Arxiv:2102.02048V2 [Gr-Qc] 9 Feb 2021

arXiv:2102.02048v2 [gr-qc] 9 Feb 2021 tto nteguig(oaiig fsaetm sym- space-time of (localizing) [ gauging metries the in itation h akrudsaetm erca oprdt the to compared the as of manner, modifications metric this into space-time In excite translate background to fermions the torsion. as of by so density fed enough spin environments dynamics strong those is new in density the spin phenomenology the new where couplings for matter new look these of to advantage takes far so ture considered. be [ also threshold can couplings Cartan’s scales minimal the at cor- than manifest imple- the and different be in densities, present can effects Lagrangian derivatives fields responding covariant The fermionic the and density. via bosonic mented Cartan’s rel- in torsion by become torsion the given that of feeds scales equations, field tensor at metric evant cur- spin the the of into theory source effects to EC the Analogously in being absent. vature, tensor is which energy-momentum torsion fields, matter when the the present in not connec- couplings affine are new the for in allows encoded freedom tion new Relativity the General but of (GR), Einstein-Hilbert the by given mally emtyi ie yteEnti-atn(C theory (EC) Einstein-Cartan RC the the by of [ given implementation simplest is The geometry connection) present. curvature, affine the also to of is addition part antisymmetric in (the gauging where torsion the Poincaré group, to the linked of geometry, (RC) Riemann-Cartan ∗ ‡ † 4 ,weetecrepniggaiainlato sfor- is action gravitational corresponding the where ], [email protected] [email protected] [email protected] o-imningoere aual peri grav- in appear naturally geometries Non-Riemannian otrsac are u ntecrepniglitera- corresponding the in out carried research Most mrnsfo imn-atnsaetm nteeeg lev energy the on space-time Riemann-Cartan a from Imprints 1 – 3 1 i diinlculnswt h etradailfermioni axial and vector tor the background with the couplings to model- coupled additional a minimally via using fermions of geometry case Riemann-Cartan the a within fermions of omlg n oi tt hsc sn urn capabiliti s current the using to physics related state (pa effects solid torsion the and of include cosmology detection to the extension on its elaborate betwee and couplings theory non-minimal determine Einstein-Cartan discuss and also fe th We quantities), of to torsion transitions. levels respect of energy with combination the aligned/anti-aligned specific on are they effect ca whether Zeeman-like the on for a and find zero-curvature, we of torsion, limit the In interactions. .Apriual neetn aei the is case interesting particularly A ]. nttt eAto´sc ieca oEspa¸co, Faculdad Ciências do Astrof´ısica e de Instituto nti ok eivsiaeteeet ftetorsion-fermi the of effects the investigate we work, this In .INTRODUCTION I. rnic Cabral Francisco nvria opues eMdi,E200Mdi,Spain Madrid, E-28040 Madrid, de Complutense Universidad dfıi 8 ap rne -7906Lso,Portugal Lisbon, P-1749-016 Grande, Campo Edif´ıcio C8, ID 2 eatmnod ´sc eoiaadIPARCOS, F´ısica Teórica and de Departamento , 1, ∗ 5 rnic .N Lobo N. S. Francisco – 7 ,wienon- while ], Dtd eray1,2021) 10, February (Dated: nti epc,a tcnpoieabekn ftede- the of breaking and a gravity [ for provide theories space-time can different relevance it these between utmost as generacy of respect, geom- is this space-time phenomena in the gravitational with in fields etry fermionic of teraction em sicue nteato fteter see.g. (see theory the these of of action any the upon in built [ included object is GR scalar terms of lowest implementations the equivalent when interpreta- physically alternative but geom- geometrically tions metric-affine for a non- allowing as and con- known etry), torsion, affine is curvature, (what the its pieces into general, metricity split in RC be that, a can recall nection on of us behaviour propagating