CERN–TH/2001-061 Non-Baryonic Dark Matter1 Leszek Roszkowski1,2 1Department of Physics, Lancaster University, Lancaster LA1 4YB, England 2TH Division, CERN, CH-1211 Geneva 23, Switzerland Abstract There exist several well-motivated candidates for non-baryonic cold dark matter, including neutralinos, axions, axinos, gravitinos, Wimpzillas. I review the dark matter properties of the neutralino and the current status of its detection. I also discuss the axino as a new interesting alternative. CERN–TH/2001-061 February 2001 1Invited plenary review talk given at 6th International Workshop on Topics in Astroparticle and Underground Physics (TAUP 99), 6-10 September, 1999, Paris, France. 2 Non-Baryonic Dark Matter † Leszek Roszkowski Department of Physics, Lancaster University, Lancaster LA1 4YB, England There exist several well-motivated candidates for non-baryonic cold dark matter, including neutralinos, axions, axinos, gravitinos, Wimpzillas. I review the dark matter properties of the neutralino and the current status of its detection. I also discuss the axino as a new interesting alternative. 1. Introduction will also comment about recent results regarding axinos. The puzzle of the hypothetical dark matter Neutrinos, the only WIMPs that are actually (DM) in the Universe still remains unresolved [1]. known to exist, do not seem particularly attrac- While there may well be more than one type of tive as DM candidates. It has long been believed DM, arguments from large structures suggest that that their mass is probably very tiny, as suggested a large, and presumably dominant, fraction of by favored solutions to the solar and atmospheric DM in the Universe is made of massive particles neutrino problems, which would make them hot, which at the time of entering the epoch of mat- rather than cold DM. This picture has recently ter dominance would be already non-relativistic, been given strong support by first direct evidence or cold. From the particle physics point of view, from Superkamiokande for neutrinos’ mass [4]. cold DM (CDM) could most plausibly be made While the new data only gives the µ τ neu- of so-called weakly interacting massive particles −3 2 trino mass-square difference of 2.2 10− eV ,it (WIMPs). it very unlikely that there would exist× two mas- There exist several interesting WIMP candi- sive neutrinos with cosmologically relevant mass dates for CDM that are well-motivated by the of 5 to 40 eV and such a tiny mass difference. underlying particle physics. The neutralino is The strength with which the WIMP candidates considered by many a “front-runner” by be- listed above interact with ordinary matter spans ing perhaps the most natural WIMP: it comes many orders of magnitude. For the neutrali- as an unavoidable prediction of supersymmetry 2 5 nos it is a fraction ( 10− 10− )ofthe (SUSY). The axion is another well-motivated can- weak strength. Interactions∼ −∼ of axions and axi- didate [2]. But by no means should one forget 2 16 nos are suppressed by (mW /fa) 10− ,where about some other contenders. While some old 10 ∼ fa 10 GeV is the scale at which the global picks (sneutrinos and neutrinos with mass in the Peccei-Quinn∼ U(1) symmetry is broken. Interac- GeV range) have now been ruled out as cosmolog- tions of gravitinos and other relics with only grav- ically relevant CDM by LEP, axinos (SUSY part- itational interactions are typically suppressed by ners of axions) have recently been revamped [3]. 2 33 (mW /MPlanck) 10− . One may wonder how Another well-known candidate is the fermionic such vastly different∼ strengths can all give the partner of the graviton, the gravitino. Most of relic abundance of the expected order of the crit- the review will be devoted to the neutralino but I ical density. The answer lies in the different ways they could be produced in the early Universe: the yInvited plenary review talk given at 6th International neutralinos are mainly produced “thermally”: at Workshop on Topics in Astroparticle and Underground some temperature they decouple from the ther- Physics (TAUP 99), 6-10 September, 1999, Paris, France. 