astro-ph0409447/ms.tex
1: \documentclass[12pt,preprint]{aastex}
2: \usepackage{xspace}
3: \received{}
4: \revised{}
5: \accepted{}
6: \newcommand\mw{\ensuremath{M_w}\xspace}
7: \newcommand\msun{\ensuremath{M_\sun}\xspace}
8: \newcommand\mcrit{\ensuremath{M_{\rm crit}}\xspace}
9: \newcommand\teff{\ensuremath{T_{\rm eff}}\xspace}
10: \newcommand\logg{\ensuremath{\log g}\xspace}
11: \newcommand\ubv{\ensuremath{U\!BV}\xspace}
12: \newcommand\ebv{\ensuremath{E(\bv)}\xspace}
13: 
14: \slugcomment{Accepted for publication in the Astrophysical Journal Letters} 
15: \shorttitle{Massive white dwarfs in NGC 2168}
16: \shortauthors{Williams, Liebert \& Koester}
17: 
18: \begin{document}
19: 
20: \title{An empirical initial-final mass relation from hot,
21:   massive white dwarfs in NGC 2168 (M35)}  
22: 
23: \author{Kurtis A. Williams}
24: \affil{Steward Observatory}
25: \affil{933 N. Cherry Ave., Tucson, AZ 85721}
26: \email{kurtis@as.arizona.edu}
27: \author{M. Bolte}
28: \affil{UCO/Lick Observatory}
29: \affil{University of California, Santa Cruz, CA 95064}
30: \email{bolte@ucolick.org}
31: \and
32: \author{Detlev Koester}
33: \affil{Institut f\"ur Theoretische Physik und Astrophysik}
34: \affil{Universit\"at Kiel, 24098 Kiel, Germany}
35: \email{koester@astrophysik.uni-kiel.de}
36: \begin{abstract}
37: The relation between the zero-age main sequence mass of a star and its
38: white-dwarf remnant (the initial-final mass relation) is a powerful
39: tool for exploration of mass loss processes during stellar evolution.
40: We present an empirical derivation of the initial-final mass relation
41: based on spectroscopic analysis of seven massive white dwarfs in NGC
42: 2168 (M35).  Using an internally consistent data set, we show that the
43: resultant white dwarf mass increases monotonically with progenitor mass for
44: masses greater than 4\msun, one of the first open clusters to show this
45: trend.  We also find two massive white dwarfs foreground to the
46: cluster that are otherwise consistent with cluster membership.  These
47: white dwarfs can be explained as former cluster members moving
48: steadily away from the cluster at speeds of $\lesssim 0.5$ km/s since
49: their formation and may provide the first direct evidence of the
50: loss of white dwarfs from open
51: clusters. Based on these data alone, we constrain the upper mass
52: limit of WD progenitors to be $\geq 5.8\msun$ at the 90\% confidence level
53: for a cluster age of 150 Myr. 
54: \end{abstract}
55: \keywords{white dwarfs --- open clusters and associations: individual
56:   (NGC 2168)}
57: 
58: \section{Introduction}
59: White dwarfs (WDs) are the final state of stellar evolution for
60: the vast majority of intermediate- and low-mass stars.  The upper mass
61: limit of WD progenitor stars, \mcrit, is also the lower mass
62: limit of core-collapse supernova progenitors.  The current best observational
63: estimate for \mcrit comes from spectroscopic analysis of WDs in the open cluster
64: \objectname{NGC 2516} \citep{Koester1996} and is 
65: $\mcrit\approx 8\msun \pm 2\msun$.  Due to the steepness of the
66: initial-mass function (IMF), however, this range results in a factor
67: of $\approx 2$ uncertainty in the number of supernovae and duration of
68: supernova-driven winds resulting from bursts of star formation. This
69: uncertainty in turn has a large impact on understanding the
70: star-formation rate in galaxies 
71:  \citep[e.g.][]{Somerville1999}, the evolution of
72: starbursts \citep[e.g.][]{Leitherer1999}, and  the fate
73: of low-mass dwarf galaxies at early times \citep[e.g.][]{Dekel1986}.
