NASA Technical Reports Server (NTRS) 20120013106: Discovery of an Unidentified Fermi Object as a Black Widow-Like Millisecond Pulsar

0.16 MB·English

by  NASA Technical Reports Server (NTRS)

Draftversion January19,2012
PreprinttypesetusingLATEXstyleemulateapjv.5/2/11
DISCOVERY OF AN UNIDENTIFIED FERMI OBJECT AS A BLACK WIDOW-LIKE MILLISECOND PULSAR
A. K. H. Kong1,10, R. H. H. Huang1, K. S. Cheng2, J. Takata2, Y. Yatsu3, C. C. Cheung4, D. Donato5,6,
L. C. C. Lin7, J. Kataoka8, Y. Takahashi8, K. Maeda8, C. Y. Hui9, P. H. T. Tam1
Draft version January 19, 2012
ABSTRACT
The Fermi Gamma-ray Space Telescope has revolutionized our knowledge of the gamma-raypulsar
2 population, leading to the discovery of almost 100 gamma-ray pulsars and dozens of gamma-ray
1 millisecond pulsars (MSPs). Although the outer-gap model predicts different sites of emission for
0 the radio and gamma-ray pulsars, until now all of the known gamma-ray MSPs have been visible in
2
the radio. Here we report the discovery of a “radio-quiet” gamma-ray emitting MSP candidate by

n using Fermi, Chandra, Swift, and optical observations. The X-ray and gamma-ray properties of the
a sourceareconsistentwithknowngamma-raypulsars. Wealsofounda4.63-hrorbitalperiodinoptical
J and X-ray data. We suggest that the source is a black widow-like MSP with a ∼ 0.1M(cid:2) late-type

7 companion star. Based on the profile of the optical and X-ray light-curves, the companion star is
1 believed to be heated by the pulsar while the X-ray emissions originate from pulsar magnetosphere
and/orfromintra-binaryshock. No radiodetection ofthe source has been reportedyetand although

] nogamma-ray/radiopulsationhasbeenfound,weestimatedthatthespinperiodoftheMSPis∼3−5
E
ms based on the inferred gamma-ray luminosity.
H
Subject headings: binaries: close – Gamma rays: stars – pulsars: general – stars: individual
. (1FGLJ2339.7–0531,SDSSJ233938.74-053305.2)– X-rays: stars
h
p
-
o 1. INTRODUCTION ray pulsars identified from the energetic class using a
r The Fermi Large Area Telescope (LAT) has detected blind fourier search of the LAT data (see Pletsch et al.
t 2011andreferencestherein),onlyfourweresubsequently
s 1873 γ-ray point sources during its first 24 months of
a identified as radiopulsarsby folding the radiodata with
operation(Abdoetal. 2011). Themajorityoftheextra-
[ the pulsar ephemerides (Camilo et al. 2009;Saz Parkin-
galactic sources have been identified as Active Galactic
son et al. 2010; Pletsch et al. 2011). This indicates
1 Nuclei (AGN; Abdo et al. 2010a), while many of the
thatthere is a populationof“radio-quiet”energeticpul-
v Galactic sources has been identified as γ-ray emitting
sarsas seen from the Earth. However,all 21 gamma-ray
9 pulsars (Abdo et al. 2010b). Gamma-ray pulsars are
emitting MSPs are “radio-loud”. Based on current ob-
2 typically either young and energetic, such as the Crab
servations, we can classify the current population of γ-
6 andVelapulsars,orveryshortperiodmillisecondpulsars
3 (MSPs). Throughoutthispaperwewillrefertothesetwo ray emitting pulsars into “radio-loud” energetic pulsars,
. classes as the energetic and millisecond pulsars. In this “radio-quiet” pulsars and “radio-loud” MSPs. However,
1
itisunclearifthereisaclassof“radio-quiet”γ-rayemit-
0 secondsource catalog(2FGL), there aremore than hun-
ting MSPs.