Let the when other geometry. in themselves of interested fields favour are fermionic in we torsion removed here the effectively observables, where been background have space-time fields the the of instead in Indeed, effects torsion. by induced effects physical edaekonsnetesmnlwrsb akr[ Parker by works seminal the since gravitational one-electron known arbitrary are free-falling an field in on curvature levels by induced energy atoms the to cations observable direct effects. have physical can non-vanishing) (if connection widely [ been characterized early have the and dynamics or explored cosmological objects applications, compact late-time of many and astrophysics consequence, the a as such As predictions. GR tandwti h el onfil fmultimessenger of field con- born be [ newly subsequently astronomy the these can within predictions process, strained their this and Through strong models the regime. in physics field new GR gravitational of their existence the- the with these for dis- compared observational criminators of constitute be predictions could that can new and that counterparts, find torsion to with is ories analysis this of 19 ID h anamo hswr st okfrmr direct more for look to is work this of aim main The h xsec fosral mrnso h modifi- the on imprints observable of existence The o nitrsigdsuso nti oi) h in- The topic). this on discussion interesting an for ] , 1, eCˆnisd nvriaed Lisboa, de Universidade Ciências da de e es. e fcntn n peial symmetric spherically and constant of ses † urnswihicueparity-breaking include which currents c emoi ed n oso ihnthe within torsion and fields fermionic n h orsodn n-tutr energy fine-structure corresponding the ltigo nrylvl nastrophysics, in levels energy of plitting xa etrpr ftetrin(rto (or torsion the of part vector axial e n ig Rubiera-Garcia Diego and iybekn)Hlttr.Fnlywe Finally term. Holst rity-breaking) ina ela o-iia extensions non-minimal as well as sion ncitrcino h nrylevels energy the on interaction onic mosadat-emosdepending anti-fermions and rmions needn prah econsider We approach. independent 18 3 .Ide,tedffrn icso h affine the of pieces different the Indeed, ]. ]. l fDrcspinors Dirac of els 8 – 17 .Temi outcome main The ]. ID 2, ‡ 20 – 2

22]. On the other hand, the presence of non-metricity where ψ denotes spinors and ψ¯ = ψ+γ0 its adjoint, while has been recently found to yield measurable effects via jλ = qψγ¯ λψ is the U(1) charge current density vector. new 4-fermion contact interactions that can be used to The (Fock-Ivanenko) covariant derivatives in this space- put constraints on the scale of non-metricity, for in- time are given by stance, in electron-positron scattering in particle acceler- 1 ab ators [23, 24]. In this paper, we show that both minimal D˜ µψ = ∂µψ + wãbµσ ψ , (2) and non-minimal couplings of torsion to Dirac fermions 2 1 ab in different backgrounds also yield measurable effects, via D˜ µψ¯ = ∂µψ¯ wãbµψσ¯ , (3) a splitting in the energy levels of fermions/antifermions − 2 (driven by the axial vector part of the torsion tensor in ab 1 a b 1 [a b] while the matrices σ 4 γ ,γ = 2 γ γ are the minimal couplings and also by the trace-vector part for Lorentz group generators≡ in the spinorial representation. specific non-minimal extensions), depending on the rela- The effect of space-time curvature is encoded in the Levi- tive orientation between spin and torsion vector/pseudo- Civita 1-form spin connection,w ãbµ, of the Riemann ge- vector quantities, which resembles a Zeeman-like effect, 1 ometry. allowing for transitions between such different levels . The corresponding Dirac equation is obtained by vary- This work is organized as follows: In Sec. II, we estab- ing the action S = 1 d4x√ g ˜ of the Dirac La- 2κ2 − LDirac lish the main equations of fermions minimally coupled to grangian (1) with respect to ψ¯ and reads as the torsion field, and find that they only interact via the R