74: 
75: The best observational constraints on \mcrit are obtained from studies of
76: WD populations in open clusters with ages $\lesssim 150$ Myr
77: \citep{Williams2002}.  \objectname{NGC 2168} (M35) is one of the
78: richest, compact and nearby open clusters in this age range, with age
79: determinations ranging from $\sim 100$ Myr \citep{vonhippel2000} to
80: $\sim 200$ Myr \citep{Sung1999}.  The WD cooling sequence of
81: NGC 2168 has been discussed often in the literature, with recent photometric
82: analyses by \citet{vonHippel2002} and \citet{Kalirai2003}.  
83: 
84: With spectroscopy of cluster WD candidates, it is possible to
85: determine unambiguously if the objects are WDs and, for the bona fide WDs, to 
86: determine \teff and \logg, and to derive
87: the cooling age ($\tau_{\mathrm cool}$), mass and luminosity.  
88: For those WDs with cooling ages smaller than the cluster age and
89: distance modulus consistent with cluster membership, subtraction 
90: of the WD cooling age from the age of the
91: open cluster results in the lifetime of the progenitor star, and
92: stellar evolutionary models can then be used to determine the
93: progenitor star mass.  \citet{Reimers1988} applied this
94: algorithm to two NGC 2168 WDs candidates identified on photographic plates by
95: \citet{Romanishin1980}; however, their signal-to-noise was too low to
96: determine the WD properties precisely.
97: 
98: As part of our ongoing program to identify and spectroscopically
99: analyze the WD populations of open clusters, the Lick-Arizona White
100: Dwarf Survey \citep[LAWDS,][]{Williams2004}, we
101: have obtained high signal-to-noise spectra of eight candidate
102: massive WDs in NGC 2168.  Six or seven of the observed WDs are hot, high-mass
103: WDs likely to be cluster members.  This sample doubles the number of
104: known open cluster WDs with high-mass progenitors.
105: A detailed photometric and spectroscopic 
106: analysis of the entire WD population of this cluster will be presented
107: in a later paper. In this Letter we present the constraints on \mcrit and
108: the upper end of the initial-final mass relation for WDs based on the 
109: high-mass WDs already analyzed.
110: 
111: \section{Observations \& Analysis}
112: \ubv imaging of NGC 2168 was obtained in September 2002 with
113: the Lick Observatory Shane 3-m reflector and the PFCam prime-focus
114: imager; additional \ubv imaging of a larger field center on the
115: cluster was obtained in
116: January 2004 with the KPNO 4m MOSAIC camera.  PSF-fitting photometry
117: was obtained using the DAOPHOT II program \citep{Stetson1987}.
118: Candidate WDs were selected by their blue excess in \ubv color space.
119: Figure~\ref{fig.cmd} shows the
120: color-magnitude diagram of all objects detected in $U$, $B$, and $V$
121: across the entire MOSAIC field.
122: Several very blue, faint objects are observed in the diagram; these 
123: are our candidate WDs. All four WD candidates of
124: \citet{Romanishin1980} are recovered, as 
125: is the WD candidate in \citet{vonHippel2002}.  Astrometry and
126: photometry for the WD candidates in this Letter are presented in
127: Table~\ref{tab.phot}. 
128: 
129: Spectroscopic observations of selected WD candidates were obtained
130: with the blue camera of the LRIS spectrograph on the Keck I
131: 10-m telescope \citep{Oke1995}. A 1\arcsec-wide longslit at
132: parallactic angle was used with the 400 l mm$^{-1}$, 3400\AA~blaze
133: grism for a resulting spectroscopic resolution of $\sim 6$\AA.  
134: The spectra were extracted and a relative spectrophotometric
135: calibration applied using standard \emph{IRAF}
136: \footnote{\emph{IRAF} is
137:   distributed by the National Optical Astronomy Observatories, which
138:   are operated by the Association of Universities for Research in
139:   Astronomy, Inc., under cooperative agreement with the National
140:   Science Foundation} 
141: routines.  
142: 
143: The \teff and \logg were determined for each WD using simultaneous
144: Balmer-line fitting \citep{Bergeron1992}.  The model spectra are
145: updated versions of those in \citet{Finley1997}.  The WD
146: evolutionary models of \citet{Wood1995} were used to calculate the
147: mass ($M_{\rm WD}$) and cooling age ($\tau_{\rm WD}$) of each WD.  A distance
148: modulus to each WD was measured by comparing the observed $V$-magnitude 
149: to the absolute magnitude $M_V$ calculated from the
150: best-fitting model atmosphere and the appropriate WD cooling model.