2 dred identified energetic and millisecond pulsars. About
It is well known that the radio emissions of pulsars
1 half ofthem with gamma-raypulsations were discovered
are associated with the activity of polar cap accelera-
: previously by using radio pulsation search(e.g. Ransom
v tor. On the other hand, the high-energy emissions from
et al. 2011;Keith et al. 2011;Cognardet al. 2011;Car-
i the pulsar magnetosphere have been studied with po-
X aveo2010). In addition, althoughthere were 35 gamma-
lar cap model (Ruderman & Sutherland 1975), slot gap
r model(Muslimov & Harding 2003)andouter gapmodel
a 1Institute of Astronomy and Department of Physics, (Takata, Wang & Cheng 2010a). The polar cap model
National Tsing Hua University, Hsinchu 30013, Taiwan;
assumes the emission region is close to the stellar sur-
[email protected]
2DepartmentofPhysics,TheUniversityofHongKong,Hong face and above the polar cap, and therefore the model
Kong implies radio-loudγ-raypulsarsare muchmorecommon
3Department of Physics, Tokyo Institute of Technology than radio-quiet γ-ray pulsars. The outer gap and slot
2-12-1,Ookayama,Meguro,Tokyo,152-8551, Japan
4National Research Council Research Associate, National gap models assume an acceleration region extending to
AcademyofSciences,Washington, DC20001, residentatNaval the outer magnetosphereand indicates radio-quietγ-ray
ResearchLaboratory,Washington, DC20375,USA pulsarsarethemajorpopulationofγ-raypulsars. Based
5CRESST and Astroparticle Physics Laboratory onthe spectralshape in the GeV bands andthe popula-
NASA/GSFC,Greenbelt, MD20771,USA
tionofradio-quietγ-raypulsarsfoundwithFermi(Abdo
6Department of Astronomy, Universityof Maryland, College
etal. 2010b),theoutergaporslotgapregionismorefa-
Park,MD20742,USA
7General Education Center, China Medical University, vorableasthe originofthe γ-raysfromenergeticpulsars
Taichung40402, Taiwan (Takata, Wang & Cheng 2010).
8Research Institute for Science and Engineering, Waseda
ForMSPs,“radio-quiet”millisecondγ-raypulsarshave
University,3-4-1Okubo,Shinjuku,Tokyo169-8555, Japan
9Department of Astronomy and Space Science, Chungnam not been identified so far. This may be due to following
NationalUniversity,Daejeon,RepublicofKorea reasons: (1) the present sensitivity of blind frequency
10GoldenJadeFellowofKendaFoundation, Taiwan
2
search prevents a detection of millisecond pulsation in 053111.
Fermi data; (2) most of known radio MSPs are in bi-
naries that gamma-ray searches would be insensitive to, 3. MULTIWAVELENGTHIDENTIFICATION
and (3) γ-ray emissions from MSPs always accompany 1FGLJ2339.7–0531 is one of the 205 bright gamma-
radio emissions. On the other hand, Venter, Harding ray sources detected with Fermi LAT during its first
& Guillemot (2009) found that the pulse profiles of six three months of operation (Abdo et al. 2009). It re-
MSPsdetectedbytheFermicanbe fittedbythegeome- mainsasanunidentifiedsourceinthesecondFermi-LAT
trieswithoutergaportheslotmodels. Thisimpliesthat source (2FGL) catalog. In the 1FGL catalog (Abdo et
the emission regions of the γ-rays are different from the al. 2010c), 1FGLJ2339.7–0531 has a variability index
radioemissionsiteforsomeoftheMSPsandthusweex- of 9.2 and a curvature index of 22.7. By comparing
pect that there are “radio-quiet” γ-ray MSPs for which the 1FGL and 2FGL catalog (Abdo et al. 2011), the
the radio beam is outside the line of sight. It is worth gamma-ray flux of 1FGLJ2339.7–0531 is constant and
notingthat“radio-quiet”heredoesnotmeannointrinsic near3×10−11 ergss−1 cm−2. Withbetterstatistics,the
radio emission, just none beamed in our direction. second source catalog shows that a log parabola spec-
In this Letter, we report a multiwavelength identifica- trum can provide a better spectral fit comparing to a
tion of a “radio-quiet” γ-ray MSP candidate. In a sepa- power-law model (Abdo et al. 2011). With its high
rate paper, Romani and Shaw (2011) arrived at similar Galactic latitude (−62◦) and unidentified nature in all
conclusions via optical spectroscopy. Fermi catalogs, 1FGLJ2339.7–0531 is therefore one of
the best targets to search for “radio-quiet” γ-ray MSPs.
