Evolving to type la supernovae with long delay time
Wen-Cong Chen ^'^ and Xiang-Dong Li ^
ABSTRACT
O
O
(N
D
tin
00
>
00
(N
(N
O
o
:^
o
U
X
Recent investigations on the delay time of type la supernovae have set useful
constraints on the progenitors of type la supernovae. Here we have calculated
the evolution of close binaries consisting of a white dwarf and a main- sequence
or subgiant companion. We assume that, once Roche lobe overflow occurs a
small fraction of the lost mass from the system forms a circumbinary disk, which
extracts the orbital angular momentum from the system through tidal torques.
Our calculations indicate that the existence of circumbinary disk can enhance the
mass transfer rate and cause secular orbital shrinkage. The white dwarf can grow
in mass efficiently to trigger type la supernovae even with relatively low-mass
(< 2M0) donor stars. Thus this scenario suggest a new possible evolutionary
channel to those type la supernovae with long delay time ~ 1 — 3 Gyr.
Subject headings: stars: general — binaries: close — stars: mass loss — stars:
evolution — circumstellar matter
1. Introduction
Many works have suggested that type la supernovae (SNe la) can be used as the stan-
Riess et al.
1998;
Perlmutter et al.
1999
Riess et al.ll2004j ). It is widely believed that SNe la are thermonuclear explos ions of accreting
CO w hite dwarfs (WDs) when their masses grow beyond a critical mass ( iHoyle fc Fowlei
19601 ). However, the nature of the progenitors and the related accretion processes have still
remained unclear. Several possible evolutionary scenarios for SN la explo sions have been pro-
posed so far. At present, there exist Chandrasekhar ( Ch) mass mode l (IWoosley &: Weaver
19861 ) and sub-Chandrasekhar (sub-Ch) mass model (iNomotd Il982bl ) for the critical mass
"'^Department of Astronomy.
chenwc@nju.edu.cn, lixd@nju.edu.cn
Nanjing University, Nanjing
5vc@nju.edu.cn, lixd@nju.edu.cn
^Department of Physics, Shangqiu Normal University, Henan, 476000, China
210093,
China;
-2-
of th e exploding WDs . Furthermore, both double degenerat e model fDDl (llben &: Tutukov
1984 : IWeebbinkI Il984l ) and single degenerate model (SD) jNomotol (19821 have been also
proposed as the candidates for progenitors of SNe la (for a review see [Branch et al.lll995l ).
For the DD model, previous works indicated that the expected accretion rates may cause
the accretion- induce d collapse (AIC) of the WDs and the formation of neutron stars instead
of SN la explosions (INomoto fc Ibenlll985l : ISaio fc Nomotdll985l : iTimmes. Woosley fc Taam
1994h .
Binary evolution investigations adopting the SD/Ch scenarios obtained two types of
progenitors of SNe la. One is close binaries with a main-seque nce or subgiant companion
of mass ~ 2 — 3.5Mf7^ and an initial orbital p eriod of ~ 1 d ( Li fc van den Heuvell 119971 :
Langer et al.l l2000l : iHan fc Podsiadlowskil |2004| ) . The required accretion rate for steady nu-
clear burning onto WDs could be obtained with thermal-timescale mass transfer from the
more massive donor stars. The other is wide binaries with a red giant of masses ~ IMf^ and
an in itial orbital period of ~ 100 d (IHachisu. Kato fc Nomotd Il996l : iLi fc van den Heuvel
19971 ). However, mass accretion i n these systems may be unstable due to the thermal- viscous
instability in the accretion disks (1Cannizzdll993l ).
The various progenitor scenarios can be tested by comparison of the expected and
measured distribution of the time delay between the formation of the progenitor systems and
their explosio ns. Recent high-^ supernova observations suggested the mean delay time of
t ~ 2 -4 Gvr Jstroker et al.booi boosi : ICal-Yam fc Maojbooi Isee however Barris fc Tonrv
20061 : iForster et al. II2006I ). However, the main-sequence li fetime of > 2Mf7^ star should be <
1 Gyr in most of the SD/DD SNe la progenitor scenarios (IHan fc Podsiadlowskil |2004| ). and
the mass transfer timescale ('< lOOMyr) between the components of binary can be neglected.
