The Formation of Low-Mass Double White Dwarfs through an Initial Phase of Stable Non-Conservative Mass Transfer

The Formation of Low-Mass Double White Dwarfs through an Initial Phase of Stable Non-Conservative Mass Transfer

T. E. Woods    N. Ivanova    M. van der Sluys    S. Chaichenets

Although many double white dwarfs (DWDs) have been observed, the evolutionary channel by which they are formed from low-mass/long-period red-giant–main-sequence (RG-MS) binaries remains uncertain. The canonical explanations involve some variant of double common-envelope (CE) evolution, however it has been found that such a mechanism cannot produce the observed distribution. We present a model for the initial episode of mass transfer (MT) in RG-MS binaries, and demonstrate that their evolution into double white dwarfs need not arise through a double-CE process, as long as the initial primary’s core mass () does not exceed . Instead, the first episode of dramatic mass loss may be stable, non-conservative MT. We find a lower bound on the fraction of transferred mass that must be lost from the system in order to provide for MT, and demonstrate the feasibility of this channel in producing observed low-mass (with ) DWD systems.

Double White Dwarfs, Binary Evolution


address=University of Alberta, 1322 89 Ave, Edmonton, AB, T6G 2E7, Canada,, thanks=This work was commissioned by the AIP

address=University of Alberta, 1322 89 Ave, Edmonton, AB, T6G 2E7, Canada,,

address=University of Alberta, 1322 89 Ave, Edmonton, AB, T6G 2E7, Canada, altaddress=Radboud University Nijmegen, P.O. Box 9010 NL-6500 GL Nijmegen, The Netherlands ,, homepage= sluys/

address=University of Alberta, 1322 89 Ave, Edmonton, AB, T6G 2E7, Canada, email=,

In a CE phase, both components of a binary become engulfed in the envelope of the more evolved star. The envelope is then expelled at the cost of dramatically shrinking the orbit [Paczynski(1976), Webbink(1984)]. It has been shown [Nelemans et al.(2000), van der Sluys et al.(2006)] that consideration of two subsequent CE events in which each phase obeys the standard -prescription (based on energy considerations, with as an efficiency factor), as well as stable conservative MT followed by a single such CE event, failed to explain the observed DWD systems, proving inconsistent with the observed distribution of periods and mass ratios (mean mass ratio approximately 1). An alternative description considering angular momentum balance (the -formalism) was proposed [Nelemans and Tout(2005)]; however, this does not allow for strong constraints on the evolution of the system [Webbink(2008), Zorotovic et al.(2010)], proving unstable against small perturbations [Woods et al.(2010)]. We revisit the possibility that the first phase of MT occurs as stable, non-conservative Roche-lobe overflow (RLOF), and demonstrate that this channel can provide an adequate model for a number of observed DWDs when considering progenitor systems in which both components are approximately 1 - 1.3 —significantly smaller than previously considered [Han(1998)].

0.1 Stability Criteria

In order for MT to proceed in a stable fashion, we require that the radius of the donor shrinks faster due to mass loss than its Roche lobe does (that is, given ). The response of the Roche lobe is given by:


(with donor mass , accretor mass , Roche radius , and orbital separation ) where depends on the system geometry, and the other terms depend on the masses as well as the conservation factor (, the fraction of mass lost by that is accreted onto ). A low mass star on the red-giant branch (RGB) can be well modelled as a condensed polytrope, and from this we can approximate the adiabatic response of the donor’s radius as a simple function of the core mass. This implies that for given system parameters (, , , and to some extent ) we have a maximum fraction of mass lost by the primary which can be accreted onto the secondary while maintaining stability. As and both depend monotonically on mass loss, we need only consider stability at the onset of MT. For this work we adopt , as it provides for generally stable MT given the above condition and our initial parameters.

0.2 Subsequent Evolution

Using a variant of Eggleton’s detailed stellar-evolution code ev (referred to as STARS) [Eggleton(1971), Eggleton and Kiseleva-Eggleton(2002), Stancliffe and Glebbeek(2008)], we evolve a set of 1.2+1.1 systems with a range of initial periods through RLOF of the primary on the RGB. Our simulations demonstrate that MT proceeds stably until the envelope has been almost entirely stripped. The subsequent RLOF of the (original) secondary can then be expected to take place in an unstable manner, as the mass ratio has been significantly increased. Given an ensuing CE (in which the central concentration of the donor envelope can be calculated from our stellar models) with , we find that our results fit within observations nicely, however further studies will be needed in order to determine if we can fully reproduce the observed distribution (in particular, those systems with q ¡ 1). This, as well as higher mass systems, may require some intermediate mechanism in the initial phase, removing the majority of the envelope with little orbital decay, as suggested by [Nelemans et al.(2000), Nelemans and Tout(2005)]. Nonetheless, the stable non-conservative MT + CE channel provides an elegant solution for an initial “envelope-removal event” with a modest increase in period.

Figure 1: Circles indicate (taken as the total helium mass) and orbital period after the initial phase of stable MT. Stars indicate the final (core) mass of the secondary and orbital period after the second, unstable episode of mass loss (a CE-phase with ). Solid lines connect the two relevant points for each system. Dotted lines mark evolution of total donor mass during initial MT phase. Triangles mark observed parameters (orbital period and mass of the inferred secondary, taken from [van der Sluys et al.(2006)]).
We thank P. P. Eggleton and E. Glebbeek for making their binary-evolution code available to us, as well as Gijs Nelemans for helpful discussion. NI acknowledges support from the Canada Research Chairs and NSERC. MvdS acknowledges support from a CITA National Fellowship to the University of Alberta.


  • [Paczynski(1976)] B. Paczynski, “Common Envelope Binaries,” in Structure and Evolution of Close Binary Systems, edited by P. Eggleton, S. Mitton, & J. Whelan, 1976, vol. 73 of IAU Symposium, p. 75.
  • [Webbink(1984)] R. F. Webbink, ApJ 277, 355–360 (1984).
  • [Nelemans et al.(2000)] G. Nelemans, F. Verbunt, L. R. Yungelson, and S. F. Portegies Zwart, A&A 360, 1011–1018 (2000).
  • [van der Sluys et al.(2006)] M. V. van der Sluys, F. Verbunt, and O. R. Pols, A&A 460, 209–228 (2006).
  • [Nelemans and Tout(2005)] G. Nelemans, and C. A. Tout, MNRAS 356, 753–764 (2005).
  • [Webbink(2008)] R. F. Webbink, “Common Envelope Evolution Redux,” in Astrophysics and Space Science Library, edited by E. F. Milone, D. A. Leahy, & D. W. Hobill, 2008, vol. 352 of Astrophysics and Space Science Library, p. 233.
  • [Zorotovic et al.(2010)] M. Zorotovic, M. R. Schreiber, B. T. Gänsicke, and A. Nebot Gómez-Morán (2010), 1006.1621.
  • [Woods et al.(2010)] T. E. Woods, N. Ivanova, M. van der Sluys, and S. Chaichenets, in preparation (2010).
  • [Han(1998)] Z. Han, MNRAS 296, 1019–1040 (1998).
  • [Eggleton(1971)] P. P. Eggleton, MNRAS 151, 351 (1971).
  • [Eggleton and Kiseleva-Eggleton(2002)] P. P. Eggleton, and L. Kiseleva-Eggleton, ApJ 575, 461–473 (2002).
  • [Stancliffe and Glebbeek(2008)] R. J. Stancliffe, and E. Glebbeek, MNRAS 389, 1828–1838 (2008).
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description