observed only for Crab and Vela,is now assumed as a standard property of NSs(Ng&Roman, 2004).
Bow-shocks(for a review see Pellizzoni et al.2005, Gaensler et al2004)have a remarkably simpler,“velocity-driven”morphology.They are seen frequently in Hαas arc-shaped structures tracing the forward shock, where the neutral ISM is suddenly excited.In other cases, X-ray emission(and/or radio emission on larger scales)is seen,with a cometary shape elongated behind the neutron star,due to synchrotron radiation from the shocked NS particles downstream(only in the case of PSR B1957+20 both the Hαand the X-ray structures have been observed, Stappers et al.,2003).According to the lower energetics of the central,older INS,bow shocks are typically fainter than static PWNe and proximity is a key parameter for their observation.
Since we aim at tracing the high energy particle escap-ing the INS magnetosphere,we concentrate on the X-ray PWNe.Inspecting the list of Gaensler et al(2004),we ?nd only PSR B1951+32in common with the gamma-ray database,leaving little hope to?nd an object displaying all the aspects of the acceleration tree.
However,recent observations of Geminga,combined with previous ones by XMM-Newton,have unveiled the presence of a bona?de PWN with complex diffuse fea-tures trailing the pulsar perfectly aligned with its well known proper motion(De Luca et al.,2005b;Caraveo et al.2003).
Thus,the combined EGRET,XMM-Newton and Chan-dra results on Geminga make this source the most suitable example for our end-to-end test of particle acceleration. For a review on the multiwavelength phenomenology of Geminga,see Bignami&Caraveo(1996).
4.GEMINGA AS A TEST CASE
Fig.3summarizes all the observational evidence collected so far on the presence of high energy electrons/positrons in the magnetosphere of Geminga.First,the EGRET light-curve whose>100MeV photons could not have been produced without high energy particles and mag-netic?elds.
Next,the contribution of a100ksec XMM-Newton ob-servation which yielded both
a)the evidence for the presence of minute hot spot(s) varying throughout the pulsar phase(Caraveo et al., 2004)
b)the detection of two elongated tails,trailing the pulsar in its supersonic motion through the ISM and perfectly aligned with the proper motion direction.The?at spectral shape of the tails’X-ray photons suggests a synchrotron origin which,combined with the typical magnetic?eld present in a shocked ISM,implies the presence of～1014 eV electrons/positrons,i.e.of particle at the upper limit of the energy range achievable for an INS like Geminga. Moreover,the lifetime of such electrons(or,more pre-cisely,the time it takes for them to lose half of their en-
Figure2.Energy-resolved light curves of PSR B0656+14,PSR B1055-52and Geminga in different energy ranges.To ease the comparison of the behaviour of the three INSs,all light curves have been plotted setting phase0to the X-ray maximum.Pulsed fractions(computed as the ratio between the counts above the minimum and the total number of counts)are as follows:PSR B0656+1412.3±0.4%in0.15-0.6keV,16.9±2.3%in0.6-1.5keV,75±20%in1.5-7.0 keV;PSR B1055-5216.7±0.6%in0.15-0.7keV,67±3%in0.7-1.5keV,90±10%in1.5-6.0keV;Geminga28.4±0.6%in
0.15-0.7keV,54.5±2.4%in0.7-2.0keV,33±5%in2.0-6.0keV.
Figure3.The acceleration tree as observed for Geminga.Top:the gamma-ray light curve.Left:the XMM-Newton average spectrum as well as the results of phase-resolved spectroscopy,showing the evolution of the black-body emitting regions as a function of the INS rotational phase.The shaded area mark the phase intervals corresponding to theγ-ray peaks observed by EGRET.The highestγ-ray peak occurs at phase0.25±0.15,the second one at phase0.75±0.15(the uncertainty is due to the extrapolation of the EGRET ephemeris to the epoch of the XMM-Newton observation).Right: Geminga as seen by Chandra and XMM-Newton(from De Luca et al.,2005b).The Chandra image,rebinned to a pixel size of2”has been superposed on the XMM-Newton/MOS image obtained by Caraveo et al.(2003).Surface brightness contours for the XMM image have been also plotted.The ACIS?eld of view is limited to a rectangular region1arcmin
wide.The pulsar proper motion direction is marked by an arrow.
ergy)in the bow-shock magnetic?eld is～800years.On
the other hand,Geminga’s proper motion(170mas/year) allows one to compute the time taken by the pulsar and
its bow shock to transit over the apparent length of the
x-ray structures in the sky(3’from the central source).
