Northwestern University, 2145 Sheridan Rd., Evanston, IL ... · Kamann7, Nathan W. C.Leigh8, Robert...

arX

iv1

703

1016

7v1

[as

tro-

phS

R]

29

Mar

201

7Draft version April 9 2018Preprint typeset using LATEX style emulateapj v 082209

ON THE ORIGIN OF SUB-SUBGIANT STARS I DEMOGRAPHICS

Aaron M Geller12daggerlowast Emily M Leiner3 Andrea Bellini4 Robert Gleisinger56 Daryl Haggard5 SebastianKamann7 Nathan W C Leigh8 Robert D Mathieu3 Alison Sills9 Laura L Watkins4 David Zurek810

1Center for Interdisciplinary Exploration and Research in Astrophysics (CIERA) and Department of Physics amp AstronomyNorthwestern University 2145 Sheridan Rd Evanston IL 60201 USA

2Adler Planetarium Dept of Astronomy 1300 S Lake Shore Drive Chicago IL 60605 USA3Department of Astronomy University of Wisconsin-Madison WI 53706 USA

4Space Telescope Science Institute 3700 San Martin Drive Baltimore MD 21218 USA5Department of Physics McGill University McGill Space Institute 3550 University Street Montreal QC H3A 2A7 Canada

6Department of Physics and Astronomy University of Manitoba Winnipeg MB R3T 2N2 Canada7Institut fur Astrophysik Universitat Gottingen Friedrich-Hund-Platz 1 37077 Gottingen Germany

8Department of Astrophysics American Museum of Natural History Central Park West and 79th Street New York NY 100249Department of Physics and Astronomy McMaster University Hamilton ON L8S 4M1 Canada

10Visiting Researcher NRC ndash Herzberg Astronomy and Astrophysics National Research Council of Canada 5071 West Saanich RoadVictoria British Columbia V9E 2E7 Canada

Draft version April 9 2018

ABSTRACT

Sub-subgiants are stars observed to be redder than normal main-sequence stars and fainter thannormal subgiant (and giant) stars in an optical color-magnitude diagram The red straggler starswhich lie redward of the red giant branch may be related and are often grouped together with thesub-subgiants in the literature These stars defy our standard theory of single-star evolution and areimportant tests for binary evolution and stellar collision models In total we identify 65 sub-subgiantsand red stragglers in 16 open and globular star clusters from the literature 50 of these including 43sub-subgiants pass our strict membership selection criteria (though the remaining sources may alsobe cluster members) In addition to their unique location on the color-magnitude diagram we findthat at least 58 (2543) of sub-subgiants in this sample are X-ray sources with typical 05-25 keVluminosities of order 1030minus31 erg sminus1 Their X-ray luminosities and optical-to-X-ray flux ratios aresimilar to those of RS CVn active binaries At least 65 (2843) of the sub-subgiants in our sampleare variables 21 of which are known to be radial-velocity binaries Typical variability periods are 15days At least 33 (1443) of the sub-subgiants are Hα emitters These observational demographicsprovide strong evidence that binarity is important for sub-subgiant formation Finally we find thatthe number of sub-subgiants per unit mass increases toward lower-mass clusters such that the openclusters in our sample have the highest specific frequencies of sub-subgiantsSubject headings open clusters and associations ndash globular clusters ndash binaries (including multiple)

close ndash blue stragglers ndash stars evolution ndash stars variables general

1 INTRODUCTION

Sub-subgiant stars are defined empirically as sourcesthat fall redward of the normal main sequence (MS) andare fainter than the subgiant and giant branches in anoptical color-magnitude diagram (CMD) Stars in thisregion were first noted in the open cluster M67 (NGC2682) by Belloni et al (1998) and then discussed in de-tail in Mathieu et al (2003) Around this same timesimilar stars were noted in the globular cluster 47 Tuc(NGC 104) by Albrow et al (2001) Since then a largenumber of sources that fall into this region in a CMDhave been pointed out in various papers in both openand globular clusters and analyzed to varying degreesThese stars are unexplained by the standard theory ofsingle-star evolution and despite the growing number ofsuch sources their origin remains a mysteryThis is the first paper in a series studying the origin of

these stars formed within both open and globular clus-ters (and perhaps also in the Galactic field) In this paperwe gather these sources from the many disparate refer-ences in the literature examine their demographics and

lowastNSF Astronomy and Astrophysics Postdoctoral FellowElectronic address daggera-gellernorthwesternedu

begin to discuss their possible formation channelsFirst we address a conflicting naming convention that

appears throughout the literature in reference to thesestars Belloni et al (1998) introduced the name ldquosub-subgiantrdquo while Albrow et al (2001) used the nameldquored stragglerrdquo The motivation behind the name sub-subgiants is empirically driven as these stars are fainterthan the normal subgiant branch The term red straggleris also empirically motivated as these stars are indeedredder than stars of similar luminosity in these clustersThe term red straggler is no doubt also influenced by themore well known ldquoblue stragglersrdquo and ldquoyellow strag-glersgiantsrdquo1 (both of which are generally brighter thanthe normal subgiant branch)In short depending on the specific reference a star

in the same region of the CMD may be called by eithername and therefore there is confusion in the literatureabout which CMD domain sub-subgiants and red strag-

1 Previously the term red straggler was also used in the literatureto refer to stars found in between the blue straggler and red giantregions on the CMD sometimes described as evolved blue strag-glers (eg Eggen 1983 Eggen amp Iben 1988 1989 Landsman et al1997) More recently these stars are referred to as ldquoyellow strag-glersrdquo or more commonly ldquoyellow giantsrdquo

2 Geller et al

glers occupyTo clarify this nomenclature we choose to use the term

ldquosub-subgiantrdquo (hereafter SSG) to refer to stars that arefainter than the normal subgiants and redder than thenormal MS stars shown as the dark-gray regions in Fig-ure 1 In addition to these SSG stars there are also starsthat are observed to be redder than the normal red gi-ants but brighter than the normal subgiants (found inthe light-gray regions of Figure 1) We choose to use theterm ldquored stragglerrdquo (hereafter RS) to refer to these starsand we encourage the community to adopt this divisionof the SSG and RS nomenclature moving forwardWe divide the SSG and RS stars in a given photome-

try band (and isochrone family) by the magnitude at thebase of the giant branch as shown in Figure 1 This def-inition depends on the position of the isochrone whichmight vary if defined by different photometric studiesor isochrone families We take cluster parameters fromthe literature (see Table 1) and use PARSEC isochrones(Bressan et al 2012) here For much of our sample thisdefinition is sufficient to place a star into either the SSGor RS category for all of the optical photometry we col-lected from the literature (and provide in Table 4) How-ever some stars reside in the SSG region in one color-magnitude combination and in the RS region for a differ-ent combination (see Figure 1 and Table 4) We discussthis phenomenon in some detail for specific sources be-low Briefly some stars move around dramatically rela-tive to the isochrone with different filter choices perhapsdue to spot activity and photometric variability (see alsoeg Milliman et al 2016) Indeed photometric variabil-ity appears to be a defining characteristic of the SSG andRS stars For the SSG analysis presented here we in-clude all stars that fall in the SSG region in at least oneoptical color-magnitude combination This is the mostinclusive definition of SSGs which may be importantgiven the photometric variability We will also accountfor potential field star contamination within this SSGsample (Section 4) prior to performing our demographicanalysesIn this paper we gather the observations from the SSG

and RS stars in each cluster in Section 2 and provide asummary of the observations of these stars in Table 4In Sections 3 and 4 we discuss possible observational bi-ases in our sample and the potential for field-star con-tamination respectively In Section 5 we investigate theaggregate characteristics of the sample We close with abrief discussion and conclusions in Section 6 Subsequentpapers in this series will investigate in detail a set oftheoretical formation channels for SSGs (some of whichpredict that SSG and RS stars may be related throughformation andor evolution) and evaluate the formationrates of each channel and their abilities to create SSGswith the demographics identified here

2 OBSERVED SUB-SUBGIANTS

In Table 4 we compile all of the SSG and RS starsthat have been identified in the literature in both openand globular clusters We provide the clusterrsquos NGCname another common name where appropriate theSSGRS ID (see the relevant paragraph below for thespecific references for IDs and other values) the RAand Dec proper-motion and radial-velocity membershipsstatuses (PPM and PRV where available) our estimate

of the probability that this star is a field star (Pfieldsee Section 4) the radial distance from the cluster cen-ter in units of core radii (rc) the available observedUBV RIJHK photometry2 an available X-ray luminos-ity (LX) and the band of the X-ray observation theradial-velocity orbital period (ldquoPerRVrdquo) and photometricperiod (ldquoPerphotrdquo) where available (we mark variableswhose periods are yet to be determined as ldquovarrdquo) and fi-nally the number of color-magnitude combinations givenin this table (from the literature) that place the givenstar in the SSG or RS regions or neither (ldquoSSGRSNrdquo)In Table 4 and throughout this paper we identify

sources with velocities that are gt 3σ from the clustermean (if that is the only available membership measure-ment) or with formal membership probabilities lt50 asnon-members We exclude definite non-members fromTable 4 In some cases authors identify stars as non-members at lt 3σ from the mean (eg at gt 2σ orgt 25σ) We choose to include such stars with ques-tionable membership within Table 4 and indicate thisuncertain membership with ldquordquo but we do not includethese in our subsequent analysis We discuss additionalmembership indicators in Section 4Observations of the SSG stars reveal an intriguing mix-

ture of properties Broadly SSGs

1 are redder than normal MS stars but fainter thannormal giants in an optical color-magnitude dia-gram

2 have X-ray luminosities LX of order 1030 minus1031 erg sminus1 in both open and globular clusters

3 are often Hα emitters (where measurements havebeen made)

4 exhibit photometric variability with periods 15days (where available) and

5 where possible are mostly identified as radial-velocity binaries

We return to these aggregate properties in Section 5First we briefly discuss the observations from each of theindividual star clusters listed in Table 4 CMDs for eachof these clusters showing the SSG and RS stars alongwith an isochrone (for reference) and the SSGRS regionsare plotted in Figure 1 and relevant cluster parametersare shown in Table 1

21 Open Cluster Observations

NGC 188 mdash This old (sim6-7 Gyr) open cluster has threeSSGs and two RSs (IDs from Geller et al 2008) The op-tical and IR photometry for these stars in Table 4 comefrom Stetson et al (2004) and 2MASS respectively Allof these stars have proper motions (Platais et al 2003)consistent with cluster membership and where measure-ments are possible these stars are also radial-velocitymembers (Geller et al 2008) All but one of these stars

2 All optical photometry are given in Johnson-Cousins filters(and have not been de-reddened) Where necessary HST mag-nitudes are converted to ground-based Johnson-Cousins followingHoltzman et al (1995) and Sirianni et al (2005) All JHK in-frared photometry come from 2MASS (Skrutskie et al 2006)

Demographics of Sub-subgiant Stars 3

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 188

00 05 10 15(V-R)0

0

1

2

3

4

5

MV

NGC 2158

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 2682

(M67)

10 15 20(V-I)0

2

3

4

5

6

7

MV

NGC 6791

05 10 15(B-V)0

0

1

2

3

4

5

MV

NGC 6819

05 10 15 20(B-V)0

1

2

3

4

5

6

MV

NGC 7142

05 10 15(V-I)0

1

2

3

4

5

6

MV

NGC 104

(47 Tuc)

05 10 15 20(B-R)0

0

1

2

3

4

5

MR

NGC 5139(ω Cen)

05 10 15(V-I)0

1

2

3

4

5

6

MV

NGC 6121

(M4)

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 6218

(M12)

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 6366

05 10 15 20(B-R)0

0

1

2

3

4

5

MR

NGC 6397

05 10 15(B-V)0

0

1

2

3

4

5

6

MV

NGC 6752

05 10 15(B-V)0

0

1

2

3

4

5

MV

NGC 6809

(M55)

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 6838

(M71)=SSG

=RS

filled X-ray source=

purple Securemembership

=

orange Less securemembership

=

Fig 1mdash Color-magnitude diagrams for the SSG and RS stars in each respective cluster Open clusters are plotted first (with namescolored black) followed by the globular clusters (with names colored dark green) All magnitudes are given in Table 4 and where necessarywe choose the color-magnitude combination for a given cluster that allows us to plot the largest number of sources For reference we alsoplot with black lines PARSEC isochrones (Bressan et al 2012) using the ages distance moduli and reddening values from Table 1 TheSSG region is shown in the dark-gray filled area while the RS region is shown in the light-gray filled area (both as defined in Section 1and note that we exclude the normal binary locus from these regions) We plot sources that appear in the SSG region in at least onecolor-magnitude combination with circles and the RS stars in squares Filled symbols show known X-ray sources and open symbols showthose without detected X-ray emission Highly likely cluster members are plotted with purple symbols those with less secure membershipare plotted in orange (see Section 5) Note that NGC 6652 is not shown here as we only have an estimate of the sourcersquos V magnitude(due to high-frequency variability)

(1141 a RS) have radial-velocity variations indicative ofa binary companion Two of these five sources (one SSGand one RS) are detected as X-ray emitters (Belloni et al1998 Gondoin 2005) Gondoin (2005) suggests that theX-rays from 4289 (a SSG) are due to rapid rotation andresulting chromospheric activity This source lies veryclose to the base of the giant branch in a V vs B minus VCMD but is farther removed from the ldquonormalrdquo stars

(and isochrones) in CMDs using other photometry com-binations particularly in combination with I band (giventhe photometry from Stetson et al 2004) Furthermore4289 is a member of a binary with a period of 1149 days(Geller et al 2009) which is consistent with the hypoth-esis that the X-rays result from rapid rotation in thesynchronized primary member of the binary 1141 is aRS and also an X-ray source (Belloni et al 1998 X5)

4 Geller et al

TABLE 1Cluster Parameters

Cluster age Mcl (mminusM)V E(B minus V ) [FeH] nH (rrc)max

[Gyr] [M⊙ ] [1020cmminus2]

Open Clusters

NGC 188 62 1500 1144 009 00 66 71NGC 2158 2 15000 1451 055 -06 419 79NGC 2682 4 2100 96 001 00 33 72NGC 6791 8 4600 1338 01 04 107 106NGC 6819 24 2600 123 01 00 193 125NGC 7142 36 500 1286 029 01 326 22

Globular Clusters

NGC 104 131 10times106 1337 004 -072 54 36NGC 5139 115 22times106 1394 012 -153 88 51NGC 6121 125 13times105 1282 035 -116 142 33NGC 6218 127 14times105 1401 019 -137 77 89NGC 6366 133 48times105 1494 071 -059 140 55NGC 6397 127 77times104 1237 018 -202 114 508NGC 6652 129 79times104 1528 009 -081 92 400NGC 6752 118 21times105 1313 004 -154 49 411NGC 6809 123 18times105 1389 008 -194 94 65NGC 6838 120 30times104 1384 025 -078 212 78

Note References for the values in this table are as follows For the open clusters NGC 188 Sarajedini et al

(1999) Meibom et al (2009) and Chumak et al (2010) NGC 2158 Carraro et al (2002) NGC 2682 Geller et al

(2015 and references therein) NGC 6791 Stetson et al (2003) Carney et al (2005) and Tofflemire et al (2014)

