From nuclei to atoms and molecules: the chemical history of the early Universe

New Astronomy Reviews 46 (2002) 709–723www.elsevier.com/ locate/newar

F rom nuclei to atoms and molecules: the chemical history ofthe early Universe

a,b c ,*Denis Puy , Monique SignoreaInstitute of Theoretical Physik, University of Zurich, Zurich, Switzerland

bPaul Scherrer Institut, Villigen, SwitzerlandcLaboratoire de Radioastronomie, Observatoire de Paris, 61, Av. de l’ Observatoire, 75014 Paris, France

Dedicated to Dennis William Sciama

Abstract

The dark age of the Universe is generally pointed out as the period between the recombination epoch (z |1000) and thehorizon of current observations (z |5–6). The arrow of time in the cosmic history describes the progression from simplicityto complexity, because the present Universe is clumpy and complicated unlike the homogeneous early Universe. Thus it iscrucial to know the nature of the constituents, in order to understand the conditions of the formation of the first boundobjects. In this paper we analyse the chemical history of thisdark age through the creation of the primordial nuclei to theformation of the first atoms and molecules. Then we will describe the consequences of the molecular formation on the birthof the proto-objects. In this context we will mention the important works of Dennis W. Sciama who influenced a largenumber of theorists—cosmologists and astronomers—on this new field of research dedicated to primordial molecules. 2002 Elsevier Science B.V. All rights reserved.

PACS: 01.30Keywords: Early Universe

1 . Introduction against molecular formation, it turns out that thetemperature is small enough for this formation

At early times the Universe was filled up with an occurs.extremely dense and hot gas. Various important The existence of a significant abundance of mole-physical processes occurred as a consequence of the cules can be crucial on the dynamical evolution ofexpansion. For example, Universe cooled below the collapsing objects. Because the cloud temperaturebinding energies of light elements leading to the increases with contraction, a cooling mechanism canformation of the nuclei which recombined with the be important for the structure formation, by loweringelectrons in order to form the first atoms. After pressure opposing gravity, i.e. allowing continuedrecombination, although the density decrease acts collapse of Jeans unstable protoclouds. This is

particularly true for the first generation of objects.Interactions between primordial molecules and the

*Corresponding author. Tel.:133-101-4051-2112; fax:133-101-cosmic microwave background radiation (CMBR)4051-2002.could be important if the molecular abundances areE-mail addresses: [email protected] (D. Puy),

[email protected](M. Signore). sufficient (Dubrovich, 1977; Melchiorri and Mel-

1387-6473/02/$ – see front matter 2002 Elsevier Science B.V. All rights reserved.PI I : S1387-6473( 02 )00240-3

mailto:[email protected]

mailto:[email protected]

710 D. Puy, M. Signore / New Astronomy Reviews 46 (2002) 709–723

chiorri, 1994; Signore et al., 1997). In particular, a plasma, where most of its energy is on the form ofresonant molecular scattering between CMBR photons at high temperature and relativistic particles.photons and molecules could, on one hand, smear The weak processes such as:out primary CMBR anisotropies and, on the other 2 1n 1n↔ p 1 e , n 1 e ↔ p 1n, andhand, produce secondary anisotropies.

2 ]n↔ p 1 e 1nIn this paper we will describe the chemical historyof the Universe from the formation of the first atoms maintained the ratio of neutrons to protons at itsuntil the formation of the first objects. Thus in thermal value, i.e.n /p | 1 (where n define the

2 1Section 2 we will recall the nucleosynthesis then the neutrons, p the protons,e the electrons,e theimportant influence of the neutrinos physics in positrons andn the neutrinos).Section 3. The period of recombination will be At about one second, the temperature of the

10described in Section 4 which will lead to the universe is around 10 K, the above weak processesdescription of the chemical chemistry or formation of became ineffective. Then /p ratio froze out at aboutprimordial molecules in Section 5. We will discuss in 1/6, leading to the collisions between neutrons andSection 6 the influence of the molecules on the protons which form deuterium:formation of the first objects, and on the CMBR in

n 1 p → D1g.Section 7. The possible reionization of the Universecould be some importance consequences on the

The collisions with deuterium led to the formationchemical history of the Universe, in Section 8 weof the helium 3 and tritium nuclei:will describe it, and we will give some conclusions

3 3and prospects in Section 9. n 1D → H, p 1D → He,4and finally to He via the reactions:

2 . Primordial nucleosynthesis 3 4 3 4n 1 He→ He, p 1 H → He,4D1D → He.The Standard Big Bang Nucleosynthesis model

(SBBN) has been analysed first by Peebles (1966), At about 100 s, the nucleosynthesis epoch was over,4Wagoner et al. (1967) and Wagoner (1969, 1973). thus most neutrons were in He nuclei, and most

Nevertheless theChicago Group gave strong im- protons remained free while smaller amounts of D,3 7pulse to this field, see Yang et al. (1979, 1984), Copi He and Li were synthetized. The low densities, the

et al. (1995), Schramm and Turner (1998), Schramm Coulomb barriers and stability gaps at masses 5 and(1998) and Olive et al. (1981, 2000), and theOxford 8 worked against the formation of heavier elements.Pole led by Sarkar (1996, 1999). Here we only The chemical composition of the Universe remainssummarize the main features of the SBBN model unchanged until the formation of the first stars. Thethat we have already presented elsewhere (Signoreyields of primordial nucleosynthesis are shown as aand Puy, 1999). function of the baryon density in Fig. 1 (from Burles

The standard model of the very early Universe is et al., 1999).quite simple and determined by the three important It is a considerable challenge to measure the actualmilestones of the early Universe: the expansion primordial abundances because of uncertainties ingoverned by General Relativity, the particle interac- measuring present day abundances and because oftions governed by Standard Model and particle uncertainties in modeling the nuclear evolution sincedistribution governed by statistical physics. The the big bang. The measured abundances of helium 4,SBBN depends on just one parameter which is the deuterium, lithium 7 and helium 3 (Burles et al.,baryon-to-photon ratioh or the dimensionless quanti- 1999) are:ty of baryon densityV .b

Y 50.24460.002In the framework of the SBBN model at times p

much less than one second after the big bang, the10 25Universe was a hot (T 4 10 K) rapidly expanding (D/H)5 (3.460.3)3 10p

D. Puy, M. Signore / New Astronomy Reviews 46 (2002) 709–723 711

4Fig. 1. Relative abundances, by number: A/H, except for He in mass fraction:Y . Concordance intervals for each element (2s uncertainty)p

and baryon density predicted (vertical line) by D measurements are also shown (from Burles et al., 1999, astro-ph/9903300).

