SUPPLEMENTS
ELSEVIER NuclearPhysicsB (Proc. Suppl.) 110 (2002) 137-143
Big Bang Nucleosynt hesis
A.D. Dolgov B b
a INFN, sezione di Ferrara, Via Paradii, 12 - 44100 Ferrara, Italy
b ITEP, B. Cheremushkinskaya 25, Moscow, 117259, Russia
A review of Big Bang Nucleosynthesis (BBN) is presented. Observations of deuterium and helium-4 are dis-
cussed. Some BBN restrictions on non-standard physics, especially on neutrino properties and time-variation of
fundamental constants are given.
1. INTRODUCTION In the next section physics of BBN and essen-
tial parameters and inputs are described. In sec-
Big Bang Nucleosynthesis (BBN) is known to tion 3 observational data are analyzed (looking
be one of three solid pillars on which the Standard from outside by a non-expert). In section 4 mod-
Cosmological Model (SCM) stands. The other ifications of the standard scenario are discussed.
two include General Relativity (GR) and Cos- Conclusion is presented in the last section 5.
mic Microwave Background Radiation (CMBR).
An agreement of BBN calculations of light ek 2. PHYSICS OF BBN
ment abundances with observations presents the
strongest proof in favor of the statement that 12- Physical processes which were essential for cos-
14 billion years ago the universe was indeed hot mological creation of light elements took place at
with the temperatures in MeV range and that the the temperatures, T, in the range of a few MeV
entropy per baryon is huge, about log. down to 60-70 keV. The corresponding time, t,
According to the theory, light elements 2H, interval was from a fewxO.1 set up to lo3 sec. At
3He, 4He, and 7Li have been created in the early this time the universe was dominated by relativis-
universe during first few hundred seconds of her tic matter and the rate of cosmological cooling
existence. The abundances of these elements span was determined by the expression:
9 orders in magnitude and are in excellent, good,
or reasonable agreement with the observational (&) ($g2 =0.74 f!g2 (1)
data, depending upon the moment when the com-
parison of theory with the data is taken and upon where g+ is the effective number of particle species
the personal point of view of a researcher. in the cosmic plasma. In the standard model
The theory of BBN is robust, well defined, and the factor g* includes contributions from photons
quite precise. The largest uncertainty is intro- equal 2, from e*-pairs equal 7/2, and the contri-
duced by the values of the cross-sections of nu- bution from three neutrino flavors equal 3 .7/4.
clear reactions. Theoretical accuracy is better Any additional, non-standard form of energy is
than 0.1% for 4He, better than 10% for 2H and parametrized in terms of effective number of neu-
is about 20-30% for 7Li [1,2]. In all the cases trino species:
theoretical uncertainty is much smaller than ob-
servational precision. Observations of light ele- g* = 10.75 + a (N, - 3)
ments encounter two serious problems: system-
atic errors and evolutionary effects. We will dis- Since the cosmological cooling rate depends upon
cuss them below. g+ it is clear that BBN is sensitive to any form
of matter/energy present in the primeval plasma
0920-5632/02/$ - see front matter0 2002 Elsevier Science B.V All rights reserved.
PII SO920-5632(02)01470-6
138 A. D. Dolgoo /Nuclear Physics B (Proc. Suppl.) 1 IO (2002) 13 7-l 43
in the temperature interval (- MeV) - 0.06 keV. be (a few) x 10m5 by number. A strong sensitivity
This effect was noticed in refs. [3]; detailed calcu- of deuterium abundance to the baryon-to-photon
lations of the effect were pioneered in papers [4]. ratio n makes this element a very convenient tool
The first step in creation of light elements is to measure the baryonic charge of the universe.
“preparation” of neutrons. Their number density That’s why primordial deuterium is called “bary-
is determined by the reactions: ometer”, the term suggested by David Schramm.
An absence of a (quasi)stable nuclei with A = 5
n+e+ * P+ve, inhibits production of 7Li, because the latter
n + ye * p+e- (3) should be produced by fusion of two lighter nu-
clei and not in a sequence of reactions of free pro-
Since the reaction rate is proportional to T5 and
tons or neutrons with nuclei. Correspondingly
the expansion rate is H N T* thermal equilibrium
the fraction of 7Li is very small, (a few) x lo-lo.
is maintained at high temperatures and nIpratio
The abundances of light elements as functions
follows the equilibrium curve:
of the baryon number density, r)ic = lO”ns/n,
n/p = exp[--(6m + CLe)/TI (4) are presented in fig. 1, taken from the review [5].