NGC 6819 Kalirai et al (2001) and Hole et al (2009 and references therein) NGC 7142 Sandquist et al (2013

and references therein) and Straizys et al (2014) For the globular clusters we take the age from Marın-Franch et al

(2009 using the ldquoG00CGrdquo values and normalized using the age of 47 Tuc from Thompson et al 2010) (m minusM)V

E(B minus V ) [FeH] and Mcl (calculated assuming a mass-to-light ratio of 2) from Harris (1996 2010) For NGC

6838 we take the age (m minus M)V E(B minus V ) from Di Cecco et al (2015) All nH values are derived from NASArsquos

HEASARC nH tool (httpsheasarcgsfcnasagovcgi-binToolsw3nhw3nhpl) which uses Dickey amp Lockman

(1990) and Kalberla et al (2005) Finally note that NGC 6397 and NGC 6752 are core-collapsed clusters the

radial limits of these surveys in units of half-mass radii are 09 and 37 respectively

though the authors of that study do not venture to guessthe source of the emission We note that Geller et al(2008) incorrectly matched 3118 (a RS) to a Belloni et al(1998) X-ray source to our knowledge 3118 does not havedetected X-ray emission 3118 is a double-lined spec-troscopic binary (SB2) with a period of 119 days anda mass ratio of 08 (Geller et al 2009) InterestinglySSG 4989 is identified as a W UMa photometric vari-able V5 by Zhang et al (2002) which in general arethought to be contact binaries containing two MS stars(Robertson amp Eggleton 1977)

NGC 2158 mdash As part of their search for transitingplanets in this intermediate age (sim2 Gyr) open clus-ter Mochejska et al (2004 2006) identified five photo-metric variables in the clusterrsquos SSG region IDs andoptical photometry in Table 4 for these stars are fromMochejska et al (2004 2006) One of these sources(V90) is in the catalog of Dias et al (2014) with a94 proper-motion membership The remaining fourhave proper motions from Kharchenko et al (1997) allare gt50 proper-motion members when considering thecluster ldquocoronardquo stellar distribution though all but oneof these fall to lt50 when considering the ldquocorerdquo distri-bution (Kharchenko et al 1997 describe the cluster as acombination of two distributions with the ldquocoronardquo hav-ing a characteristic radius of twice that of the ldquocorerdquo)The photometric periods for these sources range from lt1day to sim13 days To our knowledge there is no published

X-ray survey of NGC 2158

NGC 2682 (M67) mdash Mathieu et al (2003) performedan extensive observational analysis of the two SSGs inthe old (sim4 Gyr) open cluster M67 and we refer thereader to this paper for more information In shortthese stars were first noted by Belloni et al (1998) andboth sources are high-probability cluster members fromproper motions (Girard et al 1989) and radial veloci-ties (Mathieu et al 2003 Geller et al 2015) Both areradial-velocity binaries (see Table 3) the shorter-periodsource S1113 (IDs for both sources from Sanders 1977)is an SB2 with a companion that is likely a 09 M⊙ MSstar while S1063 the longer-period source is a single-lined spectroscopic binary (SB1) Both of these sourcesare photometric variables (van den Berg et al 2002) andX-ray sources (Belloni et al 1998 van den Berg et al2004) (The optical and IR photometry in Table 4for these stars is from (Montgomery et al 1993) and2MASS respectively) As noted in Mathieu et al (2003)both stars show strong Ca II H and K emission indica-tive of chromospheric activity and both also show Hαemission (Pasquini amp Belloni 1998 van den Berg et al1999) Finally Mathieu et al (2003) note that theycould not find a self-consistent solution for the stars ofS1113 that accounts for all of the observations

NGC 6791 mdash Platais et al (2011) identify five stars inthe SSG region and van den Berg et al (2013) identify


two additional SSGRS stars (6371 and 7011) as opti-cal counterparts to X-ray sources in this old (sim8 Gyr)metal rich ([FeH]= 04) open cluster The IDs and opti-cal photometry for these sources are from Stetson et al(2003) and the IR photometry is from 2MASS All ofthese sources have proper-motion membership proba-bilities of PPM ge 96 (Platais et al 2011) Four ofthe five Platais et al (2011) candidates (83 746 362615561) were confirmed to be radial-velocity membersby Milliman et al (2016) The authors are currentlycollecting radial-velocity measurements for the othertwo SSG candidates through the WIYN Open ClusterStudy (WOCS Mathieu 2000) Though not publishedin Milliman et al (2016) these WOCS radial velocitiesindicate that 6371 and 7011 are both short-period SB2binaries with radial velocities spanning the cluster dis-tribution we have not yet been able to derive orbital so-lutions for these stars so we cannot yet provide conclu-sive radial-velocity membership probabilities The fivePlatais et al (2011) sources are also short-period radial-velocity binaries Milliman et al (2016) published or-bital solutions for 746 3626 and 15561 (see Table 3)Five of these NGC 6791 sources (four SSGs and one RS)are short-period photometric variables X-ray sourcesand Hα emitters (de Marchi et al 2007 Mochejska et al2002 Kaluzny 2003 Bruntt et al 2003 Mochejska et al2005 van den Berg et al 2013 Milliman et al 2016)For the variable SSG and RS sources in NGC 6791photometric variability occurs on periods similar to theradial-velocity orbital period and has been attributed tospot modulation (van den Berg et al 2013) All signspoint to these stars being RS CVn-type binaries withchromospheric activity On the other hand star 83 showsno signs of a binary companion photometric variabilityHα emission or X-ray emission so it appears qualita-tively different than the other SSGs in the cluster In-terestingly RS 6371 falls to the red of the RGB in theV vs V minus I CMD of Stetson et al (2003) but appearsto be a normal cluster giant in the gprime vs gprime minus rprime CMD ofPlatais et al (2011) and is thus not reported as an SSGin the Platais et al sample 6371 is also a known photo-metric variable (de Marchi et al 2007 V9) identified asan eclipsing binary within the Kepler field and also anHα emitter (van den Berg et al 2013)

NGC 6819 mdash Gosnell et al (2012) identify 52004 withtheir X-ray source X9 in the intermediate-age (sim24 Gyr)open cluster NGC 6819 The optical and IR photometrygiven in Table 4 is from (Kalirai et al 2001) and 2MASSrespectively This star has a proper-motion membershipprobability of 99 from Platais et al (2013) Though ithas many spectroscopic observations fromWIYNHydrathe source is a rapid rotator and therefore reliable ra-dial velocities are difficult to obtain (Hole et al 2009)Gosnell et al (2012) find the X-ray and optical proper-ties of this source to be consistent with an active binaryand note that it is similar to an RS CVn The sourceis clearly in the SSG region in a V vs B minus V CMD(see Figure 1) though it is not an obvious outlier in theultra-violet CMD presented in Gosnell et al (2012 theirFigure 5b)

NGC 7142 mdash Sandquist et al (2011) performed a pho-tometric variability study of this intermediate age (sim36

Gyr) open cluster and discovered the source V4 to havevariability on multiple timescales and amplitudes in-cluding trends of sim001 mag (particularly in B andV ) in a given night plus longer timescale variationsover tens of days at about 05 mag (in B V and R)(We take the ID and optical photometry for this starfrom Sandquist et al (2011) and IR photometry from2MASS) Investigation of their best-seeing images doesnot indicate any binary companion They note thatthe location of V4 in the CMD is reminiscent of theSSGs in M67 though V4 has higher-amplitude photo-metric oscillations We place V4 in the RS category(see Figure 1) Dias et al (2014) find V4 to have a 91proper-motion membership probability As also notedby Sandquist et al (2011) radial-velocity observationsfor V4 would be very important to confirm cluster mem-bership and investigate for a binary companion To ourknowledge there is no published X-ray survey of NGC7142

22 Globular Cluster Observations

NGC 104 (47 Tuc) mdash Shortly after the discovery ofthe SSGs in M67 Albrow et al (2001) noted a popu-lation of six photometric variable stars that reside inthe SSG region in a CMD of the globular cluster 47Tuc (and called them ldquored stragglersrdquo) Edmonds et al(2003) later added four additional sources to this list intheir analysis of Chandra X-ray observations Opticalphotometry for these stars in Table 4 are converted tothe ground-based system from the HST magnitudes fromAlbrow et al (2001) and Edmonds et al (2003) We areable to match six of these ten sources to the HSTPROMOcatalog3 (Bellini et al 2014 and include their positionsin Table 4 which have an average epoch of observa-tions of 20062) and can therefore evaluate their proper-motion memberships (The other sources were rejectedfrom the proper-motion pipeline due to contaminationfrom nearby stars or poor PSF fitting in one or moreepochs or they were simply outside of the field of viewof the proper-motion catalog) After careful cleaning andanalysis of the full cluster data set we find that all buttwo of these six sources have relative velocities within 3σof the clusterrsquos mean motion and we therefore identifythese four stars as proper-motion members WF4-V17is clearly a proper-motion non-member of 47 Tuc and isnot included in our table (see Figure 2 where the proper-motion errors for the six target stars are smaller thantheir colored symbols) Instead WF4-V17 is a red giantin the Small Magellanic Cloud WF4-V18 has a 2D ve-locity 375σ from the cluster mean but also has a ratherpoor χ2 value for the linear fit defining its proper-motion(see Bellini et al 2014 Watkins et al 2015) and alsolarge uncertainties on the proper-motion This sourceappears to be a MUSE radial-velocity member (see Fig-ure 3) so we keep it in our table However as withother sources with uncertain membership we will not in-clude WF4-V18 in further analyses Three more of thesesources were also observed by the MUSE multi-epochradial-velocity survey (Kamann et al 2013 2016) whoconfirm their cluster membership based on both velocityand metallicity Two of these sources show Hα emission

3 HSTPROMO draws from the ACS Survey for Globular Clus-ters httpwwwastroufledu~atapublic_hstgc

6 Geller et al

The MUSE radial velocities indicate that all four of thesesources show strong radial-velocity variability indicativeof binary companions with the strongest radial-velocityvariable (WF2-V32) reaching an amplitude of gt30 kmsminus1 We currently do not have sufficient epochs of ra-dial velocities to derive orbital solutions and thereforecenter-of-mass radial velocities Thus we do not quoteradial-velocity membership probabilities for these starsin Table 4 we show their mean velocities relative tothe rest of the 47 Tuc MUSE sample in Figure 3 Sevenof these nine candidate cluster members are detected inX-rays (Grindlay et al 2001 Edmonds et al 2003) andAlbrow et al (2001) note that their X-ray luminositiesare consistent with that expected for a chromospheri-cally active subgiant star in an RS CVn type systemOf additional interest PC-V11 (also known as W36 inEdmonds et al 2003 and AKO 9 Auriere et al 1989) isa known CV in the SSG region (eg Grindlay et al 2001Knigge et al 2002)

NGC 5139 (ω Centauri) mdash ω Cen has (at least)seven distinct sequences apparent in the opticalIRCMD (Villanova et al 2007 Bellini et al 2010) andalso has very sensitive Chandra imaging (Haggard et al2009 2013) Of particular interest here the ldquoanoma-lous RGBSGBrdquo (sequence D from Villanova et al2007 and also known as RGBSGB-a Lee et al 1999Pancino et al 2000 Ferraro et al 2004) contains sim10of the subgiant stars has a subgiant branch that is sig-nificantly fainter than the other subgiant branches anda red-giant branch that is significantly redder than theother red-giant branches In other words the anoma-lous RGBSGB runs through the SSG and RS regionsof the ω Cen CMD relative to the bluer and brightersequences Villanova et al (2007) find the anomalousRGBSGB to be old (sim 13 Gyr) and metal rich ([FeH]sim -11) in comparison to the other branches Interest-ingly Cool et al (2013) discovered eight X-ray sourceslying within the anomalous RGBSGB on the CMD(with IDs containing numbers and a letter in Table 4)Three of these sources all candidate SSGs (22e 32fand 43c) are identified as Hα ldquoBrightrdquo by Cool et al(2013) We are able to match six of these eight sourcesto either the Bellini et al (2009) proper-motion study orthe MUSE radial-velocity survey all of which are con-sistent with cluster membership (though 32f has onlya 43 proper-motion membership probability) TheseCool et al (2013) stars form a relatively tight sequenceon the CMD (UBV RI magnitudes for the ω Cen stars inTable 4 are from Bellini et al (2010) where available ex-cept for the Cool et al (2013) sources where we converttheir HST magnitudes to ground-based B and R and usethese instead) Similarly tight sequences of SSG andorRS stars are not immediately apparent in other clusters(see Figure 1 though the number of SSG and RS stars inmost clusters is perhaps too small to discern a sequencein a CMD) Without metallicity measurements it is un-clear whether these Cool et al (2013) X-ray sources areassociated with the anomalous RGBSGB or with somedifferent branch If they are associated with the anoma-lous RGBSGB then they may not be SSG (or RS) starsaccording to the CMD definition from Section 1 (unlessall of the anomalous RGBSGB are SSG and RS stars inrelation to some different branch) However in this sce-

nario the anomalous RGBSGB would have a factor of5-15 times more X-ray sources (above the detection lim-its) than the other branches that dominate the clustermass and would indicate a strong preference for X-raysources at higher metallicities in ω Cen a trend not ob-served elsewhere (Cool et al 2013) On the other handif these X-ray sources are instead associated with a dif-ferent branch (by metallicity) then they would fit ourdefinition of SSG and RS stars We choose to includethem here and future metallicity measurements for thesestars will be very important to confirm their identity asSSGRS stars Regardless of the nature of the Cool et al(2013) sources Rozyczka et al (2012) identify an addi-tional 13 3D kinematic cluster members in the SSG andRS regions in the cluster (not confined to the anomalousRGBSGB) (Note SSG candidate 23 lies blueward ofthe R vs BminusR isochrone in Figure 1 but inside the SSGregion in other filter combinations) This sample includesthree radial-velocity variables plus four additional ldquosus-pected radial-velocity variablesrdquo (all of which we identifyas ldquovarrdquo in Table 4) and one W UMa photometric vari-able Furthermore Rozyczka et al (2012) state that itis conceivable that the remaining 5 objects may also bethus far undetected binaries

NGC 6121 (M4) mdash Bassa et al (2004) identify two can-didate SSGs in their Chandra X-ray survey of this globu-lar cluster CX8 and CX10 For source CX8 we take theoptical and IR photometry from Stetson et al (2014) and2MASS respectively and for CX10 we convert the HSTphotometry from Bassa et al (2004) to ground-based Vand I Zloczewski et al (2012) find both to be likelyproper-motion members (each with a proper-motion lt25σ from the cluster mean) Both CX8 and CX10 fallwithin the MUSE sample but there is only one epoch ofobservations for this cluster (and therefore radial-velocityvariability is unknown) CX10 appears to be a memberby both radial velocity and metallicity (Figure 3) CX8appears separated from the cluster distribution BothCX8 and CX10 show Hα emission in their spectra whichmay bias the metallicity measurements Also if CX8is a binary the one radial-velocity epoch may not re-flect the center-of-mass motion Bassa et al (2004) notethat CX8 coincides with the photometric variable V52from Kaluzny et al (1997) which they classify as a BYDra system (generally thought to contain MS stars withvariability arising from spots and chromospheric activ-ity) with a period of sim078 days Kaluzny et al (2013)continue to monitor V52 and note that the periodicityremains coherent over 14 years which they take as in-dication that V52 is a binary star However contraryto Bassa et al (2004) Kaluzny et al (2013) specificallystate that none of the X-ray sources from Bassa et al(2004) coincide with V52 (which indeed they find puz-zling given the expected chromospheric activity of sucha star) Nascimbeni et al (2014) confirm the sim078 dayphotometric period and also associate this star with CX8(and with their 7864) They categorize this source asldquounclassifieduncertainrdquo We choose to provide the pho-tometric period for V52 in Table 4 as related to CX8 Im-portantly based on radio observations Strader et al (inprep) suggest that CX8 has a compact object compan-ion with a high likelihood that the companion is a blackhole Due to the uncertainty in binarity (and hence the