151: Errors in the fits were determined empirically
152: by adding the noise measured for each spectrum to the best-fitting
153: model spectrum convolved with the instrumental response. These
154: simulated spectra were fit by the same method; 
155: nine iterations were used to calculate  the
156: scatter in \teff, \logg, $M_{\rm WD}$ and $\tau_{\rm WD}$.  The fitting
157: procedure is discussed in depth in an upcoming paper on the open
158: clusters \objectname{NGC 6633} and \objectname{NGC
159:   7063} \citep{Williams2004}.  The atmospheric 
160: fits and derived WD masses and ages are given in
161: Table~\ref{tab.spec}. The Balmer line fits are shown in
162: Figure~\ref{fig.linefits}.  
163: 
164: A systematic error in the fits of the
165: hottest ($\teff\gtrsim 50000{\rm K}$) WDs became apparent, as it was
166: not possible to simultaneously fit all the Balmer lines.  In these
167: cases, the fits were limited to H$\beta$, H$\gamma$, and
168: H$\delta$. For LAWDS 22, no satisfactory convergence was achieved with
169: these limited fits; the best-fit models are used in the
170: analysis. WDs with $\teff\gtrsim 50000$K are known to exhibit metals
171: in the atmosphere and NLTE effects \citep[e.g.][]{Napiwotzki1992,Holberg1998}, 
172: neither of which are included in
173: the models.
174: 
175: The progenitor mass for each WD was calculated by subtracting the WD
176: cooling age from the cluster age.  The age difference is the total
177: lifetime of the progenitor star.  The lifetimes of stars as a
178: function of mass and metallicity are calculated from the stellar
179: isochrones of \citet{Girardi2002}.  The progenitor mass for each
180: WD likely to be a cluster member
181: is given in Table~\ref{tab.masses} for an assumed cluster age of 150
182: Myr and for
183: three different stellar evolutionary models: $Z=0.008$ and $Z=0.019$,
184: both with modest convective overshoot, and $Z=0.019$ without
185: convective overshoot. Upper and lower errors are for 1$\sigma$
186: differences in $\tau_{\rm WD}$.
187: 
188: \section{Discussion\label{discussion}}
189: All eight of the observed WDs are much more massive than the typical
190: WD mass of $\approx 0.56\msun$ \citep{Bergeron1992}, and five of the
191: WDs have apparent distance moduli $(m-M)_V\approx 10.5$.  
192: It is therefore reasonable to
193: assume that at least five of these objects are members of NGC 2168.
194: LAWDS 11 is almost certainly \emph{not} a cluster member.  Its age is
195: likely older than that of the  cluster as a whole, and while the
196: uncertainties in the spectral fits leave open the possibility that it
197: is younger (which would require it to be hotter), the observed colors
198: are more consistent with the cooler (and older) interpretation.  The
199: distance modulus of this WD is foreground to the cluster by little
200: more than $1\sigma$, but the cooler temperature favored by the
201: \bv ~color again favors the foreground interpretation.
202: 
203: LAWDS
204: 15 has a mass and age consistent with cluster membership, but the
205: calculated distance modulus is inconsistent with that of the other
206: WDs by $\sim 4\sigma_{M_V}$.  Assuming that the distance modulus is correct
207: and that there is no difference in reddening between LAWDS
208: 15 and NGC 2168, LAWDS 15 is $\sim 185$ pc closer than the cluster.
209: Based on the spectral fits, LAWDS 15 has a cooling age of $\sim 50$
210: Myr.  If the WD has been moving away from the cluster since its
211: formation at a steady rate of only $0.4$ km/s, it will have covered
212: this distance.  Therefore, it is possible that LAWDS 15 was
213: once a cluster member and has escaped the cluster. 