As we have seen, the abundances of light ele-
where Am = 1.293 MeV is the neutron-proton ments essentially depend upon the following pa-
mass difference and he is the chemical potential rameters:
of electronic neutrinos. In the standard model the
latter is assumed to be vanishingly small. Baryon-to-photon ratio, ~10 = 101OnB/n,.
At T = 0.6 - 0.7 MeV expansion became faster Only two years ago this parameter was de-
than reactions (3) and the n/pratio would tend termined from BBN itself but now there is
to a constant, if not the neutron decay. Be- an independent way to measure it through
cause of the decay the ratio slowly decreases as the spectrum of angular fluctuations of
exp(-t/Tm) with T, = 889 sec. This behavior CMBR. According to the BOOMERANG
lasts approximately till TNS = 65 keV when light and DASI measurements:
element formation abruptly begins and all free
neutrons quickly disappear from the plasma form- RBh2 = 0.021 f 0.004 [6],
ing deuterium, helium, and lithium. The value 0.022+0,.;;; (6)
[717
of TNS is determined by the binding energies of
light nuclei, in particular deuterium, and by the where fiB = PB/Pc& is the fraction of
baryon-to-photon ratio, q = nB/n,. baryon mass density in terms of cosmo-
Practically all neutrons that remained in the logical critical energy density and h =
plasma to the moment when the temperature H/lOOkm/sec/Mpc is the dimensionless
dropped down to TNS ended their lives in 4He Hubble parameter. fl&* = 0.022 cor-
because the latter has the largest binding energy. responds to 710 = 5.7. A higher value
The mass fraction of the primordial 4He should is given by MAXIMA-I measurements [8],
be about 25%. A little deuterium and 3He sur- flBh* = 0.0325f0.007 which disagrees with
vived because they were not able to find a nu- the other groups by two standard devia-
cleon, N = p,n, for further transformation into tions.
4He through the chain of reactions:
Rate of the reactions (3). It is expressed
N + *H -+ (3H, 3He) through neutron life-time which is pretty
(3H, 3He) + N d4 He (5) well known now, TV = 886.7 f 1.9 [9].
As a result the abundance of primordial deu- Total cosmological energy density during
terium would quickly decrease with rising baryon BBN. This quantity is usually parametrized
number density or, better to say, parameter v, as the number of additional neutrino fam-
and its output with respect to hydrogen would ilies, AN = N, - 3. This is a precise
A.D. Dolgov/Nuclear Physics B (Proc. Suppl.) 110 (2002) 137-143 139
tain an upper bound on their magnitude is
BBN. Chemical potentials of v,, and’r+ can
be described by AN, since their role is only
to increase the energy density of relativis-
tic matter at BBN. Element abundances are
much more sensitive to chemical potential
of V, because the latter not only changes the
canonical energy density but also directly
shifts the n/pratio (4).
3. OBSERVATIONAL DATA
3.1. Observations of Deuterium
The present day abundance of deuterium can
be noticeably different from the primordial value
because deuterium can be burnt in stars produc-
ing 3He. Exact evolutionary effects are uncer-
tain. One can rigorously say only that the ob-
served today deuterium presents a lower bound
to the primordial one. For more detail see e.g. re-
view [5]. Fortunately in recent years observation
of deuterium in large 2 low metallicity (< l/50
of the solar) clouds of cold neutral gas (HI) be-
came possible [lO-141. Such clouds are presum-
ably not contaminated by stellar processes and
the observed deuterium may be the primordial
one; see however ref. [15] where this statement is
questioned.
Since the deuterium isotope shift corresponds
Figure 1. The predicted abundances of light ele- to velocity of only (-82) km/set the clouds with a
ments as a function of ~10. simple velocity structure are necessary for reliable
data. In addition to the uncertainty induced by
the unknown velocity structure, ionization correc-
parametrization if the additional energy is tions and a possible “interloper” further increase
in the form of relativistic particles with the the systematic errors. A discussion of these ef-
equation of state p = p/3. However it is fects and a list of references can be found e.g.
not so for matter with a different equa- in [16,17].
tion of state, e.g. nonrelativistic matter or Because of these problems, there are only a
vacuum-like energy. In this case the varia- few measurements of deuterium at high red-shifts
tion of different abundances would be differ- available now. They are summarized in Table 1.
