Hard X-ray Emission Associated with White Dwarfs. IV. Signs of Accretion from Sub-stellar Companions
You-Hua Chu, Jesus A. Toala, Martin A. Guerrero, Florian Bauer, Jana Bilikova, Robert A. Gruendl
aa r X i v : . [ a s t r o - ph . S R ] F e b Draft version February 10, 2021
Typeset using L A TEX twocolumn style in AASTeX63
Hard X-ray Emission Associated with White Dwarfs. IV. Signs of Accretion from Sub-stellarCompanions
You-Hua Chu, Jes´us A. Toal´a, Mart´ın A. Guerrero, Florian F. Bauer, Jana Bilikova, andRobert A. Gruendl Institute of Astronomy and Astrophysics, Academia Sinica (ASIAA), No.1, Sec. 4, Roosevelt Road, Taipei 10617, Taiwan, ROC Instituto de Radioastronom´ıa y Astrof´ısica (IRyA), UNAM Campus Morelia, Apartado postal 3-72, 58090 Morelia, Michoacan, Mexico Instituto de Astrof´ısica de Andaluc´ıa, IAA-CSIC, c/Camino Bajo de Hu´etor 50, E-18008 Granada, Spain Astronomy Department, University of Illinois, Urbana, Illinois 61801, USA
ABSTRACTKPD 0005+5106, with an effective temperature of ≃ ROSAT unexpectedly detected “hard” ( ∼ Chandra observations that confirm the spatial coincidence of this hard X-ray source withKPD 0005+5106. We have also obtained
XMM-Newton observations of KPD 0005+5106, as well asPG 1159 −
035 and WD 0121 − ROSAT at 3 σ –4 σ levels. The XMM-Newton spectra of the three WDs show remarkably similarshapes that can be fitted by models including a blackbody component for the stellar photosphericemission, a thermal plasma emission component, and a power-law component. Their X-ray luminositiesin the 0 . − . × to 4 × erg s − . The XMM-Newton
EPIC-pn soft-band (0 . − . . − . XMM-Newton and
Chandra hard-band lightcurves finds a convincing modulation (false alarmprobability of 0.41%) with a period of 4.7 ± Keywords: (stars:) white dwarfs – X-rays: stars – infrared: stars – stars: individual (KPD 0005+5106,PG 1159 − − INTRODUCTIONThree types of X-ray sources are known to be asso-ciated with white dwarfs (WDs): (1) photospheric X-ray emission from a WD itself, (2) accretion of materialfrom a close binary companion, as in cataclysmic vari-ables, and (3) coronal X-ray emission from a late-typebinary companion, such as dMe stars. The latter twotypes of sources require the WDs to be in binary sys-tems and their observed X-ray spectra commonly peaknear 1 keV or higher, consistent with thermal emissionfrom plasma with temperatures of a few × K. Incontrast, the photospheric X-ray emission from a WD issoft and detectable only at photon energies < L X < . × ergs s − (Weisskopf et al.2007).The above conventional wisdom was challenged bythe detection of hard (photon energy ∼ −
216 (the cen-tral star of the Helix Nebula; Guerrero et al. 2001) and
Chu et al.
KPD 0005+5106 (WD 0005+511), which show a softphotospheric component and a distinct hard compo-nent peaking near 1 keV (Paper I). In the case ofWD 2226 − < M ⊙ binary com-panion with an orbital period of 2.77 days (Aller et al.2020). In the case of KPD 0005+5106, the lack of IRexcess and H α emission associated with coronal activ-ity has been used to exclude the existence of a late-typecompanion with a corona (Chu et al. 2004a).In Paper I and Paper III, similar hard compo-nent peaking near 1 keV has also been detected to-ward WD 0339 −
451 at a 6 σ level, and WD 0121 − − − − − σ –4 σ levels;moreover, hard X-ray emission manifested in the softX-ray emission extending to harder 0.5–1.0 keV pho-ton energy range is observed in WD 1234+481 andWD 1254+223. None of these WDs have known bi-nary companion or show IR excess indicative of a late-type stellar companion. Interestingly, at least four ofthese WDs have stellar effective temperatures greaterthan 100,000 K – KPD 0005+5106, WD 0121 − − − ∼
10 photons with energies greater than 0.5keV were detected in the ROSAT PSPC observations ofWD 0121 −
756 and PG 1159 −
035 (Paper I, Paper III).Better X-ray observations are needed.To investigate physical properties of hard X-rayemission from apparently single WDs, we have ob-tained Chandra X-ray Observatory ACIS-S observa-tion of KPD 0005+5106, but its very luminous softphotospheric emission caused pile-up effects; there-fore, we have obtained XMM-Newton observations ofKPD 0005+5106 for spectral and temporal analyses.We have also acquired XMM-Newton observations ofWD 0121 −
756 and PG 1159 −
035 to confirm their hardX-ray emission and to carry out spectral analyses. Thispaper reports our analyses of these Chandra and XMM-Newton observations: Section 2 describes the X-ray ob-servations as well as complementary IR observations,Sections 3 and 4 report the spectral and temporal anal-yses of the X-ray data, Section 5 discusses the implica-tions of the X-ray results on the physical origin of the hard X-ray emission, and finally Section 6 summarizesthe conclusion of our study. OBSERVATIONS2.1.