Fig 2mdash Proper-motion diagrams for NGC 104 (47 Tuc top)and NGC 6397 (bottom) Stars from our sample are plotted incolored symbols while the rest of the stars in the direction of eachcluster respectively are plotted in black points Uncertainties onthe proper-motion measurements for the stars in our SSG sampleare smaller in size than the colored dots In both clusters themembers are easily distinguished visually and confirmed throughour more detailed analyses as those having velocities consistentwith the bulk motion of the cluster stars (where here the meancluster motion is shifted to the origin) Further details for thesestars in both clusters are provided in the text of Section 2

center-of-mass radial velocity) and the probable bias inthe metallicity measurement we do not suggest a radial-velocity membership for CX8 Finally for completenesswe note that Bassa et al (2004) identify another sourceCX24 that also falls to the red of the standard MS butis fainter than the typical SSG region as defined hereThe optical counterpart to CX24 varies by more than 1mag in brightness between their HST observation datesBassa et al (2004) suggest that CX24 is a foreground ob-ject and we therefore do not include this source in ourtable

NGC 6218 (M12) mdash Lu et al (2009) identify an X-raysource (their CX2) in NGC 6218 with a ldquorelatively hard

X-ray colorrdquo for which they find three potential opti-cal counterparts One of these potential optical coun-terparts CX2b falls in the SSG region on an opticalCMD and we convert their HST photometry for thissource to ground-based B V and R for Table 4 (Theother two fall on or possibly to the red depending on thecolor choice of the main-sequence and to the blue of themain sequence respectively) Zloczewski et al (2012)find this source to be gt 2σ from the mean proper mo-tion of the cluster and therefore categorize this star as anon-member We include this star in our table thoughas with other similar sources we will not include this starin our subsequent analysis

NGC 6366 mdash Bassa et al (2008) identify one candidateSSG CX5 in their Chandra X-ray survey of this globu-lar cluster To our knowledge there is no proper-motionmembership probability available for this source in theliterature but Bassa et al (2008) conclude that this isa probable cluster member based on the observed X-ray luminosity and their optical photometry (which weprovide in Table 4)

NGC 6397 mdash Four Chandra X-ray sources fromCohn et al (2010 U12 U18 U42 and U92) reside inthe cluster SSG region (We convert the Cohn et al(2010) HST magnitudes to ground-based B and R forTable 4) All of these sources are found in the HST-PROMO catalog and we use the HSTPROMO positionsin Table 4 which have an average epoch of observationsof 20064 All but one of these sources are probable mem-bers from both HSTPROMO proper motions and MUSEradial velocities and metallicites The remaining sourceU42 falls well outside of the cluster distribution (see Fig-ures 2 and 3) We therefore exclude U42 from our sam-ple (and also note that U42 is somewhat redder thanmost of the SSGs in our sample) All three membersare photometric variables (Kaluzny et al 2006) and allshow remarkably high amplitude radial-velocity variabil-ity Each were observed twice within about 24 hours bythe MUSE team and have radial velocities that differ by70 to 200 km sminus1 Importantly U12 whose radial ve-locity was observed to vary by 200 km sminus1 in a day isa known millisecond pulsar (MSP DrsquoAmico et al 2001Ferraro et al 2003) and Bogdanov et al (2010) suggestthat U18 is also a MSP Kaluzny et al (2006) attributethe short-period photometric variability for both of thesesources (U12 = V16 U18 = V31) to ellipsoidal variations(though they are somewhat uncertain about that charac-terization for U18) MUSE spectra show that both U12and U18 have Hα in emission Finally source U92 (V7)is an eclipsing W UMa binary and shows Hα in absorp-tion in the MUSE spectra For completeness we alsonote that U63 U65 U86 are located redward of the MSbut fainter than the SSG region discussed here All areX-ray sources and U65 and U86 each have an Hα excessHowever Cohn et al (2010) show that these three starshave proper motions that are inconsistent with both thecluster and field distributions and therefore their mem-bership status is unknown They suggest these threestars may be foreground active binaries and we do notinclude them in our sample (or Table 4)

NGC 6652 mdash Source B is one of two known LMXBsin the globular cluster NGC 6652 (Heinke et al 2001)

8 Geller et al

minus14 minus12 minus10 minus08 minus06 minus04 minus02 00

[MH]

minus100

minus80

minus60

minus40

minus20

0

20

40

v rad[kms]

NGC104

WF4-V18PC1-V48WF2-V31WF2-V32

minus4 minus3 minus2 minus1 0 1

[MH]

180

200

220

240

260

280

300

v rad[kms]

NGC5139

13b

minus25 minus20 minus15 minus10 minus05 00

[MH]

minus100

minus50

0

50

100

150

v rad[kms]

NGC6121

CX10CX8

minus35 minus30 minus25 minus20 minus15 minus10 minus05 00 05

[MH]

minus200

minus150

minus100

minus50

0

50

100

150

200v rad[kms]

NGC6397

U12U92U18U42

Fig 3mdash Radial-velocity (vrad) plotted against metallicity ([MH]) for NGC 104 (47 Tuc top left) NGC 5139 (ω Cen top right) NGC6121 (bottom left) and NGC 6397 (bottom right) Stars from our sample are plotted in colored symbols with error bars The rest ofthe stars surveyed for each cluster are shown in gray points Only stars with robust vrad and [MH] measurements are shown Verticalerror bars in the plots for 47 Tuc ω Cen and NGC 6397 account for the radial-velocity variability NGC 6121 has only one epoch ofobservations and the vertical error bars show the (much smaller) uncertainties on individual measurements Again we caution that thesingle radial-velocity epoch for NGC 6121 may not show the true center-of-mass motion of binary stars (as may be the case for CX8)We use open symbols for stars that show significant Hα emission in the MUSE spectra that likely affects our metallicity measurementsFurther details for all of these stars are provided in the text of Section 2

and was studied in detail by Coomber et al (2011)and Engel et al (2012) This source ldquoflickersrdquo ontimescales less than 75 seconds (the exposure time forthe Engel et al 2012 observations) with amplitudes ofsim 1 mag in grsquo and sim05 mag in rrsquo and therefore theoptical photometry and particularly an optical color ishighly uncertain In Table 4 we provide an estimateof the V magnitude from Heinke et al (2001) for ref-erence Engel et al (2012) suggest a color potentiallyredder than the MS akin to the SSGs in other clus-ters Deutsch et al (2000) find a possible photometric

period of 436 minutes though Heinke et al (2001) andEngel et al (2012) suggest that this period is spuriousThe source also flares in X-rays on timescales down to100s and can range from LX(05-100keV)lt 2times1033 ergsminus1 up to LX(05-100keV)sim 1035 erg sminus1 with no de-tectable periodicity though its long-term LX is observedto be relatively constant since 1994 ROSAT observations(Coomber et al 2011) The high peak in LX is strong ev-idence for a neutron star or black hole companion How-ever the variability and somewhat low X-ray luminosityis unusual for typical LMXBs Because the optical pho-


tometry is so uncertain we cannot reliably classify thissource as an SSG or RS we include this source in Table 4for reference but do not include it in our subsequentanalysis

NGC 6752 mdash Kaluzny amp Thompson (2009) find threephotometric variables in the SSG region in this globu-lar cluster (IDs and photometry for these three starsin Table 4 are from Kaluzny amp Thompson 2009) Twoare roughly the same V magnitude as the base of the gi-ant branch (V19 and V20) while the other is fainterV19 is quite red somewhat similar in this regard toU42 in NGC 6397 (which appears to be a non-member)Kaluzny amp Thompson (2009) propose that the photo-metric variability for these sources is due to binarityand possibly ellipsoidal variations with a degeneratecompanion To our knowledge these sources are notdetected in X-rays Zloczewski et al (2012) find thatV19 and V20 are gt 2σ from the mean of the clusterproper-motion distribution and categorize them as non-members Again as this is below our 3σ limit we chooseto include these stars in our table as candidates but wewill not include them in our subsequent analysis

NGC 6809 (M55) mdash We find three candidate SSGs fromthe literature in this globular cluster CX7 and CX29from Bassa et al (2008) and V64 from Kaluzny et al(2010) Both CX7 and V64 are cluster members whileCX29 is likely a non-member (at gt 3σ) from proper-motion measurements (Zloczewski et al 2011) and istherefore not included in our sample (In Table 4the IDs and photometry for CX7 are from Bassa et al2008 while the ID and photometry for V64 are fromKaluzny et al 2010) CX7 is detected in X-rays byBassa et al (2008) who note that this source is likelya magnetically active binary that has no significant Hαemission CX7 is also a photometric variable fromKaluzny et al (2010 V65) who posit that the photomet-ric variability is either due to a contact binary observedat low inclination or ellipsoidal variations suggesting adegenerate companion Moreover Kaluzny et al (2010)suggest that the coherence of the photometric variationsfor both V64 and V65 (CX7) are indicative of binarycompanions Lane et al (2011) find that CX7 is a clus-ter non-member from radial-velocity observations butgiven the evidence for binarity and therefore the un-known center-of-mass velocity (without an orbital so-lution) we suggest that the radial-velocity membershipstatus is uncertain

NGC 6838 (M71) mdash Huang et al (2010) identify twoSSG candidates in this globular cluster Their source s02has an X-ray to optical flux ratio indicative of an activebinary and X-ray variability that likely indicates flaringfrom a chromospherically active star (Elsner et al 2008)The X-ray source s19 has three potential HST opticalcounterparts within the Chandra error circle All three ofthese potential counterparts would have X-ray to opticalflux ratios consistent with active binaries (though s19ccould also be interpreted as a CV) Huang et al (2010)suggest that s19a is the most likely counterpart to thesource evidently because this would place the star inthe SSG region Photometry for both NGC 6838 sourcesare converted from the Huang et al (2010) HST filtersto ground-based magnitudes for Table 4

23 Kepler ldquoNo-Manrsquos-Landrdquo Stars

Batalha et al (2013) and then Huber et al (2014)identify a subset of roughly 5 of the Kepler targets(nearly 10000 stars) with (photometric) surface gravitiesand temperatures that are inconsistent with the expec-tations for normal stars from standard isochrones Morespecifically these are generally stars of G or K spec-tral type that have surfaces gravities that are too highand temperatures that are too cool to be consistent withany isochrone less than 14 Gyr (even at extremely highmetallicities eg see Figure 14 in Huber et al 2014)Moreover these stars fall in the SSG and RS regionsAlmost definitely some of these stars simply have in-correct surface gravities andor temperatures HoweverHuber et al (2014) follow up a subset of these starswith spectroscopic classifications from the SEGUE cat-alog (Yanny et al 2009) and find that even these moreaccurate surface gravities and temperatures do not moveall stars out of the ldquoNo-Manrsquos-Landrdquo regime Indeedthey state that ldquoa considerable number of SEGUE clas-sifications remain in the rsquoNo-Mans-Landrsquo zonerdquo We sug-gest that there may be a substantial population of fieldSSG stars within the Kepler ldquoNo-Manrsquos-Landrdquo sampleIf even a subset of the ldquoNo-Manrsquos Landrdquo stars are in-

deed SSGs then their discovery in the field suggests thatSSGs can form through channels that do not require dy-namical encounters within star clusters (This may notbe surprising since it is also well known that blue strag-glers exist in clusters and the field and can form throughmechanisms mediated by dynamics as well as throughisolated binary evolution) We point out these stars hereto motivate further observations and analyses that mighthelp confirm whether or not these stars are indeed SSGs

24 Summary

In summary we compile a sample of 65 stars in 16star clusters identified in the literature as either SSG orRS stars We classify 56 of these stars as SSGs basedon our CMD definition described above (and shown inFigure 1) In the following sections we describe andattempt to characterize the biases and incompletenessin this sample and we discuss the cluster membershipstatus of these stars From our analysis presented inthe following sections considering the proper motionsradial velocities photometric variability and X-ray lu-minosities we conclude that the vast majority of thesestars are indeed cluster members We select these highlylikely cluster members when investigating the SSG de-mographics in Section 5

3 OBSERVATIONAL BIASES IN THE SAMPLE

We do not attempt to formally correct for the selectioneffects or incompleteness that is likely present in the sam-ple of SSGRS stars in Table 4 We will however limittheir impact in the analysis of these data by includingonly the most likely cluster membersMost of the sources from the open clusters listed in Ta-

ble 4 were identified from comprehensive radial-velocityand proper-motion membership surveys Most of theglobular cluster sources on the other hand were initiallyidentified in X-ray surveys (without comprehensive mem-bership surveys)The detection limit of most of these X-ray surveys is

10 Geller et al

of the order of 1030minus31 erg sminus1 which appears to bethe characteristic X-ray luminosity of these SSG starsTherefore (a) there may be more SSGRS stars in clus-ters with X-ray luminosities below sim1030minus31 erg sminus1 thathave not been identified in the literature and (b) theremay be unidentified SSGRS stars in clusters that cur-rently have less sensitive X-ray observations not reaching1030minus31 erg sminus1 The large frequency of X-ray emittingSSG and RS stars may be simply due to the discoverymethod although the open cluster sample suggests oth-erwise Proper-motion surveys of a large sample of glob-ular clusters are nearing completion which will help toidentify non-X-ray-detected SSGRS stars (and help tofurther eliminate non-members from our SSGRS sam-ple)X-ray surveys of globular clusters have often targeted

the most dynamically active clusters since observa-tions indicate a trend of increasing frequency of X-raysources with increasing collision rate (Pooley et al 2003Bahramian et al 2013) Therefore the globular clustersample here is likely biased toward the more massiveand dense clusters Indeed NGC 6397 and NGC 6752are core-collapsed clusters in the Harris (2010) catalogOpen clusters that have a particularly large number ofstars are also often selected for radial-velocity photomet-ric and X-ray surveys and therefore again our sampleof open clusters is likely biased toward the most massiveclusters at a given ageAlso not all sources listed in Table 4 were monitored

for photometric variability and not all sources (particu-larly those in the globular clusters) were observed forradial-velocity variability Therefore the frequency ofvariables in our sample is a lower limitFurthermore as with most studies of periodic data

we expect that the ability to detect periodicity in thissample decreases with increasing period Therefore theperiod distribution of the known variable SSGRS stars(eg Figure 5) may be biased toward shorter periodsLastly as these clusters are all at different distances

and have been observed using different telescopes andinstruments the radial coverage of the clusters variesacross our sample We show the maximum radial ex-tent (rrc) of the relevant SSG discovery survey for eachcluster in Table 1 The cluster with the smallest radialcoverage that contains SSG stars is NGC 6121 (one of theclosest globular clusters in this group) with a maximalradial coverage of 33 core radii note that NGC 7142with a smaller coverage does not have any SSG starsWhen necessary to reduce the effect of radial incom-pleteness amongst our surveys we will limit our sampleto only include those SSGs within 33 core radii fromtheir respective clusters for the analysis in Section 5