Such a time is close to1,000years.Thus,Geminga’s
tails remain visible for a time comparable to the electron
synchrotron X-ray emission life time after the pulsar pas-
sage.The comet-like structure seen by Chandra is as lu-
minous as the larger and fainter tails and its spectrum is
equally hard.
Hot spot(s),elongated,faint tails and short,brighter trail
have roughly the same luminosity,corresponding to～10?6of its˙E rot.
We note that the morphology and hard spectrum of the
Trail is reminescent of the jet-like collimated out?ows
structures seen in the cases of Crab and Vela(Helfand
et al.,2001,Pavlov et al.2003,Willingale et al.,2001,
Mori et al.2004)and associated to the neutron stars spin
axis direction.In particular,the small Geminga’s Trail can be compared to the“inner counterjet”of the Vela
PSR(Pavlov et a.,2003),characterized by a similar spec-trum(photon index～1.2)and ef?ciency(L X～10?6˙E). The projected angle between Geminga proper motion and
its backward jet is virtually null,which implies that also
the pulsar spin axis should be nearly aligned with them.
Geminga would thus be the third observed neutron star
having its rotational axis aligned with its space velocity,
after the cases of the Crab and Vela.
The whole scenario,encompassing both the large Tails
and the small Trail,could therefore?t in the frame of
an anisotropic wind geometry.It includes jet structures
along the spin axis and relativistic shocks in the direction
of the magnetic axis where most of the wind pressure is
concentrated due to the near radial out?ow from magne-
tosphere open zones.
The coupling of the jet-like Trail seen by Chandra with
the larger,arc-shaped Tails seen by XMM has no similar-
ity with other pulsars.
5.CONCLUSION
The particle acceleration going on in an INS magneto-
sphere can now be traced from end-to-end.While gamma
ray emission probes directly the particle population in the
magnetosphere,using the current generation of X-ray ob-
servatories we are now able to follow the destiny of the
particles traveling up and down the magnetic?eld lines
through the study of hot spots on the star surface and of PWNe.The same process responsible for the copi-ous gamma-ray emission of Geminga would thus also be responsible for the appearance of the hot spots on its sur-face(Halpern&Ruderman,1993).Such a strong link between the X-and gamma-ray behaviour of the source could be exploited to map the relative positions of the re-gions responsible for the different emissions.A precise comparison of the source X and gamma-ray light curves is crucial at this point,but the lack of operating high en-ergy gamma ray telescope makes it impossible.Simulta-neous observations performed by XMM-Newton and by Agile and/or GLAST(foreseen to be operational in the coming years)will add important pieces of information to test INSs’capability to accelerate particles. ACKNOWLEDGMENTS