Chandra X-ray Observation
The Advanced CCD Imaging Spectrometer (ACIS) onboard the Chandra X-ray Observatory was used to ob-serve KPD 0005+5106 on 2008 March 19 (Obs. ID 8942;PI: Y.-H. Chu). KPD 0005+5106 was positioned at theaim-point of the ACIS-S array on the back-illuminatedCCD S3 and observed in the FAINT mode for a totalof 48.0 ks. The data were processed and analyzed us-ing the Chandra Interactive Analysis of Observations(CIAO) software package (version 4.11; Fruscione et al.2006). The observations were not affected by any periodof high background, and no time intervals had to be ex-cised. After dead time correction, the final exposuretime was 47.4 ks.To refine the astrometric accuracy of this Chandraobservation, we identified seven X-ray sources with op-tical counterparts and used their optical positions tocalibrate the X-ray astrometry. The final astrometricaccuracy of the Chandra observation is better than 1 ′′ .A bright X-ray point source is detected at the positionof KPD 0005+5106 with a background-subtracted countrate of 0.157 ± − in the 0.15-4.5 keV energyband.The background-subtracted ACIS-S spectrum ofKPD 0005+5106, shown in Figure 1, exhibits a brightpeak near 0.2 keV, a secondary peak near 0.4 keV anda much fainter tertiary peak near 0.6 keV. The brightprimary peak corresponds to the soft photospheric emis-sion from KPD 0005+5106. The very high count rate ofthe primary peak and the locations of the secondary andtertiary peaks suggest that the spectrum is affected by apileup of photons with energies near the primary peak.We have thus modeled the pileup effects followingguidelines in The Chandra ABC Guide to Pileup pro-vided by the Chandra X-ray Center. This model, plottedover the spectrum in Figure 1, indicates that the pileupcontributions are still noticeable up to 0.75 keV.The pileup-removed, background-subtracted spec-trum of KPD 0005+5106, shown in the bottom panelof Figure 1, dips to nearly zero at 0.5 keV, peaks near1 keV, and diminishes above ∼ ∼ ± − in the 0.6–4.5 keV band. Becauseof the large uncertainties in Chandra ACIS-S calibrationbelow 0.3 keV, the pileup correction is not ideal and thepileup-corrected spectrum of KPD 0005+5106 would not http://cxc.harvard.edu/ciao/download/doc/pileup abc.ps ard X-ray Emission from WDs Figure 1.
Modeling the pileup in the Chandra ACIS-S spec-trum of KPD 0005+5106. In the top panel, the data pointswith error bars show the background-subtracted raw spec-trum, the solid curve shows a pileup model for the spectrum,and the pileup model multiplied by a factor of 10 is plottedin a dotted curve to show the low-level effect near 0.6 keV.The bottom panel shows the pileup-removed background-subtracted ACIS-S spectrum. be suitable for spectral analysis. We will neverthelessuse the Chandra ACIS-S data to extract lightcurves inthe hard X-ray band ( ≥ XMM-Newton X-ray Observations
We obtained pointed XMM-Newton observationsfor KPD 0005+5106, WD 0121 − − − cause pileup.In addition, we find in the XMM-Newton archive twoobservations that include PG 1159 −
035 in the EPIC’sfield of view. These two observations’ target was theSeyfert-type galaxy Mrk 1310, but PG 1159 −
035 is lo-cated ∼ − evselect . Toextract the spectra of a WD, a circular source aper-ture of 20 ′′ and a surrounding or nearby backgroundregion devoid of sources were used. Their correspond-ing calibration matrices, redistribution matrix file (rmf)and auxiliary response file (arf), were produced with theSAS tasks rmfgen and arfgen . Background-subtractedsource count rates of each camera obtained from differ-ent observations are listed in the last three columns ofTable 1.EPIC spectra extracted from different cameras werecombined to produce a single EPIC spectrum for eachWD. This has been done by making use of the SAS task epicspeccombine . This task creates a single background-subtracted spectrum for each source plus correspond-ing calibration matrices . The resultant combinedEPIC spectra of KPD 0005+5106, PG 1159 − − Infrared Observations
The Multiband Imaging Photometer for Spitzer(MIPS; Rieke et al. 2004) on board the Spitzer SpaceTelescope was used to image KPD 0005+5106 in the24 and 70 µ m bands (Program ID 40953; PI: Y.-H.Chu). Images were obtained in photometry mode withthe small offset scale and 10 s exposure time for threecycles at both 24 and 70 µ m. The Basic CalibratedData processed by the Spitzer Science Center’s pipelinesoftware were used to construct mosaics in each bandusing utilities in the MOPEX software package. Priorto building each mosaic, bad pixels and latent imagesof bright point sources were flagged and removed, andthe background brightness offsets between the individ-ual frames were adjusted and removed.KPD 0005+5106 was not detected at either 24 µ m or70 µ m. We used the PHOT package in IRAF to obtainestimates for the 3 σ detection limit in these two MIPS See the epicspeccombine
Chu et al.
