eclipsing,likelysubstellarcompanionsdetectedby - arXiv · Using a large (17-px) ﬁxed aperture, we extracted the light curve, ﬁt out a quadratic function to 3-day windows, and

MNRAS 000 1ndash11 (2017) Preprint 5 September 2018 Compiled using MNRAS LATEX style file v30

Two white dwarfs in ultrashort binaries with detached

eclipsing likely substellar companions detected by K2

S G Parsons1⋆ J J Hermes2dagger T R Marsh3 B T Gansicke3 P-E Tremblay3

S P Littlefair1 D I Sahman1 R P Ashley3 M Green3 S Rattanasoon14

V S Dhillon15 M R Burleigh6 S L Casewell6 D A H Buckley7 I P Braker6

P Irawati4 E Dennihy2 P Rodrıguez-Gil58 D E Winget910 K I Winget910

Keaton J Bell910 and Mukremin Kilic11

1 Department of Physics and Astronomy University of Sheffield Sheffield S3 7RH UK2 Department of Physics and Astronomy University of North Carolina Chapel Hill NC 27599-3255 USA3 Department of Physics University of Warwick Coventry CV4 7AL UK4 National Astronomical Research Institute of Thailand 191 Siriphanich Bldg Huay Kaew Road Chiang Mai 50200 Thailand5 Instituto de Astrofısica de Canarias Vıa Lactea sn La Laguna E-38205 Tenerife Spain6 Department of Physics and Astronomy University of Leicester Leicester LE1 7RH UK7 South African Astronomical Observatory PO Box 9 Observatory 7935 Cape Town South Africa8 Universidad de La Laguna Departamento de Astrofısica E-38206 La Laguna Tenerife Spain9 Department of Astronomy University of Texas at Austin Austin TX 78712 USA10 McDonald Observatory Fort Davis TX 79734 USA11 Homer L Dodge Department of Physics and Astronomy University of Oklahoma Norman OK 73019 USA

Accepted 2017 June 22 Received 2017 June 16 in original form 2017 April 28

ABSTRACT

Using data from the extended Kepler mission in K2 Campaign 10 we identify twoeclipsing binaries containing white dwarfs with cool companions that have extremelyshort orbital periods of only 712 min (SDSS J1205minus0242 aka EPIC201283111) and725 min (SDSS J1231+0041 aka EPIC248368963) Despite their short periods bothsystems are detached with small low-mass companions in one case a brown dwarfand the other case either a brown dwarf or a low-mass star We present follow-upphotometry and spectroscopy of both binaries as well as phase-resolved spectroscopyof the brighter system and use these data to place preliminary estimates on thephysical and binary parameters SDSS J1205minus0242 is composed of a 039 plusmn 002M⊙

helium-core white dwarf which is totally eclipsed by a 0049 plusmn 0006M⊙ (51 plusmn 6MJ)brown dwarf companion while SDSS J1231+0041 is composed of a 056 plusmn 007M⊙

white dwarf which is partially eclipsed by a companion of mass 0095M⊙ In thecase of SDSS J1205minus0242 we look at the combined constraints from common-envelopeevolution and brown dwarf models the system is compatible with similar constraintsfrom other post common-envelope binaries given the current parameter uncertaintiesbut has potential for future refinement

Key words binaries eclipsing ndash stars fundamental parameters ndash stars low mass ndashwhite dwarfs ndash brown dwarfs

1 INTRODUCTION

Roughly 75 per cent of main-sequence binaries are bornwide enough that they evolve essentially as single stars

⋆ sgparsonssheffieldacukdagger Hubble Fellow

(Willems amp Kolb 2004) However for the remaining 25 percent the expansion of the more massive star at the end of itsmain-sequence lifetime causes the two stars to interact of-ten initiating a common-envelope event The frictional forcesexperienced by the stars during the common-envelope phaseresult in a dramatic reduction in the separation of the twostars down to as low as a few solar radii Angular momen-

copy 2017 The Authors

2 S G Parsons et al

Table 1 Journal of observations The eclipse of the white dwarf occurs at Phase 1 2 etc

Date at Telescope Filter Start Orbital Exposure Number of Conditionsstart of run Instrument (UT) phase time (s) exposures (Transparency seeing)

SDSS J1205minus0242

20160713 K2 Campaign 10 Kepler 0209 17653 2656 Space-based20170103 McDonaldProEM r prime 0946 084 minus 350 300 380 Clear sim17 arcsec20170119 TNTULTRASPEC iprime 1943 068 minus 231 145 481 Clear sim15 arcsec20170119 TNTULTRASPEC kg5 2059 064 minus 119 85 277 Clear sim10 arcsec20170122 TNTULTRASPEC kg5 2015 078 minus 127 85 244 Clear sim12 arcsec20170125 TNTULTRASPEC kg5 1931 081 minus 214 100 575 Clear sim15 arcsec20170126 SOARGoodman - 0736 099 minus 221 3000 17 Clear sim14 arcsec20170129 TNTULTRASPEC clear 1745 080 minus 134 100 614 Some clouds sim20 arcsec20170218 TNTULTRASPEC zprime 2025 076 minus 118 128 142 Clear sim20 arcsec20170219 TNTULTRASPEC zprime 1741 068 minus 105 128 121 Clear sim20 arcsec20170222 TNTULTRASPEC zprime 1645 053 minus 109 150 163 Clear sim20 arcsec20170304 SALTSALTICAM iprime 2200 022 minus 073 100 216 Clear sim14 arcsec20170327 GTCOSIRIS - 2327 044 minus 187 2405 25 Clear sim12 arcsecSDSS J1231+0041

20160713 K2 Campaign 10 Kepler 0210 17653 2649 Space-based20170120 TNTULTRASPEC kg5 2115 022 minus 169 200 329 Clear sim10 arcsec20170122 TNTULTRASPEC kg5 2053 056 minus 240 200 410 Clear sim12 arcsec20170221 TNTULTRASPEC kg5 1718 040 minus 220 150 529 Clear sim20 arcsec20170222 TNTULTRASPEC kg5 1745 053 minus 214 150 548 Clear sim20 arcsec20170301 SOARGoodman - 0533 054 minus 242 6000 13 Clear sim15 arcsec

tum loss via magnetic braking and gravitational radiationdrives the resulting post-common-envelope binary (PCEB)to shorter periods eventually creating a cataclysmic variablesystem

Large-scale surveys have led to an explosion inthe number of known PCEBs (Silvestri et al 2006Rebassa-Mansergas et al 2010 2016) with morethan 100 systems having measured periods (egNebot Gomez-Moran et al 2011 Parsons et al 2015)The vast majority of these systems have M-dwarf compan-ions with the spectral type distribution of the secondariespeaking near M3ndashM4 in good agreement with the peak inthe initial mass function of single stars (Chabrier 2003)Just six PCEBs are known to be composed of whitedwarfs with brown-dwarf companions (Dobbie et al 2005Burleigh et al 2006 Casewell et al 2012 Steele et al 2013Littlefair et al 2014 Farihi et al 2017)