214: 
215: The likelihood that a massive, hot WD
216: would be found foreground along the line of sight to the cluster
217: can be estimated from the luminosity function in Figure 16 of
218: \citet{Liebert2004}. The luminosity function gives the space density
219: of WDs with $M>0.8\msun$ and $\tau_{\rm cool}\leq 100$ Myr as $\sim
220: 10^{-5.3} \,{\rm pc}^{-3}\,0.5\,{\rm mag}^{-1}$. This results
221: in an estimated 0.1 hot, massive WDs in the $\sim 30\arcmin\times
222: 30\arcmin$ MOSAIC field to a distance of 1 kpc.  Therefore it is
223: unlikely, but not impossible, that LAWDS 15 is a field WD.   Based on
224: these arguments, we will consider LAWDS 15 
225: to be a cluster WD for this discussion.
226: For similar
227: arguments, we retain LAWDS 6 ($\sim 2\sigma_{M_V}$ closer than the
228: other cluster WDs) as a likely cluster member.
229: 
230: Figure~\ref{fig.ifm} shows the initial-final mass relation of these
231: seven cluster members, along with that of WDs from Hyades, Praesepe and
232: the Pleiades \citep{Claver2001} and from NGC 2516
233: \citep{Koester1996}.  Also shown are theoretical and semi-empirical
234: data from plots in \citet{Claver2001} and sources therein.  For the sake of
235: consistency, the initial and final masses of each WD from the
236: literature have been re-determined using our WD models and the
237: published \teff and \logg.  
238: 
239: From the figure, it can be seen that the NGC 2168 WDs form a monotonic
240: sequence of more massive WDs originating from more massive
241: progenitors.  This conclusion is robust, as changes in the
242: assumed age of NGC 2168 do not affect the relative positions of the
243: points, only the absolute initial masses. 
244: Figure~\ref{fig.ifm} also shows that the NGC 2168 initial-final
245: mass relation agrees with the empirical relation derived 
246: from previous clusters.
247: This need not be expected \emph{a priori}, as NGC 2168 has a
248: significantly lower metallicity than the other clusters in the
249: diagram.  Model core masses decrease with increasing metallicity, 
250: and the efficiency of mass loss processes could change with metallicity.  
251: 
252: LAWDS 15 and LAWDS 6, the potential escaped cluster WDs discussed above,
253: fit the observed initial-final mass relation,
254: providing additional evidence that they are cluster members.  If these
255: WDs are indeed escaped cluster members, they are crucial
256: pieces of evidence that WDs can receive velocity kicks during mass loss
257: and perhaps explain the observed deficit of WDs in other open clusters
258: \citep[see][and references therein]{Williams2004a}.  This and other
259: potential explanations for these objects (e.g. binarity) will be discussed more fully in
260: the later paper on WDs in NGC 2168.
261: 
262: The WD in Fig.~\ref{fig.ifm} with the apparently low final mass
263: is LAWDS 22.  As mentioned above and visible in Fig.~\ref{fig.linefits},
264: the spectral fitting did not converge satisfactorily, despite the
265: high signal-to-noise of the spectrum.  This star is a close visual
266: double with a redder companion ($V=19.12$, $\bv=1.27$, $\ub=0.93$)
267: 2\arcsec to the north.
268: While resolved, this double is close enough that the spectrum of LAWDS
269: 22 is likely contaminated by light from the neighbor star, resulting
270: in the 
271: unsatisfactory fit.  Light from the neighboring star may also be
272: contaminating the photometric colors of LAWDS 22, which would also explain
273: why the star lies redward of the 1\msun cooling track in Fig.~\ref{fig.cmd}.  
274: Given its measured \teff, LAWDS 22 likely suffers from an
275: extension of the systematic issue in the high-\teff WD 
276: spectral fits described above.  Other explanations for
277: the location of this point could include magnetic fields (although no
278: splitting is observed) or a low-mass, unresolved companion, but 
279: contamination by the neighboring star seems to be the most likely cause.
280: 
281: Based on the NGC 2168 data alone, it is possible to place lower limits
282: on the value of \mcrit.  Making the assumption that errors in the WD
283: ages in Table~\ref{tab.spec} are Gaussian, we calculate that the 
284: oldest cluster WD ages are $\log \tau_{\rm cool}\geq 7.76$
285: with 90\% confidence.  For $\log \tau_{\rm cl}=8.15$ and $Z=0.008$, this
286: corresponds to $\mcrit\geq 5.81\msun$ at a 90\% confidence level.