axial vector part of the torsion tensor. In Sec. III, we µ µ i~γ D˜ µψ + (qγ Aµ m)ψ =0 . (4) find the energy levels for fermions/anti-fermions in the − simplified regime of zero-curvature, taking the ansatze of In a flat (Minkowski) space-time, and taking the quasi (static) constant and spherically symmetric torsion, re- non-relativistic limit (leaving only terms up to (v/c)2), spectively, and find the frequency of the corresponding one finds the time independent equation in the static ex- transitions. These results are extended in Sec. IV to ternal electromagnetic potential Aµ = (φ, A~ ) (here we the case of non-minimal couplings between the fermionic reinsert the speed of light c for convenience): currents (vector and axial currents) and the torsion (the axial and trace-vector irreducible parts) where some in- 1 2 pˆ4 q~2 1 ˆ ˆ ~pˆ qA~ + qφ + ∂ φ S~ L~ teraction terms break the parity symmetry. In Sec. V 2m − − 8m3c2 4m2c2 r r · we briefly discuss the topic of non-minimal couplings q~ ˆ ˆ q~2 within the EC theory and its generalization to include S~ B~ ∂ φ ∂ E ψ(~r)=0 , (5) − m · − 4m2c2 r r − the parity-breaking Holst term. We conclude in Sec. VI with an extended discussion on different types of physi- with E mc2 and qφ mc2, and spherical symmetry, cal effects connected to the splitting of fermionic energy φ = φ(r≪), is assumed.≪ The solution to these equations levels driven by torsion in astrophysics and cosmology. gives the four-spinor ψ = ψ(~r)e−iEt/~, which corresponds to the eigenfunction of the Hamiltonian with energy E. ~ˆ ~ II. TORSION-FERMIONS COUPLINGS In the expression above, S ~σ/2 is the intrinsic angular momentum (spin), and ~σ =∼ (σ1, σ2, σ3) is the Pauli matrices spatial vector. As is well known, the second term A. Dirac fermions in a flat space-time on the left-hand side of Eq. (5) is a relativistic correction to the three-momentum, the forth and the fifth terms Let us start our analysis, for self-consistency and self- give the spin-orbit and Zeeman-effect magnetic energy, completeness, by reviewing the corrections to the non- respectively, and the sixth term is the so-called Darwin relativistic Schrödinger equation induced by the Dirac term correction. equation. Accordingly, consider the Dirac Lagrangian in If we consider the case of an electron in the Coulomb curved space-time and minimally coupled to the electro- potential φ = Ze2/r, then the corresponding energy 2 magnetic field, given by levels of this system− are given by i~ 2 2 4 4 ˜ ¯ µ ˜ ˜ ¯ µ ¯ λ 2 Z α Z α n 3 Dirac = ψγ Dµψ (Dµψ)γ ψ mψψ + j Aλ , E = mc 1 L 2 − − − 2n2 − 2n4 j +1/2 − 4 (1) + (Z6α6) , (6) O where n is the principal quantum number, j is the to-

1 tal angular momentum quantum number, and α is the For further discussions on the experimental manifestations of tor- fine structure constant. The first term correction inside sion in the interaction between spinors, torsion, and electromagnetic fields see e.g. [25–28]. the brackets is the relativistic correction of the energy 2 From now on hatted quantities will denote computation with associated to the mass of the electron, the second term the standard Levi-Civita connection of the curvature (i.e., the corresponds to the Bohr energy levels, while the next Christoffel symbols of the metric). term is the fine structure (spin-orbit) correction. As 3 an example, the fine structure between the energy levels where ˜ is given by Eq. (1). LDirac (nlj )2P3/2 and 2P1/2 corresponds to an energy difference Using the canonical properties of the tetrads we can 2 2 4 c µ µ d ∆E = mc Z α /32. write the contractions Kabµ = ϑ µKabc and γ = ed γ , | If| one considers, instead of the Minkowski limit, the so that the Lagrangian density above can also be written full curved space-time background of a Riemannian ge- as ometry, then there will be gravitational metric-induced ~ i c a b corrections to the energy levels [22], which should be- Dirac = ˜Dirac + Kabcψ¯ γ ,γ γ ψ . (13) L L 8 { } come non-negligible for strong gravitational fields. It is thus natural to wonder whether new physical effects will Next, by using the identities γc,γaγb = 2γ[cγaγb] = cabd 5 { } manifest if one generalizes the Riemann geometry to in- 2iǫ γdγ we can rewrite it as clude torsion in a RC space-time. As we will see, in − ~ the minimal coupling scenario a clear analogy with the i [c a b] Dirac = ˜Dirac + Kabcψγ¯ γ γ ψ , (14) Zeeman-effect term can be recognized. L L 4