2. UNIDENTIFIEDFERMISOURCEASA“RADIO-QUIET” 1FGLJ2339.7–0531 was observed with Chandra and
γ-RAYEMITTINGMILLISECONDPULSAR Swift/XRT for 21ks on 2009 October 13 (PI: Cheung),
To identify suitable targets of “radio-quiet” γ-ray and 3.2ks on 2009 November 4, respectively. The Chan-
MSPs, we first selected candidates from the Fermi LAT dra imaging was operated with ACIS-I while the XRT
first source catalog (1FGL; Abdo et al. 2010c) based on was in the photon counting imaging mode. We repro-
four criteria: 1) No known association at other wave- cessedwithupdatedcalibrationfiles. HEASOFTversion
lengths; 2) source variability; 3) Galactic latitude, and 6andCIAOversion4.2wereusedfordatareductionand
4) gamma-ray spectral shape. analysis. Within the 95% error circle of Fermi (based
We used the variability index in the 1FGL catalog to on the refined position in the 2FGL catalog), there is
characterize source variability. Gamma-ray pulsars have only one relatively bright X-ray source in both obser-
always been found to be steady sources of gamma-ray vations. The brightest X-ray source (CXOU J233938.7-
emission (Abdo et al. 2010c,2011). This property can 053305) near the center of the error circle has an X-ray
therefore be used to help identify which of the uniden- flux of 3×10−13 ergs s−1 cm−2 (0.3–10 keV) based on
tified sources are probably pulsars. The 1FGL catalog a spectral fitting using an absorbed power-law with a
definesavariabilityindex,forwhichavaluegreaterthan best-fit photon index of 1.1 (based on the Chandra ob-
23.21 means that there is less than a 1% probability of servation). This indicates that the ratio between the X-
beingasteadysource. Wethereforeselectedsourceswith ray and γ-ray (> 100 MeV) flux (Fγ ∼ 3×10−11 ergs
a variability index less than 23. s−1 cm−2) as measured by Fermi (Abdo et al. 2011)
MSPsareingeneralolderthanenergeticγ-raypulsars. become FX/Fγ ∼ 0.01, which is consistent with typi-
Young objects like energetic γ-ray pulsars (τ < 1 Myr)
cal observed values for γ-ray pulsars. It is worth not-
are likely located in the Galactic plane, while a frac-
ing that there are 9 much fainter Chandra sources in
tionof MSPs shouldbe at higherGalactic latitudes. We
theerrorcircleandtheir X-ray-to-gamma-rayfluxratios
thusselectedFermisourceswithhighGalacticlatitudes
are less than 0.1%. Although we cannot totally rule out
(|b| > 40◦). The choice of this value is based on Monte-
their associationwith the gamma-raysource, such a low
Carlo simulations for the Galactic population of MSPs,
flux ratio is not typical. We therefore identified CXOU
which show that no energetic γ-ray pulsar exists above
J233938.7-053305 as the potential X-ray counterpart to
|b|=40◦ (Takata, Wang & Cheng 2011).
1FGLJ2339.7–0531.
Finally, we identified potential candidates from the γ-
Within the Chandra error circle (0.6 arcsec at 90%
ray spectra. Although only a power-law spectrum is level) of CXOU J233938.7–053305, there is a R ∼ 19
listed in the first-year catalog, it also has a curvature
star from the USNO catalog (Monet et al. 2003). The
index to indicate how good of a power-law fit. For γ-
sameopticalsourceisalsoseenintheSDSSDataRelease
ray pulsars, their γ-rayspectra is usually described by a 8 images as SDSSJ233938.74-053305.2 with u(cid:4) = 20.85,
power-lawplus an exponentialcutoff model (Abdo et al. g(cid:4) = 19.0, r(cid:4) = 18.61, i(cid:4) = 18.25, and z(cid:4) = 18.23, as
2010b). According to the 1FGL catalog, the curvature
well as in the GALEX images with NUV=22.88 (177–
index, for which a value greater than C = 11.34 indi-
283 nm). Swift/UVOT observations taken simultane-
cates less than 1% chance that the power-law spectrum
ously with the X-ray observations detected the source
is a good fit. Hence we chose objects with a curvature
with U =19.58, B =19.53, and V =18.88. The X-ray-
index larger than 12.