ent from that induced by extra neutrinos. The entries in Table 1 are situated in order of
increasing the (D/H)-values. Entries in columns
4. Neutrino degeneracy. It is assumed usually 3 and 5 correspond to the same system but the re-
that the charge asymmetry of diierent neu- sult was changed to a larger value after its reanal-
trino flavors is vanishmgly small. Thus their ysis in ref. [21] where a complex velocity structure
chemical potentials are zero or negligible. suggested by S II and Hz absorption lines was
Strictly speaking their values are unknown taken into account.
and the best way to determined them or ob- The data seem to be grouped around three cen-
140 A. D. Dolgov/Nuclear Physics B (Proc. Suppl.) I IO (2002) 13 7-I 43
Table 1
Abundance of deuterium in high red-shift Lyman alpha absorption systems
2 3.514 2.0762 3.025 2.536 3.025 3.572 2.504 0.701
Ref. 1181 P71 PJI PO1 WI PI P31 1241
105(D/H) nJ 1.5 1.659~0.35 2.24f0.67 2.54f0.23 3.2f 0.4 3.25 f0.3 4.0 910.65 20 f 5
tral points. The majority of the data are accu- linear function with the slope n = dY/dZ where
mulated around lo5 (D/H) = 3. There are two Y is the observed mass fraction of 4He and 2 is
measurements indicating low deuterium about 1.5 that of metals. Analysis by different groups give
and one showing a huge value, 20 f 5. Taken by significantly different values K = 6.7 f 2.3 [29],
the face value, there is an indication that there are 6.9 f 1.5 [30], and K = 2.4 f 1.0 [31]. Though IC
systems with high, normal, and low deuterium is not accurately determined (or it may have dif-
abundances. If thii is true, it would be a con- ferent values for different systems), the systems
firmation of the model of ref. [25] where spa- where 4He is studied have a low metallicity and
tial variation of primordial abundances are pre- the net effect on primordial abundance of 4He is
dicted. According to this work, two thirds of the not very strong.
sky should have normal abundance of deuterium, The values of primordial mass fraction of 4He,
l/6 should have twice smaller ratio of D/H, and Yp, measured by several different groups are pre-
another l/6 should have about 5 times larger than sented in table 2. The discrepancy between the
normal abundance of deuterium. On the other results can be possibly prescribed to different
hand, as is argued in ref. [26], the line observed treatment of correction factors: 1) ionization cor-
in the z = 0.701 absorption system [24] which in- rection factor (icf) which determines how much
dicates a high deuterium abundance most proba- neutral hydrogen (invisible) is in the object under
bly is not deuterium. This statement is based on scrutiny, 2) temperature correction factor (tcf)
a new measurement of the velocity dispersion of which describes non-uniform temperature distri-
the neutral hydrogen. bution, and 3) density of electrons. A recent
analysis of the correction has been performed in
ref. [34], where it was shown in particular that
3.2. Observations of helium-4
combined icf and tcf would diminish the result
Helium-4 is very strongly binded nuclei so it
by 0.002-0.004. It is worth noticing in this con-
cannot be destroyed by stellar processes but only
nection that the reanalysis of the data of ref. [31]
produced in stars together with other heavier el-
in the paper [33] where icf has been taken into ac-
ements, “metals”, e.g. oxygen (0) and nitrogen
count, has led to a noticeably smaller result (see
(N) - all elements heavier than 4He are called
table 2).
“metals” in astronomy. Thus the observed mass
Even the largest mass fraction of 4He pre
fraction of 4He, should be larger than the pri-
sented in table 2 corresponds to a rather low
mordial one, Yp. 4He is observed in hot ionized
baryon number density, ~10 M 4.2 which is no-
gas regions (H II) in emission optical recombi-
ticeably smaller than the value determined from
nation lines. In contrast to deuterium 4He has
CMBR (6). Possibly it means that systematic
been observed only at relatively small distances,
uncertainties and correction factors are underesti-
maximum, I = 0.045 [27]. Thii regions are con-
mated and more work is necessary to obtain more
taminated by stellar processes and to infer pri-
precise results. This problem is discussed in the
mordial abundance from the data one should ex-
recent paper [35]. It seems that at the present
trapolate to zero metallicity, i.e. to zero abun-
time the accuracy in determination of primordial
dances of 0 or N. A detailed discussion of these
mass fraction of 4He is not sufficiently good.
issues can be found e.g. in the book [28]. The ex-
trapolation to zero metallicity is realized by the
A.D. DolgooINuclear Physics B (Proc. Suppl.) 110 (2002) 137-143 141
Table 2
Primordial mass fraction of 4He according to the results of different groups.