This led iBelczynski et al.l ( l2005l ) to conclude that only SN la progenitors in the DD scenario
has a characteristic delay time of ~ 3 Gyr. Therefore, if the SD/Ch scenario really works
for (at least part of) SNe la, a progenitor system with a relatively low-mass donor star is
needed.
Supersoft X-ra .y sources (SSS), originally discover ed by the Einstein satellite in the Large
Magellanic Cloud (iLong. Helfand &: Grabelskvlll98l!). may be the observational evidence of
the progenitors of SNe la (jKahabka fc van den Heuvel 1997). A steady-state nuclear burning
model on the surface of t he accreting WD in a bina ry was used to interpreted the SSS
CAL 83 and CAL 87 by Ivan den Heuvel et al.l (119921 ). who suggested that a near- main-
sequence secondary star with a mass of 1.3 — 2.5M0 could provide the required accretion
rate for steady nuclear burning via thermal-timescale mass transfer. However, the catalysmic
variable SSS J0439.8-6809, J0537. 7-7304 and IE 0035 4-7230 have very small mass ratios
and orbital periods (~ 3 — 4 hr) (ISpruit &: Taamll200ll . and references therein), which are
3
significantly less tha n that required for a b i nary with thermal-timescale mass transfer at a
rate ~ 10 ^ ^©yi' ^- Ivan Teeseling &: Kind (119981 ) proposed that there may exist an efficient
mechanism (irradiation-driven mass loss) to enhance the mass transfer rates in these SSS.
The purpose of this paper is to explore the possible progenitor systems for SNe la
with delay times of a few Gyr in the SD/Ch mo del. Enlightened by the original works of
Spruit fc TaamI (120011 ) and iTaam fc SpruitI (120011 ) . we consider the orbital angular momen-
tum loss through the tidal interaction of binary system with a circumbinary (CB) disk during
the mass transfer in a close binary. Our previous works indicate that the CB disk is an ef-
ficient mechanism extr acting angular momentum from the binary system (jChen. Li fc Qian
20061 : IChen fc Lil |2006| ) , which may enhance the mass transfer rates and help mass accumu-
lation on the WD. The prescriptions of binary evolution calculations are described in section
2. In section 3, we present the numerically calculated results for the evolutionary sequences
of the WD binaries. Finally, we discuss and summarize our results in section 4.
Model
We consider a binary system consisting of a CO WD (of mass Mwd), and a main-
sequence or subgiant companion (of mass Md) with solar chemical composition (X = 0.7,
Y = 0.28, Z = 0.02). Mass transfer will occur via Roche lobe overflow of the companion due
to nuclear expansion of the star or orbital shrinkage. We have calculated the ev olution of
WD binarie s adoptin g an updated ver s ion of the stellar evolution code developed by lEggleton
(1l97ll . ll972l ) (see also lHan et al.lll994l : IPols et al.lll995l ). In the calculations we take the ratio
of the mixing length to the pressure scale height to be 2.0. We include the following mass
loss processes and orbital angular momentum loss mechanisms during the mass exchange.
2.1. Mass accumulation efficiency
The key factor for the growth of the WD mass is the accumulation ratio a of the accreted
hydrogen converted into heavier elements. Unfortunately, there are large uncertainties in es-
timating the values of a. For the fraction qh of t he tra nsfer red mass during hydrogen bu rn-
ing, we adopt the prescription by lHachisu et al.l (119991 ) and lHan &: Podsiadlowskil (120041 ). If
the mass transfer rate Md is higher than a critical value Mcr, we assume that hydrogen is
converted into helium at a rate limited to due to the strong optically thick winds from
the WD. The critical mass transfer rate can be written as
1.7 -X\ /Mwd
M„ -
5.3 X 10
-7
X
IM^
- 0.4 Moyr-
-4-
where X is the hydrogen mass abundance of the accreted matter. Below this value, all
hydrogen is assumed to be burned into helium, unt il the mass transfer rate becomes less
than Mcr/8, and strong hydrogen shell flashes occur (IKovetz fc Prialniklll994l ). Thus
-Md > Mcr,
Mcr > -Md > Mcr/8,
-Md < Mcr/8.