PCEBs with brown-dwarf companions are difficult toidentify from optical data alone However infrared sur-veys have demonstrated that these systems are intrinsicallyrare with only 04minus2 per cent of white dwarfs having asub-stellar companion (Farihi et al 2005 Girven et al 2011Steele et al 2011 Debes et al 2011) including both widebinaries that never interacted (eg Becklin amp Zuckerman1988 Steele et al 2009 Luhman et al 2011) and PCEBsThe small number of white dwarfs with brown dwarfcompanions reflects the rarity of sub-stellar objects bothin the field (Kirkpatrick et al 2012) and in binaries(Grether amp Lineweaver 2006)

The short orbital periods of PCEBs provide for manydeeply eclipsing binaries which offer a unique opportunity todirectly probe the structures of both components by allow-ing for model-independent high-precision mass and radiusmeasurements (eg Parsons et al 2012b) This is especiallyuseful for uncommon objects For example there are very

few known eclipsing binaries composed of at least one browndwarf

There is only one double-lined eclipsing browndwarf binary known to date 2MASSJ05352184minus0546085(Stassun et al 2006) although another has tentatively beenidentified (David et al 2016) Both of these systems areyoung (lt10Myr) however which will affect their radii asbrown dwarfs contract throughout their lifetime The re-mainder of brown dwarfs with direct measurements of theirradii are in systems where they are highly irradiated Forexample Kelt-1b is a 27MJ brown dwarf in a 29-hr or-bit around an F star (Siverd et al 2012) it is known tobe highly inflated at the 10σ level compared to modelsHowever Wasp-30b (Anderson et al 2011) is a 61MJ browndwarf orbiting an F8 star every 416 d and has a radius thatagrees with model predictions SDSSJ14112620+2009111the only known brown dwarf to be eclipsing in a detachedPCEB also has a radius which is consistent with model pre-dictions (Littlefair et al 2014)

There is thus considerable value in finding more eclips-ing PCEBs containing a brown dwarf As part of a searchfor transits and variability in white dwarfs observed dur-ing K2 Campaign 10 we have discovered two new eclips-ing PCEBs composed of a white dwarf and a likely browndwarf companion K2 (Howell et al 2014) is an extensionof the Kepler planet-hunting mission (Borucki et al 2010)in which a number of fields along the ecliptic are continu-ously observed with high photometric precision over a pe-riod of approximately 75 days hence it is ideal for detect-ing eclipsing PCEBs We report here follow-up photometryand spectroscopy for these two new eclipsing systems andfurthermore detail and constrain their binary and stellar pa-rameters

MNRAS 000 1ndash11 (2017)

Two ultrashort eclipsing binaries from K2 3

2 OBSERVATIONS AND THEIR REDUCTION

A full journal of observations is given in Table 1

21 Target selection

We have proposed multiple Guest Observer programs tosearch for transits and variability from hundreds of knownand candidate white dwarfs in every campaign of the K2

mission As part of an analysis of targets observed withlong-cadence (294-min) exposures during K2 Campaign 10we flagged two spectroscopically confirmed white dwarfswith high-amplitude short-period variability The firstSDSSJ12051580minus0242226 (aka EPIC2012831111 here-after SDSSJ1205minus0242) showed variability near 712minvery near the Nyquist frequency of our dataset The otherSDSSJ12312714+0041329 (aka EPIC2483689632 hence-forth SDSSJ1231+0041) showed variability at a similarlyshort period of 725min

22 K2 photometry

We examined preliminary extractions of our known and can-didate white dwarfs using light curves produced by the Ke-

pler Guest Observer (GO) office (Van Cleve et al 2016)available through the Barbara A Mikulski Archive for SpaceTelescopes (MAST) The Kepler bandpass covers roughly4000minus9000 A Each K2 long-cadence observation representsa co-add of 270times602 s exposures

We improved our extraction of SDSSJ1205minus0242(Kp=188mag) by downloading the processed target pixelfile from MAST and using the PyKE software tools(Still amp Barclay 2012) Using a large (17-px) fixed aperturewe extracted the light curve fit out a quadratic function to3-day windows and corrected for K2 motion artifacts us-ing the kepsff software package (Vanderburg amp Johnson2014) Subsequently we clipped by-hand any highly dis-crepant points All data obtained in K2 Campaign 10 sufferfrom a large gap caused by the failure of a CCD module on-board the spacecraft roughly 7 d into the campaign whichpowered off the photometer for roughly 14 d Still our final6912-d light curve of SDSSJ1205minus0242 has 2656 points anda duty cycle exceeding 78 per cent

For SDSSJ1231+0041 (Kp=200mag) we saw little im-provement with our custom PyKE extraction and used thelight curve produced by the GO office for our final datasetwhich was extracted with a 2-px aperture After clippingthe final 6912-day light curve of SDSSJ1231+0041 has 2644points

23 McDonald+ProEM photometry

We obtained the first follow-up data of SDSSJ1205minus0242on 2017 January 3 using the ProEM frame-transfer cameramounted at the Cassegrain focus of the 21-m Otto Struvetelescope at McDonald Observatory in West Texas The datawere collected through an r prime filter We performed differential

1 Proposed by K2 Guest Observer programs led by PIs Kilic(GO10006) Hermes (GO10018) and Burleigh (GO10019)2 Targeted by the program led by PI Kilic (GO10006)

circular aperture photometry by extracting the target anda nearby comparison star using the IRAF task CCD HSP

(Kanaan et al 2002) and applied a barycentric correctionusing the WQED software package (Thompson amp Mullally2013)

24 TNT+ULTRASPEC photometry

We observed both our targets with the high-speed frame-transfer EMCCD camera ULTRASPEC (Dhillon et al2014) mounted on the 24-m Thai National Telescope (TNT)on Doi Inthanon Thailand in January and February 2017Our observations were made using the iprime band filter a broaduprime+g

prime+r prime filter known as kg5 (as described in Dhillon et al2014 see also the appendix of Hardy et al 2017) as wellas the zprime band and ldquoclearrdquo (fused silica) filters Exposuretimes were adjusted depending upon the conditions Thedead time between each exposure is 15ms All of these datawere reduced using the ULTRACAM pipeline software Thesource flux was determined with aperture photometry us-ing a variable aperture scaled according to the full width athalf maximum Variations in observing conditions were ac-counted for by determining the flux relative to a comparisonstar in the field of view