to-optical flux ratio (390) is much too high for a fore-
If a 1FGL source satisfied all four criteria, we short-
ground star and an AGN (e.g. Green et al. 2004, Laird
listed as a potential candidate. From the short list,
et al. 2009). Moreover, the spectral energy distribution
we further searched for X-ray imaging data (Chandra,
from UV to optical indicates that it is unlikely to be an
XMM-NewtonandSwift)frompublicarchiveandlooked
for X-ray sources within the error circles of Fermi as 11 Wewillusethe1FGLnamingconversionthroughoutthepa-
the first step for multiwavelength investigation. In this per,andthesourceiscalled0FGLJ2339.8-0530and2FGLJ2339.6–
paper, we focus on one of our targets, 1FGLJ2339.7– 0532inthebrightsourcecatalogand2FGLcatalog,respectively.
3
(cid:0)
of 1.1. Such a hard spectrum is not typical for X-ray bi-
(cid:2) nary, AGN, and foreground star. However, some MSPs
(cid:11) show similar spectral characteristics (e.g., Archibald et
(cid:21)
(cid:13)(cid:9)(cid:14)(cid:20) (cid:3) al. 2010; Tam et al. 2010) and such a hard spectrum
(cid:19) could be from a pulsar wind nebula. We propose that
(cid:18)(cid:15) (cid:4) (cid:29)(cid:8)(cid:29)(cid:29)(cid:17)(cid:30)(cid:5)(cid:4)(cid:4)(cid:31)(cid:4) !"(cid:3)#(cid:7)(cid:2)(cid:4)(cid:4)(cid:7)(cid:2)!(cid:5) 1FGLJ2339.7–0531 is a gamma-ray emitting MSP in a
(cid:15)(cid:16)(cid:17)
(cid:13)(cid:14)(cid:9) (cid:5) binary system and the optical emission is from a late-
(cid:11)(cid:12)(cid:11) type companionstar with contribution from the heating
(cid:8)(cid:9)(cid:10)(cid:10) (cid:6) on the stellar surface.
$%(cid:18)&(cid:15)(cid:12)(cid:9)(cid:23)%(cid:13)(cid:17)(cid:23)(cid:14)(cid:15)(cid:12) Theopticallight-curveshowsclearmodulationatape-
(cid:7)
riod of 4.6342 hours (Figure 2). It is beyond any doubt
(cid:6)(cid:7) (cid:6)(cid:6) (cid:6)(cid:5) (cid:6)(cid:4) (cid:6)(cid:3) (cid:6)(cid:2) (cid:6)(cid:0)
that the periodicity is associatedwith the orbitalperiod
(cid:22)(cid:9)(cid:18)(cid:11)(cid:17)(cid:23)(cid:9)(cid:13)(cid:24)(cid:11)(cid:17)(cid:5)(cid:7)(cid:6)(cid:7)(cid:17)(cid:25)(cid:24)(cid:14)(cid:17)(cid:4)(cid:6)(cid:17)(cid:26)(cid:27)(cid:22)(cid:28)
ofthesystem. Giventhelargevariation(about3–4mag-
nitudes) in the optical light-curve, the inclination must
Figure 1. Representative optical (white light) light-curve of the
optical counterpart to 1FGLJ2339.7–0531, as observed by the be high. Better photometric data in the near infrared
Lulin 1m telescope in Taiwan. The differential magnitudes are and light-curve modeling will be able to constrain the
derived by comparison with several comparison stars in the field. inclination in the future. Assuming the companion has
Periodicityonatimescaleof4–5hoursisclearlyseen. Alsoshown
filleditstheRochelobe,themasslimitofthecompanion
is the light-curve of a comparison star (with an average error of
0.01–0.02magnitudes). Thetimingresolutionis2min. starcanbe estimatedfromthe orbitalperiod. Fora 4.6-
hr orbital period, the mass for a normal main-sequence
AGN (e.g. Richards et al. 2002). Instead, it looks like a star is <0.5M(cid:2) (Wilson et al. 1999).