7 210( Li /H) 5 (1.760.15)3 10 band). From the latest determined primordial Dp

abundance (Tytler et al., 2000):3 25( He/H) 5 (0.361)3 10 25p (D/H) 5 (3.360.5)3 10 (95% c.l.).p

Fig. 1 summarizes the measured abundances withBurles et al. (2000) give the following precisetheir uncertainties. Since 1980 and until recently, determination of the baryon density:cosmologists introduced a concordance interval for

2V h 5 0.018960.0019 (95% c.l.).the baryon density where the predicted and measured b

abundances for all the four light elements areUntil this year, the agreement between the SBBN

consistent; see for instance Copi et al. (1995) whopredictions and observed abundances makes the

derivedSBBN a cornerstone of the Big Bang cosmology But

2 the recent cosmic microwave background measure-0.007,V h , 0.024.bments of BOOMERANG and MAXIMA (De Ber-nardis et al., 2000; Jaffe et al., 2000) favor theA change occurred in 1998–1999 with the accur-following value:ate determination of the primordial deuterium abun-

2 10.009dance which led to the most accurate determination V h 5 0.032 (95% c.l.)b 20.008of the baryon density: Fig. 1 shows the concordanceintervals for each element and the baryon density which, a priori, cannot be accomodated within thepredicted by the deuterium measurement (vertical SBBN model (Burles et al., 2000). Future satellite


experiments (MAP & PLANCK surveyor) will give epoch of nucleosynthesis was controlled by theus a better cosmic microwave background determi- number of particle degrees of freedom: more neut-

2nation ofV h . Meanwhile, the SBBN model seems rino species imply a higher energy density and ab4to be in trouble. . . and cosmologists search for non- faster expansion, i.e. more neutrons and more He

4standard cosmology or non standard big bang nu- production. Therefore the He abundance can becleosynthesis. used to constrain the number of neutrinos species.

For instance, Steigman et al. (1977) gave the SBBNlimit of 7 neutrino species while the laboratory limitwas about 5000! At this time, the SBBN limit was a3 . Neutrinos and nucleosynthesisvery good constraint.

The helium curve in Fig. 1 was computed assum-In cosmology, neutrinos have a strong impact oning N 5 3. Burles et al. (2000) get the upper limitbig bang nucleosynthesis and therefore important n

N ,3.2 while Lisi et al. (1999) giveN , 4 depend-consequences on the chemistry of the universe. n n

ing on what observational constraints one uses.Recently an extensive literature discussed the pos-More generally, the limits onN can be translatedsible generation ofsterile neutrinos in the early n

into limits on other types of particles (numbers oruniverse as a result of theiroscillations with activemasses) that would affect the expansion rate justneutrinos. Dennis W. Sciama was not only concernedbefore the nucleosynthesis.with active neutrinos and decaying neutrinos since a

long time, but also with sterile neutrinos. In an3 .2. Sterile neutrinosEuropean conference (Sciama, 1999), he gave a

1review and pointed out the importance of this fieldSterile neutrinos are neutrinos which lack theand described the more recent developments on the

standard electroweak interactions possessed by thesterile neutrinos and its relevance for both SBBN andactive neutrinos:n , n andn . In particular, they dothe CMBR. e m t

onot contribute to the decay width ofZ ; they wouldThus, we briefly recall how SBBN can constraingravitate; therefore, if they existed in the earlythe properties of active neutrinos and in particularuniverse, their gravitational effects would influencehow helium abundance can be used to count neut-the rate of expansion and therefore also the SBBN.rinos.

Different studies have been recently done andshowed the role played by neutrino oscillations in3 .1. Limits to the number of neutrinosSBBN; in particular, there are some controversiesabout the possible generation of lepton asymmetryIn the standard model, neutrinos possess onlyby oscillations between active and sterile neutrinosweak interactions, and are coupled to the inter-

o and its consequences on the observed baryonmediate bosonsZ . Measurements of the bosonicasymmetry of the universe, on primordial abun-decay width allow to conclude that the total number

4dances (especially He and also D), on the CMBRof different neutrino species is(Hannestad and Raffelt, 1999) and on the decaying

N 53.0760.12n neutrinos (see later in Section 8). Look, for instance:Bell et al. (1998), Sciama (1998, 1999), Mohapatrawhile the combined fit to LEP data gives the moreand Sciama (1998), Dolgov (2000) and referencesaccurate results:therein.N 5 2.99460.01.n

In the framework of the SBBN model the precise 3 .3. BOOMERANG, MAXIMA and the number ofneutrinos

1He argued:students of the CMBR have now to learn somethingAs already said in Section 2, the pre-new, thanks to the famous announcement by the Super-

BOOMERANG and MAXIMA consensus for theKamiokande team as Neutrino 98, confirming the existence of an2

atmospheric neutrino anomaly . . . (see Sciama, 1999). central value forV h deduced from SBBN isb


2V h 5 0.0189 Dicke (1968) based, at this epoch, on theprimeval-b

fireball picture of Gamow (1948) for the evolutionwhile the central value implied by BOOMERANG of the Universe.and MAXIMA measurements (Jaffe et al., 2000) is: The cosmological recombination process was not

2 instantaneous because the electrons, captured intoV h 5 0.03.bdifferent atomic energy levels, could not cascade

The difference between the SBBN and the cosmic instantaneously down to the ground state. Atoms2microwave background central values forV h has reached the ground state either through the cos-b

already triggered discussions, explanations in a lot of mological redshifting of the Lymana line photonspapers. Among them: Hu et al. (2000), Tegmark and or by the 2s 2 1s two photons process. NeverthelessZaldarriaga (2000), White et al. (2000), Kinney et al. the Universe expanded and cooled faster than recom-(2000), Kurki-Suonio and Sihvola (2000), Kurki- bination could completed, and small fraction of freeSuonio (2000), Hannestad et al. (2000) and Mel- electrons and protons remained.chiorri and Griffiths (2000).

Let us only mention that an alternative solution 4 .1. Recombination of hydrogencan be lepton asymmetry; for instance, Esposito et al.(2000) find the best fit to the BOOMERANG and The principles of calculations of the primordialMAXIMA data with N 5 9 or suggest the best recombination have been mentioned initially byn

compromiseN 5 13 for cosmic microwave back- Shklovskii (1967), Novikov and Zel’dovich (1967)n

ground data and light elements abundances; Lesgour-and Takeda and Sato (1968). Peebles (1968) was thegues and Peloso (2000) find areasonable fit with first to present a theory in which the very compli-

2V h 5 0.028 and an effective number of neutrinos cated recombination process is reduced to simplerb

2N 56. Active-sterile neutrino oscillations can also terms . The methodology of the calculations consistn

explain the current observations of cosmic micro- to take into account a three level atom with a groundwave background and light elements abundances (Di state, first excited state and continuum, representedBari and Foot, 2000). BOOMERANG and MAX- by a recombination coefficient. A single ordinaryIMA data and the number of neutrinos are a very differential equation is derived to describe the ionisa-active field of study. One must point out that more tion fraction (see Peebles, 1968). The assumptionsrecent data (April 2001) from BOOMERANG (Net- are:terfield et al., 2001), MAXIMA (Lee et al., 2001)and DASI (Halverson et al., 2001) show that there is • Hydrogen excited states are in equilibrium withmore difference between BBN and CMBA estimates the radiation

2for V : V h 50.0.2. See, in particular, Signore and • Stimulated de-excitation is negligibleb b

Puy (2001) for a review of the new constraints on • Collisional processes are negligibleneutrino physics brought by this new value ofV .b

The approximations are based on the rate oftransfer of energy per unit volume between radiation

4 . Recombination epoch and free electrons developed by Kompaneets (1957)then Weymann (1965, 1966) and coupled with the

In the hot Big Bang picture, after the nu- expansion of the Universe.cleosynthesis period the natural question concernsthe possibility to form atoms by recombination of

2primordial nucleus with free electrons, and the By means of a number of approximations valid in the conditionsof interest, so that the physical situation can be described by asecond question concerns the degree to which thefew variables, and the recombination equation can be integratedrecombination is inhibited by the presence of recom-by a few variables, and the recombination equation can be

bination radiation. integrated by hand, without having to abandon any of theThe reason for which these questions are central essential elements of the problem as written by Peebles himself

were mentioned by Peebles (1968) and Peebles and in 1968.