25 SOAR+Goodman spectroscopy

To better constrain the atmospheric parameters of theprimary white dwarfs in both systems we obtained low-resolution spectra of the upper Balmer series using the high-throughput Goodman spectrograph (Clemens et al 2004)on the 41-m SOAR telescope at Cerro Pachon in Chile Weused a 930 line mmminus1 grating and our setup covers roughly3600minus5200A at a resolution of roughly 4A set by the seeing

Using a 169primeprime slit we obtained 18times300 s exposures ofSDSSJ1205minus0242 on 2017 January 26 The data were op-timally extracted (Marsh 1989) using the pamela pack-age within starlink and flux calibrated using the stan-dard Feige 67 The final summed spectrum has a signal-to-noise (SN) of 65 per resolution element in the con-tinuum around 4600A We obtained 13times600 s exposures ofSDSSJ1231+0041 on 2017 March 1 using a 30primeprime slit Theoptimally extracted spectra were flux calibrated againstLTT2415 and have a summed SN of 24 per resolution el-ement around 4600A SOAR spectroscopy of both targetswere obtained at minimal airmass

26 SALT+SALTICAM photometry

We obtained time-series photometry of SDSSJ1205minus0242using the high-speed camera SALTICAM(OrsquoDonoghue et al 2006) mounted on the 10-m SouthernAfrican Large Telescope (SALT) on 2017 March 4 Weused SALTICAM in the frame-transfer mode whereby themoving mask occults half the CCD (the storage array) andwe took 10 s exposures with 4times4 binning yielding a platescale of 056 arcsecpixel All SALTICAM observations hadessentially zero deadtime (lt6ms) between frames

4 S G Parsons et al

Figure 1 The averaged spectrum of SDSS J1205minus0242 (left) andSDSS J1231+0041 (right) obtained with the Goodman spectro-graph on the 41-m SOAR telescope Our best fits to the Balmerlines (shown in red) yield updated atmospheric parameters forthe primary white dwarfs in each system

27 GTC+OSIRIS spectroscopy

We observed SDSSJ1205minus0242 with the Optical System forImaging and low-Intermediate-Resolution Integrated Spec-troscopy (OSIRIS) on the 104-m Gran Telescopio Canarias(GTC) on La Palma We used the R2500R grism with a 06primeprime

slit centred on the Hα line giving a resolution of R≃2500We used exposure times of 240 s and recorded a total of 25spectra of SDSSJ1205minus0242 as well as one spectrum of thespectrophotometric standard star Hiltner 600

The data were optimally extracted using pamela Anarc spectrum was used to wavelength calibrate the data Intotal 34 lines (mostly neon) were fitted with a sixth-orderpolynomial giving an rms of 001A We then applied addi-tional pixel shifts to each exposure (03 pixels maximum)based on the positions of three skylines (6300A 6863A and7276A) to correct for instrument flexure Finally the in-strumental response was removed using the spectrum of thestandard star

3 A 712-MIN BINARY SDSS J1205-0242

SDSSJ1205minus0242 (g = 185mag) was classified as a whitedwarf based on a serendipitous spectrum from the fourthdata release of the Sloan Digital Sky Survey (SDSS) byEisenstein et al (2006) The SDSS spectrum shows nosign of a companion and no obvious red excess but anautomated fit to the spectrum yields a mass of 039 plusmn

003M⊙ (Kleinman et al 2013) which is extremely low fora white dwarf implying a binary origin (Marsh et al 1995Rebassa-Mansergas et al 2011)

To confirm if this white dwarf genuinely has a lowmass we analyzed our follow-up higher-SN spectrum fromSOAR using the fitting procedures and pure-hydrogen one-dimensional model atmospheres for white dwarfs describedin Tremblay et al (2011) The fit to our SOAR spectrumis consistent with the SDSS fit giving a temperature ofTeff = 23680 plusmn 430K and surface gravity of log g = 7374 plusmn

Figure 2 Top Phase-folded binned K2 light curve ofSDSS J1205minus0242 It is strongly smeared since each long-cadenceexposure lasts 294min of the 712-min orbit Still a sinusoidalvariation and eclipse stand out Bottom Phase-folded ULTRA-SPEC kg5 light curve of SDSS J1205minus0242 with higher time sam-pling The deep eclipse of the white dwarf is clear as is the out-of-eclipse reflection effect There is no detection of the companionduring the eclipse Over-plotted in red is the best-fit model light

curve (phase smearing is accounted for in the K2 plot at top)

0057 implying a mass of 039 plusmn 002M⊙ and cooling age of50Myr calculated using the helium-core white dwarf mod-els of Panei et al (2007) The Balmer lines and best fit toSDSSJ1205minus0242 are shown in the left panel of Fig 1

The K2 light curve (Fig 2) shows strong variations ona period of 004946539(9) days (712min) The K2 data wereobtained in long-cadence mode (294-min exposures) so anysharp features in the light curve are significantly smearedout Nevertheless a clear reflection effect is seen along withsteep eclipse features implying that the system is likely tobe fully eclipsing

We display in the bottom panel of Fig 2 our high-speedfollow-up ULTRASPEC light curve of SDSSJ1205minus0242which shows a deep eclipse of the white dwarf lasting just5min from the first to fourth contact points The reflectioneffect is also evident out of eclipse We recorded a total ofseven eclipses although due to the low SN of the threezprime band eclipses we excluded these from our ephemeris cal-culations The four remaining eclipse times are listed in Ta-

Table 2 Mid-eclipse times

Cycle MJD(BTDB) Source

SDSS J1205minus0242

minus3061 57616626169(18) K2 Campaign 10minus234 577564644612(78) McDonald r prime band97 57772837463(30) ULTRASPEC iprime band118 577738762091(87) ULTRASPEC kg5 band158 577758548125(63) ULTRASPEC kg5 band218 577788227308(78) ULTRASPEC kg5 band301 577829283510(78) ULTRASPEC clearSDSS J1231+0041

minus2575 57616774090(86) K2 Campaign 10546 57773928207(40) ULTRASPEC kg5 band585 57775892041(30) ULTRASPEC kg5 band1178 57805751893(35) ULTRASPEC kg5 band1198 57806758914(31) ULTRASPEC kg5 band

ble 2 We fitted each of these eclipses with a code specificallydesigned for binaries containing at least one white dwarf(Copperwheat et al 2010) in order to determine the mid-eclipse times From these we determined the ephemeris forthe system to be

MJD(BTDB) = 57768039311(3) + 0049465250(6)E (1)

where E is the orbital phase with E = 0 corresponding tothe centre of the white dwarf eclipse

To further improve the physical constraints of each com-ponent of the binary we obtained GTC+OSIRIS time-seriesspectroscopy of the Hα line (left-hand panel of Fig 3) Thespectroscopy shows both a clear absorption component fromthe white dwarf as well as an emission component moving inanti-phase that is strongest around Phase 05 and disappearsnear the eclipse the classic signature of irradiation-inducedemission lines from the inner face of the companion to thewhite dwarf