F- or G-type star. In a separatepaper (Romani & Shaw The large optical variability and hints of color vari-
2011), they show via optical spectroscopy that the stel- ation (Fig. 2) suggest that the companion is being
lar object is consistent with a late-type star. We also heated by the pulsar. The irradiation of pulsar γ-ray
had a long-term optical monitoring program with the emissions onto the companion star can produce the or-
MITSuME 50cm telescope (Kotani et al. 2010) located bital modulation of the optical emission from the bi-
at Akeno Observatoryin Japanfrom 2010 September to nary system (e.g. Takata, Cheng & Taam 2010, 2011).
2010 November. MITSuME equips with a tricolor cam- In this model, the optical maximum occurs at the in-
era that can perform simultaneous imaging in the g(cid:4), R, ferior conjunction, where the pulsar is located between
andI bands. ItisclearfromTable1thatallthreebands the companion star and the Earth. Furthermore, dur-
show noticeable variability. Based on the above obser- ing the optical maximum, the color of the star tends
vations, we believe that the gamma-ray/X-ray source is to be bluer because of the heating effect. There is
likely a binary system with optical emission from the some indication from the MITSuME data that the color
irradiationofthecompanionstar,whilethecentralcom- of the star has become bluer during the optical maxi-
pact object is responsible to the X-ray and gamma-ray mum (Fig. 2). The maximum luminosity of the optical
emission. emission can be estimated as δLopt ∼ (πθ2/δΩ)Lγ ∼
Because 1FGLJ2339.7–0531/CXOU J233938.7– 1031(θ/0.1rad)2(δΩ/3rad)−1(Lγ/1033 ergss−1) ergss−1,
053305 may be a new class of interesting object, we
where θ is the angular size of the companion star mea-
carried out an intensive optical monitoring campaign
sured from the pulsar and δΩ is the solid angle of the
using the 1m telescope at the Lulin Observatory in
γ-raybeam. TheobservedmaximumR-bandmagnitude
Taiwan and the 0.8m Tenagra Telescope in Arizona
of 1FGLJ2339.7–0531is about 18 (see Table 1) which is
(see Table 1 for the observation log). All images are
consistent with the above estimation.
flat-field and bias corrected and we performed relative
It has been suggestedthat X-ray emissions from black
photometry by comparingwith severalcomparisonstars
widow systems originate from pulsar magnetosphere
in the field. In all 5 nights, the light-curves clearly show
and/or from intra-binary shock due to the interaction
variability on a timescale of 4–5 hours (see Figure 1
between the pulsar wind and the injected material from
& 2). We then performed a timing analysis by using
the low mass star (e.g., Kulkarni et al. 1992, Stappers
the Lomb-Scargle periodogram on the combined optical
et al. 2003). If the X-raysare produced in the magneto-
data and found a period of 4.6342(9) hours. Figure
sphere, the orbital modulation will be produced due to
2 shows the folded light-curves of all the optical data
the physicaleclipse of the pulsar by the companionstar.
(upper panel) and Chandra data (lower panel) at a
In this case, the minima of the X-ray and optical light-
period of 4.6342 hours. The phase zero is chosen at
curves should be at the same phase, and we indeed ob-
MJD 55500. In addition, we folded the optical colors
servedthat(seeFig. 2). Alternatively,iftheX-raysorig-
(g(cid:4) − I) obtained with MITSuME and X-ray hardness
inate fromthe intra-binaryshock region,aneffect of the
ratio (1.5–8keV/0.3–1.5keV)at the same period (Figure
Doppler boosting relating with the post-shock flow can
2).
produce the orbital modulation (Arons & Tavani 1993).