287: This value is in agreement with that obtained by \citet{Koester1996}.
288: 
289: Improved constraints on the initial-final mass relation, including its
290: intrinsic scatter and metallicity dependence, require improvements on
291: existing observations.  First and foremost, the ages of open clusters
292: such as NGC 2168 must be determined to higher precision.  Alternatives
293: to main-sequence fitting such as lithium depletion studies or
294: activity/rotation studies may provide the necessary constraints.
295: Second, large samples of WDs from individual open clusters are needed to
296: reduce the effect of systematics (such as errors in assumed ages) that
297: plague the comparison of inter-cluster samples.  The WD sample
298: presented here is a start to that end, and the initial-final mass
299: relation derived from these stars alone provides dramatic confirmation
300: of the existence of an initial-final mass relation, an idea that was
301: strongly hinted at from previous open cluster studies and from other
302: theoretical and observational work.  Assuming that the majority of the
303: remaining massive NGC 2168 WD candidates are cluster WDs,
304: planned spectroscopic observations of these objects will soon result
305: in a sample of nearly a dozen of WDs originating from a
306: single stellar population, permitting, for the first time, studies of
307: the intrinsic dispersion of the initial-final mass relation at high
308: masses.  
309: 
310: \acknowledgements
311: M.B.  and K.W. are grateful for support for this project in the
312: form of the National Science Foundation AST-0307492.
313: Any opinions, findings, and conclusions or recommendations expressed
314: in this material are those of the author(s) and do not necessarily
315: reflect the views of the National Science Foundation.
316: 
317: The authors would like to thank Matt Wood for freely providing
318: the WD evolutionary models and for many useful comments regarding this paper.
319: The authors also thank the anonymous referee for a well-considered
320: report that resulted in several improvements to this Letter.
321: 
322: Some of the data presented herein were obtained at the W.M. Keck
323: Observatory, which is operated as a scientific partnership among
324: the California Institute of Technology, the University of California
325: and the National Aeronautics and Space Administration. The Observatory
326: was made possible by the generous financial support of the W.M.
327: Keck Foundation.
328: 
329: 
330: The authors wish to recognize and acknowledge the very significant
331: cultural role and reverence that the summit of Mauna Kea has always
332: had within the indigenous Hawaiian community.  We are most fortunate
333: to have the opportunity to conduct observations from this mountain.
334: 
335: \begin{thebibliography}{}
336: \bibitem[Bergeron, Saffer, \& Liebert(1992)]{Bergeron1992} Bergeron, P., 
337: Saffer, R.~A., \& Liebert, J.\ 1992, \apj, 394, 228 
338: \bibitem[Claver et al.(2001)]{Claver2001} Claver, C.~F., Liebert, J., 
339: Bergeron, P., \& Koester, D.\ 2001, \apj, 563, 987 