We fitted the Hα line with the following components(1) A second-order polynomial representing the continuumof the white dwarf which is scaled according to the light-curve model during phases affected by the eclipse (2) Afirst-order polynomial representing the irradiation Its levelis modulated as (1 minus cos φ)2 where φ is the orbital phase(3) Three Gaussian absorption components for the whitedwarf that change position according to γ1 + K1 sin (2πφ)(4) Two Gaussian emission components from the compan-ion star with strengths modulated in the same way as theirradiation component and that change position accordingto γ2+Kem sin (2πφ) We also take into account the smearingof the lines caused by the finite exposure times and the bestmodel was found using the Levenburg-Marquardt minimisa-tion method

Our best-fit model is shown in the centre panel ofFig 3 with the residuals of the fit plotted in the right-handpanel Our best fit parameters were γ1 = 385 plusmn 35 km sminus1K1 = 483 plusmn 51 km sminus1 γ2 = 319 plusmn 26 km sminus1 and Kem =

3450plusmn44 km sminus1 The offset between the radial-velocity am-plitudes of the two stars γ1minusγ2 = 66plusmn43 km sminus1 is effectivelythe gravitational redshift of the white dwarf For a 039M⊙

white dwarf the expected gravitational redshift is 112 km sminus1

(Panei et al 2007) Correcting this value for the redshift ofcompanion star the difference in transverse Doppler shiftsand the potential at the companion owing to the white dwarf

reduces this to 100 km sminus1 within 1σ of the measured valueThis provides an external consistency check on the spectro-scopically determined mass

The implied mass ratio of the binary is q = M2M1 =

K1Kem = 014 Assuming a white dwarf mass of 039M⊙

gives a companion mass of 0055M⊙ or 57MJ However sincethis emission line originates only from the heated hemisphereof the companion Kem does not represent its true centre-of-mass velocity but rather a lower limit on K2 the trueradial velocity semi-amplitude of the companion Therefore0055M⊙ represents an upper limit on the mass of the com-panion the companion in SDSSJ1205minus0242 is therefore def-initely substellar

The radial velocity amplitude of the companionrsquoscentre-of-mass K2 is related to Kem via the formula

K2 =Kem

1 minus f (1 + q)R2

where R2a is the radius of the brown dwarf scaled by theorbital separation (a) and f is a constant between 0 and1 which depends upon the location of the centre of light(Parsons et al 2012a) We assume a value of f = 05 whichroughly corresponds to optically thick emission from the in-ner hemisphere which is what has been found for Hα emis-sion in similar systems (Parsons et al 2012b) This can becombined with the light curve fit to determine a more accu-rate value of K2

The combination of the eclipse light curve and the ra-dial velocity information enables us to place constraintson the stellar and binary parameters When fitting dataof the white dwarf eclipse alone there is a degeneracy be-tween the inclination and both stellar radii (scaled by theorbital separation) However we can establish the relation-ship between the masses and radii as a function of inclina-tion To do this we fitted the phase-folded light curve witha binary model (see Copperwheat et al 2010 for details ofthe code) We fixed the mass ratio to 014 (maximum valuefrom the spectroscopy) and the temperature of the whitedwarf was fixed at 24 000K We used Claret 4-parameterlimb-darkening coefficients for a 25000K log g = 75 whitedwarf (Gianninas et al 2013) in the kg5 filter3 The limb-darkening parameters of the brown dwarf have a negligibleimpact on the eclipse profile and so were fixed at the lin-ear value for a 2400K log g = 50 star in the SDSS r band(Claret et al 2012) The brown dwarf temperature was alsofixed at 2400K again this makes no difference to the eclipsefit since it is undetected during totality We then varied theinclination from 90 to 84 degrees in steps of 1 degree andallowed the scaled radii R1a and R2a to vary At eachinclination we then used the value of R2a to compute K2

via Eq 2 and combined this with K1 and the inclination todetermine the two masses as well as a and hence the tworadii

In addition to the eclipse of the white dwarf we also de-tect the transit of the white dwarf across the irradiated faceof the brown dwarf as shown in Fig 4 The depth of thisfeature is strongly dependent upon the ratio of the radii andcan be combined with the primary eclipse to place stringent

3 Limb-darkening parameters in the kg5 filter kindly provided byAlex Gianninas

6 S G Parsons et al

Figure 3 Trailed spectra of the Hα line of SDSS J1205minus0242 with time running upwards The left-hand panel shows our GTCOSIRISdata (the eclipse of the white dwarf occurs during spectrum 10) The centre panel shows our best fit model to the line including boththe absorption from the white dwarf and the emission from its companion The right-hand panel shows the residuals of the fit

Figure 4 SALTICAM iprime band light curve of the transit of thewhite dwarf in front of the heated face of the brown dwarf (iethe secondary eclipse) Also shown are models with inclinationsof 90 (blue) 87 (red) and 85 (green) Inclinations lower than87 predict a secondary eclipse that is too shallow

constraints upon the inclination (eg Parsons et al 20102012a) In this case the 90 model had a χ2 of 369 (fitting217 points) Models with inclinations lower than 87 hadχ2 values higher than 434 (χ2DOF gt 2) since they pre-dict eclipses that are too shallow Additionally the shape ofthe brown dwarf is more distorted in the lower inclination

models (since it is closer to Roche lobe filling) leading tomuch poorer fits (eg the 85 model in Fig 4) Thereforeour secondary eclipse data place a lower limit of 87 on theinclination

Our final constraints on the stellar parameters areshown in Fig 5 and Fig 6 for the white dwarf and browndwarf respectively and are listed in full in Table 3 We foundthat the minimum inclination of the system is 85 below thisthe radius of the brown dwarf needs to be so large to fit theeclipse width that it fills its Roche lobe This places a hardupper limit on the mass of the white dwarf of 043M⊙ Ourresults from the secondary eclipse (i gt 87) further constrainthis upper limit to 040M⊙ The uncertainty on the K1 mea-surement dominates the error on the brown dwarfrsquos massleading to a mass range of 0049 plusmn 006M⊙ (51 plusmn 6MJ)

Fig 5 shows that the measured radius of the white dwarfis fully consistent with theoretical predictions for helium-core white dwarfs (red line) by Panei et al (2007) at aninclination of 90 and is slightly oversized at lower incli-nations although this is still well within the uncertain-ties The radii predicted by the carbon-oxygen core models(blue line) of Fontaine et al (2001) are significantly smallercompared to our measurements Both models have canon-ically thick hydrogen-layer masses and we conclude thatSDSSJ1205minus0242 has a helium core At the highest incli-nations the fit is also consistent with the white dwarf pa-rameters found from the SOAR spectroscopy implying thatthe true inclination is somewhere close to 90 Fig 6 showsthat the brown dwarfrsquos (volume-averaged) radius is consis-tent with theoretical predictions if it is older than 25Gyr