It has been suggested for the pulsar and massive stellar
4. DISCUSSION
system (e.g. PSR B1259-63/LS2883 system, Bogovalov
Using multi-wavelength data, we found an X- et al. 2008), the post-shock flow can be accelerated into
ray/optical counterpart to 1FGLJ2339.7–0531. The X- relativistic regime because of an explosion in the down-
ray-to-gamma-rayfluxratioisconsistentwithagamma- stream region (see also Tam et al. 2011). The Doppler
raaypopwuelrs-alrawwhwilheicthheisgacmommma-ornalyysspeeecntriunmMdSePvsia.tTeshferoXm- boosting increases the intensity as Iν ∝ D3+αIν(cid:2), where
(cid:2)
I is the intensity in the comoving frame, D is Doppler
rayspectrumisveryhardwithapower-lawphotonindex ν
4
Table 1
OpticalObservationlogof1FGLJ2339.7–0531
DateandTime(UT) Telescope Filter Exposurea
2010-10-2514:44–15:57 Lulin1m V 2min
2010-10-2613:33–15:59 Lulin1m V 2min
2010-10-3110:23–15:53 Lulin1m White 2min
2010-11-0111:33–16:27 Lulin1m White 2min
2010-11-1101:45–06:20 Tenagra0.8m White 5min
2007-09-1322:32:08 GALEX NUV=22.88 1642sec
2008-11-2005:25–05:29 SDSS u(cid:2)=20.85,g(cid:2)=19.0,r(cid:2)=18.61,i(cid:2)=18.25,z(cid:2)=18.23 54sec
2009-11-0407:56–16:07 UVOT U =19.58,B=19.53,V =18.88 1144,838,1144sec
2010-09-1314:52:26 MITSuME0.5m g(cid:2)=18.51,R=18.04,I=17.53 60secb
2010-09-1714:46:37 MITSuME0.5m g(cid:2)=18.54,R=17.98,I=17.63
2010-09-2013:17:30 MITSuME0.5m g(cid:2)=18.15,R=17.33,I=17.38
2010-10-0415:55:20 MITSuME0.5m g(cid:2)=18.51,R=17.93,I=17.53
2010-10-1014:28:06 MITSeME0.5m g(cid:2)=19.12,R=18.53,I=17.94
2010-11-0913:21:22 MITSeME0.5m g(cid:2)=18.43,R=17.82,I=17.21
2010-11-1013:22:43 MITSeME0.5m g(cid:2)=18.98,R=18.12,I=17.51
2010-11-1113:19:40 MITSeME0.5m g(cid:2)>20.19,R=19.44,I=19.04
2010-11-1211:22:28 MITSeME0.5m g(cid:2)=18.43,R=17.70,I=17.44
2010-11-1412:39:55 MITSeME0.5m g(cid:2)=19.00,R=18.16,I=17.84
2010-11-1512:40:11 MITSeME0.5m g(cid:2)=18.07,R=17.91,I=17.51
2010-11-1812:53:05 MITSeME0.5m g(cid:2)>20.02,R=21.77,I=19.16
2010-11-2112:46:57 MITSeME0.5m g(cid:2)=18.70,R=17.61,I=17.49
a Exposuretimeforeachframe.
b Someframeswerecombinedtoincreasethesignal-to-noiseratio.
factorandαisthespectralindexintheco-movingframe. 1FGLJ2339.7–0531 being one of the brightest unidenti-
Anorbitalmodulationbyafactoroffivemaysuggestthe fied pulsar-like Fermi sources located at high Galactic
Doppler factor D ∼1.6 with α∼0.5. latitude, no radio detection has been reported. Based
The X-ray hardness ratio roughly correlates with the on a recent pulsation search with the Green Bank Tele-
X-raylight-curve(Figure2). Itmaysuggestthatthepul- scope, 1FGLJ2339.7–0531was not detected (Ransom et
sar(whichhasaveryhardX-rayspectrum)isphysically al. 2011). It may indicate that 1FGLJ2339.7–0531 is
eclipsed by the companion and that the soft emission “radio-quiet”,or the radio emission must be very weak.