3 mal bath. But in addition there exist also several due to assumed R-parity. A perfect candidate for mechanisms of “non-thermal” production. This aWIMP! is how CDM axinos, gravitinos can be produced The neutralino is a well-motivated candidate. (in addition to “thermal” production), as well as It is an inherent element of any phenomenologi- Wimpzillas [5]. cally relevant SUSY model. Being neutral, it is It is obvious that WIMPs do not necessarily a natural candidate for the LSP (although one have to interact only via weak interactions per should remember that this is often only an as- se. WIMPs are generally required to be electri- sumption); it couples to ordinary matter with a cally neutral because of stringent observational weak-interaction strength (reduced by mixing an- constraints on the abundance of stable charged gles) which is within the range of sensitivity of relics. On the other hand, they could in princi- present-day high energy, as well as dark matter, ple carry color charges. (For example, a stable detectors. Finally, as a bonus, it naturally gives 2 gluino above 130 150 GeV [6], or in an experi- Ωχh 1 for broad ranges of masses of SUSY mentally allowed− window 25 35 GeV [7], could particles∼ below a few TeV and for natural ranges still be the lightest SUSY particle− (LSP) although of other SUSY parameters. its relic abundance would be very small [6].) In a Much literature has been devoted to the neu- halo they would exist as neutral states by form- tralino as DM, including a number of comprehen- ing composites with gluon or quarks. We can see sive and topical reviews (see, e.g., Refs. [9,10]). that generally WIMPs are expected to have sup- Here I will only summarize the main results and pressed effective couplings with ordinary matter. comment on some recent developments and up- Otherwise, they would dissipate their kinetic en- dates. ergy. Neutralino properties as DM and ensuing im- 2 What is the WIMP relic abundance Ωχh = plications for SUSY spectra are quite model de- ρχ/ρcrit [8]? Current estimates of the lower pendent but certain general conclusions can be bound on the age of the Universe lead to drawn. First, its cosmological properties are very 2 ΩTOTALh < 0.25. Recent results from rich clus- different depending on a neutralino type. The ters of galaxies and high-redshift supernovae type relic abundance of gaugino-like (mostly bino-like) Ia imply Ωmatter 0.3. The Hubble parameter is neutralinos is primarily determined by the (light- now constrained' to 0.65 0.1. Big Bang nucle- est) sfermion exchange in the LSP annihilation ± 2 ¯ 2 4 2 osynthesis requires Ωbaryonh < 0.015. Assuming into ff:Ωh m /m . In order to have Ωχ 1, ∝ f χ ∼ that most matter in the Universe∼ is made of CDM the lightest sfermion cannot be too light (below WIMPs, one therefore obtains 100 GeV) nor tooe heavy (heavier than a few ∼ 2 hundred GeV) [11] which is a perfectly natural 0.1 < Ωχh < 0.15 (1) ∼ ∼ range of values. or so. At the very least, requiring that a dominant On the other hand, higgsino-like χ’s are fraction of CDM is located only in galactic halos strongly disfavored. Firstly, due to GUT relations 2 gives Ωχh > 0.025. among gaugino masses, higgsinos correspond to ∼ rather large gluino mass values, above 1 TeV, 2. Neutralinos which may be considered as unnatural [11]. Fur- thermore, higgsino-like neutralinos have been The DM candidate that has attracted perhaps shown to provide very little relic abundance. For the most wide-spread attention from both the mχ >mZ ,mW ,mt the χ pair-annihilation into theoretical and experimental communities is the those respective final states (ZZ, WW, tt¯)is neutralino. It is a neutral Majorana particle, the very strong. But both below and above those lightest of the mass eigenstates of the fermionic thresholds, there are additional co-annihilation partners of the gauge and Higgs bosons: the bino, 0 [12] processes of the LSP with χ1± and χ2, which wino and the two neutral higgsinos. If it is the are in this case almost mass-degenerate with the lightest SUSY particle, it is massive and stable, 4 2 LSP. Co-annihilation typically reduces Ωχh be- the theoretical criterion of naturalness: expecting low any interesting level although it has been re- that no SUSY masses should significantly exceed cently argued that the effect of co-annihilation is that value. Again, GUT relations among gaugino not always as strong as previously claimed [13]. masses cause the above constraint to provide a Higgsino-like LSPs are thus rather unattractive, much more stringent (although still only indica- although still possible DM candidates, especially tive) bound of about 150 200 GeV. Remarkably, − in the large mass (mχ > 500 GeV) regime. One in the scenario with additional GUT unification also arrives at the same∼ conclusions in the case of spin-zero mass parameters, just such a typi- of ‘mixed’ neutralinos composed of comparable cal upper bound derives from the cosmological 2 fractions of gauginos and higgsinos. This is be- constraint Ωχh < 1. Only in a relatively nar- cause, even without co-annihilation, in this case row range of parameters where the neutralino be- the neutralino pair-annihilation is less suppressed comes nearly mass degenerate with a SUSY part- and one invariably finds very small Ωχ there [11]. ner of the tau, τR, the effect of the neutralino’s Remarkably, just such a cosmologically pre- co-annihilation opens up the allowed range of mχ ferred gaugino type of neutralino typically up to about 600e GeV [17].