4 EVIDENCE FOR CLUSTER MEMBERSHIP

41 Probability of Field Star Contamination

The majority of the SSGs in our sample have proper-motion measurements indicative of cluster membershipMany of the sources also have radial-velocity measure-ments indicative of cluster membership However thequality of these kinematic data varies between clustersand eleven of our sources do not have any kinematicmembership data Therefore in this section we firstconsider other membership indicators by examining the

number of expected field stars within the SSG regionin each cluster CMD (Table 2) We will then includekinematic membership information for the individualSSGRS stars to provide an estimate of the probabil-ity that each SSG or RS star is a field star (Equation 2and Pfield in Table 4)First active galactic nuclei (AGN) are a well known

contaminant in X-ray surveys of star clusters How-ever to our knowledge none of these SSGs are notedas extended sources in the literature Also AGN areknown to show non-periodic stochastic flux variationson timescales of months to years (see eg Simm et al2015 and references therein) which is inconsistent withthe photometric variability seen for the SSGs in our sam-ple Therefore here we investigate stellar contaminantsSpecifically we investigate the probability that we wouldobserve any galactic field stars field-star X-ray sources orshort-period stellar photometric variables respectivelyin the SSG region of the CMD in each clusterWe utilize star fields from the Besancon model of the

Milky Way (Robin et al 2003) within the maximum sur-vey radius for each cluster (see Table 1) that containsSSGs in Table 4 We use the same color and magnitudecombination for each cluster as shown in Figure 1 respec-tively and identify an approximate region around the ob-served SSGs that extends from the bluest portion of theregion shown in Figure 1 to the color of the reddest SSGin the cluster plus 005 and the magnitude of the faintestSSG in the cluster plus 05 (We also perform a similaranalysis for the RS stars for Equation 2 below limitingthe region to the color of the reddest RS in the clusterplus 005 and to the magnitude of the brightest RS in acluster minus 05) These offsets in color and magnitudeare somewhat arbitrary but do not significantly affectthe results presented here For each cluster region wethen count the number of field stars expected to fall inthe SSG region of the appropriate CMD and give thisnumber as Ne in Table 2 About 93 of these expectedfield stars are dwarfs and 94 are spectral types G Kor MWe do not account for kinematic information in this

analysis in Table 2 (which is partly responsible for thelarge numbers of stars predicted for the open clusters)whereas the SSG and RS stars are mostly drawn fromsamples of known kinematic cluster members The ex-pected numbers of Galactic field stars given in Table 2are upper limitsWe estimate the number of short-period binaries ex-

pected to be in the SSG region for each cluster NeV

in Table 2 following the results from Raghavan et al(2010) More specifically we assume the field binary-star fraction is a function of spectral type from theirFigure 12 and assume all binaries follow the samelog-normal period distribution as the solar-type stars(peaked at a mean value of logP = 503 and withσlogP = 228 defining P in days) This orbital perioddistribution should serve adequately for these foregroundGKM dwarfs The log-normal period distribution pre-dicts about 4 of binaries should have orbital periodslt15 days We then use this percentage and the appro-priate binary fractions to estimate the numbers of ex-pected binaries with periods lt15 days in the SSG regionof the CMD for each respective cluster and give thesenumbers NeV in Table 2 Note that this is an overesti-


TABLE 2Photometric Field Contaminant Probabilities

Cluster No NoV NoX Ne NeV NeX P PV PX

[] [] []

Open Clusters

NGC 188 3 2 1 12 0251 0099 999 27 94NGC 2158 5 4 0 90 1879 0743 1000 122 middot middot middot

NGC 2682 2 1 2 50 1044 0396 1000 648 61NGC 6791 5 4 4 253 5281 2089 1000 772 159NGC 6819 1 0 1 331 6910 2419 1000 middot middot middot 911Total(Open Clusters) 16 11 8 736 15364 5747 1000 898 222

Globular Clusters

NGC 104 7 5 6 0 0000 0000 00 00 00NGC 5139 19 1 7 28 0585 0231 970 443 00NGC 6121 2 1 2 0 0000 0000 00 00 00NGC 6366 1 0 1 20 0342 0017 1000 middot middot middot 16NGC 6397 3 3 3 1 0021 0008 80 00 00NGC 6752 1 1 0 30 0626 0182 1000 465 middot middot middot

NGC 6809 1 1 0 6 0125 0041 998 118 middot middot middot

NGC 6838 2 0 2 1 0021 0008 264 middot middot middot 00Total(Globular Clusters) 36 12 21 86 1720 0487 1000 00 00

Note The expected numbers of sources do not account for kinematic information

mate of the number of binaries that would be detectedas we have not accounted for the expected random incli-nations the potential for low-mass companions below adetection limit observing cadence etcNext we estimate the number of stellar X-ray sources

expected for each cluster NeX in Table 2 Stellar X-raysources can be either single or binary stars Howeversingle X-ray emitters with luminosities of 1030minus1031 ergsminus1 have ages 10 Myr (see Preibisch amp Feigelson 2005and particularly their Figure 4) The Besancon modelpredicts no stars of this young age in the fields of any ofthe clusters in this study RS CVn and BY Dra activebinaries also have similar X-ray and optical properties asthe SSGs (Figure 6) Eker et al (2008) provide a catalogof known chromospherically active binaries From theircatalog we find that the BY Dra systems that have X-ray luminosities of order 1030 minus 1031 erg sminus1 all containGK dwarfs and have orbital periods between about 05and 4 days RS CVn systems within this X-ray luminos-ity range contain primarily GK subgiant and giant starsand have orbital periods between about 05 and 100 daysIf we consider only the portion of the Raghavan et al(2010) period distribution that would be occupied by BYDra and RS CVn binaries (and assume that this samedistribution holds for evolved stars) we find that about1 and 4 of (field) stars should have appropriate or-bital periods respectively Applying these percentagesand the cuts in spectral types and luminosity classes tothe Besancon model we find a total of about six field X-ray active binaries expected in the directions of all openclusters studies here and a total of less than one in thedirections of all of the globular clusters in our sampleNearly all of the expected active binaries are BY DrasystemsAlso in Table 2 we provide the observed numbers of

SSGs (No ie those with at least one color-magnitudecombination that resides in the SSG region and do nothave ldquoNMrdquo in Table 4) the number of observed SSGs

in binaries with orbital periods lt15 days (NoV and herewe assume that any short-period photometric variabilityis due to a binary companion of similar orbital period)and the number of observed SSG X-ray sources with lu-minosities of order 1030 minus 1031 erg sminus1 (NoX)We then calculate the cumulative Poisson probability

P of observing at least No SSGs (for those with at least1 SSG) with an expected number of Ne (found from theBesancon model)

P = 1minus eminus(Ne)Nominus1sum

x=0

(Ne)x

x (1)

We also calculate the probability PV that we would de-tect at least NoV short-period binaries in the SSG re-gion when NeV are expected (by replacing No and Ne

in Equation 1 by NoV and NeV respectively) Similarlywe calculate the probability PX that we would detect atleast NoX X-ray sources in the SSG region when NeX areexpected (by replacing No and Ne in Equation 1 by NoX

and NeX respectively) We provide all of these probabil-ities in Table 2 including total probabilities consideringthe summed number of each type of star in the calcula-tions for the open and globular clusters respectivelyWithout any constraints on X-ray emission or bina-

rity (or kinematic membership) the number of expectedstars in the SSG region of the CMD Ne is similar to orgreater than the number of true SSGs observed No inmost clusters and therefore the probability of observingat least No stars in this region of the CMD in most clus-ters (P ) is high For some clusters and especially forthe globular clusters taking also the photometric andradial-velocity variability (ie binarity) into account isenough to provide a high confidence level that the SSGsare not all field stars For the entire globular clustersample we find a 0 probability (PV ) that we would ob-serve NoV = 12 or more stars when NeV sim 2 stars areexpected Therefore even without accounting for kine-

12 Geller et al

matic membership information it is exceedingly unlikelythat all variable SSGs in the globular clusters could befield starsThe X-ray sources provide an even more stringent con-

straint Each respective globular cluster is predicted tohave NeX lt 1 field X-ray source with an X-ray luminos-ity between 1030 and 1031 erg sminus1 in the SSG CMD re-gion (considering both single and binary X-ray sources)For every globular cluster respectively we find that itis extremely unlikely that we would observe NoX X-raysources in the SSG CMD region when NeX lt 1 are ex-pected Summing over all observed and expected sourcesfor the globular clusters (bottom line of Table 2) we pre-dict NeX lt 1 field X-ray source with the appropriateluminosity in our entire globular cluster sample Consid-ering that we observeNoX = 21 X-ray sources in the SSGCMD region in globular clusters we find a zero percentprobability that all the X-ray-detected SSGs in globularclusters are field starsFor the open clusters if we do not consider any kine-

matic membership information there is a non-zero prob-ability that all SSGs could be field stars However againthe open clusters have the most comprehensive proper-motion and radial-velocity membership informationIn addition to this analysis of the full SSG popula-

tion we also estimate a probability that each SSG or RSindividually could be a field star (Pfield in Table 4) asfollows

Pfield =(1minus PPM)times (1minus PRV)times

PV (NoV = 1)times PX (NoX = 1) if var or X-ray

P (No = 1) otherwise(2)

where PPM and PRV are the fractional kinematic mem-bership probabilities given in Table 4 We assume a 50membership probability for stars labeled ldquoMrdquo and a 0probability for stars labeled ldquoNMrdquo For stars without aPPM or PRV value in Table 4 we simply set the appropri-ate PPM or PRV equal to 0 (thereby excluding that mem-bership indicator from the calculation) PV (NoV = 1)PX (NoX = 1) and P (No = 1) are calculated followingEquation 1 by setting the observed number NoV NoX

or No equal to unity and using the expected numberNeV NeX or Ne given in Table 2 for the cluster Wederive expected numbers for the RS stars in a similarmanner as described above for the SSGs If a given SSGor RS star is either not a photometric variable or X-ray source (with LX lt 1032 erg sminus1) we simply set therespective probability PV (NoV = 1) or PX (NoX = 1)to unity (thereby excluding that membership indicatorfrom the calculation) If the star is neither an X-raysource with the appropriate luminosity or a photometricor radial-velocity variable we replace these probabilitieswith P (No = 1) Here we assume that these probabilitiesare independent and we give each membership indicatorequal weightThese Pfield values provide an estimate of the member-

ship status of each source individually From examina-tion of the distribution of these membership values weselect a fairly strict cutoff of Pfield = 10 above whichwe exclude the source from our member sample used forour following analyses We also exclude the few sourcesthat have Pfield lt 10 but are labeled as ldquoNMrdquo 77

-10

-05

00

05

10

15

log 1

0( N

A

rea

)

-10 -08 -06 -04 -02 00 02 04log10(rrc)

00

02

04

06

08

10

frac

tion

with

in lo

g 10(

rr c

)

Fig 4mdash Radial distribution of the SSGs from our sample (iethose that fall in the SSG region in at least one color-magnitudecombination) shown in a histogram of surface density (top) as wellas the cumulative distribution (bottom) In both panels the blacklines and symbols (both open and filled) show SSGs that have thehighest likelihood of cluster membership (see Section 5) and inthe top panel we show the additional SSG candidates with thegray lines In this figure we exclude SSGs detected at radii beyondthe minimum completeness radius of all surveys (of 33 core radiiin NGC 6121) In the bottom panel openglobular cluster sourcesare plotted with openfilled circles Also for comparison in the bot-tom panel with the solid line we show the cumulative distributionexpected for a population of sources that have a uniform surfacedensity as a function of radius as expected for a field sample

(4356) of the SSG and 88 (78) of the RS sources fromthe literature pass this criteria The remaining have lesscertain membership status though may still be clustermembersIn conclusion even without considering kinematic

membership information (as in Table 2) the probabil-ity is remote that all these candidate SSG X-ray sourcesand photometric variables could be field stars (especiallyin the globular clusters) We include kinematic informa-tion in Equation 2 The product of these Pfield valuesgiven in Table 4 provides a more informed estimate ofthe probability that all sources could be field stars andis vanishingly small (for all sources and for the SSG andRS stars separately) We conclude that it is exceedinglyunlikely that all of the SSG sources listed in Table 4 (X-ray-detected or otherwise) could be field stars and moveforward with a sample of highly likely cluster membersfor our subsequent analyses


42 Radial Distributions Another Indicator ofCluster Membership

In Figure 4 we plot the radial distribution of the SSGswithin our sample of the most likely cluster membersThe specific values plotted in this figure are given in Ta-ble 4 and are calculated using the cluster centers fromGoldsbury et al (2010) Because the surveys become in-creasingly incomplete at larger radii (see Section 3) welimit this analysis to only include SSGs found at cluster-centric radii within the minimum completeness radius ofall surveys in this study (of 33 core radii in NGC 6121)Moreover the survey of SSGs is essentially complete inradius within this limit in each cluster (though in someof the globular clusters the surveys were offset from thecluster center so portions of the clusters may still re-main unobserved in X-ray andor optical) 93 of theSSGs that have the highest likelihood of cluster member-ship are found within 33 core radii from their respectivecluster centers and about 60 are found within 1 coreradiusFor comparison in the bottom panel of Figure 4 we

show the cumulative distribution expected for a popu-lation of sources distributed uniformly in surface den-sity as would be expected for a field population AKolmogorov-Smirnov (K-S) test shows that we can re-ject the hypothesis that these SSGs were drawn froma field population at very high confidence with a K-Sstatistic of 17times10minus6 (We also performed the same testbut excluding the ω Cen sample and reach the sameconclusion) Thus this comparison provides further ev-idence that these sources are indeed cluster membersFurthermore the full sample of SSGs within the 33 coreradii completeness limit (and without taking any selec-tion on membership likelihood) is equally unlikely to bedrawn from the field (with a K-S statistic of 17times10minus8)The SSGs are centrally concentrated with respect to theircluster centers as would be expected for any stellar pop-ulation in a star cluster

5 AGGREGATE EMPIRICAL CHARACTERISTICSOF SSGS

We list the general properties for this population ofSSGs at the beginning of Section 2 Here we discuss infurther detail the SSG binary properties and period dis-tribution (Figure 5) a comparison of X-ray and opticalflux (Figure 6) and the frequency of SSGs as a functionof host cluster mass (Figure 7) We will focus here onthe most secure cluster members of this SSGs sample asdiscussed in Section 4

51 Binary Orbital Parameters and PhotometricPeriods

The open clusters NGC 188 NGC 2682 NGC 6791and NGC 6819 are the only clusters with comprehen-sive and complete multi-epoch radial-velocity measure-ments capable of detecting binary companions and de-riving orbital solutions within the relevant regimes inmagnitude and orbital period (Geller et al 2008 20092015 Hole et al 2009 Milliman et al 2014 2016) Eightof the eleven SSGs in these clusters (73) are detectedas binary stars and six have secure kinematic orbital so-lutions (excluding 3259 in NGC 188 whose orbital pe-riod is uncertain) We provide the orbital parameters

