(Note that the above equation does not even include reactions such as
X + p ® Y + n, X + n ® Y + a, etc.) Much of the
work in setting up the network of equations is in deciding which reactions
are important, and which can be neglected.)
At oxygen burning temperatures, nuclei such as §32, 31 P, and 30 Si
are all destroyed in favor of 28 Si, which is the most tightly bound
nucleus in the intermediate mass range. Once the temperature becomes high
enough to photodisintegrate 28 Si (T9 ~ 3), a host of
(a,g), (p,g), (n,g) reactions (and their inverses)
begin simultaneously. This is the Silicon Burning phase. Some of the
fastest reactions run to equilibrium, but most reactions will not reach this
stage. Thus, the results of silicon burning depend on how far towards
completion the reactions run. During this time, there is also a slow
leakage of nuclei from the intermediate-mass region to the iron group
(which are the most tightly bound nuclei).
As silicon burning nears completion, the nuclei come closer and closer to
establishing nuclear statistical equilibrium. In this case, the nuclear Saha
equation defines the abundance of each species. This has a very simple
form. Consider that, from the Saha equation, the abundance ratio of species
\Am1,Z to A,Z is
|
| |
|
|
= |
Gn G(\Am1,Z)
G(A,Z)
|
|
æ ç
è
|
|
2 pmk T
h2
|
|
ö ÷
ø
|
3/2
|
exp |
æ ç
è
|
- |
Q1
k T
|
|
ö ÷
ø
|
|
| | |
|
= |
2 G(\Am1,Z)
G(A,Z)
|
Q |
æ ç
è
|
|
A - 1
A
|
|
ö ÷
ø
|
3/2
|
exp |
æ ç
è
|
- |
Q1
k T
|
|
ö ÷
ø
|
|
|
| |
|
where G is the nuclear partition function (2 for the neutron), Q1
is the binding energy of \Am1,Z, and
Q = (2 pma k T / h2)3/2 . Similarly, we can write an
expression for the ratio of species \Am2,\Zm1 in terms of
\Am1,Z
|
|
N(\Am2,\Zm1) Np
N(\Am1,Z)
|
= |
2 G(\Am2,\Zm1
G(\Am1,Z)
|
Q |
æ ç
è
|
|
A - 2
A - 1
|
|
ö ÷
ø
|
3/2
|
exp |
æ ç
è
|
- |
Q2
k T
|
|
ö ÷
ø
|
|
|
where Q2 is the binding energy of this new species.
If you multiply these two equations, then you obtain a relation between
A,Z, and \Am2,\Zm1
|
|
|
|
N(\Am2,\Zm1) Nn Np
N(A,Z)
|
= |
|
|
|
22 G(\Am2,\Zm1)
G(A,Z)
|
Q2 |
æ ç
è
|
|
A - 2
A
|
|
ö ÷
ø
|
3/2
|
|
| | |
|
exp |
æ ç
è
|
- |
Q1 + Q2
k T
|
|
ö ÷
ø
|
|
|
|
| |
|
This progression can be taken all the way to Z = A = 1, with the result
|
N(A,Z) = |
G(A,Z)
2A
|
A3/2 NpZ NnA-Z Q1-A exp |
æ ç
è
|
- |
Q
k T
|
|
ö ÷
ø
|
(PD5) |
|
where
|
Q = ( Z mH + (A-Z) Mn - M(A,Z) ) c2 (PD6) |
|
is the binding energy of the nucleus A,Z. Note the resultant abundances
are a function only of the temperature, the binding energy of the species,
and the number density of protons and neutrons. Note also that the latter
two quantities are constrained. Since the density of the star is known,
we have from (CC1)
|
NA r = Np + Nn + |
å
| Ni Ai (PD7) |
|
Moreover, if we parameterize the system with the quantity [`Z] / [`N],
i.e., the average proton to neutron ratio, then
Thus (PD5), (PD7), and (PD8) form a set of three equations with three
unknowns, and r, T, and [`Z] / [`N] described the entire set
of nuclear abundances.