Peebles (1968) computed the rate of recombina- recombination is much slower than previouslytion of the plasma from the rate coefficient for thought, and it is delayed until just before H re-recombination to excited states of hydrogen tabulated combines.by Boardman (1964), and showed that the residualionization of the hydrogen (or fractional ionization 4 .3. Recombination of lithium

25x ) is below 5.3310 for a flat cosmological modele

at redshift below 1500. Jones and Wyse (1985) The ionization potentialI 5 5.392 eV of LiI isLi

developed two important analytic approximations: shorter than that of hydrogen (I 5 13.6 eV). Dal-H

one, valid at the redshiftz .1500 is a modification garno and Lepp (1987) suggested the existence of1of the standard Saha equation, and the other, valid Li ions after the period of the hydrogen recombina-

for 800, z ,1500. They found that the fractional tion, and showed that the radiative recombination is24ionization is roughly 6.75310 at the redshiftz 5 again effective at the redshiftz | 450. This last point

555 much higher than the Peebles values. More is particularly important for the lithium chemistry asrecently Seager et al. (1999, 2000) have presented a we will see.refined treatment of the recombination through acomplete code which performs appproximate calcu-

3lations history . 5 . Primordial chemistry

4 .2. Recombination of helium From the pioneer works of Zwicky (1959) on themolecular hydrogen content of the Universe, pre-

4We have seen that He is the second most sented at the San Francisco meeting of the As-abundant element in the Universe after hydrogen. tronomical Society of the Pacific in June 1959, theZel’dovich et al. (1968) and Matsuda et al. (1969) idea of molecular formation at the immediate post-were the first to point out that the cosmological recombination took form.recombination of helium differs from that of hydro- The studies on the primordial chemistry (or post-gen. The conditions for the helium recombination are recombination chemistry) have been the source of asuch that in both cases it occurs in accordance with tremendous increase of the literature. The first com-the Saha equation (Bernshtein and Dubrovich, 1977; plete description of the hydrogen chemistry wasDubrovich and Stolyarov, 1997). Recombination developed by different japanese groups: Hirasawa ethistory of early Universe consist of three stages but al. (1969), Takeda et al. (1969) and Matsuda et al.helium recombination occurs in two steps: (1969) in the context of pre-galactic gas clouds or

formation of galaxies. The importance of the• at z |6000 (or when the temperature became less deuterium chemistry through the formation of the

than 16 000 K) the first electron recombines and HD molecule was suggested by Lepp and Shullsingly ionized HeII forms from doubly ionized (1983) as lithium chemistry through the formation ofHeIII. the molecule LiH, completed by the works of

• Then at z | 2700 (or temperature close to 7300 Dalgarno and Lepp (1987) on the helium chemistryK), HeII recombines into a neutral state. and the lithium chemistry. Latter a complete com-

• from z | 1500 recombination of hydrogen took prehensive chemistry of the lithium, at the post-place. recombination epoch, was presented by Stancil et al.

(1996).More recently Seager et al. (2000) have developed The first chemical network including the primor-

an improved recombination calculation of H, He dial molecules (such as H , HD and LiH) and ions2

(with HeII and HeIII) that involves a line-by-line was carried out by Lepp and Shull (1984), Latter andtreatment of each atomic level. They found that HeI Black (1991), Puy et al. (1993) and more recently by

Galli and Palla (1998), and Stancil et al. (1998). Thedevelopment of the primordial chemistry studies owe

3Seehttp: / /www.astro.ubc.ca/people/scott / recfast.html. to Dalgarno an important contribution concerning the

http://www.astro.ubc.ca/people/scott/recfast.html.











calculations of the reaction rates (see Lepp and Shull, with the reaction rates calculated by MacDowell1983; Dalgarno and Lepp, 1987; Puy et al., 1993; (1961) from remarks on the quantum theory of theGalli and Palla, 1998 and references therein). negative hydrogen ion emphasized by Chandrasekhar

(1944), in which the electrons act only as catalysts.Thus two possible ways of formation were pointed5 .1. Hydrogen chemistryout.

Takeda et al. (1969) were the firsts to study theAfter the important works of the recombinationevolution of molecular hydrogen abundance in theperiod, it was plausible to imagine that molecularcosmological medium (i.e. post-recombination Uni-hydrogen plays a role in the early evolution of largeverse) in contrast with the works by Saslaw andclouds. NeverthelessH was not considered as a2Zipoy (1967) and Peebles and Dicke (1968) incomponent of the pre-galactic medium.which they calculated the products of H in denseIn a review article dedicated to the hydrogen 2

clouds. They considered the two ways of H forma-molecules in astronomy Field et al. (1966), showed 2

tion and electron and proton as a kind of catalyzer.that the times scales for appreciable amounts to form2The photodetachment of H and the photodis-in three-body reactions or in radiative association

1sociation of H by the background radiation fieldwere always greater than the Hubble age. In the 2

restrict the abundance of molecular hydrogen formedcosmological context there are no grains which canat early stages, although the photodestruction ofcatalyze the reaction of formation, thus any H2molecular hydrogen is negligible. The destruction isformed in the uniform background is dissociated bydue to collisional dissociation (Lepp and Shull,the radiation, until the density is too low to produce

4 1983).it. Saslaw and Zipoy (1967) then Shchekinov and´Entel (1983) showed the importance of the charge

5 .2. Deuterium chemistrytransfer reactions:1 1H 1H → H 1H (1)2 2 Deuterium chemistry in a early Universe could

play an important role in the sense that it could giveinitiated by the radiation associationsome explanations on the controversy observations

1 1H1H → H 1 hn. (2)2 of fractional abundance in high redshift Lya (Son-1 gaila et al., 1994; Burles and Tytler, 1996). AlthoughThey pointed out that H ion is converted to H as2 2

1 similar processes contribute to the formation of HD:soon as it is formed and the H concentration never21 1becomes large. D 1H → HD 1 hn (5)

Peebles and Dicke (1968), in a scenario con-1 1cerning the origin of the globular star clusters, HD 1H → H 1H , (6)2

showed that these clusters may have originated asits formation also proceeds through:gravitationally bound gas clouds before the galaxies

1 1formed, and suggested that some molecular hydrogenH 1D → D 1H (7)can form, mainly by way of negative hydrogen by

1 1´the reactions (see also Shchekinov and Entel, 1983)D 1H → H 1HD. (8)2

2 2H1 e → H 1 hn (3) Thus the formation of HD is carried out in two steps.Nevertheless when the abundance of H is sufficient,2followed by the reactionthe second way of formation is dominant (Eq. (8)).