Figure 5 Constraints on the mass and radius of the white dwarfin SDSS J1205minus0242 based on our radial velocity and eclipse fit-ting The black line shows how the mass and radius varies as afunction of inclination The red line shows the theoretical mass-radius relationship for a 24 000K helium-core white dwarf witha canonically thick surface hydrogen layer (Panei et al 2007)

The blue line shows the same theoretical mass-radius relation-ship but for a carbon-oxygen core white dwarf (Fontaine et al2001) also with a canonically thick surface hydrogen layer Thevertical dashed line marks the white dwarf mass as determinedfrom the Balmer-line fits to our SOAR spectrum with the shadedarea showing the the 1σ uncertainties on this fit The hatchedregion shows the inclinations excluded by the secondary eclipsedata The uncertainty on the white dwarfrsquos mass from the radialvelocity data is shown on the 90 model

(gt35Gyr for the 90 solution) We did not detect the browndwarf in our zprime band light curves placing an upper limit onits spectral type of L0 consistent with its classification as abrown dwarf

We estimate the distance to the white dwarf by fittingthe SDSS photometry with the Panei et al (2007) modelsWe sample the posterior probability distributions for theparameter set log g Teff E(g minus i) d using a Markov-ChainMonte-Carlo (MCMC) analysis Posteriors on log g Teff

come from the SOAR spectral fits whilst the posterior onE(g minus i) is uniform between 0 and the maximum extinctionalong the line of sight To minimise the effects of contami-nation by the irradiated companion we only fit the uprimegprimer prime

photometry and find a distance of 720plusmn40 pc Fitting the fulluprimegprimer primeiprimezprime dataset does not change the distance significantly

By combining this distance with an estimate ofthe proper motion based upon SDSS and PanSTARRS(Tian et al 2017) and adopting γ2 as an estimate of theradial velocity of the system we can calculate its Galac-tic space velocity relative to the local standard of restas UVW = (36minus19 35) plusmn (7 6 4) km sminus1 We adopt the con-vention that the sign of U is positive towards the Galac-tic anti-centre Following Bensby et al (2014) we find thatSDSSJ1205minus0242 is ten times more likely to belong to thethin disk than the thick disk and 50000 times more likely tobelong to the thin disk than the halo justifying the adoptionof solar metallicity models for the brown dwarf

Figure 6 Constraints on the mass and radius of the brown dwarfin SDSS J1205minus0242 The black line shows how the mass and ra-dius varies as a function of inclination given our observationalconstraints Note that the radius measurements correspond tothe volume-averaged radius of the brown dwarf Inclinations lessthan 85 are ruled out as the brown dwarf would fill its Rochelobe The hatched region shows the inclinations excluded by thesecondary eclipse data Also shown in red are theoretical mass-radius relationships for solar metallicity brown dwarfs of differentages (Baraffe et al 2003) The uncertainty on the brown dwarfrsquosmass from the radial velocity data (which dominates over the in-clination uncertainty) is shown on the 90 model The implicationof these models is that SDSS J1205minus0242 has a total system agebetween 25 ndash 10Gyr

4 A 725-MIN BINARY SDSS J1231+0041

The second of our two systems SDSSJ1231+0041 is a faint(g = 201) white dwarf identified from a serendipitous SDSSspectrum Spectroscopic fits by Rebassa-Mansergas et al(2016) to the SDSS spectrum found the white dwarf to bea Teff = 38740 plusmn 2680K log g = 707 plusmn 041 white dwarf theirsolution suggested a possible photometric excess at longerwavelengths sufficient for them to classify it as a possiblewhite dwarf plus main-sequence star system

We have fit our higher-SN SOAR spectrum to betterconstrain the white dwarf atmospheric parameters as wedid for SDSSJ1205minus0242 Our updated SOAR fits find theprimary white dwarf has Teff = 37210plusmn1140 K log g = 777plusmn

015 which yields a white dwarf mass of 056plusmn007 M⊙ usingthe models of Fontaine et al (2001) The SOAR spectrumand best fit are shown in the right panel of Fig 1 Followingthe same method as for SDSSJ1205minus0242 we estimate adistance of 1500 plusmn 200 pc by fitting the carbon-oxygen corewhite dwarf models of Fontaine et al (2001) to the SDSSuprimegprimer prime photometry

The K2 light curve of SDSSJ1231+0041 shows clearvariations on a period of 0050353815(28) days (725 min)displayed in Fig 7 However since these data were takenin long-cadence mode (with 294-min exposures) it was notimmediately clear whether this is the true binary period orif the period is twice this value If this is the binary periodthen the light curve variations must be the result of repro-cessed light on the inner hemisphere of the companion tothe white dwarf (ie reflection effect) The period could alsobe double this value with the variations then caused by the

8 S G Parsons et al

Figure 7 Top Phase-folded binned K2 light curve ofSDSS J1231+0041 As with SDSS J1205minus0242 the data aresmeared since each long-cadence exposure comprises more than40 per cent of the 725-min orbital period Still it shows a strongreflection effect as well as slight evidence for an eclipse Bottom

Phase-folded ULTRASPEC kg5 light curve of SDSS J1231+0041with higher time sampling The partial eclipse of the white dwarfis clear as is the out-of-eclipse reflection effect A model lightcurve is over-plotted in red (and smeared in the top panel tomatch the K2 exposures)

Roche-distorted companion presenting different surface ar-eas throughout the orbit (ie ellipsoidal modulation) Thelong exposure times of the K2 data relative to the variabilitymake it difficult to distinguish between these two possibil-ities There is also marginal evidence for an eclipse in theform of a steeper curve just before and after the minimum

Our ground-based follow-up high-speed photometry(Fig 7) shows that the true binary period is 725 min andestablishes that the system is eclipsing albeit only partially(we do not detect any secondary eclipse) In total we coveredfour eclipses of the white dwarf (see Table 2) From these wedetermined the ephemeris for the system to be

MJD(BTDB) = 57746435076(24) + 0050353796(23)E (3)

Since SDSSJ1231+0041 is only partially eclipsing de-termining accurate parameters from the light curve is com-plicated due to the extra level of degeneracy However withsuch a short orbital period we can place an upper limit on

