A Curious Case of 4U 1700 37 · Introduction 4U 1700{37 is a massive eclipsing High Mass X-ray binary, with a O supergiant star in a short orbital period of 3.412 days. The nature

A Curious Case of 4U 1700–37

Nazma Islam

Nicholas Copernicus Astronomical Centre, Warsaw&

Biswajit PaulRaman Research Institute, Bangalore, India

Nazma Islam A Curious Case of 4U 1700–37 1 / 17

Introduction

• 4U 1700–37 is a massive eclipsing High Mass X-ray binary, with a Osupergiant star in a short orbital period of 3.412 days.

• The nature of the compact object is uncertain due to lack of any definitiveindication of a surface like pulsations, X-ray bursts or cyclotron scatteringfeature. However, it is likely to be a neutron star due to the similarities in thespectra with the other accreting X-ray pulsars.

• The mass of the compact object ∼ 2.4 M�, either a high mass neutron staror low mass black hole.

• Orbital evolution is studied for this system using eclipse timing due to lack ofX-ray pulsations and presence of eclipses.


Orbital evolution in X-ray binaries

• Orbital evolution in HMXBs occurs due to• Wind driven angular momentum loss from the high mass stars• Tidal interactions

• In addition to the orbital evolution, the elliptic orbits of the X-ray binariesalso undergo apsidal motion.

Eclipse Timing Analysis

For eclipsing X-ray binaries, mid-eclipse epochs are used as fiducial markers tostudy the change in orbital period of the binary system.


Present Work

We utilised the mid-eclipse epoch measurementsof 4U 1700–37 from archival observations as wellas the new measurements from all sky monitors

• 16 years of RXTE–ASM light-curves in 5–12keV energy band.

• 10 years of Swift–BAT light-curves in 15–50keV energy band.

• 5 years of MAXI light-curves in 5–20 keVenergy band

• 11 orbital cycles coverage of eclipse withRXTE–PCA

−2000 −1500 −1000 −500 0−0.3

−0.2

−0.1

0

0.1

Del

ay (

days

)

Number of Orbits

Uhuru

Copernicus

EXOSATWatch

BATSE

Figure : Plot of delay inmid-eclipse epochs with respect toa constant orbital period. Previousmeasurements of the orbitalevolution by Rubin et al.(1996),using 20 years of eclipse data,found an orbital period decay of10−6 yr−1


Eclipse profile of 4U 1700–37

0 0.5 1 1.5 20

0.5

1

1.5

Nor

mal

ised

Int

ensi

ty

Orbital Phase0.65 0.7 0.75 0.8

0

0.5

1

1.5

Nor

mal

ised

Int

ensi

ty

Orbital Phase

Figure : Left panel shows the orbital intensity profile of 4U 1700-37 created out of onesegment of Swift–BAT lightcurves. Right panel shows the zoomed in eclipse state alongwith ingress and egress. The eclipse state is modelled by an empirical box function withdifferent slopes for ingress and egress. Blue line denotes the current mid-eclipse epochand the red line denotes the expected mid-eclipse epoch from the orbital period decaymeasurements from Rubin et al.(1996).


Systematic error with proportional counters

• Presence of flares outside the eclipse leadsto an uncertainity in the mid-eclipse epoch.

• The error of 0.003 days quoted on theEXOSAT mid-eclipse epoch in Haberl et al.(1989), is an underestimate of the actualerror.

• The mid-eclipse times measured from thelong light-curves of the all sky monitors donot suffer from the uncertainties due toindividual flares, because of averagingeffects of continuous observations.

0.9 1 1.1

0

2

4

Nor

mal

ised

Int

ensi

ty

Orbital Phase

Figure : We have extracted theEXOSAT GSPC 8- 14 keVlight-curve and created the orbitalintensity profile (black line), whichis overlaid on the orbital intensityprofile created by 11 orbital cycleseclipse observation by RXTE PCA(red line)


Orbital period decay

−2000 −1000 0 1000 2000

−0.1

0

0.1

Del

ay (

days

)

Number of Orbits

Uhuru

Copernicus

EXOSAT

WATCH

BATSE

RXTE−ASM

RXTE−PCA

INTEGRAL

Swift BAT

MAXI

• Plot of delay in mid-eclipse epochs with respect to a constant orbital period.The black solid line represents the quadratic portion of the best fit to theepochs, corresponding to an orbital period decay P /P = −(4.7± 1.9)× 10−7

yr−1, smaller compared to its previous estimates of −3 × 10−6 yr−1 (blueline: Rubin et al. 1996) and −1.6 × 10−6 yr−1 (magenta line: Falanga et al.2015).

• The orbital period decay is smaller than that seen in other similar HMXBslike Cen X–3, SMC X–1, which have P /P ∼ 10−6 yr−1

Islam & Paul 2016, MNRASNazma Islam A Curious Case of 4U 1700–37 7 / 17

Mid-eclipse times as a function of apsidal motion

• Mid-eclipse timing measurements havepreviously been used to determine the rateof apsidal motion and other orbitalparameters in case of eccentric opticaleclipsing binaries.

• For a system having small eccentricity andundergoing apsidal motion, the mid-eclipsetimes can be written as (Gimenez &Garcia-Pelayo 1983 )

TN = T0 + PN +ePaπ

cos(ω0 + ∆ωN)

where ∆ω is the change in ω in one orbitalcycle.