Tha analysis of XMM-Newton as well as Chandra data is supported by the Italian Space Agency(ASI). REFERENCES
[1]Arons,J.&Scharlemann,E.T.,1979,ApJ231,854
[2]Becker,W.;Aschenbach, B.,2002,MPE Report278. Edited by W.Becker,H.Lesch,and J.Trmper.Garching bei Mnchen:Max-Plank-Institut fr extraterrestrische Physik,64 [3]Bignami,G.F.&Caraveo,P.A.1996,Ann.Rev.Astron.As-trophys.34,331
[4]Caraveo,P.A.2000in”Pulsar Astronomy-2000and Be-yond”,ASP Conference Series,Ed.s M.Kramer,N.Wex,and N.Wielebinski,202,289
[5]Caraveo,P.A.,De Luca,A.,Mereghetti,S.,Pellizzoni,A., Bignami,G.F.,2004,Science305,376
[6]Caraveo,P.A.,Bignami,G.F.,De Luca,A.,et al.,2003, Science301,1345
[7]Caraveo,P.A.,De Luca,A.,Mignani,R.P.,Bignami,G.F., 2001,ApJ561,930
[8]Caraveo,P.A.,Mignani,R.P.,1999,A&A344,367
[9]Chatterjee,S.&Cordes,J.M.,2002,Ap.J.575,407
[10]De Luca,A.et al.2005a Ap.J,623,1051
[11]De Luca, A.et al.2005b A&A in press (astro-ph/0511185)
[12]Dyks,J.and B.Rudak,2003,Ap.J.,598,1201
[13]Gaensler,B.M.,van der Swaluw,E.,Camilo,F.,et al., 2004,ApJ616,383
[14]Gaensler,B.M.,2005,“Young Neutron Stars and Their En-vironment”,ASP Conf.Proc.,eds F.Camilo and B.M.Gaensler, p.1517)
[15]Goldreich,P.Julian,W.H.,1969,Ap.J.,157,869
[16]Halpern J.P.and M.Ruderman,1993,Ap.J.,415,286
[17]Harding,A.K.and J.K.Daugherty,1998,Adv.Space Res.21,251
[18]Harding,A.K.2005,Proceedings of the22nd Texas Sym-posium on Relativistic Astrophysics at Stanford University, Stanford,2004,edited by P.Chen et al.,eConf C041213
[19]Hartman,R.C.et al.,1999,Ap.J.S.,123,79
[20]Helfand,D.J.,Gotthelf,E.V.,Halpern,J.P.,2001,ApJ 556,380
[21]Lai,D.,Chernoff,D.F.,Cordes,J.M.,2001,ApJ549, 1111
[22]Manchester,R.N.et al.,2005,A.J.,129,1993
[23]P.F.Michelson SPIE,2003,4851,1144 https://www.wendangku.net/doc/c315967005.html,/
[24]Mignani,R.P.,de Luca,A.,Caraveo,P.A.,2004,IAU Symposium no.218Ed.s Fernando Camilo and Bryan M. Gaensler.Astronomical Society of the Paci?c,2004,391
[25]Mori et al.,2004,ApJ,609,186
[26]Muslimov,A.G.and A.K.Harding,2003,Ap.J588,430
[27]Ng,C.-Y.,Romani,R.W.,2004,ApJ601,479
[28]Pavlov,G.G.,Teter,M.A.,Kargaltsev,O.,Sanwal,D., 2003,ApJ591,1157
[29]Pellizzoni,A.,Mattana,F.,De Luca,A.,et al.,2005,“High Energy Gamma-Ray Astronomy”,eds.F.A.Aharonian,
H.J.Vlk,D.Horns,AIP Conference Proceedings,745,p.371
[30]Romani,R.W.,1996,Ap.J470,469
[31]Rudak,B.and J.Dyks,1999,Mon.Not.R.Astron.Soc. 303,477.
[32]Ruderman,M.A.&P.G.Sutherland1975Ap.J196,51
[33]Slane,P.,2005,Adv.Sp.Res.35,1092
[34]Stappers,B.W.,Gaensler,B.M.,Kaspi,V.M.,van der Klis, M.,Lewin,W.H.G.,2003,Science299,1372
[35]Sturrock,P.A.,1971Ap.J164,529
[36]Tavani,M.et al.,2003,SPIE,4851,1151 https://www.wendangku.net/doc/c315967005.html,r.it
[37]Thompson,D.J.,2001in High Energy Gamma-Ray As-tronomy,eds.F.A.Aharonian,H.Volk,AIP Conf.Proc.558, p.103112,2001.
[38]Zhang,B.and A.K.Harding,2000,Ap.J532,1150
[39]Willingale et al.,2001,A&A,365,L212