Unfortunately, the analysis above is not completely accurate, since
it fails to account for reactions that involve the weak nuclear force,
i.e., beta decays. Since the neutrinos from these decays leave the star,
there is no inverse reaction associated with beta-decays, and no
equilibrium condition. To first order, however, the effect of beta-decays
can be modeled through their impact on [`Z] / [`N]. Thus,
|
|
æ ç ç ç
ç ç è
|
|
|
|
ö ÷ ÷ ÷
÷ ÷ ø
|
t + Dt
|
» |
æ ç ç ç
ç ç è
|
|
|
|
ö ÷ ÷ ÷
÷ ÷ ø
|
t
|
+ Dt |
¶
¶t
|
|
æ ç ç ç
ç ç è
|
|
|
|
ö ÷ ÷ ÷
÷ ÷ ø
|
t
|
(PD9) |
|
The computation of beta-decay rates is a bit tricker than normal, since
many of the decays come from excited nuclear states, and the high
densities increase the probability of electron captures by a factor of
~ 100 over terrestrial conditions. However, these reactions are
important, in that a small change in [`Z] / [`N] greatly affects
the composition of the iron group. Clearly, the faster the reactions
proceed, the less time there is for beta-decay, and the closer
[`Z] / [`N] will be to 1.
The question of beta-decays and the value of [`Z] / [`N] has
important consequences for supernova calculations and for cosmic
abundance determinations. For example, consider silicon burning,
where 28 Si is converted to iron peak elements. Let's examine the
case where the two dominant resultant species are 54 Fe and 56 Ni.
These nuclei are in statistical equilibrium via the reactions
®
¬
If 28 Si burns to 56 Ni, then the reaction is exothermic with an
net energy release of 10.9 MeV and the star can continue burning; on the
other hand, if the result is 54 Fe + 2 p, then -1.3 MeV is lost
in the reaction, and the star will collapse.
From (PD4), the ratio between 54 Fe and 56 Ni is
|
|
N(54 Fe)
N(56 Ni)
|
Np = 22 |
G(54 Fe)
G(56 Ni)
|
|
æ ç
è
|
|
54
56
|
|
ö ÷
ø
|
3/2
|
Q2 exp |
ì í
î
|
- |
Q(55 Co) + Q(56 Ni)
k T
|
|
ü ý
þ
|
|
|
The ratios of the partition functions for these two species is ~ 1,
so plugging in the numbers yields
|
|
N(54 Fe)
N(56 Ni)
|
Np = T93 1068.13 - 62.09/T9 (PD10) |
|
We can now substitute for Np using (PD8); if [`Z] / [`N] = 1
|
|
|
= 1 » |
26 N(54 Fe) + 28 N(56 Ni) + Np
28 N(54 Fe) + 28 N(56 Ni)
|
Þ Np » 2 N(54 Fe) |
|
Substituting this in (PD10) gives
|
|
N(54 Fe)3
N(56 Ni)
|
= |
NA2 X(54 Fe)3 A(56 Ni)
X(56 Ni) A(54 Fe)3
|
r2 = T93 1067.53 - 62.09/T9 |
|
Thus, for X(54 Fe) = X(56 Ni),
r2 » T93 1024.09 - 62.09 / T9
Figure
The graph for [`Z] / [`N] = 1
shows that at low temperatures, 28 Si will fuse to 56 Ni
(which then beta-decays to 56 Fe); this reaction provides energy to
supports the star. However, if 28 Si burns at a high temperature,
54 Fe will result, and the star will collapse.
This process of nucleosynthesis is call the e-process, for
equilibrium process.
In general, the peak iron element will have approximately the same
proton to neutron ratio as given by [`Z] / [`N]. Thus,
In nature, the most abundant isotope in the iron group is 56 Fe, with
26 protons and 30 neutrons. This suggests that the e-process usually either
occurs very rapidly at low temperatures, so that 56 Fe is produced via
the decay of 56 Ni, or proceeds very slowly so that beta-decays make
[`Z] / [`N] » 0.87.
File translated from TEX by TTH, version 1.98.
On 3 Mar 1999, 12:57.