0

1

2

3

4

5

N

0 5 10 15 20period [days]

00

02

04

06

08

10

frac

tion

Fig 5mdash Distribution of the photometric and radial-velocityperiods (in days) of the SSGs from our sample (ie those that fallin the SSG region in at least one color-magnitude combination)shown in histogram (top) and cumulative (bottom) distributionforms In both panels the black lines and symbols (both open andfilled) show SSGs that have the highest likelihood of cluster mem-bership (see Section 5 and at any cluster-centric radius) and inthe top panel we show the additional SSG candidates with the graylines Openglobular cluster sources are plotted with openfilledsymbols in the bottom panel Here triangles show periods fromradial-velocity binary orbital solutions and circles show periodsfrom photometry For sources with both photometric and radial-velocity periods we plot both

for the six SSGs and one RS (NGC 188 ID 3118) withsecure orbital solutions in Table 3 All have orbital peri-ods of less than 20 days with a mean SSG period ofabout 10 days (and a standard deviation of about 5days) The longest-period binary with an orbital solu-tion NGC 2682 ID 1063 is also the only eccentric binary(with e = 0206plusmn0014) Mathieu et al (2003) note thatthis eccentricity is typical for normal main-sequence andsubgiant binaries in M67 with periods of sim20 days butargues against a phase of mass transfer or a large evolvedprimary as these would both tend to circularize the or-bit (in the absence of any sufficiently recent dynamicalencounter or a tertiary companion that could have in-creased the eccentricity)In Figure 5 we plot the distribution of measured pho-

tometric andor radial-velocity periods for the SSGs withthe highest likelihood of cluster membership We do notattempt to correct for incompleteness and therefore thisdistribution (at least for the photometric periods) maybe biased toward shorter periods We simply note here

14 Geller et al

TABLE 3Orbital Elements for Sub-subgiant and Red Straggler Radial-Velocity Binaries

Cluster ID PerRV Orbital γ K e ω T0 a sin i f(m)dagger q σ Nobs

(days) Cycles km sminus1 km sminus1 deg (HJDminus2400000 d) 106 km M⊙ km sminus1

NGC 188 4289 114877 1162 -4262 408 0012 200 507445 644 807times10minus2middot middot middot 111 25

plusmn00009 middot middot middot plusmn023 plusmn03 plusmn0010 plusmn50 plusmn17 plusmn005 plusmn20times10minus3 middot middot middot middot middot middot middot middot middot

NGC 188 3118A 119022 1338 -4323 436 0011 20 509003 714 0656 0795 168 44plusmn00004 middot middot middot plusmn020 plusmn03 plusmn0006 plusmn30 plusmn10 plusmn006 plusmn0012 plusmn0009

3118B middot middot middot middot middot middot middot middot middot 549 middot middot middot middot middot middot middot middot middot 898 0521 middot middot middot 198 36middot middot middot middot middot middot middot middot middot plusmn04 middot middot middot middot middot middot middot middot middot plusmn007 plusmn0009 middot middot middot middot middot middot middot middot middot

NGC 2682 1063 18396 middot middot middot 3430 200 0206 95 4748219 00330 143times10minus2middot middot middot 099 28


NGC 2682 1113A 2823094 middot middot middot 334 606 0 middot middot middot 48916368 00157 0544 0703 315 18plusmn0000014 middot middot middot plusmn04 plusmn09 middot middot middot middot middot middot plusmn0004 plusmn00003 plusmn0012 plusmn0012 middot middot middot middot middot middot






NGC 6791 15561 77812 489 -454 271 0015 0 569898 289 16times10minus2 middot middot middot 135 16plusmn00012 middot middot middot plusmn04 plusmn06 plusmn0019 plusmn90 plusmn20 plusmn006 plusmn10times10minus3

middot middot middot middot middot middot middot middot middot

We provide the orbital period (PerRV) the center-of-mass radial velocity (γ) the radial-velocity amplitude (K) the eccentricity (e) the longitude of periastron

(ω) a Julian Date of periastron passage (T0) the projected semi-major axis (a sin i) the mass functiondagger (f(m)) the mass ratio (q if available) the rms residual

velocity from the orbital solution (ω) and the number of RV measurements (N) Orbital parameters for NGC 188 come from Geller et al (2009) NGC 2682 from

Mathieu et al 2003 and NGC 6791 from Milliman et al 2016 All spectroscopic binaries included here are single-lined except for NGC 2682 ID 1113 which is double

lined we provide two sets of parameters for 1113 on subsequent rows for the primary and secondary stars respectively

daggerFor the two SB2s NGC 2682 ID 1113 (SSG) and NGC188 ID 3118 (RS) we provide M1sin3(i) and M2sin3(i) for the primary (A) and secondary (B) respectively

in place of the mass function

that about 90 (1921) of the SSGs with measured pho-tometric andor radial-velocity periods show periodicityat 15 daysFinally the SB2 NGC 2682 S1113 is the only SSG

that has a dynamically measured mass ratio Its mass ra-tio of 07 and the spectroscopic temperature and surfacegravity measurements of Mathieu et al (2003) point toa main-sequence companion to the SSG However againMathieu et al (2003) caution that they could not find aself-consistent solution for S1113 that accounts for all ofthe observations The RS star NGC 188 3118 is also anSB2 and has a mass ratio of 08 If 3118 has a mass sim-ilar to normal giants in the cluster (114 M⊙ ) then thesecondary would have a mass of 09 M⊙ (Geller et al2009) and would most likely be a main-sequence starnear the turnoff (though further analysis is desirable)

52 X-ray Emission

Next we turn to the X-ray and optical properties of theSSG and RS stars compared in Figure 6 58 (2543)of the SSGs are observed to be X-ray sources which islikely a lower limit There is no obvious preference inCMD locations for SSGs detected in X-ray (see Figure 1where filled symbols mark X-ray sources)The majority of the cluster X-ray data originate from

the Chandra X-ray Observatory because Chandrarsquos highspatial resolution enables the most accurate associationsbetween X-ray and optical sources particularly in densecluster environments The X-ray luminosity and obser-vation band from the literature are quoted in Table 4but for the purpose of comparing clusters we convertthe quantities to a common unabsorbed 25ndash5 keV X-ray band X-ray fluxes are converted with the Portable

Interactive Multi-Mission Simulator (PIMMS)4 assum-ing a 1 keV thermal bremsstrahlung spectral model(Edmonds et al 2003 Heinke et al 2003 Haggard et al2009 Cool et al 2013) and the nH given in Table 1We plot flux ratios for sources with measurements in

the V and R Johnson-Cousins filters where we have thelargest number of individual SSG measurements We ac-count for reddening using the standard relation with the(m minus M)V and E(B minus V ) values listed in Table 1 andARAV = 0749 (from Cardelli et al 1989 for redden-ing by the diffuse interstellar medium) We convert thereddening-corrected magnitudes into flux densities withthe zero point fluxes from Bessell (1979) The conversionto flux also requires the flux density to be multiplied bythe bandwidth of the filter for which we use 415 A in Rand 360 A in V In Figure 6 we combine the X-ray and optical infor-

mation for a subset of the SSGs (colored symbols) andseveral comparison samples (grey symbols) in plots of05ndash25 keV X-ray luminosity vs X-ray-to-optical fluxratio The comparison sources are members of the sameclusters as our SSG sample (specifically from 47 Tuc M4M55 and M67 in the V -band and M55 M67 and ω Cenin the R-band see references in Section 2) and containa variety of X-ray-emitting objects including quiescentlow mass X-ray binaries (qLMXB) cataclysmic variables(CV) active binaries that are found at or fainter than theMS turnoff (ldquoAB (MS)rdquo eg BY Dra stars) active bina-ries found beyond the MS turnoff (ldquoAB (evol)rdquo eg RSCVn stars) blue straggler stars (BSS) andWUMa starsWe note that the comparison samples are dominated byω Cen in R (Haggard et al 2013 Cool et al 2013) and

4 PIMMS is maintained by HEASARC and is available herehttpsheasarcgsfcnasagovdocssoftwaretoolspimmshtml


minus4minus3minus2minus10123

Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838

Fig 6mdash Comparison between X-ray luminosities (05 - 25 keV) and X-ray to optical flux ratios in the V and R optical filters Weplot the SSG and RS stars that pass our membership criteria in solid colored circles and hexagons respectively and those with uncertainmembership in open symbols for all clusters that have measurements in either V or R Each cluster sample is plotted with a differentcolor For comparison in the gray points we also plot other X-ray sources that have optical counterparts from the literature drawingfrom the same clusters as our SSG sample (specifically from 47 Tuc M4 M55 and M67 in the V -band and M55 M67 and ω Cen in theR-band) The types of sources and related plot symbols for these comparison samples are defined in the figure legend For clarity ldquoAB(evol)rdquo are active binaries which appear above the main-sequence turn-off (eg RS CVn stars) and ldquoAB (MS)rdquo are those which appearbelow the turnoff (eg BY Dra stars)

47 Tuc in V (Grindlay et al 2001 Edmonds et al 2003)The SSGs occupy the same region as active binaries on

these X-ray-to-optical diagrams Furthermore the com-parison sources that occupy the most similar locus arethe active binaries that contain evolved stars ie RSCVn binaries In both panels the SSGs follow a trackthat is separated from most of the BY Dra ABs CVsand qLMXBs in the comparison samples The RS starsalso fall in the same region as the SSGs This overlapwith the active binaries hints that heightened chromo-spheric activity may be responsible for the X-ray activityobserved for at least some and perhaps many SSG andRS stars The diminished overlap between the SSG andBY Dra stars (which are also chromospherically active)is due to the generally lower optical luminosities of mostBY Drarsquos in the comparison samples relative to the SSGsThe X-ray luminosities of the SSGs agree well with boththe BY Dra and RS CVn active binary samples

53 Frequency of Sub-subgiants Across Different StarClusters

Figure 7 shows the number (top) and specific frequency(bottom) of SSGs as a function of the cluster mass for allclusters with SSGs in our sample The black points inthis figure show only the SSGs that are the most proba-ble cluster members and are within our minimum radiuscompleteness limit and gray points show all SSGs ThePearsonrsquos correlation coefficient relating cluster massand number of SSGs for the full sample of highly likelycluster members is 0925 However this value is largebecause of ω Cen and 47 Tuc Excluding these two clus-ters (the only clusters in our sample with masses above10 6M⊙ ) results in a Pearsonrsquos correlation coefficient of-0278 Furthermore a χ2 test indicates that this limitedsample of clusters (still excluding ω Cen and 47 Tuc) isconsistent with a flat line at the mean value of 〈N〉=17

SSGs This result holds even if we include ω Cen and47 Tuc in this test (using 〈N〉=17) with a reduced χ2

value of χ2red=191 We conclude that for this sample

there is no significant correlation between the number ofSSGs and the cluster massThis lack of a correlation in NSSG vs Mcl results in a

clearly visible trend in the specific frequency panel thenumber of SSGs per unit mass increases toward lower-mass clusters A χ2 test shows that the full sample ofSSGs is inconsistent with a flat line at the mean value atvery high significance (P (χ2) lt 10minus7) The same resultholds true for the open and globular clusters individuallyThe open clusters have a significantly higher specific

frequency of SSGs than do the globular clusters If wetake the mean specific SSG frequency for the open clus-ters of sim 10minus3 M⊙

minus1 and extrapolate that to the glob-ular cluster regime this would predict about 100 SSGsin a globular cluster with a total mass of 105 M⊙ Themaximum number of SSGs known in any cluster is 19 inω Cen followed by eight in 47 Tuc (both of which havemasses ge 106 M⊙ ) A hint of this trend was also notedby Cool et al (2013) in comparison of their ω Cen SSGsample to those in M67 and NGC 6791 This is now veri-fied with the larger sample of SSGs studied here and wediscuss the implications of this result in Section 6

54 Summary of Demographics

We conclude that the SSG region has an elevated fre-quency of variable stars and X-ray sources relative tostars of similar luminosities in the normal regions ofa CMD The X-ray and optical luminosities of thesesources are most similar to those of active binaries andspecifically RS CVn stars Likewise at least 14 of theSSGs in our sample are Hα emitters (and not all havebeen observed for Hα emission) another common charac-teristic of RS CVn stars Importantly the open clusters

16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2

Fig 7mdash Number (top) and specific frequency (bottom numberof SSGs NSSG divided by the cluster mass Mcl) of SSGs as afunction of the cluster mass for all clusters with at least one SSGin our sample OpenGlobular clusters are plotted in openfilledsymbols Black points include only those SSGs with the highest-likelihood of cluster membership and within the same radial com-pleteness limit as shown in Figure 4 Gray points show all SSGsError bars show the standard Poisson uncertainties on NSSG (andwe truncate the lower error bars for cases with NSSG = 1)

NGC 188 NGC 2682 NGC 6819 and NGC 6791 havecomprehensive three-dimensional membership analysesthese long-term multi-epoch radial-velocity surveys re-veal a 73 (811) binary frequency Thus at least forthe open clusters we conclude that the SSGs have a highfrequency of binaries relative to the normal stars (whichfor these clusters typically have measured spectroscopicbinaries frequencies with the same completeness limitsin orbital periods closer to sim25) Taking the entiresample of SSG members we find 65 (2843) to be pho-tometric andor radial-velocity variables 21 of which areradial-velocity binaries Of perhaps equal importance tothe binary and X-ray properties is the trend of increas-ing specific frequency of SSGs toward lower mass clus-ters Apparently open clusters are more efficient perunit mass at producing SSGs than globular clusters

6 DISCUSSION AND CONCLUSIONS

This is the first paper in a series studying the origin ofthe sub-subgiant (SSG) stars found redward of the MSand fainter than the normal giant branch (see Figure 1)Here we identify from the literature a sample of 56 SSGs(plus one candidate SSG discussed in the literature butwithout reliable color information due to short timescalevariability) and 8 red stragglers (RSs) in 16 star clus-ters including both open and globular clusters (Table 4)We identify 43 SSGs from this sample with the highestlikelihood of cluster membership This sample has the

following important empirical characteristics

1 They occupy a unique location on a CMD redwardof the normal MS stars but fainter than the sub-giant branch where normal single-star evolutiondoes not predict stars

2 ge58 (2543) of the SSGs are observed to be X-raysources with typical luminosities of order 1030minus31

erg sminus1 consistent with active binaries

3 ge33 (1443) of the SSGs are known to exhibit Hαemission (including gt50 in the MUSE sample)

4 ge65 (2843) of the SSGs are known photometricandor radial-velocity variables with typical peri-ods of 15 days