and noting that K[αβγ] = Tαβγ, we arrive at the final expression B. Fermions minimally coupled to torsion λ Dirac = ˜Dirac +3T˘ s˘λ , (15) Let us consider a free Dirac fermionic field mini- L L mally coupled to the RC space-time geometry (for a where more detailed analysis of fermions in RC and metric- ~ λ ¯ λ 5 affine geometries see for example [29–34]). The tor- s˘ ψγ γ ψ , (16) λ ≡ 2 sion tensor of the affine connection Γµν is defined by 1 λ λ is the Dirac axial spin vector current. Note that in this Tµν (Γ Γ ), with the irreducible components ≡ 2 µν − νµ expression we have reinserted the space-time (holonomic) T λ = T¯λ + 2 δλ T + gλσǫ T˘ρ, and the traceless µν µν 3 [ν µ] µνσρ indices. This simple expression, which is valid for any ¯λ λµνρ ¯ tensor obeying T µλ = 0 and ǫ Tµνρ = 0, while Tµ is Dirac field minimally coupled to a RC space-time geom- the trace vector and etry (regardless of the gravitational theory chosen) means that, in the minimal coupling case, Dirac fermionic fields ˘λ 1 λαβγ T ǫ Tαβγ , (7) only interact with the axial vector part of torsion. ≡ 6 The axial vectors ˘λ in Eq. (16) can be understood as is the pseudo-trace (axial) vector, which will play a key representing the density of fermionic spin (spin/volume role in our work. or energy/area, in c = 1 units). To see this more explic- The minimally-coupled fermionic (Dirac) piece, Eq. itly, let us consider the γaγ5 matrices for a = 0, 1, 2, 3, (77), in presence of torsion, is given by that is

i~ i = ψγ¯ µD ψ (D ψ¯)γµψ mψψ¯ , (8) a 5 0 I σ 0 Dirac 2 µ µ γ γ = , i , (17) L − − ( I 0 0 σ ) − − which is formally equal to Eq. (1), but where the generalized Fock-Ivanenko covariant derivatives of spinors are (with i = 1, 2, 3), respectively, and with I representing now defined as the 2 2 identity matrix. Since× the eigenvalues of the Pauli matrices are λ = 1 ab 1 [a b] 1 for the spin up/down configurations, and using the Dµψ = ∂µψ + wabµσ ψ = D˜ µψ + Kabµγ γ ψ , (9) 2 4 fact± that in the usual Pauli-Dirac representation the σ3 and matrix is already diagonal, we can use this direction as the one relative to which we define the up and down spin 1 ab 1 [a b] states. Then one can show that Dµψ¯ = ∂µψ¯ wabµψσ¯ = D˜µψ¯ Kabµψγ¯ γ , (10) − 2 − 4 ~ s˘3 n , (18) where the Lorentzian spin connection (wabν = wbaν ) of | |∼ 2 the RC space-time can be written as the spin connection− where n is a normalization constant giving the number of the Riemann geometry plus the so-called contortion of particles (or anti-particles) per volume. tensor K T +2T , that is αµν ≡ αµν (µν)α The torsion-spin interaction term in Eq. (15) actually wabµ =w ãbµ + Kabµ . (11) resembles a Zeeman-like effect with the axial spin vector playing the role of an external magnetic field: Replacing the expressions of the covariant derivatives (9) T˘ ˘s . (19) and (10) in the Lagrangian density (8) yields Lts ∼ · i~ where bold letters indicate a product of two (axial) vec- = ˜ + K ψ¯ γµ,γaγb ψ , (12) LDirac LDirac 8 abµ { } tors. 4