340: \bibitem[Dekel \& Silk(1986)]{Dekel1986} Dekel, A.~\& Silk, J.\ 1986, \apj, 
341: 303, 39 
342: \bibitem[Finley, Koester, \& Basri(1997)]{Finley1997} Finley, D.~S., 
343: Koester, D., \& Basri, G.\ 1997, \apj, 488, 375 
344: \bibitem[Girardi et al.(2000)]{Girardi2000} Girardi, L., Bressan, A., 
345: Bertelli, G., \& Chiosi, C.\ 2000, \aaps, 141, 371 
346: \bibitem[Girardi et al.(2002)]{Girardi2002} Girardi, L., Bertelli, G., 
347: Bressan, A., Chiosi, C., Groenewegen, M.~A.~T., Marigo, P., Salasnich, B., 
348: \& Weiss, A.\ 2002, \aap, 391, 195 
349: \bibitem[Holberg, Barstow, \& Sion(1998)]{Holberg1998} Holberg, J.~B., Barstow, M.~A., \& Sion, E.~M.\ 1998, \apjs, 119, 207
350: \bibitem[Kalirai et al.(2003)]{Kalirai2003} Kalirai, J.~S., Fahlman, G.~G., 
351: Richer, H.~B., \& Ventura, P.\ 2003, \aj, 126, 1402
352: \bibitem[Koester \& Reimers(1996)]{Koester1996} Koester, D.~\& Reimers, D.\ 
353: 1996, \aap, 313, 810 
354: \bibitem[Leitherer et al.(1999)]{Leitherer1999} Leitherer, C., et al.\ 
355: 1999, \apjs, 123, 3 
356: \bibitem[Liebert, Bergeron, \& Holberg(2004)]{Liebert2004} Liebert, J., Bergeron,
357:   P., \& Holberg, J.~B.~ 2004, \apjs, in press (astro-ph/0406657)
358: \bibitem[McCook \& Sion(1999)]{McCook1999} McCook, G.~P.~\& Sion, E.~M.\ 
359: 1999, \apjs, 121, 1 
360: \bibitem[Napiwotzki(1992)]{Napiwotzki1992} Napiwotzki, R.\ 1992, 
361: Lecture Notes in Physics, Berlin Springer Verlag, 401, 310 
362: \bibitem[Oke et al.(1995)]{Oke1995} Oke, J.~B., et al.\ 1995, \pasp, 107, 
363: 375 
364: \bibitem[Reimers \& Koester(1988)]{Reimers1988} Reimers, D.~\& Koester, D.\ 
365: 1988, \aap, 202, 77 
366: \bibitem[Romanishin \& Angel(1980)]{Romanishin1980} Romanishin, W.~\& 
367: Angel, J.~R.~P.\ 1980, \apj, 235, 992 
368: \bibitem[Somerville \& Primack(1999)]{Somerville1999} Somerville, R.~S.~\& 
369: Primack, J.~R.\ 1999, \mnras, 310, 1087 
370: \bibitem[Stetson(1987)]{Stetson1987} Stetson, P.~B.\ 1987, \pasp, 99, 191 
371: \bibitem[Sung \& Bessell(1999)]{Sung1999} Sung, H.~\& Bessell, M.~S.\ 1999, 
372: \mnras, 306, 361 
373: \bibitem[von Hippel et al.(2000)]{vonhippel2000} von Hippel, T., 
374: Kozhurina-Platais, V., Platais, I., Demarque, P., \& Sarajedini, A.\ 2000, 
375: ASP Conf.~Ser.~198: Stellar Clusters and Associations: Convection, 
376: Rotation, and Dynamos, 75 
377: \bibitem[von Hippel et al.(2002)]{vonHippel2002} von Hippel, T., 
378: Steinhauer, A., Sarajedini, A., \& Deliyannis, C.~P.\ 2002, \aj, 124,
379: 1555
380: \bibitem[Weidemann(2000)]{Weidemann2000} Weidemann, V.\ 2000, \aap, 363, 
381: 647  
382: \bibitem[Williams(2002)]{Williams2002} Williams, K.~A.\ 2002, Ph.D.~Thesis
383: \bibitem[Williams(2004)]{Williams2004a} Williams, K.~A.\ 2004, \apj, 601, 
384: 1067 
385: \bibitem[Williams et al.(2004)]{Williams2004} Williams, K.~A., Bolte,
386:   M., Koester,  D., \& Wood, M.~A. 2004, in preparation  
387: \bibitem[Wood(1995)]{Wood1995} Wood, M.~A.\ 1995, LNP Vol.~443: White 
388: Dwarfs, 41 
389: \end{thebibliography}
390: 
391: \clearpage
392: 
393: \begin{deluxetable}{llccccccc}
394: \rotate
395: \tabletypesize{\footnotesize}