Figure 8 Theoretical massndashradius relationships (solid lines) forsolar metallicity brown dwarfs and low-mass stars (Baraffe et al2003) with ages of 05Gyr (black) 1Gyr (red) 5Gyr (green)and 10Gyr (blue) The grey dashed lines show the possibleloci of Roche-lobe-filling companions to a white dwarf of massMWD = 056M⊙ for fixed orbital period (in steps of 5min) Thefigure shows that the most dense brown dwarfs are those withmasses sim0065M⊙ that could still fit within their Roche lobes atperiods as short as 45min provided they are old enough It alsoshows that to fit within its Roche lobe at a period of 725min (in-dicated by the solid grey line) the companion to the white dwarfin SDSS J1231+0041 must have a mass of less than 0095M⊙

the mass of the companion to the white dwarf based on thefact that it does not fill its Roche lobe In Fig 8 we show sev-eral mass-radius relationships for low-mass stars and browndwarfs of different ages and solar metallicity Also shown arelines of constant density at different orbital periods whicheffectively shows the Roche lobe radius at different orbitalperiods It is interesting to note how the radii of these low-mass objects are strongly related to their ages to the extentthat some binary configurations could only be possible witholder brown dwarfs For example any detached system witha period 50 minutes must be older than sim5Gyr and musthave a mass 007M⊙ With a period of 725 min we canonly say that SDSSJ1231+0041 must be older than sim1Gyrsince at this period most of the models converge We canalso place an upper limit on the mass of the companion tothe white dwarf of sim0095M⊙ anything more massive thanthis would fill its Roche lobe Therefore it is quite likelythat the companion in SDSSJ1231+0041 is a brown dwarfalthough radial-velocity data are required to confirm this

Such a low-mass companion is completely out-shone by the white dwarf at visible wavelengths butRebassa-Mansergas et al (2016) suggested a possible pho-tometric excess at red optical wavelengths We have phasedthe SDSS photometry to the ephemeris established here andfind it was all taken within 5 min of orbital phase 025 Theapparent photometric red excess is therefore likely the re-sult of irradiation since the heated face of the companionhas a much lower temperature but larger area than the whitedwarf

Our final parameters for both systems are listed in Ta-ble 3

Table 3 Stellar and binary parameters for the two systems pre-sented in this paper WD refers to the white dwarf

Parameter SDSS J1205-0242 SDSS J1231+0041

Orbital period (d) 0049465250(6) 0050353796(23)Orbital separation (R⊙) 042minus045 -Orbital inclination (deg) 87minus90 -WD mass (M⊙) 039 plusmn 002 056 plusmn 007

WD radius (R⊙) 00217minus00223 -WD Teff(K) 23680 plusmn 430 37210 plusmn 1140

WD log g 737 plusmn 005 777 plusmn 014

WD cooling age (Myr) 50 5Secondary mass (M⊙) 0049 plusmn 0006 0095

Secondary radius (R⊙) 0081minus0087 012

App magnitude (gprime) 185 201Distance (pc) 720 plusmn 40 1500 plusmn 200

5 DISCUSSION

The white dwarfs in both the systems presented in this pa-per are the remnants of giant stars that were once muchlarger than their current orbits This implies significant or-bital shrinkage and points towards their emergence fromcommon-envelopes that formed around both components ofthe binaries when mass transfer from the giant stars totheir low mass companions took place (Paczynski 1976)Common-envelope evolution is one of the most poorly un-derstood and yet significant phases of close binary evolution(Ivanova et al 2013) and well-constrained examples of itseffects are worth examination for the constraints they mayraise

In this case SDSSJ1205minus0242 offers the most interest-ing test first because it is better constrained second becauseit contains a moderately low-mass helium-core white dwarfThe helium white dwarf in SDSSJ1205minus0242 is a remnant ofthe first ascent red giant branch (RGB) As Nelemans et al(2000) pointed out the close relation between the core-massand radius of RGB stars (Refsdal amp Weigert 1970) can al-low tight constraints to be derived on the prior evolutionof binary stars containing helium white dwarfs The radiusof an RGB star rises rapidly with the helium core mass(approximately prop M4

c Iben amp Tutukov 1985) thus heliumwhite dwarfs of low mass that have emerged from common-envelopes are of particular interest since they come fromrelatively small tightly-bound RGB stars They can there-fore lead to the most stringent constraints upon the effi-ciency with which the envelope is ejected We express theeffect of the common-envelope upon the orbital separationa through the relation

GM1 f M2

2a fminus

GM1iM2

GM1i(M1i minus M1 f )

λR1i (4)

which equates a fraction of the orbital energy change onthe left with the binding energy of the RGBrsquos envelope onthe right (Webbink 1984 Dewi amp Tauris 2000) Here thesubscripts i and f refer to the initial and final values of therespective parameter when they differ The parameters αand λ encapsulate the efficiency with which orbital energy isused to eject the envelope and the internal structure of theenvelope respectively

There are alternative formulations for the binding en-ergy of the envelope (Iben amp Livio 1993) We claim no

08 10 12 14 16 18 20Initial primary mass M1i (M⊙)

αλ 035 M⊙ 039 M⊙ 043 M⊙

Figure 9 The value of the combined common-envelope RGBstructure parameter αλ (Dewi amp Tauris 2000) required to matchthe parameters of SDSS J1205minus0242 The curved lines were cal-culated for three values of white dwarf mass the value of 039M⊙

from spectroscopy and 2σ either side of it and set M2 = 0049 M⊙The shaded regions show the ranges of progenitor mass consistentwith the age of the brown dwarf for the models of Baraffe et al(2015) (upper-right shaded blue) and Saumon amp Marley (2008)(lower-left shaded orange) allowing for the white dwarfrsquos cool-ing age of 50Myr (Panei et al 2007) and using the formula forthe time taken to reach the base of the RGB from Hurley et al(2000) Solar metallicity has been assumed (Section 3)

advantage for our choice other than its popularity whicheases comparison with other studies we refer the reader toIvanova et al (2013) and Zorotovic et al (2010) for furtherdiscussions of such variations and their effect upon the out-come of the common-envelope phase Like Zorotovic et al(2010) we condense what we can deduce from the systeminto a single constraint upon the combination parameterαλ

The core mass-radius relation means that R1i is largelydefined by the mass of the white dwarf M1 f with only amodest dependence upon its progenitorrsquos mass M1i There-fore as the progenitor mass increases both terms in the nu-merator of the right-hand side of Eq 4 increase with littlechange in the denominator On the left-hand side howeverthere is relatively little change with the progenitor massas it is the first term in the brackets that dominates sinceai ≫ a f The result is that the value of αλ required to pro-duce SDSSJ1205minus0242 increases rapidly with the progeni-tor mass M1i These constraints are encapsulated in Fig 9which is based on the formulae presented by Hurley et al(2000) and Eggleton (1983) in order to calculate R1i and aifor a given choice of progenitor mass

Ranges of progenitor mass consistent with the age ofthe brown dwarf for two sets of models (Baraffe et al 2015Saumon amp Marley 2008) are highlighted in Fig 9 The rangeof αλ runs from 01 ndash 22 consistent with many of the sys-tems studied in a similar manner by Zorotovic et al (2010)If brown dwarf models can be improved there is potentialfor sharper constraints upon the common-envelope parame-ters and given the easily-detectable secondary eclipse inSDSS J1205minus0242 (Fig 4) there are good prospects fortightening the parameter constraints significantly beyondthose shown in Fig 9 There are caveats however firstare the significant existing uncertainties of cloud physics

