ЯДЕРНАЯ ФИЗИКА, 2010, том 73, № 4, с. 630-634
= ЭЛЕМЕНТАРНЫЕ ЧАСТИЦЫ И ПОЛЯ
MOST LUMINOUS SUPERNOVAE PRODUCED BY SHOCKS
© 2010 S. I. Blinnikov*
Institute for Theoretical and Experimental Physics, Moscow, Russia Received August 11,2009
The extremely luminous supernova SN2006gy is explained as some other peculiar supernovae: light is produced by a radiative shock propagating in a dense circumstellar envelope. This envelope is formed by a previous weak explosion at a stage of hydrodynamic instability due to creation of electron—positron pairs in stellar interiors. The problems in the theory and observations of supernovae created by multiple explosions are briefly reviewed.
1. INTRODUCTION: SUPERNOVA TYPES
A standard supernova (SN) explosion has kinetic energy -1051 erg, but only around ~1049 erg are emitted in photons during the first year of a normal SN. The discovery of SN2006gy [1,2] demonstrates that some SN events produce 10 or even 100 times more visible photons than other, already powerful explosions. The anomalously high power of the emission of SN2006gy demands an explanation.
Astronomical classification divides SN into type I (no hydrogen spectral lines near maximum light) and type II (with hydrogen). One may subdivide the physics of SN explosions also into two groups, but they do not coincide with spectral types.
We believe that SNeII are born when a giant star with a hydrogen-rich atmosphere has a powerful explosion in its core. This explosion may be a result of a catastrophic collapse of the stellar core. The details of the explosion of core-collapsing SNe are still not clear, in spite of many important results obtained by workers in this field. If the hydrogen envelope is lost prior to the core collapse we observe a type I SN (more rigorously, SNIb or SNIc subtypes). A pure thermonuclear explosion of a white dwarf produces an SNIa.
In spite of a huge energy of SN explosions (-1051 erg) we would not see any powerful burst of photons if all this energy were stored in pure kinetic motion of baryons injected into vacuum. In addition to the mechanism of explosion we need to find means for production of light, i.e., of the specific entropy of exploded matter, and there are different ways to achieve this. E.g., the light of SNeII is the manifestation of entropy produced during a short period (hours to days) of the shock propagation in the body of the presupernova star. The heat stored in the
E-mail: sergei.blinnikov@itep.ru
matter diffuses out during weeks and months after the explosion.
Remarkably, the details of the explosion are not important for explaining the light curves of many SNII that are born from giant stars that retain their huge hydrogen envelopes. Successful SNII light models were already constructed four decades ago (largely by the work of the Soviet group in Moscow and Riga [3]) due to this insensitivity to details.
If the hydrogen is lost by a massive star, then we have an SNIb/c. The same (still unknown in details) core-collapse mechanism may lead to their explosions, as in SNeII, however, the light is due now to entropy produced by radioactivity: the decays 56 Ni ^ ^ 56Co ^ 56Fe which lead to a slower heating of ejecta. This way of producing light is most important for SNeI of all subtypes (including the thermonuclear SNIa). However, some contribution of radioactivity is clearly present in late light curves of type II SN as well, and for SN1987A in LMC it was important already before its maximum light, a month after the explosion.
If the radioactive mechanism were responsible for the light of the most luminous SN2006gy, then the amount of 56Ni had to be higher than 10MQ [46]. This immediately implies a very large mass for the presupernova star, more than 100M©. More important, this implies unrealistically huge explosion energy, (50-80) x 1051 erg!
In rare cases the SNeII are produced by thermonuclear explosions, not by a collapse, inside hydrogen envelopes. This may be the case for SNeIIn (type II with narrow spectral lines), and especially, SN2006gy. A SN of type II shines for weeks and months because of the heat stored in its body (the shock dies quickly), while in an SNIIn the heat (entropy) may be replenished by the shock living several months and propagating in a dense circumstellar
MOST LUMINOUS SUPERNOVAE Four kinds of deaths for nonrotating massive stars
Initial mass (M©) He core (MQ) Supernova mechanism
10 < M < 95 2 < M < 40 Fe core collapse to a neutron star or a black hole
95 < M < 130 40 < M < 60 Pulsational pair-instability followed by Fe core collapse
130 < M < 260 60 < M < 137 Pair-instability supernova
260 < M 137 < M Black hole. Possible GRB?
envelope. The powerful visible light of SN2006gy can be easily explained by a radiative shock born due to a collision of SN ejecta with a dense cloud formed by a weak explosion some years before the SN.