396: \tablecolumns{9}
397: \tablecaption{WDs in the field of NGC 2168, photometry.\label{tab.phot}}
398: \tablewidth{0pt}
399: \tablehead{\colhead{LAWDS} & \colhead{McCook \& Sion} & \colhead{RA} & \colhead{Dec}& $V$ & $\sigma_V$ &  
400:   \colhead{$\bv$} & \colhead{$\sigma_{\bv}$} & \colhead{Previous} \\
401:   \colhead{ID} & \colhead{Designation\tablenotemark{a}} & \colhead{(J2000)} & \colhead{(J2000)} & & & & & \colhead{References}}
402: \startdata
403: \objectname{NGC 2168:LAWDS 1}  & \objectname{WD J0608+242}  & 6:08:38.79 & 24:15:06.9 & 20.989 & 0.019 & -0.035 & 0.028 &1\\
404: \objectname{NGC 2168:LAWDS 2}  & \objectname{WD J0608+241}  & 6:08:42.30 & 24:10:17.7 & 21.569 & 0.032 &  0.061 & 0.044 &\\
405: \objectname{NGC 2168:LAWDS 5}  & \objectname{WD J0609+244.1}& 6:09:11.54 & 24:27:20.9 & 20.065 & 0.017 & -0.128 & 0.024 &1,2\\
406: \objectname{NGC 2168:LAWDS 6}  & \objectname{WD J0609+244.2}& 6:09:23.48 & 24:27:22.0 & 19.863 & 0.016 & -0.128 & 0.023 &1,2\\
407: \objectname{NGC 2168:LAWDS 11} & \objectname{WD J0609+241}  & 6:09:42.79 & 24:11:05.4 & 21.198 & 0.025 &  0.110 & 0.037 &\\
408: \objectname{NGC 2168:LAWDS 15} & \objectname{WD J0609+240}  & 6:09:11.63 & 24:02:38.5 & 20.785 & 0.022 & -0.039 & 0.032 &\\
409: \objectname{NGC 2168:LAWDS 22} & \objectname{WD J0608+245}  & 6:08:24.65 & 24:33:47.6 & 19.657 & 0.016 &  0.008 & 0.023 &\\
410: \objectname{NGC 2168:LAWDS 27} & \objectname{WD J0609+243}  & 6:09:06.26 & 24:19:25.3 & 21.398 & 0.026 &  0.090 & 0.039 &3\\
411: \enddata
412: \tablenotetext{a}{Format as in \citet{McCook1999}}
413: \tablerefs{(1) \citet{Romanishin1980}, (2) \citet{Reimers1988}, (3) \citet{vonHippel2002}}
414: \end{deluxetable}
415: 
416: \clearpage
417: 
418: \begin{deluxetable}{lccccccccccc}
419: \rotate
420: \tabletypesize{\footnotesize}
421: \tablecolumns{12}
422: \tablecaption{WDs in the field of NGC 2168, spectral fits.\label{tab.spec}}
423: \tablewidth{0pt}
424: \tablehead{\colhead{ID} & \colhead{\teff} & \colhead{$d\teff$} & \colhead{\logg} & \colhead{$d\logg$} & 
425:    \colhead{$M_{\rm WD}$} & \colhead{$dM_{\rm WD}$} & \colhead{$\log(\tau_{\rm WD})$} & \colhead{$d\log(\tau_{\rm WD}$)} & 
426:    \colhead{$M_V$} & \colhead{$dM_V$} & \colhead{$(m-M)_V$}} 
427: \startdata
428: LAWDS 1  & 32400 &  512 & 8.40 & 0.125 & 0.888 & 0.075 & 7.386 & 0.208 & 10.454 & 0.215 & 10.535 \\
429: LAWDS 2  & 31700 & 1800 & 8.74 & 0.191 & 1.072 & 0.102 & 7.929 & 0.291 & 11.132 & 0.328 & 10.437 \\
430: LAWDS 5  & 52600 & 1160 & 8.24 & 0.095 & 0.824 & 0.051 & 6.158 & 0.025 &  9.542 & 0.184 & 10.523 \\
431: LAWDS 6  & 55200 &  897 & 8.28 & 0.065 & 0.851 & 0.036 & 6.094 & 0.027 &  9.564 & 0.108 & 10.299 \\
432: LAWDS 11 & 19900 & 2792 & 8.48 & 0.367 & 0.921 & 0.213 & 8.288 & 0.463 & 11.534 & 0.769 &  9.664 \\
433: LAWDS 15 & 29900 &  318 & 8.48 & 0.060 & 0.934 & 0.037 & 7.693 & 0.089 & 10.754 & 0.111 & 10.031 \\
434: LAWDS 22 & 54400 & 1203 & 8.04 & 0.121 & 0.721 & 0.060 & 6.169 & 0.054 &  9.155 & 0.219 & 10.502 \\
435: LAWDS 27 & 30900 &  500 & 8.58 & 0.164 & 0.995 & 0.086 & 7.760 & 0.220 & 10.866 & 0.310 & 10.532 \\
436: \enddata
437: \end{deluxetable}
