[AIT logo]

Institut für Astronomie und Astrophysik

Abteilung Astronomie

Sand 1, D-72076 Tübingen, Germany
[Uni logo]

Hinweis: Einige Seiten auf astro.uni-tuebingen.de können veraltet sein und werden nicht mehr aktualisiert.
Note: Some webpages at astro.uni-tuebingen.de may be outdated and will no longer be updated.


Dissertation Victor Doroshenko


Magnetic fields of accreting pulsars

Summary.
Accreting pulsars are rotating, highly magnetized neutron stars in binary systems which emit pulsed X-rays. This emission is powered by the gravitational energy of the plasma accreted from a non degenerate companion funneled onto the polar caps of the neutron star by the magnetic field, and thermalized to X-rays either in the impact with the surface of the neutron star, or in the so-called accretion column. Although discovered more than forty years ago many aspects of the emission form accreting pulsars remain puzzling. One of the key open issues is the magnetic field of the neutron stars in these systems and how we can measure it. The X-ray spectrum of these systems is characterized by a cut-off power-law, interpreted as the result of the comptonization of the bremstrahlung and cyclotron emission produced in the highly magnetized accreted plasma in the vicinity of the polar caps by the hot in-falling plasma. The theoretical description of the spectra is, however, far from complete. The presence of a strong magnetic field (B > 1 TG) constrains and quantizes the motion of the electrons perpendicularly to the field in the Landau levels, thus modifying the cross-section for Compton scatterings. The scattering cross-section strongly depends on the photon momentum and the magnetic field which manifests as the so-called cyclotron resonance scattering features (CRSFs), observed as line-like absorption features in the spectra of some pulsars. The centroid energy of these features is directly related to the magnetic field in the line-forming region via E[keV]~11.57 B[TG], providing a way to directly measure the magnetic field of a neutron star. The timing properties of a neutron star, and more specifically the changes of the pulse spin frequency, can be also used to estimate its magnetic field. Neutron stars are very compact objects with a low intrinsic moment of inertia and therefore their spin is completely governed by the external torques. These torques depend on the interaction between the magnetosphere, determined by the strength of the field, and the accreted matter. Therefore the study of these torques and the spin history can provide an estimate of the magnetic field.
In this thesis the role of these torques in the spin history of a sample of slowly rotating pulsars and the magnetic fields inferred from the torque theory has been studied. More specifically, I present a detailed study of the spectral and timing behavior of three slow-rotating accreting pulsars, namely 1A 1118-61, GX 301-2 and Vela X-1. Based on RXTE data, I first report on the discovery of a cyclotron scattering feature in the spectrum of 1A 1118-61. I also perform a detailed study of the timing parameters of 1A 1118-61 and suggest a possible measurement of the previously unknown orbital period of the source. Using the data from INTEGRAL and RXTE, I present an analysis of the well known High Mass X-ray Binaries GX 301-2 and Vela X-1 confirming with high statistics the already known spectral properties (including the observation of CRSFs) of the sources. Based on Suzaku data, I then report on the discovery of pulsed emission in the "off-states" of Vela X-1. This was a surprising result considering that until now it was thought that the X-ray emission was totally suppressed in the "off-states". I discuss these findings in the context of recent theories for "gated" accretion. I then focus on the central question of this work, i.e. on how the timing properties of GX 301-2 and Vela X-1 may help to constrain the torques affecting the neutron star and to estimate its magnetic field. The predictions of several existing torque models on the spin history of the sources are compared with CGRO BATSE observations. These observations reveal, for the first time, a correlation between the pulse frequency derivative and the accretion rate, as predicted in all torque models. Using this correlation, we can put a strong lower limit B > 10 TG on the magnetic field strength of the neutron stars in Vela X-1 and GX 301-2. This result contradicts the estimates of the magnetic field obtained from the ob- served energies of the CRSFs in these sources, which once again brings onto the stage the long-standing problem of the existence of luminous long-periodic accreting pulsars with moderate magnetic field. Possible ways to reconcile this contradiction are discussed. First I suggest that CRSFs may form high above the neutron star in the accretion column or the accretion stream above the polar caps where the magnetic field considerably weakens. Second, I critically revise possible modifications of the existing torque theory and from basic conservation arguments conclude that the magnetic field obtained on the basis of current torque models is probably not significantly overestimated. Finally I discuss how the result on GX 301-2 and Vela X-1 can be further verified, and why similar considerations might apply to other accreting pulsars.