10 S G Parsons et al

molecular opacity and convection in brown dwarf models(Saumon amp Marley 2008 Baraffe et al 2015) and secondit is possible that the unusual environment of rapid rotationand irradiation could affect the brown dwarfrsquos size althoughthe 50 Myr since the common-envelope is a blink of an eyecompared to the brown dwarfrsquos Kelvin-Helmholtz timescaleResolving such uncertainties is motivation for finding moreexamples of such systems

The usual aim of studies such as this is to constrain thecommon-envelope efficiency parameter α but as we haveseen it is the combination αλ that is directly constrainedUnfortunately the structure parameter λ often taken tobe 05 (de Kool 1990) is almost as ill-defined as α as it isnot known to what extent the internal (thermal) energy ofthe envelope needs to be taken into account when calculat-ing it (Han et al 1995 Dewi amp Tauris 2000 Camacho et al2014) a contribution from internal energy can increase λsignificantly The relatively tightly-bound RGB star in thisinstance should make this uncertainty relatively small com-pared to later stages of stellar evolution and once the pa-rameters of the binary are firmed up it will be worth investi-gating stellar models to define the range of λ for this specificinstance In any event it is clear that SDSSJ1205minus0242 andsimilar white dwarf brown dwarf systems have significantpotential for both common-envelope evolution and browndwarf physics

Prior to the formation of the common-envelope thebinary would have had an orbital period in the range60 minus 200days placing it within or close to the ldquobrown dwarfdesertrdquo where few brown dwarf companions to solar-typestars are seen (Marcy amp Butler 2000 Ma amp Ge 2014) Itwould be interesting to ascertain whether or not the numbersof white dwarf brown dwarf PCEBs are consistent with therarity of their progenitors in radial-velocity surveys

These systems will transfer mass in the near future(sim300Myr for SDSSJ1205minus0242) and presumably appearas cataclysmic variable stars Their existence in this formhowever may be brief if recent suggestions of the desta-bilising effects of novae on cataclysmic variables containinglow mass white dwarfs are correct (Schreiber et al 2016Nelemans et al 2016) They may then soon merge to becomesingle white dwarfs and in the case of SDSSJ1205minus0242a single white dwarf of low mass (Zorotovic amp Schreiber2017) a number of which are known (Marsh et al 1995Brown et al 2011) Their emergence from the common-envelope phase so close to Roche-lobe filling also suggeststhat had the companions been even less massive these sys-tems might not have survived the common-envelope at allbut simply have merged This exact scenario has been sug-gested as another way to form single low-mass white dwarfs(Nelemans amp Tauris 1998)

6 CONCLUSIONS

Using long-cadence photometry from the Kepler space tele-scope we have discovered two new ultrashort detachedeclipsing binaries composed of white dwarfs plus cool com-panions The binaries have such short orbital periods mdash712min and 725min mdash that the companions are likely sub-stellar on the basis of their periods alone in order that theydo not fill their respective Roche lobes

Follow-up photometry and spectroscopy significantlyconstrain both systems SDSSJ1205minus0242 contains a hot(Teff = 23680plusmn430 K) low-mass (039plusmn002M⊙ ) white dwarfwith a radius consistent with a helium-core white dwarf(00217minus00223R⊙ ) It is totally eclipsed every 712min bya 45ndash57 MJ brown dwarf companion that has a radius con-sistent with an age greater than 25Gyr (0081minus0087R⊙)Detection of secondary eclipses constrains the orbital incli-nation to gt 87 deg

SDSSJ1231+0041 contains a hot (Teff = 37210plusmn1140 K)056 plusmn 007M⊙ white dwarf that is partially eclipsed every725min by a companion of less than 0095M⊙ likely also tobe a brown dwarf Details of all physical constraints to bothsystems are listed in Table 3

The shorter-period system SDSSJ1205minus0242 placesuseful constraints upon common-envelope evolution becauseof its helium-core white dwarf and the need for the whitedwarfrsquos total age to match the age of its brown dwarfcompanion This demonstrates that ultrashort-period whitedwarf plus brown dwarf binaries can be used to test the-ories of common-envelope evolution because of the time-dependent radii of brown dwarfs although uncertainties inbrown dwarf models require clarification for this method tobe applied with confidence

Both systems were discovered as part of a search fortransits and variability among white dwarfs in K2 Campaign10 we expect to find more similar short-period eclipsing bi-naries as K2 continues surveying new fields along the eclip-tic The results here also help build expectations for thenext space-based photometric mission TESS (Ricker et al2014) which can be used to target many bright white dwarfsall-sky at 2-min cadence

ACKNOWLEDGEMENTS

SGP acknowledges the support of the Leverhulme TrustThe research leading to these results has received fund-ing from the European Research Council under the Euro-pean Unionrsquos Seventh Framework Programme (FP2007-2013) ERC Grant Agreement numbers 340040 (HiPER-CAM) and 320964 (WDTracer) as well as the EuropeanUnionrsquos Horizon 2020 Research and Innovation Programme ERC Grant Agreement number 677706 (WD3D) UL-TRACAM TRM VSD SPL are supported by the Sci-ence and Technology Facilities Council (STFC) under grantsSTL000733 and STM001350 DAHB is supported by theNational Research Foundation of South Africa Support forthis work was provided by NASA through Hubble Fellow-ship grant HST-HF2-51357001-A by NASA K2 Cycle 4Grant NNX17AE92G as well as NSF grants AST-1413001and AST-1312983 This work has made use of data ob-tained at the Thai National Observatory on Doi Inthanonoperated by NARIT the Southern Astrophysical Research(SOAR) telescope which is a joint project of the Ministerioda Ciencia Tecnologia e Inovacao da Republica Federativado Brasil the US National Optical Astronomy Observa-tory the University of North Carolina at Chapel Hill andMichigan State University the McDonald Observatory ofthe University of Texas at Austin as well as the SouthernAfrican Large Telescope (SALT) through DDT programme2016-2-DDT-006 where the assistance of Marissa Kotze is

acknowledged Data for this paper have been obtained un-der the International Time Programme of the CCI (Interna-tional Scientific Committee of the Observatorios de Canariasof the IAC) with the Gran Telescopio Canarias (GTC) op-erated on the island of La Palma in the Observatorio delTeideRoque de los Muchachos

Facilities K2 Otto Struve (ProEM) TNT (ULTRA-SPEC) SOAR (Goodman) SALT (SALTICAM) GTC(OSIRIS)