2 2H 1H → H 1 e (4) Stancil et al. (1998) presented a complete review of2

the deuterium chemistry of the early Universe.

4In 1967, William C. Saslaw was a Ph.D. student of Dennis W.5 .3. Helium chemistrySciama at the Department of Applied Mathematics and Theoret-

ical Physics of the University of Cambridge (see thefamily treein Ellis et al., 1993). The presence of helium in the early Universe gives


rise to a rich assembly of molecular processes as 5 .5. Molecular abundancespointed out by Hirasawa (1969). The main molecular

1species containing helium is HeH . Processes that In order to estimate the molecular abundances5lead to the formation and destruction of the molecu- which depend on the reaction rates , it is necessary

1lar ion HeH in astrophysical environments have to know the thermal and dynamical evolution of thebeen the subject of a number of investigations. medium. In the Einstein–de Sitter Universe theDabrowski and Herzberg (1978) were the one of the evolution of matter temperatureT is described bym

firsts to introduce the possibility that this molecular the equation:ion exists in astrophysical plasmas; it is from these dT 2T dam minitial works that Roberge and Dalgarno (1982) ]] ]]]5 ax T (T 2 T )2 (15)e r r mdt a dtpresented the formation and destruction mechanisms.More recently Zygelman et al. (1998) calculated the wherea is a constant,x the fractional ionization ofe

enhancement of the rate coefficient for the radiative the hydrogen,T the temperature of the cosmologicalr1association to form HeH : radiation temperature anda(t) the expansion parame-

ter. The first term describes the heat transfer from1 1He1H → HeH 1n. (9) radiation to the electrons when the second term1 1 characterizes the cooling due to the expansion of theThe H ions and H produce HeH by the fast2 2 Universe. The result of the integration of this im-reactions

portant equation led to the idea that there is thermal1 1 decoupling between matter and radiation. Neverthe-H 1He→ HeH 1H, (10)2

less the formation of primordial molecules such as1 1 H , HD, and LiH can involve a thermal response due2H 1He → HeH 1H. (11)2

to the excitation of the rotational levels of theseNevertheless for all of these processes, once photo-molecules. The population of the rotational levels is

1dissociation ceases to be rapid, the H ions react mainly due to collisional excitation and de-excitation2

preferentially with neutral atomic hydrogen to form with H, H and He on one hand and to radiative2

molecular hydrogen as we have seen. processes on the other hand: absorption from theCosmic Microwave Background Radiation (CMBR)and spontaneous or induced emission.5 .4. Lithium chemistry

The importance of radiative cooling rate by molec-ular hydrogen were pointed out by Takayanagi andThe lithium chemistry is initiated by the recombi-Nishimura (1960), which developed theoreticalnation of lithium, which occurred near the redshift

1 calculations of the cooling process due to the rota-z | 450. The formation of the molecular ion LiHtional excitation of H molecule. They considered2formed by radiative association processes:collisions with hydrogen atom in the context of

1 1Li 1H → LiH 1 hn, (12) interstellar clouds. Saslaw and Zipoy (1967) used theradiative function of Takayanagi and Nishimura

1 1 (1960) in the context of pre-galactic gas clouds. TheH 1 Li → LiH 1 hn (13)molecules HD and LiH play an important thermal

opens the way of the formation of the LiH molecules role, since their allowed dipole rotational transitions,through the exchange reactions see Lepp and Shull (1984). Puy et al. (1993)

1 1 calculated precisely the population of the levels andLiH 1H → LiH 1H , (14)provide a molecular thermal functionC due to H ,mol 2

which are more rapid than the formation by radiative HD and LiH. A complete description of the coolingassociation of H and Li atoms. of astrophysical media is found for H in Le Bourlot2

The complete description of the lithium chemistry5was emphasized by Stancil et al. (1996) where new The complete chemical network is described in Puy et al. (1993),

quantal rate coefficients were included. Galli and Palla (1998) and Stancil et al. (1998).


1 212 1 218 1et al. (1999), and for HD in Flower et al. (2000). H /H|1.3310 , HD /H|2.1310 , H D /2 2214 1 213 1Flower (2000) presented recently a study on the role H|5.1310 , HeH /H|6.2310 and LiH /218of HD in the thermal balance of the primordial gas H|9.4310 .

and conclude that HD is about as important as H in2

the thermal balance of the primordial gas. Notice thatthe estimation of the ortho-para H in the primordial 6 . Primordial objects2

gives a better estimation of the H cooling as Flower2

ˆand Pineau des Forets (2000) emphasized very The development of structure in the universe wasrecently. Furthermore some chemical reactions well advanced at high redshifts. For example,produce a chemical thermal functionQ through quasars have been detected nearly toz | 5, and thechem

the enthalpy of reaction. most distant galaxies to even greater distances. StillIn the cosmological context the thermal evolution unknown, however, is the epoch during which the

equation becomes first generation of objects was formed. Hoyle (1953)were one of the firsts to suggest that the importance

dT 2T 2Cda of the fragmentation process of extragalactic hydro-m m mol]] ]]] ]]5 ax T (T 2 T )2 1e r r mdt a dt 3nk gen cloud gas clouds, with particular reference to the

formation of galaxies out of hot hydrogen clouds.2Qchem]]1 (16) From these pioneer works Hunter (1962) developed3nk

an instability theory of a collapsing gas cloud.Puy et al. (1993) took into account the transfer Gravitational instability and thermal instability wereprocess between radiation and matter via the Thom- supposed to be the main processes to form condensa-son scattering, coupled with the molecular source tions in a dilute gas. In the framework of theterm and the enthalpy of the reactions in order to gravitational instability theory, each structure started

6calculate the abundances of the species . as a tiny local overdensity; nevertheless very little isAfter a transient growth all abundances of the known about the protoclouds. Fluctuations that sur-

molecules become almost constant. The final abun- vive decoupling are subject to gravitational in-dances of H , HD and LiHfreeze out due to the stabilities if they are on sufficiently large scale. The2

expansion. Moreover the final values of molecular general treatment of the gravitational instability in anabundances depend on the choice of the cosmologi- expanding Universe, developed by Lifshitz (1946)cal parameters. Galli and Palla (1998) summarized and Lifshitz and Khalatnikov (1963), showed that aand discussed the chemistry of early Universe for condensation due to gravitational instability cannotdifferent cosmological models. In the standard model grow so fast during the time scale of the Universe.defined by the following cosmological parameters: Parker (1953) suggested that the problem of con-

H 5 67 for the Hubble constant andh 5 4.5 for densation can be understood as a consequence ofo 10

the baryon-to-photon ratio in an Einstein–de Sitter instability on the thermal equilibrium of a diffuseUniverse. In this context we have at the end of the medium.recombination, the following initial abundance: Since the works of Jeans (1902) the importance of