Zusammenfassung
Akkretierende Pulsare sind rotierende Neutronensterne mit einem starken Magnetfeld, die sich in einem Doppelsternsystem befinden. Solche Systeme geben gepulste Röntgenstrahlung ab. Diese Emission wird durch die Gravitationsenergie gespeist, die freigesetzt wird, wenn Plasma von einem nicht entarteten Begleitstern durch das Magnetfeld auf die Pole des Neutronensterns geleitet wird. Das Plasma wird durch den Aufprall auf die Oberfläche des Neutronensterns oder in der sogenannten Akkretionssäule zu Röntgenstrahlung thermalisiert. Eine der offenen Fragen ist das Magnetfeld der Neutronensterne in diesen Systemen an sich und wie man jenes messen kann. Das Röntgenspektrum dieser Systeme ist gekennzeichnet durch ein Potenzgesetz mit einem Abbruch. Das Potenzgesetz wird als das Ergebnis aus der Comptonisierung der Bremsstrahlung und der Zyklotronstrahlung angesehen. Die theoretische Beschreibung des Spektrums ist jedoch bei weitem noch nicht vollständig. Die Anwesenheit eines starken Magnetfeldes (B > 1 TG) beschränkt die Bewegung der Elektronen senkrecht zum Magnetfeld in Landau-Niveaus. Diese sind quantisierte Energiezustände in denen sich die Elektronen befinden können. Der Wirkungsquerschnitt der Compton-Streuung verändert sich dadurch, da dieser von dem Photonenimpuls und dem Magnetfeld abhängt. Diese Abhängigkeit zeigt sich in Zyklotronresonanzlinien (engl. cyclotron resonance scattering features, im folgenden CRSFs). Die Schwerpunktsenergie dieser Linien steht über E[keV] ~ 11.57 B[TG] in direkter Beziehung zu dem Magnetfeld im Entstehungsbereich der Linien. Die temporalen Eigenschaften von Neutronensternen, speziell die Änderung der Spinpuls Frequenz, dienen ebenfalls zur Abschätzung des Magnetfeldes. Da Neutronensterne sehr kompakte Objekte mit einem geringen intrinsischen Trägheitsmoment sind, wird der Spin komplett von den äußeren Drehmomenten beherrscht. Die Studie dieser Drehmomente und der Spinentwicklung kann daher ebenfalls eine Abschätzung des Magnetfeldes liefern.
Ich stelle eine detaillierte Studie des spektralen und zeitlichen Verhaltens dreier langsam rotierender, akkretierender Pulsare vor. Namentlich sind dies 1A 1118-61, GX 301-2 und Vela X-1. Basierend auf RXTE Daten gehe ich zuerst auf die Entdeckung einer CRSF im Spektrum von 1A 1118-61 ein. Ebenfalls führe ich eine genaue Betrachtung der temporalen Parameter von 1A 1118-61 durch und zeige eine Möglichkeit auf um die bisher unbekannte Umlaufdauer dieses Systems zu bestimmen. Darauf folgend präsentiere ich die Analyse der Beobachtungen der bekannten High Mass X-ray Binaries GX 301-2 und Vela X-1, die mit INTEGRAL und RXTE durchgeführt wurden. Da die Beobachtungen eine gute Statistik bieten, können die bekannten spektralen Eigenschaften der Quellen bestätigt werden. Desweiteren berichte ich von der Entdeckung, basierend auf Suzaku Daten, gepulster Röntgenstrahlung im "off-state" von Vela X-1. Dies ist ein überraschendes Ergebnis, da bisher angenommen wurde, dass die Röntgenstrahlung im "off-state" komplett unterdrückt wird. Ich diskutiere diese Resultate im Zusammenhang mit aktuellen Theorien der "gated" Akkretion. Anschliessend lege ich das Hauptaugenmerk auf die zentrale Frage dieser Arbeit, wie die temporalen Eigenschaften von GX 301-2 und Vela X-1 möglicherweise helfen, den Einfluss der Drehmomente auf den Neutronstern einzugrenzen und eine Abschätzung für das Magnetfeld des Neutronensterns zu liefern. Die Vorhersagen von mehreren bestehenden Modellen des Drehmoments für die Spinentwicklung dieser Quellen werden mit Beobachtungen von CGRO BATSE verglichen. Zum ersten Mal kann eine Korrelation zwischen der Ableitung der Spinfrequenz und der Akkretionsrate gezeigt werden, die in allen Modellen vorausgesagt wird. Mit Hilfe dieses Korrelation können wir eine gesicherte untere Grenze B > 10 TG für die Magnetfeldstärke der Neutronensterne in GX 301-2 und Vela X-1 angeben. Dieses Ergebnisse widersprechen den Abschätzungen des Magnetfeldes, die mit den Energien der CRSFs der Quellen gemacht wurden. Dieser Widerspruch bringt erneut das alte Probleme der Existenz von leuchtkräftigen lang-periodischen akkretierenden Pulsaren mit moderatem Magnetfeld ins Spiel. Möglichkeiten diese Widersprüche in Einklang zu bringen werden diskutiert. Zuerst zeige ich die Möglichkeit auf, dass CRSFs hoch oberhalb des Neutronsterns in der Akkretionssäule oder in dem Akkretionsstrom oberhalb der Polkappen entstehen, in einem Bereich, in dem sich das Magnetfeld erheblich abschwächt. Zusätzlich überprüfe ich kritisch mögliche Abwandlungen bestehender Theorien der Drehmomente und folgere auf Grund von grundlegenden Erhaltungsargumenten, dass man durch die Verwendung gängiger Drehmoment-Modelle möglicherweise die Magnetfeldstärke nur unwesentlich zu hoch ansetzt. Schließlich diskutiere ich wie die Ergebnisse für GX 301-2 und Vela X-1 zusätzlich bestätigt werden können und warum ähnliche Überlegungen auch auf andere akkretierende Pulsare angewendet werden könten.
 

Online-Publikation: http://nbn-resolving.de/urn:nbn:de:bsz:21-opus-57289


[Home Page] [PhD theses / Dissertationen] [Quick Reference] [Feedback]


| Impressum
Last modified 02 Nov 2011
[Valid HTML 4.0!]