• accidental stability
• chiral model
• composite
• dark matter
• dark sector
• infrared fixed point
• non-abelian gauge theory

In this work we consider extensions of the Standard Model based on the same principles that justify its effectiveness, and featuring accidentally stable composite dark matter candidates.

The success of the Standard Model in describing all the observed microscopic phenomena can be understood if it is considered as an effective field theory with an high ultraviolet cut-off. Higher dimensional non-renormalizable operators are suppressed by powers of the cut-off scale and become irrelevant in the infrared. The renormalizable lagrangian features some accidental global symmetries, such as baryon and lepton number conservation, and custodial symmetry; these give a natural explanation to many experimental observations.
A fundamental ingredient of the Standard Model is its gauge theory structure, which provides a rich infrared dynamics while giving, nonetheless, a renormalizable field theory.

We consider such properties as paradigmatic, offering a theoretical rational for the success of the Standard Model, and try to use them as guidance principles to build possible extensions featuring a dark matter candidate.

We focus on extensions of the Standard Model based on a new non-Abelian gauge interaction (with gauge group G_{DC}) and new fermionic fields (dark quarks) charged under both G_{DC} and G_{SM}. We refer to the field content added to the Standard Model as the dark sector. Differently from technicolor and composite Higgs models, we require that the dark sector dynamics does not break the Standard Model gauge group.

Previous works have studied QCD-like models which have accidentally stable baryon-like states that can account for the dark matter. The purpose of this thesis is to generalise this construction and understand if there are different scenarios which, despite being based on the same framework, feature a different dynamics and phenomenology.

As a first step, we critically analyse the implicit assumptions of these models and suggest possible generalisations, identifying some interesting alternatives which have received little attention.

An interesting class of models is that of chiral models, based on complex representations of the gauge group. After discussing some general properties and what are the necessary conditions to have a consistent model and preserve G_{SM}, we focus on the case G_{DC}=SU(N) and argue on general ground, using 't Hooft anomaly matching, that in the absence of fundamental scalar fields these models have always light states. This considerations are compared with the case of the Standard Model, which is itself a chiral theory, giving a different viewpoint on some well-known facts.

We then consider models based on real representations and focus on the scenario in which the dark colour gauge group dynamics has an infrared fixed point, up to deformations induced by fermions mass terms. We discuss general model building issues and how the would-be conformal symmetry is broken by the mass term, inducing a confining dynamics.
We identify, in particular, a model with a perturbative infrared fixed point and consider two possible assignments of Standard Model quantum numbers as benchmark scenarios. The dark sector field content corresponds to five Weyl fermions, each one transforming as the adjoint of the gauge group G_{DC} =SU(3)_{DC}, together with the gauge bosons (dark gluons). We discuss the dynamics of these models, with an emphasis on asymptotic states below the confinement scale (gluonium bound states, i.e. glueballs, and gluon-quark bound states, i.e. gluequarks) and their accidental stability.

As a last step, we study the phenomenology of the two models, with particular attention to the consistency with cosmological observations. Assuming that all the dark quarks have the same mass, the model can be characterised by two scales, namely the confinement scale and the dark quark mass MQ. A natural hierarchy of scales arises from the structure of the model and we study the different phenomenological regimes as a function of the two scales.

After estimating the glueball lifetime with an effective field theory approach, we identify two relevant regimes for the models: one in which the glueballs are stable on cosmological scales and one in which they decay quickly.

The first scenario has a non-standard cosmology with so called "cannibalism" in the glueball sector (i.e. number changing interactions involving only glueballs, such as 3 to 2 processes). This translates into an unusual scale dependence of the dark sector temperature. We evaluate the glueballs relic density and the cosmological bounds, concluding that this parameter space region is excluded by observations.

In the second scenario the only stable relics are gluequarks. To evaluate their relic density, we compute the cross section for the annihilation of dark quarks in dark gluons, including the Sommerfeld enhancement correction. In order to reproduce the dark matter relic abundance, dark quarks with mass in the range MQ = [1,10] TeV are needed.