25 210D/H|4.3310 , and Li /H|2.2310 . thermal instability as another possible mechanism forIn this context we obtained the following final galaxy formation were made by Hoyle (1958) and

abundance atz 5 5: above all Field (1965). Kato et al. (1967) examined26 29H /H|1.1310 , HD/H|1.2310 and LiH/ the thermal instability of a dilute gas in expanding2

220H|7.2310 Universe. Sofue (1969) found that formation ofand for the main molecular ions withx | 33 galaxies could be initiated by thermal instability duee2410 : to radiative cooling, and developed by Saito (1969)

in the context of a uniformly rotating homogeneous6 medium. The formalism were definitively establishedThe numerical integration of the coupled chemical equation is an

by Kondo (1970) in an expanding medium.initial value problem for stiff differential equations (see Puy etal., 1993). Doroshkevich et al. (1967) showed that if the


cloud contracts adiabatically, hydrogen atoms in it Palla et al. (1983) examined precisely the three bodyare soon collisionally ionized again and the cloud is reactionsfinally dispersed by radiation pressure. Therefore,

H1H1H → H 1H (18)some kind of cryogen is necessary to form a bound 2

system as mentioned (Takeda and Sato, 1968). Theproblem was to find a good cryogen and so primor- H1H1H → H 1H , (19)2 2 2dial molecules offered an interesting solution. Mat-suda et al. (1969) realized that hydrogen molecule and investigated the thermal and chemical evolutioncould be a possible cooling of pre-galactic gas clouds of a collapsing spherical cloud composed of purefrom the works on the H cooling in the interstellar hydrogen gas, and various regimes of oscillation in2

clouds of Takayanagi and Nishimura (1960). the collapse can occurs as Lahav (1986) showed.Yoneyama (1972) pointed out that after hydrogen Lepp and Shull (1984) emphasized that the dipolemolecules are formed, the cooling of the gas by these rotational transitions of HD and LiH are particularlymolecules plays an important role on the further important at high density and low temperature. Incondensation and fragmentation of the cloud. Since this context Puy and Signore (1997) examined thethese first works, the hypothesis that galaxies formed evolution of primordial molecules in a context offrom density perturbations which collapse within the gravitational collapse and showed how the abun-expanding background of a Friedmann universe dances can be modified. The importance of HDappears worthy of continued investigation. Neverthe- molecules was pointed out by Puy and Signoreless all of these first models considered the collapse (1998b) where they analysed the ratio between theof non-rotating spherical clouds. Thus Hutchins molecular cooling due to HD and that due to H , and2(1976) studied the thermal effects of H molecules in2 found that the main cooling agent around 200 K isrotating and collapsing spheroidal gas clouds, and HD. This results was confirmed by Okumuraisuggested an early period of formation of objects in

(2000), Uehara and Inutsuka (2000) and Flower etthe mass range of ordinary stars; which led to the

al. (2000).model of Silk (1977) concerning the fragmentation

The cooling of LiH is less important. Lepp andof cosmic gas clouds consisting of gas of a pre-

Shull (1984) and Puy and Signore (1996) considereddominantly primordial composition. The contraction

the LiH molecules formed through only the radiativeof a such cloud is initially adiabatic.

association. Stancil et al. (1996), Dalgarno et al.Early studies focused on the chemical evolution

(1996) and Gianturco and Gori-Giorgi (1996) pro-and cooling of primordial clouds by solving a

posed a new lithium chemistry with new ways ofchemical reaction network within highly idealized

LiH formation. In this context Bougleux and Gallicollapse models (Palla et al., 1983; Mac Low and

(1997) then Puy and Signore (1998a) developed aShull, 1986; Puy and Signore, 1996; Tegmark et al.,

complete lithium chemistry in a collapsing cloud.1997). Some hydrodynamic aspects were studied in

Nevertheless at the temperature higher than 200 K,spherical symmetry and multidimensional studies by

H and HD cooling are thermically dominant as Abel2Aninos and Norman (1996) and more recently byet al. (1997) pointed out in three-dimensional nu-

Abel et al. (2000). A small fractional abundance ofmerical simulations, which lead to a scenario of the2H molecules is formed via the H way. The small2 formation of first stars in the Universe due to the

residual ionization remaining after decoupling isfragmentation of primordial gas (see also Bromm et

sufficient to produce enough H to radiate away the2 al., 1999).compressional energy of collapse. Further collapse

Thus primordial molecules play an important roleproceeds almost isothermally, and considerable frag-

in the pre-galatic gas particularly on the thermoch-mentation occurs. This process is terminated only

emical stability as mentioned by Rosenzweig et al.when at sufficiently high densities the H molecules2 (1994) and on the problem of the minimum massare destroyed by the reaction:

which virialized in order to form the first cosmologi-H 1H1H → 3H. (17) cal objects (Tegmark et al., 1997).2


28 297 . Anisotropies of CMBR and primordial 10 –3310 ) which was adopted—at that time—molecules by all the experts in primordial chemistry.

An attempt to search for LiH has even beenThe study of the CMB anisotropies (CMBA)— carried out with the IRAM 30m telescope (De

primary and secondary ones—are important tools to Bernardis et al., 1993). Because the resonant scatter-study the origin and the evolution of perturbations. ing is an elastic process, it can result in a primaryDifferent physical processes are responsible for the CMBA attenuation and in a secondary CMBAproduction of the CMBA: each one associated with a production, exactly as in the case of an earlytypical angular scale. reionisation of the universe. Note that this effect is

Among the different kinds of secondary CMBA, strongly frequency-dependent while the effect ofthe Rees–Sciama effect (Rees and Sciama, 1968) Thomson scattering in a ionized medium is notcan produce a detectable signal only if very large frequency dependent. Finally, one can show that the

16masses are involved (M . 10 M ) and can give blurring of primary CMBA by resonant scattering(28significant contributions only at large angular scales lines of LiH can occur if LiH/H.3310 for n ,

(degree scale) and inV ±1 scenarii. The Sunyaev– 60 GHz and for intermediate angular scales (seeZel’dovich effect (Sunyaev and Zel’dovich, 1972) is Maoli et al., 1994).important at cluster scales (subarcminute scale). Atvery small scales (subarcminute scale) two kinds of 7 .2. Possible molecular signals?secondary CMBA can be sizeable: thermal emissionby dust in galaxies and primordial molecular lines Just as primary CMBA could be erased, secondaryproduced by resonant elastic scattering. Here, we CMBA could be expected by molecular resonantonly consider the primordial molecular lines. scattering. In the framework of a Cold Dark Matter

(CDM) scenario—and also in a more general case—7 .1. Interaction between the CMB and primordial Maoli et al. (1996) have calculated the intensity themolecules line width of the expected signal during the three

stages of the evolution of a very simple protocloudFrom an initial idea of Zel’dovich, Dubrovich model: linear phase, turn-around epoch and non-

(1977) showed that resonant elastic scattering must linear collapse. In particular, they have shown that25be considered as the most efficient process in for standard observational conditions (10,Dn /