(2)
After the gradual increase of the mass in the helium layer on the surface of the WD, helium
ignition occu rs. A part of the envelope rn ass is expected to be blown off due to the helium-
shell flashes ( Kato. Saio fc Hachisul 1989). For the h elium mass accumulation ratio ane, we
adopt the prescription given by Hachisu et al. (jl999 ).
"He
-0.175 (logMne + 5.35)2 ^ ^
-7.3 < logMne < -5.9,
1 , -5.9 < logMne < -5.
(3)
Summarize the above prescriptions, the mass growth rate of the CO WD is Mwd = — aH^HeMd.
The mass lost rate from the binary system can then be written as M = (1 — Q;HaHe)Md.
2.2. Orbital angular momentum losses
During the mass exchange in WD binaries some fraction of the transferred matter
from the donor star may be lost from the system in various ways. For example, high ve-
loci ty outflows were obs erved from the ultraviolet spectra of RX J0513. 9-6951 and CAL
83 (IGansicke et al.lll998l ). Part of the lost matter may form a disk structure surround-
ing the binary syste m rather leave the binary system (Ivan den Heuvel fc de Loord Il973
van den Heuvellll994l ). The optical and UV spectrum of the SSS RX J0019 8+215 6 indicate
the presence of circumbinary material (iKuduz et al.l l2002l : iHutchings et al.l 1200 ll ) . Here we
assume that a small fraction 5 of the mass lost feeds into the CB disk surrounding the binary
system at its inner radius r\. Tidal torques are then exerted on the CB disk via gravitational
interaction, extracting orb ital angular momentum from the binary system fISpruit fc Taam
2001
20061 )
Taam fc Spruitll200ll ). The angular momentum loss rate via the CB disk is (jChen fc Li
CB
7
P.
orb
1/3
(4)
where 7^ = rj/a, t is the mass transfer time, Porb and a are the orbital per iod and the separa-
tion of the binary, respectively. In the standard "a viscosity" prescription (jShakura &: Sunyaev
- 5 -
1973 ). the viscous timescale tvi at the inner edge in the CB disk is given by t^i = s^assP'^ '
where /3 = H-Jri, ass and Hi are the viscosity parameter and the scale height of the CB
disk, respectively.
In addition, we assume that the other part {1 — 6) of the lost mass M is ejected in the
vicinity of the WD in the form of isotropic winds or outflows, carrying away the specific
orbital angular momentum of the WD. The mass loss in the donor's wind and its effect on
the change of orbital angular momentum are comparatively negligible.
3. Numerical Results
We incorporate the prescriptions in last section into the stellar evolution code, and
calculate the evolution of WD binaries with initial parameters of the WD mass Mwd,!, donor
mass Md^, and orbital period Porb,i- Once the WD mass Mwd grows to IAMq, we stop the
calculation an d assume a SNe la occurred. In our calculations, we set 7 = 1.3, ass = 0.01,
and p = 0.03 fIChen fc Lill2006f )
Examples of the mass transfer sequences are shown in Figs. 1 and 2. We plot the
calculated evolutionary sequences for a system with MwD,i = O.SM©, Md,i = l.SM©, and
-Porb,i = 1 d in Fig. 1. The sohd and dashed curves correspond to the cases of 6 = 0.01
and 0, respectively. If no CB disk is assumed to exist, the mass transfer rate is low enough
(~ 5 X 10"^'' — 2 X 10~^MqYt~^) that the accumulation ratio of the accreted matter a ~ 0,
and the WD mass hardly increases. The orbital period first decreases to ~ 0.8 d, then
increases to < 1.4 d. When we include the effect of the CB disk with 6 = 0.01, the mass
transfer rate maintains a relatively high value in the range of ~ 3 x 10^^ — 2 x 10^"^ MqJt^^ .