is an extra source of light (e.g. a wind shock). There The system resemblesblack widow MSP for which the
is some support for this in the assymetry of the optical MSP has a very low-mass (<< 0.1M(cid:2)) companion in
light curve. Much better X-ray and optical photometric an orbit less than a day. The optical light-curves of
data are required for serious modeling. black widow systems usually shows large orbital varia-
The source has a very low absorption (3×1020 cm−2) tion because the irradiation produces strong heating on
alongthelineofsight. IncomparisonwithGemingawith the companion facing the pulsar. It is therefore sugges-
NH = 1.5×1020 cm−2 and a distance of 150–250 pc, tivethatthecompanion(usuallyawhite dwarf)is being
and SAX J1808.4–3658with NH =4×1020 cm−2 and a evaporated by the high-energy radiation from the pul-
distance of 1.3 kpc. The distance of 1FGLJ2339.7–0531 sar. On the other hand, there is a class of black widow
is therefore likely to be between 300 and 1000 pc. If the MSPs with high mass (∼> 0.1M(cid:2)) non-degenerate com-
companion star is a late-type main sequence (e.g., a M5 panions. One classic example is PSR J1023+0038 that
dwarf), and has V ∼ 21 during optical minimum when has a 0.2M(cid:2) companion (Archibald et al. 2009). The
the heating effect is atminimum, the implied distance is X-ray/gamma-ray properties of 1FGLJ2339.7–0531 are
about 700 pc. similartothatofPSRJ1023+0038butnoradioemission
Ifweassumeadistanceof700pc,thenthegamma-ray from the former has been detected yet.
power is Lγ ∼ 1033 ergs s−1. Most of MSPs have an Hence, we suggest that 1FGLJ2339.7–0531 is a new
efficiency about 10% which gives spin-down luminosity class of black widow-like MSP system, with no or
Lsd ∼ 1034 ergs s−1. According to Takata, Cheng & very faint radio emission along the line-of-sight. The
Taam(2011),the spinperiodofthe MSPcanbe derived high-mass non-degenerate companion indicates that
from the spin-down power. Using the estimated spin- 1FGLJ2339.7–0531isinthelatestagesofrecycling. The
down power, the spin period will be ∼ 4 ms. We note absence of radio emission is the result of different emis-
thatthe distancetothesourceisuncertain;forinstance, sionregionsforradioandgamma-rays. Forinstance,the
RomaniandShaw(2011)estimatedadistanceof1.1±0.3 radioemissionisfromthepolarcapwiththeradiobeam
kpc. If we assume a distance error of a factor of 2 (i.e., out of the line-of-sight. On the other hand, the gamma-
350–1400pc), the spin period will be 3–5 ms. ray emission is from the outer magnetosphere which is
The bright Fermi source 1FGLJ2339.7–0531 is an predicted by the outer gap model.
intriguing object that is potentially the first “radio- An intensive search for radio emission and pulsation
quiet” γ-ray emitting MSP in a binary system. Despite from 1FGLJ2339.7–0531 in the future can confirm the
5
2010-10-25 (Lulin) 2010-10-31 (Lulin) 2010-11-11 (Tenagra)
-1
2010-10-26 (Lulin) 2010-11-01 (Lulin) MITSuME color
e 0
d
u
nit
ntial mag 12 (g’-I) colo
e r
er
Diff 3
4
35 8
Chandra ACIS-I
HR (1.5-8.0keV/0.3-1.5keV)
30
6
25
H
ounts/bin 1250 4 ardness R
C a
tio
10
2
5
0 0
0 0.5 1 1.5 2
Orbital Phase
Figure 2. Foldedlight-curveofopticalandChandraobservationsof1FGLJ2339.7–0531withabest-fitperiodof4.6342hr. Opticalcolors
(g(cid:2)−I) obtained with MITSuME are also plotted. The phase zero is defined as 2010 October 31 (MJD 55500). Also plotted with the
X-raylight-curveistheX-rayhardnessratio(1.5–8keV/0.3–1.5keV)withtriangles. Itisevident thatbothoptical andX-raylight-curve
showsimilarmodulation. TheX-rayhardnessratioexhibitssomevariabilitywhentheX-raylight-curveisatitsminimum. Notethatthe
Chandra dataonlycover aboutoneorbitalperiod.
above scenario. At the same time, a search for gamma- results. This project is supported by the National Sci-
ray/X-ray pulsation from the MSP will reveal the true ence Councilofthe Republic ofChina (Taiwan)through
nature of the system. grant NSC100-2628-M-007-002-MY3 and NSC100-2923-
M-007-001-MY3. A. K. H. K. gratefully acknowledges
support from a Kenda Foundation Golden Jade Fellow-
We thank the supporting staffs at the Lulin Obser-
ship. C.C.C.andD.D.weresupportedinpartbyChan-
vatory to arrange the service observations. The Lulin
dra award GO0-11022A. C. Y. H. is supported by the
ObservatoryisoperatedbytheGraduateInstituteofAs-
National Research Foundation of Korea through grant
tronomyinNationalCentralUniversity,Taiwan. Wealso
2011-0023383.
thank Roger Romani for providing insights and useful
comments as well as his quick-look optical spectroscopic
REFERENCES
Abdo,A.A.etal.2010a, ApJ,715,429 Laird,E.S.etal.2009,ApJS,180,102
Abdo,A.A.etal.2010b, ApJS,187,460 Monet,D.G.etal.2003,ApJ,125,984
Abdo,A.A.etal.2010c, ApJS,188,405 Muslimov,A.G.&Harding,A.K.2003,ApJ,588,430
Abdo,A.A.etal.2011, ApJS,submitted,arXiv:1108.1435 Pletsch,H.J.etal.2011,ApJ,inpress,arXiv:1111.0523
Archibald,A.M.etal.2010, ApJ,722,88 Ransom,S.M.etal.2011,ApJ,727,L16
Archibald,A.M.etal.2009, Science,324,1411 Richards,G.T.etal.2002,AJ,123,2945
Arons,J.&Tavani,M.1993,ApJ,403,249 Romani,R.W.&Shaw,M.S.2011,ApJ,743,L26
Bogovalov, S.V., Khangulyan, D.V., Koldoba, A.V., Ustyugova, Ruderman,M.A.&Sutherland,P.G.1975, ApJ,196,51
G.V.,Aharonian,F.A.2008, MNRAS,387,63 SazParkinson,P.M.etal.2010,ApJ,725,571
Camilo,F.etal.2009,ApJ,705,1 Stappers, B.W.,Gaensler, B.M.,Kaspi,V.M.,vander Klis,M.&
Caraveo,P.A.2010,arXiv:1009.2421 Lewin,W.H.G.2003,Science,299,1372
Cognard,I.2011, ApJ,732,47 Takata,J.,Cheng, K.S.&Taam,R.E.2010,ApJ,723,L68
Green,P.J.etal.2004,ApJS,150,43 Takata, J., Cheng, K.S. & Taam, R.E. 2011, ApJ, in press
Keith,M.J.etal.2011, MNRAS,inpress,arXiv:1109.4193 (arXiv:1111.3451)
Kotani,T.,Kawai,N.,Yanagisawa,K.,etal.2005,NuovoCimento Takata,J.,Wang, Y.&Cheng,K.S.2010,ApJ,715,1318
CGeophysics SpacePhysicsC,28,755 Takata,J.,Wang, Y.&Cheng,K.S.2011,ApJ,726,44
Kulkarni, S.R., Phinney, E.S., Evans, C.R., Hasinger, G. 1992, Tam,P.H.T.,Hui,C.Y.,Huang,R.H.H.,Kong,A.K.H.,Takata,J.
Nature,359,300 etal.2010,ApJ,724,L207
6
Tam,P.H.T.,Huang,R.H.H.,Takata,J.,Hui,C.Y.,Kong,A.K.H., Wilson,C.A.,Dieters,S.,Finger,M.H.,Scott,D.M.,vanParadijs,
Cheng,K.S.2011, ApJ,736,L10
Venter,C.Harding,A.K.&Guillemot,L.2009, ApJ,707,800 J.1999,ApJ,513,464