Table 1.
XMM-Newton Observations of Three Apparently Single WDs with Hard X-ray EmissionObject Obs. ID. Date Total Exposure Time Net Exposure Time Source Count Ratepn M1 M2 pn M1 M2 pn M1 M2(yyyy-mm-dd) (ks) (ks) (ks) (ks) (ks) (ks) (counts ks − )KPD 0005+5106 0693050401 2012-12-31 35.3 34.3 34.3 35.2 34.3 34.3 60.2 17.4 18.6WD 0121 −
756 0693050501 2012-11-26 15.0 15.1 15.1 6.0 12.0 12.5 14.0 4.1 3.5PG 1159 −
035 0693050601 2012-12-12 10.0 10.1 10.1 5.0 9.5 9.8 7.5 2.3 1.2PG 1159 −
035 0723100301 2013-12-09 . . . 51.2 51.2 . . . 33.7 31.0 . . . 0.8 1.3PG 1159 −
035 0693050601 2019-01-04 24.2 26.0 26.0 18.3 25.2 24.2 5.8 1.7 1.2Note: The three EPIC cameras are denoted as pn, M1, and M2. The EPIC-pn observation 0723100301 was performed inthe small window mode and did not encompass PG 1159 − bands and find flux density limits of < < µ m, respectively. These limits aretoo shallow to provide useful constraints on faint binarycompanions.Spitzer InfraRed Array Camera (IRAC; Fazio et al.2004) observations of KPD 0005+5106 were made in the4.5 and 8.0 µ m bands, and the flux densities were re-ported to be 297.5 ± ± µ Jy, respectively(Mullally et al. 2007). These values will be used in ouranalysis in Section 5.3.3.Near-IR
JHK s observations of KPD 0005+5106 werealso obtained with the Near Infrared Camera Spec-trometer (NICS) on the 3.58 m Telescopio NazionaleGalileo (TNG) on 2008 October 12–13. The detectorwas a HgCdTe Hawaii 1024 × . ′ × . ′ J , H , and K s bands. No systematic variations areseen among the 21 measurements in each band, and theiraverages are J = 14.07 ± H = 14.15 ± K s = 14.26 ± J = 13.95 ± H = 14.14 ± K s = 14.19 ± J magnitude shows a ∼ σ difference, but the H and K s magnitudes are constant. No large long-termvariations are present. X-RAY SPECTRAL ANALYSESThe XMM-Newton EPIC spectra, shown in Figure 2,clearly detect the hard X-ray emission peak near 1keV in all three WDs, confirming the previous ROSATPSPC’s 3 σ detection at 10 counts for PG 1159 − − ∼ tabs model as described inWilms et al. (2000) and included in XSPEC. Initially,we model the spectra with two emission components, ablackbody component for the WD’s photospheric emis-sion and an optically thin plasma emission componentor a nonthermal power-law component. None of the two-component models can successfully fit the EPIC spectra;they all result in reduced χ greater than 2. Thus, morecomplex models need to be considered.As the EPIC spectrum of KPD 0005+5106 has thehighest quality and shows line features in hard X-rays(see Fig. 2 - left panel), we start detailed spectralmodeling of this spectrum by including (1) a black-body component for the WD’s photospheric emissionwith temperature fixed at the effective temperature ofKDP 0005+5106, T eff = 2 × K (Werner et al. 2008;Wassermann et al. 2010); (2) a thermal plasma emis-sion component for the line features; and (3) a thermalplasma or power-law component to improve the spectralfits.For the three-component fits to KPD 0005+5106’sspectrum, we first consider another plasma emissionmodel for the third component, i.e., using the abovefixed-temperature blackbody emission model plus twooptically thin apec plasma emission models. The absorp-tion column density as well as the plasma temperaturesof the two apec models are left as free parameters. Thebest-fit model to the spectrum, with plasma tempera-tures of kT =0.20 keV and kT =0.62 keV and an unre- ard X-ray Emission from WDs − − − − I [ c o un t ss − k e V − ] Black bodyAPECPower law
Energy [keV] -303 ∆ I / σ − − − − Black bodyAPECPower law
Energy [keV] -30310 − − − − I [ c o un t ss − k e V − ] Black bodyAPECPower law
Energy [keV] -303 ∆ I / σ KPD 0005+5106 PG 1159-035 WD 0121-756
Figure 2.