A large fraction (~ 55%) of the transferred material is accreted by the WD, making the WD
mass grow to IAMq to trigger SNe la. Note that the binary orbit secularly shrinks until the
orbital period Porb reaches < 0.4 d. This evolutionary sequence may represent the formation
history of systems like IE 0035.4—7230 with narrow orbit, low-mass donor star and rapid
mass transfer. Figure 2 shows the evolution of the donor and WD masses with time. It is
clear that the donor star fills its Roche lobe when its age is ~ 2 Gyr for an initial donor star
of I.5M0, much longer than the ~ 20 Myr mass transfer time. This means that it is possible
for the SD / Ch model to have a few Gyr delay time.
We have calculated the evolutions of a large number of WD binaries for a wide distribu-
tion of the initial input parameters. Figure 3 summarizes the final results of our binary evo-
lution calculations for the distributions of the progenitor systems of SNe la in the Md,i — Porb.i
diagram. The bias and pane shading denote the distribution area of WD binaries with a
-6-
[M J
Fig. 1. — Evolution of the mass transfer rate M^, orbital period Porb, and WD mass Mwd
(lower curves) for a WD binary with MwD.i = O.SMq, Md,i = 1.5Mq and Porb.i = Id. The
solid and dashed curves denote the evolutionary tracks with 5 — 0.01 and 0, respectively.
- 7-
Fig. 2. — Evolution of the donor mass (solid curve) and WD mass (dashed curve) when
MwD,i = O.8M0, Md,i = I.5M0, and Porb,i = Id.
- 8 -
WD of initial mass MwD,i = 1-2 and O.8M0, respectively. Beyond these two areas, SNe la
explosions cannot occur due to either low mass accumulation rate or unstable mass transfer.
Compared with previous investigations, our CB disk model opens a possible evolutionary
channel to SNe la for WD binaries with relatively low initial mass donor star of ~ 1 — 2 Mq
and hence long delay times.
To investigate the properties of a companion star that has survived the supernova ex-
plosion we have examined the consequences of heating and mass stripping by the impact
of the supernova shell in a SN la on the secondaries. For the 24 (M-wD,i = O.8M0) and
71 (MwD,i = I.2M0) WD progenitor binaries shown in Fig. 3, we c alculate the rnass los s
fraction of the secondary star following the semianalytical method of IWheeler et al.l (Il975l ):
the total ejected mass fraction is given by ?7ej = r/st + Ar^cv, where r/st and 77ev are the stripped
and evaporated mass fraction respectively, A is an adjusted parameter we assume to include
the uncertainties in estimating r]eM- Here we take the mass Msn =.l-4A f(n and the velocity
t;sn = 8500 kms~i for the SN ejecta (IMarietta. Burrows fc Frvxelll 120001 ). and A = 0.5. We
find that the subgiant secondary (of mass 0.3 — 2.6M0) loses ~ 9% — 28% of its mass after
the SN. Generally the greater mass or binary separation, the smaller mass ejection fraction.
Figure 4 shows the distribution of the secondary stars in the HR diagram just before and
after the SN explosions. The latter could be compared with and testified by f uture obser-
vatio ns of the companions of SNe la in young, nearby supernova remnants (e.g. ICanal et al.
2OOII).
4. Discussion and Summary
Recently iMannucci. Delia Valle fc Panagial (120061 ) , on the basis of observational argu-
ments, suggest that there is a bimodal delay time distribution, in which about half of SNe
la explode soon after their stellar birth, in a time ~ 10^ yr, while the remaining half have
a much wider distribution, well described by an exponential function with a decay time
of about 3 Gyr. In the traditional SD/Ch model the rapid mass transfer required limits
the donor stars to be more massive than ~ 2Mq (ILi fc van den Heuvell 119971 : iLanger et al.