438: 
439: \clearpage
440: 
441: \begin{deluxetable}{lccc}
442: \tablecolumns{4}
443: \tablecaption{Progenitor Masses for NGC 2168 WDs.\label{tab.masses}}
444: \tablewidth{0pt}
445: \tablehead{
446:     \colhead{ID} & \colhead{$Z=0.008$} & \colhead{$Z=0.019$} & \colhead{$Z=0.019$\tablenotemark{a}}\\
447:     & (\msun) &  (\msun) &  (\msun)}
448: 
449: \startdata
450: LAWDS 1  & $5.00^{+0.31}_{-0.16}$    & $5.08^{+0.29}_{-0.15}$    & $4.86^{+0.28}_{-0.14}$ \\ 
451: LAWDS 2  & $7.02^{+\infty}_{-1.60}$  & $6.96^{+\infty}_{-1.49}$  & $6.63^{+\infty}_{-1.40}$ \\
452: LAWDS 5  & $4.628^{+0.001}_{-0.001}$ & $4.734^{+0.002}_{-0.001}$ & $4.540^{+0.001}_{-0.001}$ \\
453: LAWDS 6  & $4.625^{+0.001}_{-0.001}$ & $4.732^{+0.001}_{-0.001}$ & $4.537^{+0.001}_{-0.010}$ \\
454: LAWDS 15 & $5.57^{+0.34}_{-0.24}$    & $5.62^{+0.31}_{-0.22}$    & $5.37^{+0.30}_{-0.21}$ \\
455: LAWDS 22 & $4.629^{+0.003}_{-0.003}$ & $4.735^{+0.003}_{-0.002}$ & $4.540^{+0.003}_{-0.002}$ \\
456: LAWDS 27 & $5.81^{+1.99}_{-0.61}$    & $5.84^{+1.81}_{-0.56}$    & $5.59^{+1.46}_{-0.54}$ \\
457: \enddata
458: \tablenotetext{a}{No convective overshoot}
459: \end{deluxetable}
460: 
461: \clearpage
462: 
463: \begin{figure}
464: \plotone{f1.eps}
465: \figcaption{$\bv,\ V$ color-magnitude diagram
466: for NGC 2168. Squares
467:   indicate WDs presented in this study, both cluster members (filled)
468:   and the non-member (open).  Solid curves indicate cooling curves for WDs
469:   with masses of 0.4\msun (top curve), 0.7\msun,
470:   and 1.0\msun(bottom curve) at the
471:   distance and reddening of NGC 2168.  The dotted line indicates the
472:   location of WDs with $\log \tau_{cool}=8.15$, the assumed age of NGC
473:   2168. \label{fig.cmd}} 
474: \end{figure}
475: 
476: \begin{figure}
477: \includegraphics[angle=270,width=6.25in]{f2.eps}
478: \figcaption{Balmer line fits for WD candidates in NGC 2168. The figures
479:   show the Balmer lines from H$\beta$ (bottom) to H9 (top), with the
480:   curves showing the best-fit models. \label{fig.linefits}}
481: \end{figure}
482: 
483: \begin{figure}
484: \plotone{f3.eps}
485: \figcaption{Initial-final mass relation.  Points with error bars are WDs from
486:   this study; other points are from the Hyades and Praesepe (Claver et
487:   al.~2001, open triangles), NGC 2516 (Koester 1996, open squares) and
488:   the Pleiad WD (Claver et al.~2001, upper limit).  The starred point
489:   with error bars is LAWDS 22, discussed in the text. Curves represent
490:   the theoretical initial-final mass relation from \citet{Girardi2000}
491:   (solid), the core mass at the first thermal pulse from the same
492:   models (dashed), and the quasi-empirical relation from
493:   \citet{Weidemann2000} (dotted line). Slight horizontal offsets have been
494:   applied to the three WDs with $M_{\rm init}\approx 4.6\msun$ for
495:   the sake of clarity. Statistical error bars in the initial masses
496:   for these three WDs are smaller than the points.\label{fig.ifm}}
497: \end{figure}
498: 
499: \end{document}
500: 
501: 
502: 
503: 
504: 
505: 
506: 
507: 
508: 
509: 
510: 
511: 
512: 
513: