A Search for Pulsar Companions Around Extremely Low Mass White - - PowerPoint PPT Presentation

A Search for Pulsar Companions Around Extremely Low Mass White Dwarfs

Tilemachos M. Athanasiadis

14th BONN workshop

Supervisors: Dr. John Antoniadis, Prof. Dr. Michael Kramer

1 7 F e b . 2 0 2 0

• Usually
• Single star evolution needs more than a Hubble time to create a LMWD
• 50% of the LMWDs of ~0.4M☉ are expected to exist in binaries
• r ELM WDs should exist
• Possible dark companions:
• Another WD
• Black hole
• Optical surveys have discovered LMWDs and through optical spectroscopy orbital

parameters are measured.

LMWD+MSPs are ideal for through timing of the MSP and optical spectroscopy of the LMWD (for example Antoniadis et al. 2013). Very few NS mass measurements are currently available. Observational among the double-degenerate population. This information is a crucial missing input in stellar- evolution and population synthesis models.

Outline

ELM follow-up Survey

Effelsberg

GAIA follow-up Survey

Effelsberg

GAIA follow-up Survey

Arecibo

PROJECTS

• ELM survey

target selection:

• Sloan Digital Sky Survey (SDSS) photometric catalog by color.
• Objects with
• Our target selection:
• ELM

WDs in binaries with a dark companion with mass > 0.8M☉. for MSP companions observed by M. Berezina & L. Spitler (Effelsberg-2014)

• For these systems there are also spectroscopic observations

.

Athanasiadis et al. 2020 (in prep.)

Our SIGPROC-PRESTO pipeline are based on the HTRU-N pipeline (M. Cruces)

• The 3 most compact systems was observed

for 90 minutes each.

• 5 systems was observed multiple times

for 30 minutes per session.

• SIGPROC: Acceleration (A) and/or Jerk

(J) search applied based on the orbital period.

• Acceleration and jerk range based on

each system

• DM range: 0-100 (calc. based on NE2001)

90 min 30 min Porb*10% J0751-0141 J+A J+A A J0755+4800 n.o. A

• J0755+4906

J+A J+A A J0811+0225 n.o. A

• J1233+1602

n.o. J+A A J1443+1509 n.o. J+A A J1741+6526 J+A J+A A J2132+0754 n.o. A

• Inject simulated pulsar signals into

real Effelsberg noise.

• We have used that tool to have a

better understanding of the sensitivity of our survey.

• ~0.1 mJy error before RFI mitigation
• ~0.05 mJy error after RFI mitigation

Flux density Sv

We simulate 10000 companions for every LMWD in a specific distance d (GAIA)

using random Luminosity (Svd2) from distribution (Gonthier 2018) using a beaming fraction model (Tauris & Manchester 1998) using random Pspin from distribution (Lorimer 2015)

assuming a percentance of NSs within the companions based on the mass function

Spin period Ps Beaming Flux density Sv Flux density Sv Spin period Ps DETECTION RATE

Compare Sv with the survey sensitivity as function of Ps

MONTE CARLO SIMULATION OF COMPANIONS FOR EACH SYSTEM Assumption: The acceleration range that we use in

• ur search is enough to detect the systems that we

are looking for.

Probability for each system to host a NS.

Athanasiadis et al. (paper A) Athanasiadis et al. (paper A) Athanasiadis et al. 2020A Flux Density (mJy) Flux Density (mJy)

Athanasiadis et al. 2020A Flux Density (mJy) Flux Density (mJy)

Athanasiadis et al. 2020A Probability of PNS PNS detection mass func after obs

Precise proper motions for 1.5 billion objects in the Galaxy. ~500,000 WDs with hydrogen atmospheres, ~30,000 LMWDs with M < 0.25 solar masses Well known and small distances (<1.5 kpc) Different astrometric properties (due to the supernova kick) : high proper motion high galactic latitudes

EFFELSBERG GAIA FOLLOW UP SURVEY Telescope Effelsberg Radio Telescope Receiver 7-beam receiver (21 cm) Targets 104 selected GAIA white dwarfs

• Obs. time

2x30 min Sensitivity 0.125 mJy

NS fraction is the percentance of the LMWD/NS binaries compared to other double degenerate binaries. Binomial distribution: Upper limit based on van Leewen et al. 2007:

 

0073 . 2 / 1 1

det / 1

   P P

N NS

N NS

P P L ) 1 ( ~

det 



GAIA follow up survey 104 targets Upper limit: PNS < 0.031 ELM follow up survey 8 targets Upper limit: PNS < 0.35

Athanasiadis et al. 2020b

 

09 . 2 / 1 1

det / 1

   P P

N NS

GAIA survey ELM survey

• We have strong motivation to search for radio pulsars at the positions of low mass white

dwarfs.

• Both detections and non-detections are useful:
• Detection > Precise NS mass measurements
• Non-detections > Constraints on the LMWD/NS population
• Survey sensitivity, beaming fraction and distance are the most important factors regarding

the detection rate.

• Based on our results, we expect the fraction of LMWD/NS systems to be close to zero

and not higher than ~3%.