5 ge75 (2128) of the variable SSGs are radial-velocity binaries

6 The specific frequency of SSGs increases towardlower-mass star clusters

All of the percentages given above are lower limits (morecomplete demographics require further observations)Most of these sources have kinematic membership data

indicative of cluster membership In Sections 4 and 5we also argue that the X-ray luminosities photometricvariability and their centrally concentrated radial distri-bution with respect to their cluster centers (Figure 4)provides additional strong evidence that these sourcesare indeed cluster membersThe large fraction of short-period variables (most ei-

ther inferred or confirmed to be due to binary compan-ions) and the similarities in the X-ray characteristics be-tween the SSGs and active binaries (Figure 6) suggestthat the SSGs form through binary-mediated channelsThe Kepler ldquoNo-Manrsquos-Landrdquo stars (Section 23) if con-firmed as SSGs also indicate that SSGs can form in iso-lation through binary evolution channels Furthermorethe X-ray and optical luminosities and the Hα emissionseen in many of these SSGs are similar to RS CVn bina-riesIt is well known that dynamics can both create and

destroy binaries We find that the specific frequency ofSSGs increases toward smaller cluster masses (Figure 7)where dynamical encounters are on average less frequentand less energetic Thus this trend may indicate that dy-namics inhibits the formation of SSGs perhaps by dis-rupting or modifying the binary progenitors to SSGsToward the highest mass clusters there is a hint at aflattening to this distribution in specific SSG frequencywhich may be indicative of an increased efficiency of dy-namical productionAn important next step is to search for additional

SSGs particularly in the globular clusters that may berevealed from comprehensive proper-motion (and radial-velocity) membership analyses Still a typical globularcluster would need sim100 SSGs to be consistent with themean specific frequency of SSGs in the open clusterswhich seems exceedingly unlikely for the globular clus-ters studied hereOf additional great interest are the wide variety of evo-

lutionary states of the companions known for these SSGsWithin our SSG sample there is one MSP (and possibly


a second) in NGC 6397 a candidate black hole compan-ion in NGC 6121 (M4) a massive compact companion(either neutron star or black hole) in NGC 6652 a MScompanion near the turnoff in NGC 2682 (M67) andthree SSGs in W UMa binaries (in NGC 188 ω Cen andNGC 6397) This is far from a complete list but in otherstellar systems that do not follow standard stellar evo-lution theory the companions provide significant insightinto the most active formation channels (For examplethe white dwarf companions to blue stragglers in NGC188 found by Geller amp Mathieu 2011 and Gosnell et al2015 point directly to a past stage of mass transfer astheir formation mechanism) The compact object com-panions to these SSGs are particularly intriguing Toobtain an orbital period on the order of 10 days in a bi-nary containing a neutron star or black hole presumablyeither the system formed early and went through a previ-ous stage of mass transfer or common envelope to shrinkthe orbit or the system was formed more recently by adynamical encounter perhaps involving a tidal-capturescenario Regardless of the specific evolutionary histo-ries the diversity of companions amongst the SSGs mayoffer important guidance in developing theories of SSGformation which we discuss in subsequent papersFurther observations aimed at determining the com-

panion stars to these SSGs (and RSs) would be very valu-able Also recall that not all of these sources have therelevant observations to determine an X-ray luminosityor variability period Therefore it is possible that evenmore of these SSGs are indeed X-ray sources andor vari-ables with similar luminosities and periods respectivelyto those shown in Figures 5 and 6 Follow-up observa-tions of these sources are desirable Additionally thenearly 10000 stars in the ldquoNo-Manrsquos-Landrdquo from Ke-pler (Batalha et al 2013 Huber et al 2014) are a veryinteresting population for follow-up work These maybe field SSGs (that may have formed without the needfor cluster dynamics) and we will investigate these ldquoNo-Manrsquos-Landrdquo stars in more detail in a future paperThough we focus in the majority of this paper on the

SSG stars we return here briefly to comment on the RSstars There are eight total RS stars in this sample (inNGC 188 NGC 6791 NGC 7142 ω Cen and NGC 6752)seven of which would pass our criteria for high likelihoodof cluster membership (and one of these is outside ofthe radial completeness that we have set above) Thusthe number of sources is too small to perform any rigor-ous statistical tests Of these seven RS stars that wouldpass our membership criteria four are X-ray sourcesand three are photometric or radial-velocity variablesThus apart from their somewhat different location onthe CMD the RS stars appear to have similar character-istics as the SSGs As we will discuss in future papers

certain formation channels predict that the SSG and RSstars are related through evolution where one may bethe progenitor of the otherPrevious authors have suggested various dynamical

and interacting-binaries scenarios to explain the SSGsthough none present detailed models of SSG formation(or evolution) For instance Albrow et al (2001) positwith regards to the sources in 47 Tuc and M67 (theonly SSGs known at the time) that ldquoa plausible ex-planation for these stars is a deflated radius from sub-giant or giant origins as the result of mass transfer ini-tiated by Roche lobe contact by the evolved star forwhich the secondary has a lower massrdquo where PC-V11has a WD companion and the others have MS com-panions They liken these stars to BV Centauri whichGilliland (1982) suggests is a long-period CV variablewith a subgiant donor and an expected lifetime thatmay exceed 109 years Hurley et al (2005) briefly discussthe formation of a SSG star in their N -body model ofM67 through a common-envelope merger event (createdwithin the parameterized Binary Stellar Evolution codeBSE Hurley et al 2002) Empirically there is evidencein some systems for past and possibly ongoing masstransfer (eg the MSP system in NGC 6397) We alsonote that some of the main-sequence ndash main-sequencecollision models of Sills et al (2005) evolve through theSSG regionWe perform an in-depth investigation of specific SSG

formation routes in Leiner et al (in prep) In this andsubsequent papers we will investigate two general modesof SSG formation (i) through isolated binary evolutionand (ii) through dynamical processes (also likely involv-ing binaries) These papers will provide the first detailedmodels of SSG formation and will investigate the relativeformation frequencies that each model predicts The ob-servations and demographic information provided in thispaper solidify SSGs as a new class of stars that defiesstandard single-star evolution theory These stars mayprove to be very important test cases for both binaryevolution and star cluster dynamics modeling

AMG acknowledges support from HST grant AR-13910 and a National Science Foundation Astronomy andAstrophysics Postdoctoral Fellowship Award No AST-1302765 EML was supported by a Grant-In-Aid ofResearch from Sigma Xi the Scientific Research Soci-ety AB acknowledges support from HST grant AR-12845 SK received funding through BMBF Verbund-forschung (project MUSE-AO grant 05A14MGA) ASis supported by the Natural Sciences and EngineeringResearch Council of Canada

REFERENCES

Albrow M D Gilliland R L Brown T M Edmonds P DGuhathakurta P amp Sarajedini A 2001 ApJ 559 1060

Auriere M Koch-Miramond L amp Ortolani S 1989 AampA 214113

Bahramian A Heinke C O Sivakoff G R amp Gladstone J C2013 ApJ 766 136

Bassa C et al 2004 ApJ 609 755Bassa C G Pooley D Verbunt F Homer L Anderson

S F amp Lewin W H G 2008 AampA 488 921Batalha N M et al 2013 ApJS 204 24Bellini A et al 2014 ApJ 797 115

Bellini A Bedin L R Piotto G Milone A P Marino A Famp Villanova S 2010 AJ 140 631

Bellini A et al 2009 AampA 493 959Belloni T Verbunt F amp Mathieu R D 1998 AampA 339 431Bessell M S 1979 PASP 91 589Bogdanov S van den Berg M Heinke C O Cohn H N

Lugger P M amp Grindlay J E 2010 ApJ 709 241Bressan A Marigo P Girardi L Salasnich B Dal Cero C

Rubele S amp Nanni A 2012 MNRAS 427 127Bruntt H Grundahl F Tingley B Frandsen S Stetson

P B amp Thomsen B 2003 AampA 410 323

18 Geller et al

Cardelli J A Clayton G C amp Mathis J S 1989 ApJ 345245

Carney B W Lee J-W amp Dodson B 2005 AJ 129 656Carraro G Girardi L amp Marigo P 2002 MNRAS 332 705Chumak Y O Platais I McLaughlin D E Rastorguev A S

amp Chumak O V 2010 MNRAS 402 1841Cohn H N et al 2010 ApJ 722 20Cool A M Haggard D Arias T Brochmann M Dorfman

J Gafford A White V amp Anderson J 2013 ApJ 763 126Coomber G Heinke C O Cohn H N Lugger P M amp

Grindlay J E 2011 ApJ 735 95DrsquoAmico N Possenti A Manchester R N Sarkissian J

Lyne A G amp Camilo F 2001 ApJ 561 L89de Marchi F et al 2007 AampA 471 515Deutsch E W Margon B amp Anderson S F 2000 ApJ 530

L21Di Cecco A et al 2015 AJ 150 51Dias W S Monteiro H Caetano T C Lepine J R D

Assafin M amp Oliveira A F 2014 AampA 564 A79Dickey J M amp Lockman F J 1990 ARAampA 28 215Edmonds P D Gilliland R L Heinke C O amp Grindlay J E

2003 ApJ 596 1177Eggen O J 1983 AJ 88 813Eggen O J amp Iben I Jr 1988 AJ 96 635Eggen O J amp Iben I Jr 1989 AJ 97 431Eker Z et al 2008 MNRAS 389 1722Elsner R F et al 2008 ApJ 687 1019Engel M C Heinke C O Sivakoff G R Elshamouty K G

amp Edmonds P D 2012 ApJ 747 119Ferraro F R Sabbi E Gratton R Possenti A DrsquoAmico N

Bragaglia A amp Camilo F 2003 ApJ 584 L13Ferraro F R Sollima A Pancino E Bellazzini M Straniero

O Origlia L amp Cool A M 2004 ApJ 603 L81Geller A M Latham D W amp Mathieu R D 2015 AJ 150

97Geller A M amp Mathieu R D 2011 Nature 478 356Geller A M Mathieu R D Harris H C amp McClure R D

2008 AJ 135 2264Geller A M Mathieu R D Harris H C amp McClure R D

2009 AJ 137 3743Gilliland R L 1982 ApJ 263 302Girard T M Grundy W M Lopez C E amp van Altena

W F 1989 AJ 98 227Goldsbury R Richer H B Anderson J Dotter A

Sarajedini A amp Woodley K 2010 AJ 140 1830Gondoin P 2005 AampA 438 291Gosnell N M Mathieu R D Geller A M Sills A Leigh N

amp Knigge C 2015 ApJ 814 163Gosnell N M Pooley D Geller A M Kalirai J Mathieu

R D Frinchaboy P amp Ramirez-Ruiz E 2012 ApJ 745 57Grindlay J E Heinke C Edmonds P D amp Murray S S

2001 Science 292 2290Haggard D Cool A M amp Davies M B 2009 ApJ 697 224Haggard D Cool A M Heinke C O van der Marel R Cohn

H N Lugger P M amp Anderson J 2013 ApJ 773 L31Harris W E 1996 AJ 112 1487Harris W E 2010 ArXiv e-printsHeinke C O Edmonds P D amp Grindlay J E 2001 ApJ 562

363Heinke C O Grindlay J E Edmonds P D Lloyd D A

Murray S S Cohn H N amp Lugger P M 2003 ApJ 598516

Hole K T Geller A M Mathieu R D Platais I MeibomS amp Latham D W 2009 AJ 138 159

Holtzman J A Burrows C J Casertano S Hester J JTrauger J T Watson A M amp Worthey G 1995 PASP107 1065

Huang R H H Becker W Edmonds P D Elsner R FHeinke C O amp Hsieh B C 2010 AampA 513 A16

Huber D et al 2014 ApJS 211 2Hurley J R Pols O R Aarseth S J amp Tout C A 2005

MNRAS 363 293Hurley J R Tout C A amp Pols O R 2002 MNRAS 329 897Kalberla P M W Burton W B Hartmann D Arnal E M

Bajaja E Morras R amp Poppel W G L 2005 AampA 440775

Kalirai J S et al 2001 AJ 122 266Kaluzny J 2003 Acta Astron 53 51Kaluzny J amp Thompson I B 2009 Acta Astron 59 273Kaluzny J Thompson I B amp Krzeminski W 1997 AJ 113

2219Kaluzny J Thompson I B Krzeminski W amp

Schwarzenberg-Czerny A 2006 MNRAS 365 548Kaluzny J Thompson I B Krzeminski W amp Zloczewski K

2010 Acta Astron 60 245Kaluzny J Thompson I B Rozyczka M amp Krzeminski W

2013 Acta Astron 63 181

Kamann S et al 2016 AampA 588 A149Kamann S Wisotzki L amp Roth M M 2013 AampA 549 A71Kharchenko N Andruk V amp Schilbach E 1997

Astronomische Nachrichten 318 253Knigge C Zurek D R Shara M M amp Long K S 2002 ApJ

579 752Landsman W Aparicio J Bergeron P Di Stefano R amp

Stecher T P 1997 ApJ 481 L93Lane R R Kiss L L Lewis G F Ibata R A Siebert A

Bedding T R Szekely P amp Szabo G M 2011 AampA 530A31

Lee Y-W Joo J-M Sohn Y-J Rey S-C Lee H-C ampWalker A R 1999 Nature 402 55

Lu T-N Kong A K H Bassa C Verbunt F LewinW H G Anderson S F amp Pooley D 2009 ApJ 705 175

Marın-Franch A et al 2009 ApJ 694 1498Mathieu R D 2000 in ASPC Vol 198 rdquoStellar Clusters and

Associations Convection Rotation and Dynamosrdquo edR Pallavicini G Micela amp S Sciortino 517

Mathieu R D van den Berg M Torres G Latham DVerbunt F amp Stassun K 2003 AJ 125 246

Meibom S et al 2009 AJ 137 5086Milliman K E Leiner E Mathieu R D Tofflemire B M amp

Platais I 2016 AJ 151 152Milliman K E Mathieu R D Geller A M Gosnell N M

Meibom S amp Platais I 2014 AJ 148 38Mochejska B J Stanek K Z Sasselov D D amp Szentgyorgyi

A H 2002 AJ 123 3460Mochejska B J et al 2006 AJ 131 1090Mochejska B J et al 2005 AJ 129 2856Mochejska B J Stanek K Z Sasselov D D Szentgyorgyi

A H Westover M amp Winn J N 2004 AJ 128 312Montgomery K A Marschall L A amp Janes K A 1993 AJ

106 181Nascimbeni V et al 2014 MNRAS 442 2381Pancino E Ferraro F R Bellazzini M Piotto G amp Zoccali

M 2000 ApJ 534 L83Pasquini L amp Belloni T 1998 AampA 336 902Platais I Cudworth K M Kozhurina-Platais V et al 2011

ApJ 733 L1Platais I Gosnell N M Meibom S Kozhurina-Platais V

Bellini A Veillet C amp Burkhead M S 2013 AJ 146 43Platais I Kozhurina-Platais V Mathieu R D Girard T M

amp van Altena W F 2003 AJ 126 2922Pooley D et al 2003 ApJ 591 L131Preibisch T amp Feigelson E D 2005 ApJS 160 390Raghavan D et al 2010 ApJS 190 1Robertson J A amp Eggleton P P 1977 MNRAS 179 359Robin A C Reyle C Derriere S amp Picaud S 2003 AampA

409 523Rozyczka M Kaluzny J Pietrukowicz P Pych W Catelan

M amp Contreras C 2012 AampA 537 A89Sanders W L 1977 AampAS 27 89Sandquist E L Serio A W amp Shetrone M 2011 AJ 142 194Sandquist E L Shetrone M Serio A W amp Orosz J 2013