We can now find the Dirac equation corresponding equation to the Lagrangian density (15) for spinors and adjoint 3~ spinors as i~γk∂ ψ + m T˘3γ γ5 ψ(~r)= γ0Eψ(~r) . (25) − k − 2 3 3~ i~γµD˜ ψ mψ = T˘λγ γ5ψ , (20) µ − − 2 λ In terms of their components this equation reads ~ ~ ˜ ¯ µ ¯ 3 ˘λ ¯ 5 ~ i (Dµψ)γ + mψ = T ψγλγ , (21) ~ k II 3 ˘3 3 I − 2 i σ ∂kψ = E m T σ ψ , (26) − − − 2 respectively. In the next section we shall study specific ~ ~ k I 3 ˘3 3 II solutions of this system in order to determine the energy i σ ∂kψ = E + m T σ ψ . (27) − − 2 levels for fermions and anti-fermions. Moreover, taking into account the harmonic solution ψ(~r)= χei~k·~r = χei~p·~r/~, corresponding to a well defined III. IMPRINTS OF TORSION UPON momentum state, where χ is a constant 4-spinor, we get FERMION/ANTI-FERMION ENERGY LEVELS the system of equations for the χ components as

A. Constant background axial torsion 3~ p χ4 ip χ4 + p χ3 = E m T˘3 χ1, 1 − 2 3 − − 2 We start our analysis by taking again the zero- 3~ p χ3 + ip χ3 p χ4 = E m + T˘3 χ2, curvature limit and, moreover, we specify an axial tor- 1 2 − 3 − 2 sion vector along one specific direction (for example the 3~ z axis of a cartesian coordinate system). Under these p χ2 + ip χ2 p χ1 = E + m T˘3 χ3, − − 1 2 − 3 − 2 conditions, the Dirac equation (20) reads ~ 1 1 2 3 3 4 ~ p1χ ip2χ + p3χ = E + m + T˘ χ , ~ α 3 ˘3 5 − − − 2 i γ ∂αψ = mψ T γ3γ ψ . (22) − 2 respectively. More explicitly, using the expressions of the Pauli ma- Since χ is assumed to have constant components, the trices and recalling that γ3 = γ , one can recast this − 3 background torsion itself has to be constant too. In this equation into the dynamical system static, constant background axial torsion regime, assum- ˘3 ~ ing again that torsion is positively oriented, T > 0, there I k II 3 3 3 I i~ ∂tψ + σ ∂kψ = m + T˘ σ ψ , (23) are two independent solutions for the spinor ψ(~r, t) = 2 χei(~p·~r−Et)/~, corresponding to the free particle momen- 3~ tum eigenstates with spin up and spin down. But, as i~ ∂ ψII + σk∂ ψI = m T˘3σ3 ψII , (24) − t k − 2 opposed to Dirac theory in Minkowski space-time, in this case the presence of torsion breaks the degeneracy in en- where we have introduced the notation ψI = (ψ1, ψ2) and ergy and these two states have different (positive) energy ψII = (ψ3, ψ4), while k = 1, 2, 3, and the 2 2 identity values. As an example, consider the case of motion along × matrix I is implicit in the first terms of the left-hand side the p1 direction for this eigenstate, for which we get and in the first (mass) terms on the right-hand side. ~ From these equations one can see that the axial-axial 4 3 3 1 p1χ = E m T˘ χ , interaction between the fermionic spin density and the − − 2 background space-time torsion gives a spin-dependent 3~ p χ3 = E m + T˘3 χ2, energy (depending on the relative orientation between 1 − 2 the axial spin vector and the background space-time tor- 3~ sion). Therefore, an electron or any massive free fermion p χ2 = E + m T˘3 χ3, 1 − 2 in a well defined momentum (eigen)state will have two possible energy levels depending on the alignment/anti- 3~ p χ1 = E + m + T˘3 χ4. alignment between its spin and the axial torsion vector, 1 2 which is analogous to the Zeeman effect. Moreover, if we assume that T˘3 > 0, then the anti-alignment is preferred The two possible energy solutions for the particle are for the fermion as it corresponds to the lower energy level then given by and the same result occurs for the anti-fermion. 2 To make our analysis more concrete, let us assume the 3~ E2 = p2 + m T˘3 , (28) simpler case of a static, constant torsion field. Consider ± ± 2 then a 4-spinor ψ = ψ(~r )e−iEt/~, corresponding to the eigenfunction of a well-defined energy state. After sub- for the spin up/down, respectively. The independent so- stituting in Eq. (22), we obtain the time-independent lutions for the spin up (aligned) state and the spin down 5

(anti-aligned) state are given by and those for anti-particles:

p ip p3 1 0 1 − 2 3~ 3~ 0 1 E + m + T˘3 E + m T˘3    p   2   − 2  0 N p ,N ~ , p3 p1 + ip2 3 3 N  −  , N   .  ~  E + m T˘   3~   3~   3   − 2   ˘3   ˘3  E + m + T˘3    E + m + T  E + m T   2   0   2   − 2               0   1      respectively, where N is a normalization constant (typ-  1   0  ically chosen to satisfy ψ†ψ = 2E) given in this case    (32) In all cases the energy of the anti-aligned state is lower by N = E + (m 3~ T˘3/2) for the spin up/down ± than the aligned state. (aligned/anti-aligned)q state. From this discussion, we Let us denote by mT˘ the mass correction due to the see that not only the axial-axial torsion-spin interaction spin-torsion interaction, and consider the two possible is analogous to a Zeeman effect but also the equations energy levels E1 and E2, with E2 > E1. We therefore reveal that one could think of the fermion state with the get the expression for the energy transition spin aligned with torsion as being slightly more massive 4mmT˘ than the fermion state with the spin anti-aligned to the hν = E2 E1 = , (33) axial torsion. In the coupling to the space-time struc- − p˜+ +˜p− ture, torsion is therefore providing an effective mass to 2 2 2 wherep ˜± = p + (m mT˘) , and in the reference frame fermions that distinguishes between spin states. of the particle we obtain± Let us also note that in this regime of static constant ~ background torsion there are two more independent solu- 1 3 hν = E2 E1 = m ˘ = T.˘ (34) tions for the spinor ψ(~r)= χe−i(~p·~r−Et)/~, corresponding − 2 T 4 to the free anti-particle momentum eigenstates with spin Therefore, reinserting the speed of light in vacuum, we down or spin up, respectively. In this case, we obtain get p 3c 0 ν = T˘ (35) p 3~ 8π E + m + T˘3  3~   2  N E + m T˘3 ,N , If we consider, for instance, T˘ 10−16m−1, then we end 0  − 2    up with the prediction of a transition∼ in the ν nHz  1   0      regime. ∼  0   1         (29) respectively, with N = E + (m 3~ T˘3/2) for the spin B. Spherically symmetric torsion background ∓ down/up (anti-aligned/aligned)q states, respectively. The two corresponding energy levels are In this section, we analyze the case of a static, spherically symmetric torsion background, which is relevant for 3~ 2 astrophysical applications. The Dirac equations in this E2 = p2 + m T˘3 , (30) case are still given by Eqs. (20) and (21). To estimate the ∓ 2 effect of this scenario on the energy levels we consider the following axial torsion ansatz around some astrophysical for the spin down/up states. source: For completeness, let us also mention that in the general case of motion along any direction, with ~p = T˘µ(r)= bµf(r) , (36)

(p1,p2,p3), then we would reach similar conclusions with µ the spin up and spin down solutions for particles: where b is a constant (axial) 4-vector. If we neglect the eﬀect of curvature, the limit of the generalized Dirac equation above is 1 0 0 1 ~ ~ α 3 ˘λ 5  p3   p ip  i γ ∂αψ = mψ T (r)γλγ ψ . (37) 1 − 2 − 2 3~ 3~ N E + m + T˘3  ,N E + m T˘3  , The torsion-spin interaction can be seen as a small per-  2       − 2  turbation to an (unperturbed) time-independent Hamil-  p1 + ip2   p3     −  tonian. Using perturbation theory to ﬁrst order, we have  3~   3~  E + m + T˘3  E + m T˘3  then  2   − 2         (31) E E + ψ Uˆ ψ , (38) ≃ (0) (0) ts (0)

6 where ψ(0) are the eigenstates of the unperturbed Hamil- quantity tonian associated to the eigenvalue E(0). Again, tak- ~ − ~ 3 † ˆ ing the 4-spinor ψ = ψ(~r)e iEt/ , corresponding to the ψ Uˆ ψ = ψ (~r)T˘λ(r)γ γ5ψ (~r)d3r (0) ts (0) − 2 0 λ 0 eigenfunction of a well-defined energy state, we obtain Z 3~b the time-independent equation = 0 f(r)ψ†(~r)γ γ5ψ (~r)d3r(45) − 2 0 0 0 ~ Z k 3 λ 5 0 ~ 3 i~γ ∂kψ + m T˘ (r)γλγ ψ = γ Eψ , (39) 3 i † i 5 3 2 + b f(r)ψ (~r)γ γ ψ0(~r)d r , − − 2 0 i=1 Z ˆ X such that the torsion-spin operator Uts reads which, taking into account the quite useful general rela- ~ tions 3 ˆλ 5 Uˆts = T˘ (r)γλγ . (40) − 2 z1 z3 z1 z2 5 z2 z4 1 5 z2 z1 γ0γ = , γ γ = , Now, consider the 4-spinor state z3  z1 z3  z4 z −z z −z  4 − 2  4 − 3 1 0         z1 iz2 z1 z1 0 1 − ψ = ψ1 + ψ2 2 5 z2 iz1 3 5 z2 z2 (0) (0) 0 (0) 0 γ γ = , γ γ = − , z3  iz4  z3  z3 E 0 E 0 −     z4 