ARECIBO GAIA FOLLOW UP SURVEY Telescope Arecibo Radio Telescope Receiver L-wide (1.15-1.73 GHz) Targets 42 selected GAIA white dwarfs (some with Fermi counterparts)

• Obs. time

15 min Sensitivity 0.016 mJy for a 2ms pulsar at DM=100pc/cm3

TARGET SELECTION: High tranverse velocities High galactic latitudes Cross-matched with 3FGL Fermi catalog (<0.5o positional difference)

THANK YOU

BACKUP SLIDES

Beaming fraction model for MSPs

Tauris & Manchester (1998)

MSPs

Consistent with Kramer et al 1999 MSPs are considered to have large beaming fraction values

• Pdet= NMSPs/NNS
• We simulate a population of

NS companions for a LMWD in different (well known) parallaxes.

• We calculate how many

pulsars the survey would have detected.

• For systems with parallax

higher than 0.5 mas (d<2kpc) the Pdet is constant.

• Pdet cannot go higher than

~90% due to the beaming fraction.

Athanasiadis et al. 2020

• For each NS companion (MSP candidate), we assume

a random: from a spin distribution based on Lorimer 2015. based on Tauris & Manchester 1998. (based on Gonthier et al. 2018) within the distance error based on GAIA DR2.

• From the Luminosity (L) and distance (r) we

calculate the : (L/4πr2)

• Comparing the flux density with the sensitivity
• f the survey provide us with the

MSPs Lorimer (2008)

Athanasiadis et al. (paper A)

Athanasiadis et al. (paper A)

Li et al. 2019 Li et al. 2019

• Possible companions of LMWDs:
• Another WD
• Neutron star
• Black hole
• No hydrogen shell flashes occur during the ELM WD cooling stage in contrast with more massive WDs.
• In the latter case the hydrogen-rich envelope is loosing mass, therefore they have thiner envelopes.
• WDs have thicker envelopes which allow for significantly higher stable hydrogen burning rates,
• ELM WDs are much more luminous than their more massive companions (Driebe et al. 1999).

Athanasiadis et al. (paper A)

Athanasiadis et al. (paper A) Athanasiadis et al. (paper A)

Acceleration search on 10 min

• bservations

Sensitive to systems with Porb > 1,6 hours Acceleration range: ± 500 km/s2 DM range: 0-2000 The DATA are archived and easily accesible on HERCULES cluster. Useful scripts for easy retrieval and reprocessing FFA with acceleration search (in coop. with T. Gautam) Period range: 0.5-30 seconds random discoveries Single pulse search (PRESTO) Our SIGPROC-PRESTO pipelines are based on the HTRU-N pipeline (M. Cruces) Acceleration-Jerk on 30 min observ. Sensitive to systems with Porb > 5 hours Acceleration range: ± 100 km/s2 Jerk range: ± 4cm/s3 DM range: 0-2000

EFFELSBERG GAIA FOLLOW UP SURVEY Telescope Effelsberg Radio Telescope Receiver 7-beam receiver (21 cm) Targets 104 selected GAIA white dwarfs

• Obs. time

2x30 min Sensitivity 0.125 mJy

Test pulsars

Known MSPS-WD systems observed as tests of the pipeline:

Binary Period (ms) DM (pc*cm^-3) P_orb (hours) S1400 (mJy) J2053+4650 12.58 98.08 2.4 J1738+0333 5.85 33.77 8.5 0.67 J0751+1807 3.47 30.24 6.3 3.2 J0348+0432 3.9 40.46 2.4

Acceleration range

We need to be sure that the acceleration range that we use is enough for the

• bjects that we are looking for.

(Handbook, Lorimer & Kramer) Acceleration range depends on masses and the orbital period: Red: MSP+WD (1+0.25 solar masses) Green: NS+NS (1+2 solar masses) Black: BH+BH (10+10 solar masses)

Athanasiadis et al. 2019 (in prep.)

Expected Orbital Periods (MSPs+He WDs)

Tauris et al. 1999

ARECIBO OBSERVATIONS PART II

Graphical example of our selection method. Clustering based

• n a gaussian mixture model based on their transverse

velocity and galactic latitude. Our algorithm clusters nearby GAIA LMWDs into 4 distinct in the galactic latitute/velocity

• plane. Known MSPs (in green) belong exclusively in two of

these clusters.

Luminosity distribution for MSPs

Based on all-sky surveys carried

• ut in the 90s and for pulsars
• bserved at 430 MHz within 1.5

kpc of the Sun. Severe undersampling of low- luminosity pulsars. The observed (dashed line) and corrected (solid line) luminosity distribution for MSPs. Power law with a slope of -1 Lorimer 2008

• For every system:
• Known parameters:
• rbital period, mass function
• For random orientations:

MC simulation of companions

neutron stars 1.4 < M < 2.5 Ozel & Freire 2016

Athanasiadis et al. (paper A)

Spin Period distribution for MSPs

It's important to simulate accurately the spin period distribution We can compare with the sensitivity in specific spin periods. This distribution is gaussian centered at 4 ms Tauris 2015