REFERENCES

Anderson D R et al 2011 ApJ 726 L19

Baraffe I Chabrier G Barman T S Allard F Hauschildt P H2003 AampA 402 701

Baraffe I Homeier D Allard F Chabrier G 2015 AampA577 A42

Becklin E E Zuckerman B 1988 Nature 336 656Bensby T Feltzing S Oey M S 2014 AampA 562 A71

Borucki W J et al 2010 Science 327 977Brown J M Kilic M Brown W R Kenyon S J 2011 ApJ

730 67Burleigh M R Hogan E Dobbie P D Napiwotzki R Maxted

P F L 2006 MNRAS 373 L55Camacho J Torres S Garcıa-Berro E Zorotovic M Schreiber

M R Rebassa-Mansergas A Nebot Gomez-Moran A Gan-sicke B T 2014 AampA 566 A86

Casewell S L et al 2012 ApJ 759 L34Chabrier G 2003 PASP 115 763Claret A Hauschildt P H Witte S 2012 AampA 546 A14Clemens J C Crain J A Anderson R 2004 in Moor-

wood A F M Iye M eds Proc SPIEVol 5492Ground-based Instrumentation for Astronomy pp 331ndash340doi10111712550069

Copperwheat C M Marsh T R Dhillon V S Littlefair S PHickman R Gansicke B T Southworth J 2010 MNRAS402 1824

David T J Hillenbrand L A Cody A M Carpenter J MHoward A W 2016 ApJ 816 21

Debes J H Hoard D W Wachter S Leisawitz D T CohenM 2011 ApJS 197 38

Dewi J D M Tauris T M 2000 AampA 360 1043

Dhillon V S et al 2014 MNRAS 444 4009Dobbie P D Burleigh M R Levan A J Barstow M A Napi-

wotzki R Holberg J B Hubeny I Howell S B 2005MNRAS 357 1049

Eggleton P P 1983 ApJ 268 368Eisenstein D J et al 2006 ApJS 167 40

Farihi J Becklin E E Zuckerman B 2005 ApJS 161 394Farihi J Parsons S G Gansicke B T 2017 Nature Astronomy

1 0032Fontaine G Brassard P Bergeron P 2001 PASP 113 409Gianninas A Strickland B D Kilic M Bergeron P 2013 ApJ

Girven J Gansicke B T Steeghs D Koester D 2011 MNRAS417 1210

Grether D Lineweaver C H 2006 ApJ 640 1051Han Z Podsiadlowski P Eggleton P P 1995 MNRAS 272 800Hardy L K et al 2017 MNRAS 465 4968Howell S B et al 2014 PASP 126 398

Hurley J R Pols O R Tout C A 2000 MNRAS 315 543Iben Jr I Livio M 1993 PASP 105 1373Iben Jr I Tutukov A V 1985 ApJS 58 661Ivanova N et al 2013 AampARv 21 59

Kanaan A Kepler S O Winget D E 2002 AampA 389 896Kirkpatrick J D et al 2012 ApJ 753 156

Kleinman S J et al 2013 ApJS 204 5

Littlefair S P et al 2014 MNRAS 445 2106Luhman K L Burgasser A J Bochanski J J 2011 ApJ

730 L9Ma B Ge J 2014 MNRAS 439 2781Marcy G W Butler R P 2000 PASP 112 137Marsh T R 1989 PASP 101 1032Marsh T R Dhillon V S Duck S R 1995 MNRAS 275 828Nebot Gomez-Moran A et al 2011 AampA 536 A43Nelemans G Tauris T M 1998 AampA 335 L85Nelemans G Verbunt F Yungelson L R Portegies Zwart S F

2000 AampA 360 1011Nelemans G Siess L Repetto S Toonen S Phinney E S 2016

ApJ 817 69OrsquoDonoghue D et al 2006 MNRAS 372 151Paczynski B 1976 in Eggleton P Mitton S Whelan J eds

IAU Symposium Vol 73 Structure and Evolution of CloseBinary Systems p 75

Panei J A Althaus L G Chen X Han Z 2007 MNRAS382 779

Parsons S G Marsh T R Copperwheat C M Dhillon V SLittlefair S P Gansicke B T Hickman R 2010 MNRAS402 2591

Parsons S G et al 2012a MNRAS 419 304Parsons S G et al 2012b MNRAS 420 3281Parsons S G et al 2015 MNRAS 449 2194Rebassa-Mansergas A Gansicke B T Schreiber M R Koester

D Rodrıguez-Gil P 2010 MNRAS 402 620Rebassa-Mansergas A Nebot Gomez-Moran A Schreiber M R

Girven J Gansicke B T 2011 MNRAS 413 1121Rebassa-Mansergas A Ren J J Parsons S G Gansicke B T

Schreiber M R Garcıa-Berro E Liu X-W Koester D2016 MNRAS 458 3808

Refsdal S Weigert A 1970 AampA 6 426Ricker G R et al 2014 in Space Telescopes and Instrumenta-

tion 2014 Optical Infrared and Millimeter Wave p 914320(arXiv14060151) doi101117122063489

Saumon D Marley M S 2008 ApJ 689 1327Schreiber M R Zorotovic M Wijnen T P G 2016 MNRAS

455 L16Silvestri N M et al 2006 AJ 131 1674Siverd R J et al 2012 ApJ 761 123Stassun K G Mathieu R D Valenti J A 2006 Nature

440 311Steele P R Burleigh M R Farihi J Gansicke B T Jameson

R F Dobbie P D Barstow M A 2009 AampA 500 1207Steele P R Burleigh M R Dobbie P D Jameson R F

Barstow M A Satterthwaite R P 2011 MNRAS 416 2768Steele P R et al 2013 MNRAS 429 3492Still M Barclay T 2012 PyKE Reduction and analysis of Ke-

pler Simple Aperture Photometry data Astrophysics SourceCode Library (ascl1208004)

Thompson S Mullally F 2013 Wqed Lightcurve AnalysisSuite Astrophysics Source Code Library (ascl1304004)

Tian H-J et al 2017 preprint (arXiv170306278)Tremblay P-E Bergeron P Gianninas A 2011 ApJ 730 128Van Cleve J E et al 2016 PASP 128 075002Vanderburg A Johnson J A 2014 PASP 126 948Webbink R F 1984 ApJ 277 355Willems B Kolb U 2004 AampA 419 1057Zorotovic M Schreiber M R 2017 MNRAS 466 L63Zorotovic M Schreiber M R Gansicke B T Nebot Gomez-

Moran A 2010 AampA 520 A86de Kool M 1990 ApJ 358 189

This paper has been typeset from a TEXLATEX file prepared bythe author

1 Introduction

2 Observations and their reduction

21 Target selection

22 K2 photometry






3 A 712-min Binary SDSS J1205-0242

4 A 725-min Binary SDSS J1231+0041

5 Discussion

6 Conclusions