24 24coupling matter and radiation at high redshift. They n ,10 ; DI /I . 10 the beginning of the non-CBR

noted that the cross section for resonant scattering linear collapse phase of a protocloud, just after itsbetween CMB and molecules is several orders of turn-around epoch, is the best one for a detection ofmagnitude larger than that between CMB and elec- the first two LiH rotational lines which are observ-trons: even a modest abundance of primordial mole- able in the frequency range 30,n , 250 GHz withcules would produce significant Thomson scattering. an angular scale range 70,u , 200. They concludedAt this point, every velocity field (due to molecular that the IRAM 30-m telescope with its 100 of angularmotion or to cloud infall) would leave its imprint on resolution and its frequency channels, is well suitedCMB via Doppler shift. This technique for exploring for this kind of observations, see also Signore et al.the early universe has been revised by the group of (1997).Melchiorri (De Bernardis et al., 1993; Melchiorri andMelchiorri, 1994; Maoli et al., 1994; Signore et al., 7 .3. Discussion1994). A careful analysis of various primordialmolecules has led to the identification of LiH and The effects of LiH molecules on the CMBA-

1LiH as the best candidates (Dubrovich, 1977; De erasing of primary CMBA, creation of secondaryBernardis et al., 1993). Note only that this conclu- CMBA-strongly depend on the final LiH abundancesion was relevant because all these authors assumed which is a function of the lithium produced bya value for the LiH abundance (i.e. LiH/H|33 primordial nucleosynthesis and of the efficiency of


its conversion to LiH. With the quantal rate coeffi- rinos decayed into lighter neutrinos and UV photons.cient for the radiative association of lithium and The hypothesis of the photoionization of the inter-hydrogen given by Stancil et al. (1996) (see Section galatic medium by radiatively decaying neutrinos or6) the fractional abundance of LiH is much smaller photinos has been previously formulated by Sciamathan that of the above CMBA studies and therefore (1982a,b,c), which led also to a scenario describingleads tono erasing of the primary CMBA and no the ionization of HI regions in the galaxy (Sciama,production of secondary CMBA. But let us only 1990a) or the ionization of Hydrogen throughout theemphasize that, again at present times, the primordial Universe (Sciama, 1990b, 1993b, 1994; Sethi, 1997).lithium abundance and the percentage of lithium In this context Salati and Wallet (1984) reinvesti-converted to LiH are both quite uncertain! gated the recombination of neutral hydrogen taking

into account light neutral fermions, stable or radia-tively unstable. They found a limit on the fractional

8 . Reionization ionization:24 2343 10 , x , 2310e

In the 1960s years there were several ways which instead of the previous baryon-dominated universeled to the idea that the medium at high redshift could result:

25 24be ionized: The radiation from a very distant object 33 10 , x , 3310 .e

could be scattered by free electrons (Field, 1959; Sciama (1988) from the works of Cabibbo et al.Gunn and Peterson, 1965); radio waves of low (1981), Sciama and Melott (1982) and Sciamafrequency from extragalactic sources would be ab- (1982c, 1984) proposed a scenario of photino decaysorbed (Field, 1959; Sciama, 1964a; Ericson and in order to explain the observation of the ionizationCronyn, 1965; Sciama and Rees, 1966), and the of Lyman-a clouds at large redshifts.radiation from radio sources could suffer dispersion Bowyer et al. (1999) observed the spectrum of the(Haddock and Sciama, 1965). Sciama (1964a,b) and night sky with an extreme-ultraviolet spectrometer

˚thereafter Rees and Sciama (1966) were one of thecovering the bandpass from 350 to 1100 A on boardfirst to put forward that the intergalactic hydrogen the spanish satelliteMINISAT-1 launched in 1997could be partially ionized at large scale. April 21. The observed is far less intense than that

The absence of a Gunn–Peterson (Gunn and expected in the Sciama model of radiative decay ofPeterson, 1965) effect in the spectra of high redshift massive neutrinos, but they cannot exclude thequasars implies that the intergalactic medium is earlier model of Melott (1984), based also on thehighly ionized at low redshifts (i.e.z |5). These theory of Sciama with a decay energy which isobservations suggested that the intergalactic mediumsomewhat greater than 13 eV, and different than the

7was reionized in the redshift interval (5, z , 300) . 13.6 KeV of the Sciama model.Thus at epochs corresponding toz | 1000 the inter-galactic medium is expected to recombine andremain neutral until sources of radiation and heatdevelop that are able of reionizing it. 9 . Conclusions

A number of suggestions have been made forpossible alternative sources of intergalactic ionizing First, let us recall that the 2000 CMB resultsphotons. Sciama (1993a) gave an exhaustive descrip- require a higher baryon density than allowed bytion of the scenario of these alternatives sources. SBBN theory and observations and therefore new

The scenario through decay of dark matter par- physics—for example a strong lepton asymmetry.ticles has been proposed by different authors, the But see the latest CMB results from BOOMERANG,first was pointed by Cowsik (1977) and extented by MAXIMA, DASI and their interpretation, for in-De Rujula and Glashow (1980) from massive neut- stance in Signore and Puy (2001).

Then, we have seen that the chemistry could have7 consequences on the formation of the proto-objects.The detection of CMBR anisotropies rules out a fully ionizedintergalactic medium beyondz |300, see Scott et al. (1995). Nevertheless many important questions still await


answers. In primordial chemistry a lot of reaction of adenine (H C N ), a DNA base, may be produced5 5 5

rates are estimated only and particularly those in- during molecular cloud collapse through the chainvolved in the formation of LiH molecules. Recently reaction of HCN addition. The existence of primor-Dickinson and Gadea (2000) developed a fully dial carbon or nitrogen could also produce primordialquantal description of the radiative association in bio-molecules or pre-biotic molecules just after the

1Li 1H collisions, and showed that the rate coeffi- cosmological recombination period (see Puy, 2000a).cient is at least five orders of magnitude larger than The researches and the perpectives on primordialthat for the classical radiative association Li1H. molecules are very large. The ESA cornerstoneThus the radiative cooling due to LiH could be mission as Far InfraRed and Submillimetre Tele-changed (LiH has a strong dipole moment). scope (FIRST), or the Atacama Large Millimetre

Calculation of the rate coefficients, for rotational Array project (or ALMA), with a collecting area of2transitions induced in collisions between the primor- up 10 000 m , will offer interesting perspective of

dial molecules and atoms, is crucial. For example detection.Roueff and Zeippen (1999) presented calculations ofrotational excitation of HD molecules by He atoms,and provide a new estimation of molecular cooling A cknowledgementsand heating. In this context Le Bourlot et al. (1999)and Flower et al. (2000) showed the importance to One of the authors (D.P.) was EEC Marie Curieuse the recent quantum mechanical calculations of Post-Doc Fellowship (1994–1996) at the Internation-cross-sections for rotational transitions, in order to al School of Advanced Studies (SISSA-Trieste) onestimate respectively the H cooling and HD cool-2 the supervision of Dennis W. Sciama. Dennis W.ing. Thus recently the rate coefficients are accessedSciama gave complete freedom to choose the topics,in the software library ofCollaborative Computa- and always followed this field with constant support8 9tional Project , theUMIST data-base which provide and encouragement. The multiple discussions witha large data of reaction rates and collisional co- Dennis W. Sciama showed that a great physicist haveefficients. Primordial chemistry is still in a nascent not only a formidable knowledge of the physics, butstage: the development of computing technics and also a real sense of the human respect. This worksspecifical software for the quantum mechanics will have been supported by the Dr Tomalla Foundationopen large possibilities. and by the Swiss National Science Foundation.