2OOOI : iHan fc Podsiadlowskil I2OO4J ). The innovation in this work is taking into account the
effect of the CB disk on the mass transfer process. Our calculations indicate that the CB
disk can significantly enhance the mass transfer rates, especially for systems with low-mass
(< 2M0) donor stars, to trigger SNe la explosions. In particular, these progenitor systems
are consistent with the time delay of ~ 1 — 3 Gyr between the formation of the progenitors
and the SNe la. This scenario could also explain the existence of the peculiar SSS contain-
ing rapidly accreting white dwarf and very low-mass donor star like IE 0035.4—7230 (e.g.
-9-
Fig. 3. — Distribution of the progenitor systems of SNe la in the Md,i — -Porb,i diagram. The
bias and pane shading denote the distribution area of WD binaries with a WD of initial
mass 1.2Mq and O.SMq, respectively.
-10-
®
O
3.8 3.7
LogT^,(K)
3.5
Fig. 4. — Distribution of the donor stars in the HR diagram when and after SNe la explosions
occur. The triangles and circles denote the donor stars accompanied by an initial WD of
1.2Mq and O.SM©, while the sohd and open signs represent the cases of when and after SNe
la explosions occur, respectively.
- 11 -
Taam fc Spruitll200lh .
Obviously there exist many uncertainties in the CB disk-driven mass accretion onto
WDs in this work. First, the existence of the CB disk in supersoft X-ray bi naries needs to
be confirmed or disp roved by future infrared observations (IDubus et aLll2004l ). as in GG Tau
( jRoddier et aLlll996l ). Second, the calculated mass transfer processes depend sensitively on
the adopted value of S, which is poorly known, and likely to change with time. This makes it
difficult to estimate the contribution of the such binaries to SNe la. Third, we do not include
exceed 1AM
Howell et al.
2006)
Uenishi et al.
2003;
Yoon & Laneer
2005
progenitor scenario are more compatible with observations, we need to investigate carefully
various mass transfer processes in WD binaries.
This research was supported by Natural Science Foundation of China under grant num-
ber 10573010, 10573009, and by the Graduate Scientific Research Foundation of Jiangsu
Province, China.
REFERENCES
Barris, B. J., & Tonry, J. L. 2006, ApJ, 637, 427
Belczynski, K., Bulik, T., & Ruiter, A. 2005, ApJ, 629, 915
Branch, D., Livio, M., Yungelson, L. R., Boffi, F. R., & Baron, E. 1995, PASP, 107, 717
Canal, R., Mendez, J., & Ruiz-Lapuente, P. 2001, ApJ, 550, L53
Cannizzo, J. K. 1993, in Accretion Disks in Compact Stellar Systems, ed. J. C. Wheeler
(Singapore: World ScientiDc), 6
Chen, W.-C, Li, X.-D., & Qian, S.-B. 2006, ApJ, 649, 973
Chen, W.-C, & Li, X.-D. 2006, MNRAS, 373, 305
Dubus, G., Campbell, R., Kern, B., Taam, R. E., & Spruit, H. C. 2004, MNRAS, 349, 869
Eggleton, P. P. 1971, MNRAS, 151, 351
Eggleton, P. P. 1972, MNRAS, 156, 361
- 12 -
Forster, F., Wolf, C, Podsiadlowski, Ph., Han, Z. 2006, MNRAS, 368, 1893
Gal- Yam, A. & Maoz, D. 2004, MNRAS, 347, 942
Gansicke, B. T., van Teeseling, A., Beuermann, K., & de Martino, D. 1998, A&A, 333, 163
Hachisu, 1., Kato, M., & Nomoto, K. 1996, ApJ, 470, L97
Hachisu, I., Kato, M., Nomoto, K., & Umeda, H. 1999, ApJ, 519, 314
Han, Z., Podsiadlowski, P., & Eggleton, P. P. 1994, MNRAS, 270, 121
Han, Z., & Podsiadlowski, Ph. 2004, MNRAS, 350, 1301
Hoyle, F., & Fowler, W. A. 1960, ApJ, 132, 565
Howell, D. A. et al. 2006, Nature, 443, 308
Hutchings, J. B. et al. 1001, A J, 122, 1572
Iben, 1. Jr., & Tutukov, A. V. 1984, ApJS, 54, 335
Kahabka, P. & van den Heuvel, E. P. J. 1997, ARA&A, 35, 69
Kato, M., Saio, H., & Hachisu, I. 1989, ApJ, 470, L97
Kovetz, A., & Prialnik, D. 1994, ApJ, 424, 319
Kuduz, M., Reinsch, K., Beuermann, K. & Kube, J. 2002, in The Physics of Cataclysmic
Variables and Related Objects, ASP Conference Proceedings (ed. B. T. Gansicke, K.