XMM-Newton EPIC spectra of KPD 0005+5106, PG 1159 − − apec , and power-law components are plotted in differentshades of blue lines. Table 2.
Best-fit parameters to the EPIC spectraObject N H T eff kT Γ F X , TOT L X , TOT L apec /L X , TOT L pow /L X , TOT ( × cm − ) ( × K) (keV) (erg s − cm − ) (erg s − )KPD 0005+5106 9 +5 − +0 . − . +0 . − . . × − . × − +0 . − . . × − . × − +2 . − . . × − . × −
035 and WD 0121 − alistically low absorption column density, has a reduced χ of 2.5 and hence is still not formally acceptable.We then consider a power-law model for the thirdcomponent, i.e., using a fixed-temperature blackbodycomponent, an apec plasma emission component, anda power-law component. The best-fit model toKPD 0005+5106’s spectrum has a plasma temperatureof kT =0.84 +0 . − . keV, a power-law index of Γ = 2 . +0 . − . ,and an absorption column density of N H = (9 +5 − ) × cm , consistent with estimates from the Ly α absorp-tion profile (Werner et al. 1994); furthermore, the re-duced χ is improved to 1.16. We consider this best-fit model acceptable. The best-fit model’s three in-dividual emission components and their sum are plot-ted over the background-subtracted EPIC spectrum ofKPD 0005+5106 in the leftmost panel of Figure 2. Itcan be seen that the blackbody component of the pho-tospheric emission dominates the spectrum at energiesbelow 0.4 keV. The apec plasma emission componentcontributes to the broad peak around 1 keV and theemission line feature at ∼ VIII line. The power-law component dominates athigh energies from 2 keV up to 5–6 keV, as well as the intermediate energies at 0.4–0.6 keV. The intrin-sic (unabsorbed) flux in the 0.3–7.0 keV energy rangeis F X = 4 . × − erg s − cm − , corrresponding to anX-ray luminosity of L X = 8 . × erg s − at the dis-tance of 387 ± apec components. This results in a rea-sonably good fit ( χ =1.5) with plasma temperaturesof kT = 0 . +0 . − . keV, kT = 0 . +0 . − . keV, and kT = 2 . +1 . − . keV. The highest temperature compo-nent is needed for the X-ray emission above ∼ χ =1.1–1.3) if more plasma com-ponents are added. Based on Occam’s Razor, we donot pursue multi-temperature plasma model fits further;however, as we show in Section 4, multiple temperatures Chu et al. may be needed to explain the temporal variations of thespectral properties.Following our spectral analysis of KPD 0005+5106, wefind that the spectra of PG 1159 −
035 and WD 0121 − apec thermal plasma component, and a power-lawcomponent. These two WDs are fainter and have shorterexposure times; thus their observations detected muchfewer counts than KPD 0005+5106 and their spectra donot allow fitting many free parameters simultaneously.We have converted these two WDs’ extinction measure-ments into H column densities and adopt them as fixedabsorption column densities N H , which are 2 × cm − and 4 × cm − for PG 1159 −
035 (Dreizler, & Heber1998) and WD 0121 −
756 (van Teeseling et al. 1996), re-spectively. The blackbody components’ temperaturesare fixed at the effective temperatures of the WDs,1 . × K for PG 1159 −
035 (Dreizler, & Heber 1998)and 1 . × K for WD 0121 −
756 (Werner et al. 1996).We also adopt the best-fit Γ of KPD 0005+5106 and fixthe Γ of PG 1159 −
035 and WD 0121 −
756 at 3.The resultant best-fit models for PG 1159 −
035 ( χ =1 .
20) and WD 0121 −
756 ( χ = 1 .