AJ 146 40Sarajedini A von Hippel T Kozhurina-Platais V amp

Demarque P 1999 AJ 118 2894Sills A Adams T amp Davies M B 2005 MNRAS 358 716Simm T et al 2015 AampA 584 A106Sirianni M et al 2005 PASP 117 1049Skrutskie M F et al 2006 AJ 131 1163Stetson P B et al 2014 PASP 126 521Stetson P B Bruntt H amp Grundahl F 2003 PASP 115 413Stetson P B McClure R D amp VandenBerg D A 2004

PASP 116 1012Straizys V Maskoliunas M Boyle R P Zdanavicius K

Zdanavicius J Laugalys V amp Kazlauskas A 2014 MNRAS437 1628

Thompson I B Kaluzny J Rucinski S M Krzeminski WPych W Dotter A amp Burley G S 2010 AJ 139 329

Tofflemire B M Gosnell N M Mathieu R D amp Platais I2014 AJ 148 61

van den Berg M Stassun K G Verbunt F amp Mathieu R D2002 AampA 382 888

van den Berg M Tagliaferri G Belloni T amp Verbunt F 2004AampA 418 509

van den Berg M Verbunt F amp Mathieu R D 1999 AampA347 866

van den Berg M Verbunt F Tagliaferri G Belloni T BedinL R amp Platais I 2013 ApJ 770 98

Villanova S et al 2007 ApJ 663 296Watkins L L van der Marel R P Bellini A amp Anderson J

2015 ApJ 803 29Yanny B et al 2009 AJ 137 4377Zhang X B Deng L Tian B amp Zhou X 2002 AJ 123 1548


Zloczewski K Kaluzny J Rozyczka M Krzeminski W ampMazur B 2012 Acta Astron 62 357

Zloczewski K Kaluzny J amp Thompson I B 2011 MNRAS414 3711

20

Geller

etal

TABLE 4Sub-subgiant Summary Table

Cluster Other ID RA Dec PPM PRV Pfield rrc U B V R I J H K LX X-band PerRV Perphot SSGRSN(J20000) (J20000) [] [] [] [erg sminus1] [keV] [d] [d]

Open Clusters

NGC 188 middot middot middot 1141 00444450 +8532164 88 98 0002 434 middot middot middot 1528 1411 middot middot middot middot middot middot 121 1154 1142 102times1031 01-24 middot middot middot middot middot middot 010NGC 188 middot middot middot 3118 00322120 +8518389 34 88 0164 459 middot middot middot 1578 1465 middot middot middot middot middot middot 1251 1195 1173 middot middot middot middot middot middot 119 middot middot middot 010NGC 188 middot middot middot 3259 00345919 +8518428 71 M 3213 382 middot middot middot 1710 1612 middot middot middot middot middot middot 1429 1399 1373 middot middot middot middot middot middot 102 middot middot middot 100NGC 188 middot middot middot 4289 00424039 +8516495 98 98 0001 146 1694 1624 1530 1477 1420 1354 1300 1288 470times1030 05-20 1149 middot middot middot 1000NGC 188 middot middot middot 4989 00482265 +8515553 95 middot middot middot 1108 045 1750 1709 1614 1557 1506 1444 1384 1383 middot middot middot middot middot middot middot middot middot 059 901NGC 2158 middot middot middot V44 0607058 +2408514 4480 middot middot middot 16944 093 middot middot middot middot middot middot 1741 1670 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 28 100NGC 2158 middot middot middot V48 0607062 +2402103 4169 middot middot middot 26264 124 middot middot middot middot middot middot 1711 1641 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 634 100NGC 2158 middot middot middot V49 0607102 +2410195 2651 middot middot middot 41514 145 middot middot middot middot middot middot 1717 1648 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 673 100NGC 2158 middot middot middot V90 0607362 +2405254 94 middot middot middot 5083 232 middot middot middot middot middot middot 1658 1592 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 1281 100NGC 2158 middot middot middot V95 0607188 +2401401 5370 middot middot middot 25417 174 middot middot middot middot middot middot 1670 1592 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot var 100NGC 2682 M67 S1063 08511337 +1151402 97 98 0013 087 1556 1484 1379 middot middot middot 1259 1166 1106 1096 130times1031 03-7 184 middot middot middot 600NGC 2682 M67 S1113 0851254 +1202565 98 98 0008 336 153 1478 1377 middot middot middot middot middot middot 1167 1112 1097 73times1030 01-24 282 middot middot middot 300NGC 6791 middot middot middot 83 19201061 +3751112 96 85 0600 300 middot middot middot middot middot middot 183 middot middot middot 168 1579 1504 1478 middot middot middot middot middot middot middot middot middot middot middot middot 100NGC 6791 middot middot middot 746 19202148 +3748216 99 95 0044 203 middot middot middot 1931 1796 middot middot middot 1656 1553 1481 1471 450times1030 03-7 11419 1383 300NGC 6791 middot middot middot 3626 19203888 +3749043 99 63 0323 122 middot middot middot 1912 1796 middot middot middot 1667 1580 1525 1506 480times1030 03-7 58251 637 300NGC 6791 middot middot middot 6371 19204788 +3746372 99 M 0083 033 middot middot middot 1842 1727 middot middot middot 1586 1489 1430 1404 770times1030 03-7 var 319 030NGC 6791 middot middot middot 7011 19204986 +3745507 99 M 0436 024 middot middot middot 1933 1827 middot middot middot 1715 1643 1587 1524 230times1030 03-7 var 409 300NGC 6791 middot middot middot 15561 19212522 +3745498 99 84 0139 197 middot middot middot 1899 1765 middot middot middot 1613 1520 1450 1449 127times1031 03-7 77815 764 120NGC 6819 middot middot middot 52004 19412589 +4012236 99 middot middot middot 0911 075 middot middot middot 1649 1565 middot middot middot middot middot middot 1414 1375 1376 930times1030 02-100 middot middot middot middot middot middot 100NGC 7142 middot middot middot V4 21445598 +6545500 91 middot middot middot 7435 044 middot middot middot 1732 1543 middot middot middot 1337 1197 1100 1077 middot middot middot middot middot middot middot middot middot var 030

Globular Clusters

NGC 104 47 Tuc PC1-V11 00240465 -7204553 middot middot middot middot middot middot lt0001 015 1783 middot middot middot 1737 middot middot middot 1641 middot middot middot middot middot middot middot middot middot 750times1030 05-25 middot middot middot 1108 201NGC 104 47 Tuc PC1-V41 00240876 -7205077 M middot middot middot lt0001 055 1854 middot middot middot 1761 middot middot middot 1689 middot middot middot middot middot middot middot middot middot 210times1030 05-25 middot middot middot 502 201NGC 104 47 Tuc PC1-V48 00241046 -7205000 M M lt0001 062 1854 middot middot middot 1737 middot middot middot 1635 middot middot middot middot middot middot middot middot middot 350times1030 05-25 var 671 300NGC 104 47 Tuc WF2-V31 00241516 -7204436 M M lt0001 119 1887 middot middot middot 1756 middot middot middot 1639 middot middot middot middot middot middot middot middot middot 730times1030 05-25 var 534 300NGC 104 47 Tuc WF2-V32 00241376 -7203346 M M lt0001 231 1818 middot middot middot 1712 middot middot middot 1610 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot var 92 120NGC 104 47 Tuc WF4-V18 00241201 -7205125 NM M lt0001 094 1811 middot middot middot 1726 middot middot middot 1637 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot var 59 300NGC 104 47 Tuc W4 00241320 -7204511 middot middot middot middot middot middot lt0001 092 1820 middot middot middot 1748 middot middot middot 1663 middot middot middot middot middot middot middot middot middot 64times1030 05-25 middot middot middot middot middot middot 300NGC 104 47 Tuc W38 00240464 -7204460 middot middot middot middot middot middot lt0001 022 1804 middot middot middot 1723 middot middot middot 1635 middot middot middot middot middot middot middot middot middot 290times1030 05-25 middot middot middot middot middot middot 300NGC 104 47 Tuc W43 00240420 -7204436 middot middot middot middot middot middot lt0001 031 1846 middot middot middot 1708 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 460times1030 05-25 middot middot middot middot middot middot 010NGC 5139 ω Cen 12 13265836 -4732467 99 M 0500 093 1919 1862 1764 1718 1671 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 1000NGC 5139 ω Cen 16 13263420 -4732148 93 M 1549 086 2002 1982 1876 1858 1809 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot var middot middot middot 802NGC 5139 ω Cen 22 13270424 -4731339 96 M 2000 084 1947 1983 1883 1852 1811 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 406NGC 5139 ω Cen 23 13260826 -4731267 84 M 3541 149 1953 1914 1859 1834 1785 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 025 505NGC 5139 ω Cen 28 13270684 -4730587 95 M 1107 083 1854 1908 1798 middot middot middot 1718 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot var middot middot middot 501NGC 5139 ω Cen 35 13270825 -4730143 96 M 2000 080 1836 1863 1763 1729 1681 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 802NGC 5139 ω Cen 36 13270583 -4730027 96 M 0885 071 1817 1823 1723 1685 1621 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot var middot middot middot 901NGC 5139 ω Cen 37 13263157 -4729580 100 M lt0001 061 1855 1742 1638 1592 1539 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 082NGC 5139 ω Cen 42 13262372 -4729201 99 M 0221 084 1935 1875 1778 1738 1684 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot var middot middot middot 1000NGC 5139 ω Cen 49 13263835 -4725412 91 M 1992 072 1927 1875 1773 1734 1668 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot var middot middot middot 1000NGC 5139 ω Cen 50 13270830 -4725413 94 M 1328 099 1947 1955 1855 1829 1772 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot var middot middot middot 802NGC 5139 ω Cen 58 13263068 -4724258 99 M 0221 108 1959 1906 1808 1758 1703 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot var middot middot middot 1000NGC 5139 ω Cen 68 13261657 -4722435 94 M 3000 167 1945 1944 1833 1753 1664 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 1000NGC 5139 ω Cen 13b 13265053 -4729181 middot middot middot M 10321 016 middot middot middot 1827 middot middot middot 1714 middot middot middot middot middot middot middot middot middot middot middot middot 490times1030 05-25 middot middot middot middot middot middot 100NGC 5139 ω Cen 22e 13265993 -4728097 94 middot middot middot 1239 047 1937 1797 1723 1642 1597 middot middot middot middot middot middot middot middot middot 800times1030 05-25 middot middot middot middot middot middot 361NGC 5139 ω Cen 24f 13263729 -4729429 95 middot middot middot 0041 041 1773 1697 1596 1541 1493 middot middot middot middot middot middot middot middot middot 160times1030 05-25 middot middot middot middot middot middot 091NGC 5139 ω Cen 32f 13270533 -4728088 43 middot middot middot 11766 065 1802 1827 1712 1672 1597 middot middot middot middot middot middot middot middot middot 400times1030 05-25 middot middot middot middot middot middot 442

Dem

ographics

ofSub-su

bgiantStars

21

TABLE 4 mdash Continued


NGC 5139 ω Cen 34b 13263744 -4730533 middot middot middot middot middot middot 20643 056 middot middot middot 1877 middot middot middot 1742 middot middot middot middot middot middot middot middot middot middot middot middot 280times1031 05-25 middot middot middot middot middot middot 100NGC 5139 ω Cen 41h 13264396 -4724426 middot middot middot middot middot middot 20643 086 middot middot middot 1917 middot middot middot 1804 middot middot middot middot middot middot middot middot middot middot middot middot 140times1030 05-25 middot middot middot middot middot middot 100NGC 5139 ω Cen 42c 13270965 -4727289 75 middot middot middot 5161 084 1903 1907 1835 1783 1746 middot middot middot middot middot middot middot middot middot 140times1030 05-25 middot middot middot middot middot middot 901NGC 5139 ω Cen 43c 13270691 -4730094 79 middot middot middot 4335 075 1865 1847 1754 1712 1646 middot middot middot middot middot middot middot middot middot 180times1030 05-25 middot middot middot middot middot middot 1000NGC 6121 M4 CX8 16233148 -2630578 M middot middot middot lt0001 102 1784 1767 167 1603 1539 1457 1399 1379 170times1030 05-25 middot middot middot 077 1000NGC 6121 M4 CX10 16233505 -2631197 M M lt0001 022 middot middot middot middot middot middot 1677 middot middot middot 1525 middot middot middot middot middot middot middot middot middot 170times1030 05-25 middot middot middot middot middot middot 100NGC 6218 M12 CX2b 16471840 -0156536 NM middot middot middot 3249 071 middot middot middot 1904 1832 1754 middot middot middot middot middot middot middot middot middot middot middot middot 110times1031 03-7 middot middot middot middot middot middot 300NGC 6366 middot middot middot CX5 17274414 -0504272 middot middot middot middot middot middot 1638 011 middot middot middot 2067 1901 1811 1705 middot middot middot middot middot middot middot middot middot 420times1030 05-25 middot middot middot middot middot middot 501NGC 6397 middot middot middot U12 17404462 -5340415 M M 0004 983 middot middot middot 1740 middot middot middot 1618 middot middot middot middot middot middot middot middot middot middot middot middot 220times1031 03-8 var 135 100NGC 6397 middot middot middot U18 17404261 -5340275 M M 0004 172 middot middot middot 1748 middot middot middot 1602 middot middot middot middot middot middot middot middot middot middot middot middot 320times1031 05-6 var 131 100NGC 6397 middot middot middot U92 17404392 -5340352 M M 0004 668 middot middot middot 1823 middot middot middot 1663 middot middot middot middot middot middot middot middot middot middot middot middot 220times1029 05-6 var 027 100NGC 6652 middot middot middot B 18354365 -3259268 middot middot middot middot middot middot middot middot middot 107 middot middot middot middot middot middot 20 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 220times1036 2-10 middot middot middot var 000NGC 6752 middot middot middot V19 19111304 -5955176 NM middot middot middot 46541 1742 middot middot middot 1739 1632 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 618 100NGC 6752 middot middot middot V20 19105050 -5957372 NM middot middot middot 86466 556 middot middot middot 1683 1609 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot gt 8 010NGC 6752 middot middot middot V22 19102967 -5956241 middot middot middot middot middot middot 46541 1468 middot middot middot 1921 1844 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 176 100NGC 6809 M55 CX7 19395119 -3059010 M NM 0238 057 middot middot middot 181 middot middot middot 1675 middot middot middot middot middot middot middot middot middot middot middot middot 570times1030 05-25 middot middot middot 559 100NGC 6809 M55 V64 19394720 -3057123 M middot middot middot 5886 074 middot middot middot 1777 1702 middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot middot 1295 100NGC 6838 M71 s02 19534262 +1845472 middot middot middot middot middot middot 0822 135 middot middot middot middot middot middot 1771 middot middot middot 1632 middot middot middot middot middot middot middot middot middot 150times1031 05-25 middot middot middot middot middot middot 100NGC 6838 M71 s19a 19534885 +1846340 middot middot middot middot middot middot 0822 060 1917 1873 1767 middot middot middot 1650 middot middot middot middot middot middot middot middot middot 891times1031 05-25 middot middot middot middot middot middot 600