Primordial molecules could play a role at theturn-around period, i.e. where the expansion of the

He was a man, take him for all in all.density perturbation is maximum and begins to

I shall not look upon this like again . . .collapse, and a modification of the temperature at

(Hamlet, William Shakespeare)this turn-around point can occur (Puy, 2000b; Puyand Signore, 2001). Moreover the primordial mole-cules play a crucial role on the evolution of the first

R eferencesobjects, through the process of fragmentation. Thefirst objects could be massive stars. Thus, as soon as

Abel, T., Anninos, P., Zhang, Y., Normal, M., 1997. NewA 3, 181.stellar processes occur, proto-objects can leadAbel, T., Bryan, G., Norman, M., 2000. ApJ 540, 39.through SN explosions to the formation of otherAninos, P., Norman, M., 1996. ApJ 460, 556.

molecules species, such as CO, CI or HCN. They in Bell, N. et al., 1998. Phys. Rev. D58, 5010.turn are important sources of contamination of the Bernshtein, I., Bernshtein, D., Dubrovich, V., 1977. AZh 54, 727.medium, and thus can offer different ways of early Boardman, W., 1964. ApJS 9, 185.

Bougleux, E., Galli, D., 1997. MNRAS 288, 638.pre-biotic molecular formation. Chakrabarti andBowyer, S., Korpela, E., Edelstein, J. et al., 1999. ApJ 526, 10.Chakrabarti (2000) showed that a significant amountBromm, V., Coppi, P., Larson, R., 1999. ApJ 527, L5.Burles, S., Tytler, D., 1996. ApJ 114, 1330.

8http: / /ccp7.dur /ac.uk. Burles, S., Nollett, K., Turner, M., 1999. astro-ph/9903300.9http: / /saturn/ma.umist.ac.uk:8000. Burles, S., Nollett, K., Turner, M., 2000. astro-ph/0008495.

http://ccp7.dur/ac.uk/






http://saturn/ma.umist.ac.uk:8000/







Cabibbo, N., Farrar, G., Maiani, L., 1981. Phys. Lett. B 105, 155. Le Bourlot, J., Pineau des Forets, G., Flower, D., 1999. MNRASChakrabarti, S., Chakrabarti, S.K., 2000. A&A 354, L6. 305, 892.Chandrasekhar, S., 1944. ApJ 100, 176. Lee, A. et al., 2001. astro-ph/0104459.Copi, C., Schramm, D., Turner, M., 1995. Science 267, 192. Lepp, S., Shull, M., 1983. ApJ 270, 578.Cowsik, R., 1977. PhRvL 39, 784. Lepp, S., Shull, M., 1984. ApJ 280, 465.Dabrowski, I., Herzberg, G., 1978. Ann. NY Acad. Sci. 38, 14. Lesgourgues, J., Peloso, M., 2000. astro-ph/0004412.Dalgarno, A., Lepp, S., 1987. In: Vardya, M.S., Tarafdar, S.P. Lifshitz, E., 1946. J. Phys. USSR 10, 116.

(Eds.), Astrochemistry, Dordrecht–Reidel Eds. Lifshitz, E., Khalatnikov, I., 1963. Adv. Phys. 12, 185.Dalgarno, A., Kirby, K., Stancil, P., 1996. ApJ 458, 397. Lisi, E., Sarkar, S., Villante, F., 1999. Phys. Rev. D59, 3520.De Bernardis, P. et al., 1993. A&A 271, 683.

MacDowell, M., 1961. Observatory 81, 240.De Bernardis, P. et al., 2000. Nature 404, 955.

MacLow, M., Shull, M., 1986. ApJ 302, 585.De Rujula, A., Glashow, S., 1980. PhRvL 45, 942.

Maoli, R. et al., 1994. ApJ 425, 372.Di Bari, P., Foot, R., 2000. astro-ph/0008258.Maoli, R. et al., 1996. ApJ 457, 1.Dickinson, A., Gadea, F., 2000. MNRAS 318, 1227.Matsuda, T., Sato, H., Takeda, H., 1969. Prog. Theor. Theor. 41,Dolgov, A., 2000. hep-ph/0004032.

219.Doroshkevich, A., Zel’dovich, Ya., Novikov, I., 1967. Soviet Aj.Melchiorri, B., Melchiorri, F., 1994. Riv. Nuovo Cim. 17, 1.11, 233.Melchiorri, A., Griffiths, L., 2000. astro-ph/0011147.Dubrovich, V., 1977. SvAL 3, 128.Melott, A.L., 1984. SvA 28, 478.Dubrovich, V., Stolyarov, V., 1997. AstL 23, 565.Mohapatra, R., Sciama, D.W., 1998. astro-ph/9811446.Ellis, G., Lanza, A., Miller, J., 1993. In: The Renaissance ofNetterfield, C. et al., 2001. astro-ph/0104460.General Relativity and Cosmology. Cambridge Univ. Press.Novikov, I., Zel’dovich, Ya., 1967. Ann. Rev. Astr. Ap. 5, 627.Ericson, W., Cronyn, W., 1965. ApJ 142, 1156.Olive, K. et al., 1981. ApJ 246, 557.Esposito, S. et al., 2000. astro-ph/0007419.Olive, K. et al., 2000. PhR 333, 389.Field, G., 1959. ApJ 129, 536.Okumurai, K., 2000. ApJ 534, 809.Field, G., 1965. ApJ 142, 531.

Field, G., Somerville, W., Dressler, K., 1966. Ann. Rev. Astron. Palla, F., Salpeter, E., Stahler, S., 1983. ApJ 271, 632.Ap. 4, 207. Parker, E., 1953. ApJ 117, 431.

ˆFlower, D., Le Bourlot, J., Pineau des Forets, G., Roueff, E., 2000. Peebles, P.J.E., 1966. PhRvL 16, 410.MNRAS 314, 753. Peebles, P.J.E., 1968. ApJ 153, 1.

Flower, D., 2000. MNRAS 318, 875. Peebles, P.J.E., Dicke, R., 1968. ApJ 154, 891.ˆFlower, D., Pineau des Forets, G., 2000. MNRAS 3116, 901. Puy, D. et al., 1993. A&A 267, 337.