Eclipse duration as function of apsidal angle

• For orbital inclination of 90◦, eclipseduration is given by

∆T (ω′) =

a2(1 − e2)2

L

∫ ω′+φ2

ω′−φ1

dθ

(1 + ecosθ)2

where L = (G(M? +MC)a(1 − e2))12 .

0 100 200 300

0.16

0.18

0.2

0.22

Ecl

ipse

Dur

atio

n (p

hase

)

ω’ in Degrees

e = 0

e = 0.05

e = 0.1

e = 0.22

Figure : Plot of variation of eclipse duration as a function of ω′

for different value ofeccentricity. The ratio of maximum to minimum eclipse duration is equal to 1+e

1−e.


Constraints on eccentricity

0 500 1000

−0.01

0

0.01

Del

ay (

days

)

Number of Orbits0 500 1000

0.17

0.18

0.19

0.2

0.21

Ecl

ipse

dur

atio

n (p

hase

)

Number of Orbits

Figure : Left panel: Delay in mid-eclipse times with respect to a constant orbital periodfor 10 segments of 1 year Swift–BAT light curves, along with solid lines showing themaximum and minimum value of the delay. Right panel: Plot of eclipse duration asfunction of orbit number calculated from 10 segments of 1 year Swift–BAT light curves,along with solid lines showing the maximum and minimum value of the eclipse duration.We use these plots to derive an upper limit on the value of eccentricity of 0.008 and 0.05respectively.

Islam & Paul 2016, MNRASNazma Islam A Curious Case of 4U 1700–37 10 / 17

Eccentricity of the binary system

• Lack of pulsations makes it difficult tomeasure the orbital parameters, includingeccentricity, by pulsar timing analysis.

• The companion star is a bright O supergiantstar. The orbital parameters can beestimated by radial velocity measurements.

• Due to extreme mass loss rate and highstellar wind from the companion star,previous attempts in measuring e haveyielded inconsistent results.

Figure : Value of e estimated byradial velocity measurements canbe as high as 0.22 and as low at 0(circular orbit). Reference:Hammerschlag-Hensberge et al.2003


SALT observation of 4U 1700–37

• We proposed a SALT observation to carryout radial velocity measurements of thecompanion O supergiant star.

• 13 observations were carried out distributedover different orbital phases, with someorbital phases revisited.

• The orbital parameters estimated from theseradial velocity measurements, especially e,will provide new estimates on the mass ofthe compact object. Work in progress.

Figure : Preliminary results:Observed spectra and best fitmodel for one of the observation.


ASTROSAT observation of 4U 1700–37

• A 40 kilosec ASTROSAT–LAXPCobservation of 4U 1700–37 was done, tocarry out the deepest search for pulsationand/or cyclotron line.

• ASTROSAT consists of 5 payloads forsimultaneous multi-wavelength observations

• Ultra-violet Imaging Telescope coveringoptical to far UV (130 nm to 530 nm).

• Soft X-ray Telescope in the energy range0.3-8 keV.

• Large Area Xenon Proportional Countersin the energy range 3-80 keV.

• Cadmium Zinc Telluride Imager in theenergy range 10-150 keV.

• Scanning Sky Monitor in the energy range2-10 keV.


Large Area Xenon Proportional Counter

• 3 co-aligned Xenon proportional counterswith collimators, having a total effectivearea of 6000 cm2 at 5 keV.

• Wide energy band: 3-80 keV. Timingresolution of 10 µsec.


Preliminary analysis

500

1000

coun

ts/s

ec lxp1

500

1000

coun

ts/s

ec lxp2

0 2×104 4×104 6×104 8×104

200

400

600

800

1000

coun

ts/s

ec

Time (s)

lxp3

Start Time 17654 15:00:55:660 Stop Time 17655 15:53:15:661

Figure : LAXPC lightcurve of 4U 1700–37, with a binsize of 1 sec. The backgroundcount-rate is shown by the solid line. Work in progress


Preliminary analysis

200 400 600 800 1000

10−4

10−3

0.01

0.1

1

norm

aliz

ed c

ount

s s

−1

chan

−1

Channel

Figure : LAXPC-3 spectrum of 4U 1700–37, with along with the background spectrum.Work in progress


Summary

• We have investigated the long term orbital period evolution of 4U 1700–37 byeclipse timing and found the orbital period decay P /P = −(4.7± 1.9)× 10−7

yr−1, smaller compared to its previous estimates.

• The orbital period decay occurs due to stellar wind driven angular momentumloss coupled with strong tidal interaction between the binary components.However, the rate of orbital period decay is slower than that seen in otherX-ray binaries like Cen X–3, SMC X–1, undergoing tidal interactions ofsimilar strength.

• Using delay in mid-eclipse times and eclipse duration from 10 year ofSwift-BAT data, we have independently constrained the value of eccentricity.

• SALT observation of the companion star to determine the orbital parametersusing radial velocity measurements. Estimation of the mass of the compactobject.

• ASTROSAT LAXPC observation of 4U 1700–37, to search for the surface ofthe compact object.

• Watch out this space.


Documents

A Curious Case of 4U 1700 37 · Introduction 4U 1700{37 is a massive eclipsing High Mass X-ray binary, with a O supergiant star in a short orbital period of 3.412 days. The nature