Note References for values in this table are given in Section 2

2 Geller et al

glers occupyTo clarify this nomenclature we choose to use the term

ldquosub-subgiantrdquo (hereafter SSG) to refer to stars that arefainter than the normal subgiants and redder than thenormal MS stars shown as the dark-gray regions in Fig-ure 1 In addition to these SSG stars there are also starsthat are observed to be redder than the normal red gi-ants but brighter than the normal subgiants (found inthe light-gray regions of Figure 1) We choose to use theterm ldquored stragglerrdquo (hereafter RS) to refer to these starsand we encourage the community to adopt this divisionof the SSG and RS nomenclature moving forwardWe divide the SSG and RS stars in a given photome-

try band (and isochrone family) by the magnitude at thebase of the giant branch as shown in Figure 1 This def-inition depends on the position of the isochrone whichmight vary if defined by different photometric studiesor isochrone families We take cluster parameters fromthe literature (see Table 1) and use PARSEC isochrones(Bressan et al 2012) here For much of our sample thisdefinition is sufficient to place a star into either the SSGor RS category for all of the optical photometry we col-lected from the literature (and provide in Table 4) How-ever some stars reside in the SSG region in one color-magnitude combination and in the RS region for a differ-ent combination (see Figure 1 and Table 4) We discussthis phenomenon in some detail for specific sources be-low Briefly some stars move around dramatically rela-tive to the isochrone with different filter choices perhapsdue to spot activity and photometric variability (see alsoeg Milliman et al 2016) Indeed photometric variabil-ity appears to be a defining characteristic of the SSG andRS stars For the SSG analysis presented here we in-clude all stars that fall in the SSG region in at least oneoptical color-magnitude combination This is the mostinclusive definition of SSGs which may be importantgiven the photometric variability We will also accountfor potential field star contamination within this SSGsample (Section 4) prior to performing our demographicanalysesIn this paper we gather the observations from the SSG

and RS stars in each cluster in Section 2 and provide asummary of the observations of these stars in Table 4In Sections 3 and 4 we discuss possible observational bi-ases in our sample and the potential for field-star con-tamination respectively In Section 5 we investigate theaggregate characteristics of the sample We close with abrief discussion and conclusions in Section 6 Subsequentpapers in this series will investigate in detail a set oftheoretical formation channels for SSGs (some of whichpredict that SSG and RS stars may be related throughformation andor evolution) and evaluate the formationrates of each channel and their abilities to create SSGswith the demographics identified here

2 OBSERVED SUB-SUBGIANTS

In Table 4 we compile all of the SSG and RS starsthat have been identified in the literature in both openand globular clusters We provide the clusterrsquos NGCname another common name where appropriate theSSGRS ID (see the relevant paragraph below for thespecific references for IDs and other values) the RAand Dec proper-motion and radial-velocity membershipsstatuses (PPM and PRV where available) our estimate

of the probability that this star is a field star (Pfieldsee Section 4) the radial distance from the cluster cen-ter in units of core radii (rc) the available observedUBV RIJHK photometry2 an available X-ray luminos-ity (LX) and the band of the X-ray observation theradial-velocity orbital period (ldquoPerRVrdquo) and photometricperiod (ldquoPerphotrdquo) where available (we mark variableswhose periods are yet to be determined as ldquovarrdquo) and fi-nally the number of color-magnitude combinations givenin this table (from the literature) that place the givenstar in the SSG or RS regions or neither (ldquoSSGRSNrdquo)In Table 4 and throughout this paper we identify

sources with velocities that are gt 3σ from the clustermean (if that is the only available membership measure-ment) or with formal membership probabilities lt50 asnon-members We exclude definite non-members fromTable 4 In some cases authors identify stars as non-members at lt 3σ from the mean (eg at gt 2σ orgt 25σ) We choose to include such stars with ques-tionable membership within Table 4 and indicate thisuncertain membership with ldquordquo but we do not includethese in our subsequent analysis We discuss additionalmembership indicators in Section 4Observations of the SSG stars reveal an intriguing mix-

ture of properties Broadly SSGs

1 are redder than normal MS stars but fainter thannormal giants in an optical color-magnitude dia-gram

2 have X-ray luminosities LX of order 1030 minus1031 erg sminus1 in both open and globular clusters

3 are often Hα emitters (where measurements havebeen made)

4 exhibit photometric variability with periods 15days (where available) and

5 where possible are mostly identified as radial-velocity binaries

We return to these aggregate properties in Section 5First we briefly discuss the observations from each of theindividual star clusters listed in Table 4 CMDs for eachof these clusters showing the SSG and RS stars alongwith an isochrone (for reference) and the SSGRS regionsare plotted in Figure 1 and relevant cluster parametersare shown in Table 1

21 Open Cluster Observations

NGC 188 mdash This old (sim6-7 Gyr) open cluster has threeSSGs and two RSs (IDs from Geller et al 2008) The op-tical and IR photometry for these stars in Table 4 comefrom Stetson et al (2004) and 2MASS respectively Allof these stars have proper motions (Platais et al 2003)consistent with cluster membership and where measure-ments are possible these stars are also radial-velocitymembers (Geller et al 2008) All but one of these stars

2 All optical photometry are given in Johnson-Cousins filters(and have not been de-reddened) Where necessary HST mag-nitudes are converted to ground-based Johnson-Cousins followingHoltzman et al (1995) and Sirianni et al (2005) All JHK in-frared photometry come from 2MASS (Skrutskie et al 2006)


05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 188

00 05 10 15(V-R)0

0

1

2

3

4

5

MV

NGC 2158

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 2682

(M67)

10 15 20(V-I)0

2

3

4

5

6

7

MV

NGC 6791

05 10 15(B-V)0

0

1

2

3

4

5

MV

NGC 6819

05 10 15 20(B-V)0

1

2

3

4

5

6

MV

NGC 7142

05 10 15(V-I)0

1

2

3

4

5

6

MV

NGC 104

(47 Tuc)

05 10 15 20(B-R)0

0

1

2

3

4

5

MR

NGC 5139(ω Cen)

05 10 15(V-I)0

1

2

3

4

5

6

MV

NGC 6121

(M4)

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 6218

(M12)

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 6366

05 10 15 20(B-R)0

0

1

2

3

4

5

MR

NGC 6397

05 10 15(B-V)0

0

1

2

3

4

5

6

MV

NGC 6752

05 10 15(B-V)0

0

1

2

3

4

5

MV

NGC 6809

(M55)

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 6838

(M71)=SSG

=RS



=


=




4 Geller et al




Open Clusters


Globular Clusters


























6 Geller et al













8 Geller et al


[MH]

minus100

minus80

minus60

minus40

minus20

0

20

40

v rad[kms]

NGC104



[MH]

180

200

220

240

260

280

300

v rad[kms]

NGC5139

13b


[MH]

minus100

minus50

0

50

100

150

v rad[kms]

NGC6121

CX10CX8


[MH]

minus200

minus150

minus100

minus50

0

50

100

150

200v rad[kms]

NGC6397

U12U92U18U42












24 Summary





10 Geller et al


















[] [] []

Open Clusters



Globular Clusters











x=0

(Ne)x

x (1)





12 Geller et al











-10

-05

00

05

10

15

log 1

0( N

A

rea

)

-10 -08 -06 -04 -02 00 02 04log10(rrc)

00

02

04

06

08

10

frac

tion

with

in lo

g 10(

rr c

)












0

1

2

3

4

5

N


00

02

04

06

08

10

frac

tion




14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21






05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 188

00 05 10 15(V-R)0

0

1

2

3

4

5

MV

NGC 2158

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 2682

(M67)

10 15 20(V-I)0

2

3

4

5

6

7

MV

NGC 6791

05 10 15(B-V)0

0

1

2

3

4

5

MV

NGC 6819

05 10 15 20(B-V)0

1

2

3

4

5

6

MV

NGC 7142

05 10 15(V-I)0

1

2

3

4

5

6

MV

NGC 104

(47 Tuc)

05 10 15 20(B-R)0

0

1

2

3

4

5

MR

NGC 5139(ω Cen)

05 10 15(V-I)0

1

2

3

4

5

6

MV

NGC 6121

(M4)

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 6218

(M12)

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 6366

05 10 15 20(B-R)0

0

1

2

3

4

5

MR

NGC 6397

05 10 15(B-V)0

0

1

2

3

4

5

6

MV

NGC 6752

05 10 15(B-V)0

0

1

2

3

4

5

MV

NGC 6809

(M55)

05 10 15(B-V)0

1

2

3

4

5

6

MV

NGC 6838

(M71)=SSG

=RS



=


=




4 Geller et al




Open Clusters


Globular Clusters


























6 Geller et al













8 Geller et al


[MH]

minus100

minus80

minus60

minus40

minus20

0

20

40

v rad[kms]

NGC104



[MH]

180

200

220

240

260

280

300

v rad[kms]

NGC5139

13b


[MH]

minus100

minus50

0

50

100

150

v rad[kms]

NGC6121

CX10CX8


[MH]

minus200

minus150

minus100

minus50

0

50

100

150

200v rad[kms]

NGC6397

U12U92U18U42












24 Summary





10 Geller et al


















[] [] []

Open Clusters



Globular Clusters











x=0

(Ne)x

x (1)





12 Geller et al











-10

-05

00

05

10

15

log 1

0( N

A

rea

)

-10 -08 -06 -04 -02 00 02 04log10(rrc)

00

02

04

06

08

10

frac

tion

with

in lo

g 10(

rr c

)












0

1

2

3

4

5

N


00

02

04

06

08

10

frac

tion




14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21





4 Geller et al




Open Clusters


Globular Clusters


























6 Geller et al













8 Geller et al


[MH]

minus100

minus80

minus60

minus40

minus20

0

20

40

v rad[kms]

NGC104



[MH]

180

200

220

240

260

280

300

v rad[kms]

NGC5139

13b


[MH]

minus100

minus50

0

50

100

150

v rad[kms]

NGC6121

CX10CX8


[MH]

minus200

minus150

minus100

minus50

0

50

100

150

200v rad[kms]

NGC6397

U12U92U18U42












24 Summary





10 Geller et al


















[] [] []

Open Clusters



Globular Clusters











x=0

(Ne)x

x (1)





12 Geller et al











-10

-05

00

05

10

15

log 1

0( N

A

rea

)

-10 -08 -06 -04 -02 00 02 04log10(rrc)

00

02

04

06

08

10

frac

tion

with

in lo

g 10(

rr c

)












0

1

2

3

4

5

N


00

02

04

06

08

10

frac

tion




14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21













6 Geller et al













8 Geller et al


[MH]

minus100

minus80

minus60

minus40

minus20

0

20

40

v rad[kms]

NGC104



[MH]

180

200

220

240

260

280

300

v rad[kms]

NGC5139

13b


[MH]

minus100

minus50

0

50

100

150

v rad[kms]

NGC6121

CX10CX8


[MH]

minus200

minus150

minus100

minus50

0

50

100

150

200v rad[kms]

NGC6397

U12U92U18U42












24 Summary





10 Geller et al


















[] [] []

Open Clusters



Globular Clusters











x=0

(Ne)x

x (1)





12 Geller et al











-10

-05

00

05

10

15

log 1

0( N

A

rea

)

-10 -08 -06 -04 -02 00 02 04log10(rrc)

00

02

04

06

08

10

frac

tion

with

in lo

g 10(

rr c

)












0

1

2

3

4

5

N


00

02

04

06

08

10

frac

tion




14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21













8 Geller et al


[MH]

minus100

minus80

minus60

minus40

minus20

0

20

40

v rad[kms]

NGC104



[MH]

180

200

220

240

260

280

300

v rad[kms]

NGC5139

13b


[MH]

minus100

minus50

0

50

100

150

v rad[kms]

NGC6121

CX10CX8


[MH]

minus200

minus150

minus100

minus50

0

50

100

150

200v rad[kms]

NGC6397

U12U92U18U42












24 Summary





10 Geller et al


















[] [] []

Open Clusters



Globular Clusters











x=0

(Ne)x

x (1)





12 Geller et al











-10

-05

00

05

10

15

log 1

0( N

A

rea

)

-10 -08 -06 -04 -02 00 02 04log10(rrc)

00

02

04

06

08

10

frac

tion

with

in lo

g 10(

rr c

)












0

1

2

3

4

5

N


00

02

04

06

08

10

frac

tion




14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21













24 Summary





10 Geller et al


















[] [] []

Open Clusters



Globular Clusters











x=0

(Ne)x

x (1)





12 Geller et al











-10

-05

00

05

10

15

log 1

0( N

A

rea

)

-10 -08 -06 -04 -02 00 02 04log10(rrc)

00

02

04

06

08

10

frac

tion

with

in lo

g 10(

rr c

)












0

1

2

3

4

5

N


00

02

04

06

08

10

frac

tion




14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21





10 Geller et al


















[] [] []

Open Clusters



Globular Clusters











x=0

(Ne)x

x (1)





12 Geller et al











-10

-05

00

05

10

15

log 1

0( N

A

rea

)

-10 -08 -06 -04 -02 00 02 04log10(rrc)

00

02

04

06

08

10

frac

tion

with

in lo

g 10(

rr c

)












0

1

2

3

4

5

N


00

02

04

06

08

10

frac

tion




14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21








[] [] []

Open Clusters



Globular Clusters











x=0

(Ne)x

x (1)





12 Geller et al











-10

-05

00

05

10

15

log 1

0( N

A

rea

)

-10 -08 -06 -04 -02 00 02 04log10(rrc)

00

02

04

06

08

10

frac

tion

with

in lo

g 10(

rr c

)












0

1

2

3

4

5

N


00

02

04

06

08

10

frac

tion




14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21





12 Geller et al











-10

-05

00

05

10

15

log 1

0( N

A

rea

)

-10 -08 -06 -04 -02 00 02 04log10(rrc)

00

02

04

06

08

10

frac

tion

with

in lo

g 10(

rr c

)












0

1

2

3

4

5

N


00

02

04

06

08

10

frac

tion




14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21













0

1

2

3

4

5

N


00

02

04

06

08

10

frac

tion




14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21





14 Geller et al



























52 X-ray Emission









Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21







Log[fx fR ]

290

295

300

305

310

315

320

325

Log

[Lx

]


Log[fx fV ]

CV

AB (evol)

AB (MS)

Blue St aggle s

W UMa

qLMXB

NGC 188

NGC 2682

NGC 6791

NGC 6819

NGC 104

NGC 5139

NGC 6121

NGC 6218

NGC 6366

NGC 6397

NGC 6652

NGC 6809

NGC 6838














16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21





16 Geller et al

1

10

100N

SS

G

1

10

100

103 104 105 106 107

Mcl (M )

10-6

10-5

10-4

10-3

10-2

NS

SG

Mcl (

M -1

)

103 104 105 106 10710-6

10-5

10-4

10-3

10-2



























REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21













REFERENCES











18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21





18 Geller et al








































































20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21








20

Geller

etal



Open Clusters


Globular Clusters


Dem

ographics

ofSub-su

bgiantStars

21





Dem

ographics

ofSub-su

bgiantStars

21





Northwestern University, 2145 Sheridan Rd., Evanston, IL ... · Kamann7, Nathan W. C.Leigh8, Robert...

Documents

Transcript of Northwestern University, 2145 Sheridan Rd., Evanston, IL ... · Kamann7, Nathan W. C.Leigh8, Robert...