Ref. [29] [301 [321 ]311 1331
Y, 0.228 f 0.005 0.234 f 0.002 0.2348 f 0.0025 0.2443 f 0.0015 0.238 f 0.003
4. BIG BANG NUCLEOSYNTHESIS 0.4) . 10-s. Here & = pa/T are dimensionless
AND NON-STANDARD PHYSICS chemical potentials.
These bounds are valid in absence of neutrino
Many possible modifications of the standard oscillations. On the other hand, solar and atmo-
cosmological scenario and/or Minimal Standard spheric neutrino anomalies present a strong evi-
Model (MSM) in elementary particle physics can dence in favor of mixing between v,, up, and u,
be strongly constrained or even excluded by BBN. (see e.g. talks by Y. Tots&a and A. McDonald
As we have already mentioned, BBN permits to at this Conference). Oscillations of these neutri-
restrict the number of neutrino families. The nos do not conserve individual leptonic charges
most recent analysis [36], based on Yp = 0.238 f and a primordial lepton asymmetries in elec-
0.002 f 0.005, results in the following 95% CL tronic, muonic, and tauonic charges would be re-
upper limits: distributed by the oscillations. In particular, a
0.6 for 1710= 5.8 , large p,, or pc17would create a noticeable pe of
AN< the same sign. An analysis of active-active neu-
( 0.9 for 710 = 2.4,
trino oscillations have not yet been accurately
though in an earlier paper [IS] a much more re-
performed. Preliminary results [41] are j&j < 0.1
strictive bound, AN < 0.2, was advocated. At
for any flavor a = e, p, 7. Let us note that in the
the present stage is safe to assume that AN < 1. case of vanishing chemical potentials oscillations
Hopefully in the near future one will be able to
between active neutrinos would not have a notice
derive a stronger limit.
able impact on BBN, because the neutrinos would
Other additional parameters that can influence
practically remain in thermal equilibrium inde-
light element abundances are chemical potentials pendently of oscillations and no deviation from
of different neutrino species, pa, where a = e, ~1,r.
the standard scenario would emerge. There could
A possible role of neutrino degeneracy in big bang
be a small effect due to spectral distortion of neu-
nucleosynthesis was noted already in the pioneer-
trinos by late e+e--annihilation [42]. Oscillations
ing work by Wagoner, Fowler, and Hoyle [37]
could change the distorted spectrum and might
and after that it was analyzed in a number of
be in principle observable in BBN.
papers. A combined analysis of the effect of si-
It is still not excluded that sterile companions
multaneous variation of all three chemical poten-
of active neutrinos exist. Moreover, recent mea-
tials on BBN was recently performed in the pa-
surements [43] seem to confirm creation of 0e in
pers [38,39]. Since the effect of positive pe, that
&, beam. If this is the case, an interpretation of
diminishes n/p-ratio, can be compensated by a
all neutrino data in terms of oscillations demands
non-zero pp or ,+ the limits obtained without
an existence of one (or several) sterile neutrino(s),
any extra information are rather loose, assuming
u,, mixed with active ones.
that the mentioned above conspiracy between pe
In contrast to equilibrium active-active case,
and /.+,,7 exists. However, additional data on the
mixing of active and sterile neutrinos should have
angular spectrum of CMBR permits to obtain [40] a noticeable impact on BBN. There are three pos-
stronger bounds even if the conspiracy is allowed:
sible effects due to oscillations: 1) production of
-0.01 < &,. < 0.2, ]&,prJ < 2.6 (8) new relativistic states, ZJ,, and an increase of N,,;
2) distortion of neutrino spectrum, especially im-
These results are derived under assumptions that
portant is u, spectrum; 3) resonance generation of
primordial fraction of deuterium is D/H = (3.0 f
142 A. D. Dolgoo /Nuclear Physics B (Proc. Suppl.) 110 (2002) 13 7-l 43
a large lepton asymmetry from a very small initial Moreover there is a trend to discrepancy between
leptonic or baryonic ones [44] (only the pioneer- the observations of deuterium which indicate a
ing and most recent papers are quoted; references higher value of qrc than the observations of 4He.
to many other ones can be found therein). Hopefully it will be clear in a few coming years if
The effect of creation of additional relativis- this is a real problem or an artifact of systematic
tic species by oscillations is easily estimated and and evolutionary effects. We have not discussed
from the condition that BBN allows AN extra above primordial 7Li because the accuracy of its
neutrino families one can obtain the bound on measurements are rather low now but potentially
vacuum mixing angle and the mass difference: this element could be very important for verifi-
cation of the standard model. A recent discus-
3.16 - sion of 7Li can be found in ref. [47]. Still even
( 6m2/eV2) sin4 29 < (9)
1.74 *
with the existing level of accuracy BBN permit to
for v, v, and u,,,r us mixings respectively. This put powerful constraints on deviations from the
bounds are meaningful only if AN < 1. standard model. The number of extra neutrino
Discussion of the effects of neutrino spectrum species allowed by the contemporary observations
distortion on BBN can be found in ref. [45]. The is about unity and with this bound very little can
impact of asymmetry generation on BBN is dis- said about mixing parameters between active and
cussed in ref. [44] ( second paper) and in refs. [46]. sterile neutrinos. However, if AN could be re-
Resonance case is quite complicated and the ef- duced, say, to 0.3 the limit would be meaningful.
fect may have either sign. No simple bounds can The restriction on the time (Y is quite strong and
be presented here. may exclude some of the models predicting such
BBN permits put stringent bounds on possi- variation.
ble time-variation of fundamental constants. If
we assume e.g. that the fine structure constant REFERENCES
cz is different at BBN from its present-day value,
(Y = l/137, we should expect that the neutron- 1. G. Fiorentini, E. Lisi, S. Sarkar, and F.L. Vil-
proton mass difference should also change with lante, Phys. Rev. D58 (1998) 063506;
time. The (n - p)-mass difference is given by E. Lisi, S. Sarkar, and F.L. Villante, Phys.
bm~mm,-m~=m~-m,+am,,,Wherem~ Rev. D59 (1999) 123520.
and m, are the masses of u and d quarks and 2. S. Burles, K.M. Nollett, J.N. Truran, and
the last term describes electromagnetic loop con- MS. Turner, Phys. Rev. Lett. 82 (1999) 4176;
tribution into 6m. As we have seen above the K.M. Nollett and S. Burles, Phys. Rev. D61
n/pratio is equal to n/p = exp( -6m/Tf), where (2000) 123505.
Tf is the freezing temperature of reactions (3). 3. F. Hoyle and R.J. Tayler, Nature, 203 (1964)
This temperature is determined by the competi- 1108;
tion between the Hubble expansion and weak in- P.J.E. Peebles, Phys. Rev. Lett. 16 (1966)
teraction rates. The latter is proportional to the 410.
magnitude of the electroweak coupling constant, 4. V.F. Shvartsman, Pisma Zh. Eksp. Teor. Fiz.
which is essentially cy. Hence Tf N crb2j3. De- 9 (1969) 315;
manding that successful results of BBN are not G. Steigman, D.N. Schramm and J.R. Gunn,
destroyed one can obtain that (ACY/O)BBN < (a Phys. Lett. B66 (1977) 202.
few) x 10m2. 5. K.A. Olive, G. Steigman, and T.P. Walker,
PhysRept. 333 (2000) 389.
5. Conclusion 6. C.B. Netterfield et al., astroph/0104460.
7. N.W. Halverson et al., astro_ph/0104489;
We see that gross features of BBN well agree C. Pryke et al., astro-ph/0104490.
with observations but the latter are not yet suffi- 8. R. Stompor et al., astro-ph/0105062.
ciently accurate to make it really a precise science. 9. Particle Data Group, European Physical
A. D. Dolgou / Nuclear Physics B (Proc. Suppl.) 110 (2002) I3 7-143 143
Journal C 15 (2000) 1. Ap. J. 483 (1997) 788.
10. R.F. Carswell, M. Rauch, R.J. Weymann, 31. T. X. Thuan and Y. IIzotov, sstro-
A.J. Cooke, and J.K. Webb, MNRAS 268 ph/0012382.
(1994) Ll. 32. A. Peimbert, M. Peimbert, and V. Luridiana,
11. A. Songaila, L.L. Cowie, C. Hogan, and M. astro_ph/0107189.
Rugers, Nature, 368 (1994) 599. 33. R. Gruenwaid, G. Steigman, and S.M. Viegas,
12. D. Tytler, X.-M. Fan, and S. Burles, Nature sstro-ph/0109071.
381 (1996) 207; 34. D. Sauer and K. Jedamzik, astro-ph/0104392.
S. Burles and D. Tytler, Ap.J. 460 (1996) 584. 35. K.A. Olive and E.D. Skillman, astro-
13. J.K. Webb, R.F. Carswell, K.M. Lanzetta, ph/0007081.
R. Ferlet, M. Lemoine, A. Vidal-Madjar, and 36. R.H. Cyburt, B.D. Fields, and K.A. Olive,
D.V. Bowen, Nature 388 (1997) 250. astro_ph/0105397.
14. S. Burles and D. Tytler, Ap.J. 499 (1998) 699; 37. R.V. Wagoner, W.A. Fowler, and F. Hoyle,
S. Burles and D. Tytler, Ap.J. 507 (1998) 732. Astrophys. J. 148 (1967) 3.
15. N. Prantzos and Y. Ishimar, s&o- 38. M. Orito, T. Kajino, G.J. Mathews and
ph/0106497 . R.N. Boyd, astro_ph/0005446.
16. D. Tytler, J.M. O’Meara, N. Suzuki, and D. 39. S. Esposito, G. Mangano, G. Miele and
Lubin, astro-ph/0001318 (to appear in Phys- 0. Pisanti, JHEP 0009 (2000) 038
ica Scripta). 40. S.H. Hansen, G. Mangano, A. Melchiorri,
17. M. Pettini and D.V. Bowen, Ap. J. 560 (2001) G. Miele and 0. Pisanti, a&o-ph/0105385.
41. 41. A.D. Dolgov, S.H. Hansen, S. Pastor, S.T.
18. P. Molaro, P. Bonifacio, M. Centurion, and Petcov, G.G. RaiIelt, and D.V. Semikoz (work
G. Viadilo, astro_ph/9908060. in progress).
19. S. D’Odorico, M. Dessauges-Zavadsky, P. Mo- 42. A.D. Dolgov and M. Fukugita, JETP Lett.
laro, astroph/0102162. 56 (1992) 123 [Pisma Zh. Eksp. Teor. Fiz. 56
20. J.M. O’Meara, et al Ap. J. 552 (2001) 718. (1992) 12;
21. P. Molsro, S. A. Levshakov, M. Dessauges- A.D. Dolgov and M. Fukugita, Phys. Rev. D
Zavadsky, and S. D’Odorico, astro- 46 (1992) 5378;
ph/0111589. S. DodeIson and M. S. Turner, Phys. Rev. D,
22. D. Tytler, X. Fan and S. Burles, Nature, 381 46 (1992) 3372;
(1996) 207; S. Hannestad and J. Madsen, Phys. Rev. D,
S. Burles and D. Tytler, Ap. J. 499 (1998) 52 (1995) 1764;
699. A.D. Dolgov, S.H. Hansen and D.V. Semikoz,
23. D. Tytler and S. Burles, astro-ph/9606110; Nucl. Phys. B 503 (1997) 426; Nucl. Phys. B,
S. Burles and D. Tytler, Ap. J. 507 (1998) 543 (1999) 269.
732. 43. A. Aguilar et al, PhysRev. D64 (2001)
24. J.K. Webb, R.F. Carswell, K.M.Lanzetta, et 112007.
al. Nature, 388 (1997) 250. 44. R. Foot, M. J. Thomson and R.R. Volkas,
25. A.D. Dolgov and B.E.J. Pagel, New Astron. Phys. Rev. D 53 (1996) 5349;
4 (1999) 223. P. Di Bari and R. Foot, Phys. Rev. D 63
26. D. Kirkman et aZ, astro_ph/0103305. (2001) 043008;
27. Y.I. Izotov, T.X. Tuan, and V.A. Lipovetsky, A. D. Dolgov, Nucl. Phys. B610 (2001) 411.
ApJS, 108 (1997) 1. 45. D.P. Kirilova and M.V. Chizhov, astro-
28. B.E.J. Pagel, Nuclwsynthesti and Chemical ph/0101083, and references therein.
Evolution of Galaries, Cambridge University 46. R. Foot and R.R. Volkas, Phys. Rev. D 61
Press, 1997. (2000) 043507;
29. B.E. J. Page1 et al, MNRAS, 255 (1992) 325. R. Foot, Phys. Rev. D 61 (2000) 023516.
30. K.A. Olive, E.D. Skiiiman, and G. Steiman, 47. K.A. Olive, a&o-ph/0009475.