Galli, D., Palla, F., 1998. A&A 335, 403. Puy, D., Signore, M., 1996. A&A 305, 371.Gamow, G., 1948. Phys. Rev. 74, 505. Puy, D., Signore, M., 1997. NewA 2, 299.Gianturco, F., Gori-Giorgi, P., 1996. PhRvA 54, 1. Puy, D., Signore, M., 1998a. NewA 3, 27.Gunn, J., Peterson, B., 1965. ApJ 142, 1633. Puy, D., Signore, M., 1998b. NewA 3, 247.Haddock, F., Sciama, D.W., 1965. PhRvL 14, 1007. Puy, D., 2000. A&A in preparation, see astro-ph/0011496.Halverson, N. et al., 2001. astro-ph/0104489. Puy, D., 2000. In: Proceedings 6th Trieste Conference, astro-ph/Hannestad, S., Raffelt, G., 1999. Phys. Rev. D59, 3001. 0011435.Hannestad, S., Hansen, S., Villante, F., 2000. astro-ph/0012009. Puy, D., Signore M., 2001. In: Primordial Cosmochemistry, NovaHirasawa, T., Aizu, K., Taketani, M., 1969. Prog. Theor. Phys. 41, Eds.

835. Rees, M., Sciama, D.W., 1966. ApJ 145, 6.Hirasawa, T., 1969. Prog. Theor. Phys. 42, 523. Rees, M., Sciama, D.W., 1968. Nature 517, 611.Hoyle, F., 1953. ApJ 118, 513. Roberge, W., Dalgarno, A., 1982. ApJ 255, 489.Hoyle, F., 1958. Proceedings of the 11th Solvay Congress. Rosenzweig, P., Parravano, A., Ibanez, M., Izotov, Y., 1994. ApJHu W. et al., 2000. astro-ph/0006436. 432, 485.Hunter, C., 1962. ApJ 136, 594. Roueff, E., Zeippen, C., 1999. A&A 343, 1005.Hutchins, J., 1976. ApJ 205, 103. Saito, M., 1969. PASJ 21, 230.Jaffe, A. et al., 2000. astro-ph/0007333. Salati, P., Wallet, J.C., 1984. Phys. Lett. B 144, 61.Jeans, J., 1902. Phil. Trans. Roy. Soc. 199A, 49. Sarkar, S., 1996. Rep. Prog. Phys. 59, 1493.Jones, B., Wyse, R., 1985. A&A 149, 144. Sarkar, S., 1999. astro-ph 9903183.Kato, S., Nariai, H., Tomita, K., 1967. PASJ 19, 130. Saslaw, W., Zipoy, D., 1967. Nature 216, 967.Kinney, W., Melchiorri, A., Riotto, A., 2000. astro-ph/0007375. Schramm, D., 1998. PNAS 95, 42.Kompaneets, A., 1957. Soviet Phys. JETP 4, 730. Schramm, D., Turner, M., 1998. RvMP 70, 303.Kondo, M., 1970. PASJ 22, 23. Sciama, D.W., 1964a. QJRAS 5, 196.Kurki-Suonio, H., 2000. astro-ph/0002071. Sciama, D.W., 1964b. Nature 201, 767.Kurki-Suonio, H., Sihvola, E., 2000. astro-ph/0011544. Sciama, D.W., 1982a. MNRAS 198, 1.Lahav, O., 1986. MNRAS 220, 259. Sciama, D.W., 1982b. MNRAS 200, 13.Latter, W., Black, J., 1991. ApJ 372, 161. Sciama, D.W., 1982c. Phys. Lett. B 114, 19.


Sciama, D.W., 1984. In: Berger A. (Eds.), Big Bang: G. Lemaitre Stancil, P., Lepp, S., Dalgarno, A., 1996. ApJ 472, 102.in a Modern World, Dordrecht–Reidel Eds. Stancil, P., Lepp, S., Dalgarno, A., 1998. ApJ 509, 1.

Sciama, D.W., 1988. MNRAS 230, 13p. Steigman, G. et al., 1977. Phys. Lett. B66, 202.Sciama, D.W., 1990a. ApJ 364, 549. Sunyaev, R., Zel’dovich, Y., 1972. Comments Astroph. SpaceSciama, D.W., 1990b. ComAp 15, 71. Phys. 4, 173.Sciama, D.W., 1993a. In: Modern Cosmology and the Dark Matter Takayanagi, K., Nishimura, S., 1960. PASJ 12, 77.

Problem. Cambridge University Press. Takeda, H., Sato, H., 1968. In: Proceedings of Symposium onSciama, D.W., 1993b. ApJ 409, L25. Cosmology held at Kyoto February 1968 (in Japanese), p. 135.Sciama, D.W., 1994. PTRSL 346, 137. Takeda, H., Sato, H., Matsuda, T., 1969. Prog. Theor. Theor.Sciama, D.W., 1998. astro-ph/9811172. Phys. 41, 840.Sciama, D.W., 1999. In: 3K Cosmology, Maiani et al. (Eds.), AIP Tegmark, M., Silk, J., Rees, M., Blanchard, A., Abel, T., Palla, F.,

Conf Proc. 476, 74. 1997. ApJ 474, 1.Sciama, D.W., Rees, M., 1966. Nature 212, 1001. Tegmark, M., Zaldarriaga, M., 2000. astro-ph/0004393.Sciama, D.W., Melott, 1982. MNRAS PhRvD 25, 2214. Tytler, D. et al., 2000. astro-ph/0001318.Scott, D., Silk, J., White, S., 1995. Science 268, 829. Uehara, H., Inutsuka, S., 2000. ApJ 531, L91.Seager, S., Sasselov, D., Scott, D., 1999. ApJ 523, L1. Wagoner, R., Fowler, W., Hoyle, F., 1967. ApJ 148, 3.Seager, S., Sasselov, D., Scott, D., 2000. ApJ 128, 407. Wagoner, R., 1969. ApJS 18, 247.Sethi, S., 1997. ApJ 474, 13. Wagoner, R., 1973. ApJ 179, 343.

´Shchekinov, Y., Entel, M., 1983. SvA 27, 622. Weymann, R., 1965. Phys. Fluids 8, 2112.Shklovskii, I., 1967. In: Proceedings 4th Texas Conf. on Relativis- Weymann, R., 1966. ApJ 145, 560.

tic Ap. 5, 627. White, M. et al., 2000. astro-ph/0008167.Signore, M. et al., 1994. ApJS 92, 535. Yang, J. et al., 1979. ApJ 227, 697.Signore, M. et al., 1997. ApL&C 35, 349. Yang, J. et al., 1984. ApJ 281, 493.Signore, M., Puy, D., 1999. NewAR 43, 18. Yoneyama, T., 1972. PASJ 24, 87.Signore, M., Puy, D., 2001. astro-ph/0108515. Zel’dovich, Y., Kurt, V., Sunyaev, R., 1968. Zh. Eksp. Teor. Fiz.Silk, J., 1977. ApJ 211, 638. 55, 287.Sofue, Y., 1969. PASJ 21, 211. Zwicky, F., 1959. PASP 71, 468.Songaila, A., Cowie, L., Hogan, C., Rugers, M., 1994. Nature 368, Zygelman, B., Stancil, P., Dalgarno, A., 1998. ApJ 508, 151.

599.

Documents

From nuclei to atoms and molecules: the chemical history of the early Universe