Beuermann, and K. Reinsch, San Francisco: Astronomical Society of the Pacific), vol.
261. p. 641
Langer, N., Deutschmann, A., Wellstein, S., & Hoflich, P. 2000, A&A, 362, 1046
Li, X. -D., & van den Heuvel, E. P. J. 1997, A&A, 322, L9
Long, K. S., Helfand, A., & Grabelsky, D. A. 1981, ApJ, 248, 925
Mannucci F., Delia Valle M., Panagia N., 2006, MNRAS, 370, 773
Marietta, E., Burrows, D. J., & Fryxell, B. 2000, ApJS, 128, 615
Nomoto, K. 1982a, ApJ, 253, 798
Nomoto, K. 1982b, ApJ, 257, 780
-13-
Nomoto, K., Iben, I. 1985, ApJ, 297, 531
Perlmutter, S., et al. 1999, ApJ, 517, 565
Pols, O., Tout, C. A., Eggleton, P. P., & Han, Z. 1995, MNRAS, 274, 964
Rappaport, S., Di Stefano, R., Smith, J. D. 1994, ApJ, 426, 492
Riess, A. G. et al. 1998, AJ, 116, 1009
Riess, A. G. et al. 2004, ApJ, 607, 665
Roddier, C, Roddier, F., Northcott, M. J., Graves, J. E., & Jim, K. 1996, ApJ, 463, 326
Saio, H., Nomoto, K. 1985, A&A, 150, L21
Shakura, N. I., & Sunyaev, R. A. 1973, A&A, 24, 337
Spruit, H. C, & Taam, R. E. 2001, ApJ, 548, 900
Strolger, L. -G., et al. 2004, ApJ, 613, 200
Strolger, L.-G. et al., 2005, ApJ, 635, 1370
Taam, R. E., & Spruit, H. C. 2001, ApJ, 561, 329
Timmes, F. X., Woosley, S. E., & Taam, R. E. 1994, ApJ, 420, 348
Uenishi, T., Nomoto, K., & Hachisu, I. 2003, ApJ, 595, 1094
van den Heuvel, E.P.J. , & de Loore, C. 1973, A&A, 25, 387
van den Hcuvcl, E. P. J., Bhattacharya, D., Nomoto, K., & Rappaport, S. A. 1992, A&A,
262, 97
van den Heuvel, E.P.J. 1994 in Interacting Binaries (Saas-Fee 22), Shore, S. N., et al., eds.,
p263
van Teeseling, A., & King, A. R. 1998, A&A, 338, 957
Webbink, R. F. 1984, ApJ, 277, 355
Wheeler, J. C., lecar, M., & McKee, C. F. 1975, ApJ, 200, 145
Woosley, S. E., & Weaver, T. A. 1986, ARA&A, 24, 205
Yoon, S. -C., & Langer, N. 2005, A&A, 435, 967
- 14-
Yungelson, L. & Livio, M. 1998, ApJ, 497, 168
This preprint was prepared with the A AS IM^jX macros v5.2.