9) are listed in Table 2and plotted over their background-subtracted EPICspectra in Figure 2. It can be seen that the three com-ponents’ contributions to PG 1159 −
035 in different en-ergy ranges are very similar to those to KPD 0005+5106.WD 0121 − − kT = 2 keV, instead of the kT ∼ − TEMPORAL VARIATIONS OF HARD X-RAYSAmong the three WDs observed with XMM-Newton,only KPD 0005+5106 had high enough a count rate andlong enough an exposure time (see Table 1) to war-rant temporal variation analyses. We first extractedbackground-subtracted lightcurves from the EPIC pnobservation in three different energy ranges: the entire0.2–3.0 keV energy band that covers the bulk of X-rayemission, the soft X-ray band of 0.2–0.5 keV, and a hardband covering the 0.6–3.0 keV energy range. We havealso extracted a lightcurve from the Chandra ACIS-Sobservation. As stated in Section 2.1, the pileup effect −
10 0 10 20 30 40
Time [ks] I [ c o un t s ] EPIC-pn 0.2 - 3.0 keVEPIC-pn 0.2 - 0.5 keVEPIC-pn 0.6 - 3.0 keVACIS-S 0.75 - 2.0 keV
Figure 3.
Background-subtracted EPIC and ACIS-Slightcurves of KPD 0005+5106. Each bin corresponds to2 ks. The dashed thin lines represent the average values. is negligible compared with the WD’s hard X-ray emis-sion at photon energies greater than ∼ ∼
18 ks (= 5 hr).The ACIS-S lightcurve in the hard band (0.75–2.0 keV)shows similar variations. If the ACIS-S lightcurve isshifted to align its clearest peak with the EPIC-pn hard-band lightcurve peak at 20 ks, it can be seen that ACIS-Slightcurve also shows a lower peak at ∼ ard X-ray Emission from WDs ∼ − ),although each data set covers only 2-3 cycles of this mod-ulation. The combined data set covers 4.6 cycles, andits X-ray intensity modulation for a period of 4.7 ± ∼ . xi line emission at a 2–3 σ confidence level.Moreover, the low-state spectrum exhibits an emissionpeak near 0.5 keV that is also present in the EPIC spec-trum in Figure 2 but not seen in the high-state spectrum.It would be very interesting to analyze detailed differ-ences between the high- and low-state spectra, for exam-ple, plasma temperatures and abundances, the numberof thermal plasma components needed in spectral fits,and their relative contributions (see Section 3); how-ever, the current data do not have adequate counts andresolution for such analyses. Future deeper observationswould be desirable. DISCUSSIONWe first use the positional coincidence to affirm thephysical association between the hard and soft X-rayemission components in the WDs, then compare thespectral properties among the three WDs, and discussthe possible origins of the hard X-ray emission.5.1.
Spatial Coincidence between the Hard and SoftX-ray Emission
The superb angular resolution of Chandra enablesus to determine accurately whether a position off-set exists between the hard and soft X-ray emissionfrom KPD 0005+5106. We have extracted images ofKPD 0005+5106 in a soft band (0.2–0.5 keV) and a hardband (0.8–3.5 keV), and find that the soft and hardX-ray point sources are coincident with each other (inthe same 0 . ′′ ′′ ). These coincidences support that boththe soft and hard X-ray components are attributed toKPD 0005+5106.The angular resolution of XMM-Newton is not as goodas Chandra, yet it can still be shown from the pointedobservations of PG 1159 −
035 and WD 0121 −
756 thatboth their hard and soft X-ray components and the op-tical position of the WD are coincident with one anotherwithin ∼ ′′ . We therefore consider that the spatial co-incidences affirm the physical association between thehard and soft X-ray components and between the X-rayemission and the WD.5.2. Similarities among the Three WDs
KPD 0005+5106 is a DO WD, while PG 1159 − −
756 are PG 1159 type WDs. Both DOand PG 1159 spectral types imply a H-deficient, He-richatmosphere. Besides the similarity in composition, allthree have stellar effective temperatures greater than100,000 K: 200,000 K, 140,000 K, and 180,000 K for
Chu et al. G L S po w e r Frequency (day -1 ) Chandra XMM ALL -0.4-0.20.00.20.4 R e l a ti v e f l ux Chandra XMM fit-0.20.00.20.0 0.2 0.5 0.8 1.0 R e s i du a l s Phase
Figure 4.
Left panel: Generalized Lomb-Scargle (GLS) periodograms of Chandra ACIS-S data (blue), XMM-Newton EPIC-pndata (orange), and the combined data set (black). False alarm probabilities (FAP) of 10%, 1%, and 0.1% are indicated bydash-dotted, dotted, and dashed grey lines, respectively. Right panel: Phase-folded X-ray lightcurves of Chandra ACIS-S data(blue squares), XMM-Newton EPIC-pn data (orange dots) and a spline model fit (black dashed line). . . . . . . . . I [ c o un t ss − k e V − ] High stateLow state
Energy [keV] -303 ∆ I / σ Figure 5.
Background-subtracted EPIC pn spectra ex-tracted from time intervals in the high and low states, plottedin black and red, respectively.
KPD 0005+5106, PG 1159 − − Origin of the Hard X-ray Emission
We will consider three possible origins of the observedhard X-rays from these apparently single WDs: (1) pho-tospheric emission, (2) hidden coronal companion, and(3) accretion from a hidden companion.5.3.1.
Photospheric Emission
KPD 0005+5106 is the most well-studied and its X-ray spectrum has the highest quality among the threeWDs reported in this paper. We will thus first examineprevious studies of KPD 0005+5106 in detail.KPD 0005+5106 is such a bright soft X-ray sourcethat it was detected in the ROSAT All Sky Sur-vey. Fleming et al. (1993) showed that the spectrumof KPD 0005+5106 in the 0.1–0.4 keV range could befitted by a model of thermal plasma at a tempera-ture of 2.6 × K, and suggested that its X-ray emis-sion originated from a corona cooler than those of typ-ical main-sequence stars. A corona for KPD 0005+5106would be surprising because the star is fully ionizedand should not have a convective envelope to powera corona. Indeed, based on a Chandra Low Energy ard X-ray Emission from WDs g = 7. Theyruled out coronal emission as the X-ray source becausethe extrapolated optical continuum intensity would ex-ceed the observed photospheric continuum, and becauseof a lack of H-like and He-like C lines that should oth-erwise have been observed in X-rays. Their 120,000K photosphere model can explain the soft X-ray spec-trum, but cannot reproduce the hard X-ray emissionfrom KPD 0005+5106.The discovery of photospheric Ca X emission lines inthe Far Ultraviolet Spectroscopic Explorer (FUSE) spec-trum, in conjunction with the Ne VIII lines in UV andoptical spectra, provided evidence that KPD 0005+5106has an effective temperature of ∼ −
035 andWD 0121 −
756 as well.5.3.2.
Hidden Companions with Active Coronae
Previously, Chu et al. (2004a) assumed the hard X-rayemission of KPD 0005+5106 originated completely froma coronal companion, and showed that such compan-ion would outshine the WD itself in near-IR passbands.Here we use a similar but more systematic approach toexamine the possible presence of hidden companions.To assess the possibility that the hard X-rayemission from KPD 0005+5106, PG 1159 − −
756 originates from the coronal activity oflate-type dwarf companions, we first compare the bright-ness of the WD with a potential companion in the K band to see whether a companion can hide under-neath the bright WD emission. We have computedthe expected m K magnitudes of K0, K5, M0, M5, andM8 dwarf stars at the distances of KPD 0005+5106(390 pc), PG 1159 −
035 (550 pc), and WD 0121 − m K magnitudes of these stars, as well as the observed K magnitudes the WDs, are listed in Table 3. Obviously,a bright companion cannot hide behind a fainter WD;thus only the latest type of M dwarfs are candidates ofhidden companions.We next consider whether coronae of late-type com-panions are able to provide the hard X-ray luminosi-ties of these three WDs: 3.9 × , 3.9 × , and3.2 × erg s − for KPD 0005+5106, PG 1159 − − L X /L bol ) for the putative late-type dwarf compan-ions, and listed them in Table 3, too. The highestlog( L X /L bol ) observed in K–M dwarf stars with satu-rated coronal activity is about − − L X /L bol ) listed in Table 3 indicate thatthese late type dwarf stars would have log( L X /L bol ) ex-ceeding the range, up to about −
3, that can possibly beprovided by stellar coronae.Figure 6 is another presentation comparing the bright-ness and log( L X /L bol ) of putative late-type dwarf com-panions with those of the WD. All K0–M8 dwarf starsare either too bright in the K band and/or cannot pro-vide the observed hard X-ray luminosities. Therefore,we can rule out coronal activities of a late-type compan-ion for the origin of the hard X-ray emission detected inKPD 0005+5106, PG 1159 − − Accretion from a Hidden Companion
Finally we consider accretion-powered hard X-rayemission from the WDs. We will use KPD 0005+5106as an example for our analysis, as it has the most welldetermined physical parameters among the apparentlysingle WDs with hard X-ray emission. We adopt a massof 0.64 M ⊙ and a radius of 0.059 R ⊙ (or 6.5 R ⊕ ) forKPD 0005+5106 (Wassermann et al. 2010).We use the Spitzer IRAC observations and the2MASS, or TNG NICS, JHK s observations ofKPD 0005+5106 to place constraints on possible com-panions. The B and V magnitudes of KPD 0005+5106have been measured by Downes et al. (1985): B =13.02 and V = 13.32. These optical and IR photo-metric measurements are used to produce the spec-tral energy distribution (SED) of KPD 0005+5106shown in Figure 7, where the model spectrum fromWassermann et al. (2010) is also plotted. The extinc-tion correction has been made by using the hydrogencolumn density of N H = 5 × H-atoms cm − to-ward KPD 0005+5106 (Werner et al. 1994) and the gas-to-dust ratio of N H / E ( B − V ) = 5 . × H-atomscm − mag − (Bohlin et al. 1978).0 Chu et al.
Table 3.
Infrared and X-ray Properties of Putative Late-type Dwarf CompanionsSpectral Type KPD 0005+5106 PG 1159 −
035 WD 0121 − m k log( L X /L bol ) m k log( L X /L bol ) m k log( L X /L bol )K0 11.90 − − − − − − − − − − − − − − − ± ± ± ... Figure 6.
X-ray to bolometric luminosity ratio and in-frared excess in the K band of putative late-type dwarf com-panions of KPD 0005+5106 (black), PG 1159 −
035 (blue),and WD 0121 −
756 (red). The different spectral types havebeen connected by dotted lines for each WD and are la-beled in the track of PG 1159 − m K magnitudes of KPD 0005+5106 (14.26), PG 1159 − −
756 (16.213). Dots above the hori-zontal dashed line imply that putative late-type dwarf com-panions would have actually X-ray to bolometric luminosityratios in excess of the canonical value for saturated activityin these stars.
Comparing the observed SED with these model SEDs(Fig. 7, see figure caption for origin of data and refer-ences), we can easily rule out the existence of a com-panion of spectral type M5 or earlier. The observed 4.5and 8.0 µ m fluxes are ∼
20% lower than those expected
Figure 7.
Spectral energy distribution (SED) ofKPD 0005+5106. The solid symbols are data: B and V (Downes et al. 1985; McCook & Sion 1999), 2MASS J , H ,and K s , Spitzer IRAC 4.5 and 8 µ m (Mullally et al. 2007).The error bars are all smaller than the symbols, except the8 µ m measurement. The open symbols are models assumingM-type companions. The photometry of standard M starsare from Kirkpatrick & McCarthy (1994) and Patten et al.(2006). The solid curve is a model SED of KPD 0005+5106from Wassermann et al. (2010) normalized to the K s band. with an M8 V companion. Suffice it to say, any hiddencompanion of KPD 0005+5106 has to be fainter and lessmassive than an M8 V star.We will consider three types of hidden companions, anM9 V star with a mass of 0.075 M ⊙ and a radius of 0.08 R ⊙ (Kaltenegger & Traub 2009), a T type brown dwarfwith a mass of 0.035 M ⊙ and a radius of ∼ R ⊙ , anda Jupiter-like planet with a mass of 0.001 M ⊙ and a ra-dius of ∼ R ⊙ . We assume that the period of 4.7 hr inthe lightcurve of KPD 0005+5106’s hard X-ray emissionis the orbital period of this binary system. Using theKepler’s third law, we can determine the separation be-tween the WD and the companion, 1.27 R ⊙ for a M9 Vstar, 1.24 R ⊙ for a T brown dwarf, and 1.22 R ⊙ for a ard X-ray Emission from WDs r L of thecompanion can be approximated by: r L /a = 0 . q / . q / + ln (1 + q / ) , (1)where a is the separation between the WD and thecompanion and q is the companion to WD mass ratio(Eggleton 1983). We find Roche radii of 0.28 R ⊙ , 0.21 R ⊙ , and 0.067 R ⊙ for a M9 V star, a T brown dwarf,and a Jupiter-like planet, respectively. Comparing theseRoche radii with their respective radii, it is clear that aJupiter-like planet is larger than the Roche radius andcan channel mass to the WD easily. The atmospheres ofM9 V star and T brown dwarf, being ionized and heatedby the 200,000 K KPD 0005+5106 at a distance of ∼ R ⊙ , may be inflated and the outer edge of their atmo-spheres may exceed the Roche radius and be accretedby the the WD. We conclude that all three types of hid-den companions may be the donor providing materialfor accretion-powered hard X-ray emission, although aJupiter-like planet can do it most easily.Previously, ∼