2. PAIR INSTABILITY SUPERNOVAE
The models with a long living radiative shock running in a dense circumstellar envelope were invoked earlier [7, 8] to explain the unusual properties of other powerful SN with narrow emission lines in their spectra, the subtype SNIIn. The brilliant SN2006gy also belongs to the same IIn class. How to make the dense envelope around the SN? A natural way to this is provided by violent pulsations — weak explosions — of so-called pair instability supernovae.
One can delineate four kinds of deaths for massive stars [9], see the table. There is some uncertainty about the exact values due to mass-loss, rotation, etc. Nevertheless, stars with M > 100M© do exist, and they can have powerful explosions in their oxygen cores which experience instability due to the creation of large numbers of electron—positron pairs.
The word "instability" refers here to hydrodynamics, to mechanical equilibrium, not to the process of pair creation which is quite stable and reversible in a thermodynamic sense in stellar interiors. A massive star loses its mechanical stability when pairs are being created because the adiabatic exponent 7 goes down at T ~ 0.1 MeV belowthe critical value 4/3, see Fig. 1: the work of contraction is spent in creating new particles and not for raising the momenta of particles that already exist and which provide for the equilibrium pressure.
We can easily estimate the path that leads the star into the domain of the pair-creation instability. From the virial theorem, for a star of mass M and radius R, omitting all coefficients of order unity, PcV ~ PcR3 ~ ~ GnM2/R. Hence, the pressure Pc in the center is Pc ~ GnM2/R4, while the density pc ^ M/R3,
and they are related as Pc ~ gnm2/3p4/3. So, we have found a simple explanation why the value 4/3 is critical for 7: when 7 is too small, the physical pressure during contraction is unable to match Pc
needed for mechanical equilibrium, and the star must start collapsing. The collapse may be stopped by an explosion of a nuclear fuel (oxygen in our case). The entropy grows, hence P grows (of course, adiabatic exponent 7 is irrelevant in this growth) and in place of the contraction we see a violent expansion of the whole star. If the energy of nuclear burning is not high enough to disrupt the whole star, then a loss of a large fraction of its mass is possible. Then the contraction begins again, and new explosions may follow. This is the fate of stars in the second line of the table. We will discuss it in the next section.
If we have a classical ideal plasma with P = = RpT/^, where R is the universal gas constant, and H — mean molecular mass, we find
Tc ~ GnM2/3pl/3^/R.
For a relativistic addition of aT4 to P, the law Tc a
a pc/3 is the same (but the coefficient is a bit different and dependence on M is weaker). The right panel of Fig. 1 shows how a massive star follows the law
Tc a pl/3 until entering the domain of pair-instability.
There is a difference of "classical" pair-instability supernovae for M > 130M© (third line in the table) with the pulsational pair-instability SN. In case of M > 130M© there are no pulsations, visible in right panel of Fig. 1, there is one strong explosion. We had already computed the light curves of pair-instability SNe for M > 130M© some years ago (with S. Woosley and A. Heger), but we do not like them in the case of SN2006gy because we do not see the evidence for a tremendously high explosion energy, (50—80) x 1051 erg, produced in that case.
If the explosion energy were two orders of magnitude higher than for normal SNe, then it should be seen in very broad spectral lines. What we see in SN2006gy is different: it has narrow lines (hence, it is type IIn) superimposed on moderately broad emission component (~5 x 103 km/s [2, 12]). There is no sign of huge kinetic energy in this event.
632
BLINNIKOV
T, 109 K log(pc/g cm-3)
Fig. 1. (Left) adiabatic exponent in the low density asymptotic limit taking into account pair creation [10], see also [11]. (Right) evolution track for one of the models [9] with initial M ~ 103M© (solid line). The approximate boundary of pair-instability (where j < 4/3) is shown by the dashed line.
t, d
Fig. 2. (Left) SN1994W structure [8] prototypical for SNeIIn. (Right) light curve models for SN2006gy. Dots -observations [2], the last observed point is from [13]. The solid curve is the model discussed in the text (with numerical resolution higher than in [9]), and the dashed one where the velocity of all the ejecta has been multiplied by 2 (hence, an artificial increase in the explosion energy to 2.9 x 1051 erg). The dotted curve is for the doubled density in ejecta.
2.1. Multiple Ejections in SNIIn
Supernovae of type IIn are among the most powerful transients in visible light. No radioactive material is needed to explain their light during the first several months: the light is produced by a long living radiative shock in a dense circumstellar envelope. This is the main difference with standard SNeII where the shock breaks out into rarefied interstellar medium and di
Для дальнейшего прочтения статьи необходимо приобрести полный текст. Статьи высылаются в формате PDF на указанную при оплате почту. Время доставки составляет менее 10 минут. Стоимость одной статьи — 150 рублей.