10 WDs accreting from brown dwarfshave been reported (Longstaff et al. 2019). Only two ofthem show X-ray emission, EF Eridani (Schwope et al.2007) and SDSSJ121209.31+013627.7 (Stelzer et al.2017); however, both are cool WDs with effective tem-peratures lower than 10,000 K, and both are polarswhose strong magnetic fields channel material from thecompanion to the WDs’ magnetic poles to be accreted.The three WDs with hard X-ray emission reported inthis paper have much higher effective temperatures thanthese polars. As noted above, these hot WDs can accretematerial from either a Jupiter-like planet or an irradi-ated brown dwarf, and hence are very different from thepolars.If we assume that the hard X-ray emission is poweredby accretion from a companion, then the accretion ratewould be L X R WD / ( GM WD ), where L X is the hard X-ray luminosity, R WD and M WD are the radius and massof the WD, and G is the gravitational constant. ForKPD 0005+5106’s hard X-ray luminosity (3 × ergss − ), mass, and radius, the mass accretion rate wouldneed to be 2.3 × − M ⊙ yr − , or 1.45 × g s − .This mass accretion rate is very small that a Jupiter-like planet donor may survive for a few times 10 yr.It ought to be noted that KPD 0005+5106 is a He-richDO WD, the accreted material must be He-rich as well;otherwise the WD’s atmosphere will be over-pollutedwith H. While a Jupiter-like planet can easily providethe material to be accreted to power the hard X-rayemission, it is not clear how such a planet can survivethe stellar evolution and what its chemical composition would be. It is also not clear how an extremely low-mass star companion gets so close to a WD whose initialmass may be ∼ M ⊙ (Kalirai et al. 2005). Binary starevolution needs to be considered for the progenitor ofthis system. SUMMARY AND CONCLUSIONThree types of X-ray sources are commonly associatedwith WDs: soft photospheric emission, harder X-rayemission from a stellar companion with coronal activ-ity, and accretion powered X-ray emission. It has beenpuzzling that a small number of apparently single WDsare associated with X-ray emission peaking near 1 keV.The two most conspicuous cases are the central star ofthe Helix Nebula and KPD 0005+5106.With an effective temperature of 200,000 K,KPD 0005+5106 is a bright soft X-ray source even de-tected in the ROSAT All Sky Survey, and a pointedlonger observation reveal an additional harder X-raycomponent peaking near 1 keV. We obtained Chan-dra X-ray Observatory observations of KPD 0005+5106and confirmed the spatial coincidence of the hard andsoft X-ray emission components with the WD. The verybright soft X-ray emission caused photon pileup, ren-dering the spectral analysis of the hard X-ray emissionunreliable. We have thus obtained XMM-Newton X-rayobservations of KPD 0005+5106 and two other WDs,PG 1159 − − ∼
10 photons with energies near 1keV. The XMM-Newton observations show these threeWDs have similar X-ray spectral shapes with a brightsoft component below 0.5 keV and a harder componentpeaking near 1 keV, and all three spectra are best fittedby models consisting of a blackbody component for thestellar photospheric emission, a thermal plasma emissioncomponent, and a power-law component. The hard X-ray luminosities in the 0.6–3.0 keV band are 3.9 × ,3.9 × , and 3.2 × ergs s − for KPD 0005+5106,PG 1159 − − ∼ Chu et al. the most robust period, 4.7 ± ± M ⊙ M9 V star,a 0.035 M ⊙ T brown dwarf, and a 0.001 M ⊙ Jupiter-like planet as the companion. We find the Jupiter-likeplanet exceeds the Roche radius and can be a donorproviding material to be accreted by the WD to powerthe hard X-ray emission. The M9 V star and T browndwarf, being photoionized and heated by the 200,000 KKPD 0005+5106 at a distance of ∼ R ⊙ , may be in-flated and the outermost material may pass the Rocheradius and be accreted by the WD. We conclude thatthe hard X-ray emission from apparently single WDsis powered by accretion from sub-stellar companions orgiant planets, and is modulated by the orbital motionwith a period of 4.7 ± Software:
CIAO (v4.11; Fruscione et al. 2006), SAS(v17.0; Gabriel et al. 2004), MOPEX (Makovoz & Mar-leau 2005), IRAF (Tody 1986, Tody 1993), XSPEC(v12.10.1; Arnaud 1996) ard X-ray Emission from WDs13REFERENCES