iz3 z4 z4   −       0  0         0 0 and assuming the particle with the spin up configuration + ψ3 + ψ4 , (41) i~p·~r/~ p i~p·~r/~ (0) 1 (0) 0 in Eq. (43), with z1 = Ne , z4 = N E+m e and E E z2 = z3 = 0, we arrive at 0 1     3~b3 p2 solution to the unperturbed Hamiltonian. In configura- ψ Uˆ ψ = N 2 1+ F (r) , (0) ts (0) 2 (E + m)2 tion space, this expression becomes (0) (46) 3 3 1 0 where F (r)= b f(r)d r represents the geometrical fac- 0 1 tor coming from a spherically symmetric torsion func- ~r ψ = ψ1 (r) + ψ2 (r) R (0) (0) 0 (0) 0 tion integrated over the relevant volume of the spatial 3-dimensional hypersurfaces, for a specific space-time fo- 0 0     liation. Doing the same exercise with the spin down 0 0 i~p·~r/~ p i~p·~r/~ state with z2 = Ne , z3 = N e and 0 0 E(0)+m +ψ3 (r) + ψ4 (r) . (42) (0) 1 (0) 0 z1 = z4 = 0, we also get Eq. (46) but with a global mi- nus sign. The energy difference between these two states 0 1     thus becomes If we assume the motion to take place along a specific p2 direction, then there are four independent solutions, two δE =3~b3N 2 1+ F (r) , (47) (E + m)2 for the particle states (up/down): (0) which corresponds to the frequency 1 0 0 1 ~ ~ δE 3E(0) N   ei~p·~r/ ,N  p  ei~p·~r/ , (43) ν = = F (r) , (48) 0 h π p E + m    (0)  E + m  0   (0)    where we have used again the conventional normalization     † ψ0ψ0 = 2E0. Note that the same result for the energy and two for the anti-particle states (down/up) levels would have been obtained if we had considered all the components of the 3-momentum. 0 p p E(0) + m −i~p·~r/~ −i~p·~r/~ N E(0) + m e ,N  0  e , IV. NON-MINIMALLY COUPLED FERMIONS 1 0     TO TORSION  0   1          (44) Let us study now a fermionic Dirac Lagrangian non- 2 2 2 with N = E(0) + m and E(0) = p + m . minimally coupled to the RC geometry. Consider the Next, we need to compute from Eq. (40) the following vector and axial vector fermionic currents jλ ψγ¯ λψ p ≡ 7 and aλ ψγ¯ λγ5ψ, which are coupled to torsion via the again the solutions ψ = ψ(~r)e−iEt/~ = e−i(Et−~p·~r)/~. Ac- matter Lagrangian≡ density cordingly, we have the following time-independent Dirac equation fermions = ˜Dirac + α1T j + α2T˘ a , (49) L L · · k λ λ i~γ ∂kψ + m α1T + β2T˘ γλ where T λ T νλ is the trace vector part of torsion and − − ν h α , α some≡ constants. For α = 0 and α =3~/2 we re- α T˘λ + β T λ γ γ5 ψ(~r)= γ0Eψ(~r) . (55) 1 2 1 2 − 2 1 λ cover the case of the minimal coupling to torsion analyzed i k 5 in the previous section. The extended Dirac equation in Using the properties of the matrices γk = γ , γkγ and this case reads γ0, we can also write this equation via the− Hamiltonian matrix µ λ λ 5 i~γ D˜ µψ mψ = α1T γλψ α2T˘ γλγ ψ . (50) − − − Hψˆ (~r)= Eψ(~r) , (56) This model for the free fermion in a RC space-time geometry has parity symmetry, and can be extended into which explicitly reads as a family of parity-breaking models. To include parity- m (t0 ~τ ~σ) ~σ ~pˆ (τ 0 ~t ~σ) ψ breaking terms (for a more detailed account of parity − − · · − − · I violation in the general framework of Poincare theories ~σ ~pˆ + (τ 0 ~t ~σ) m + (t0 ~τ ~σ) ψII − · − · − · of gravity see the recent work [35]) in the Lagrangian E 0 ψI density (which are expected to be relevant in the early = , (57) 0 E ψII Universe due to the matter-antimatter asymmetry), we − consider the additional couplings T a and T˘ j, which ˆ ~~ · · where ~p = i is the 3-momentum operator, and we yields the new Lagrangian density have introduced− ∇ the following notation for the torsion quantities fermions = ˜Dirac+(α1T+β2T˘ ) j+(α2T˘ +β1T) a , (51) L L · · ν ν ν λ λ λ t α1T + β2T˘ , τ α2T˘ + β1T . (58) with new coefficients β1,β2. The corresponding general- ≡ ≡ ized Dirac equation is given by Alternatively, this system can also be written in the more convenient way i~γµD˜ ψ mψ = α T λ + β T˘λ γ ψ µ − − 1 2 λ ˆ m E ~σ ~p ψI λ λ 5 − · α2T˘ + β1T γλγ ψ , (52) ~σ ~pˆ m + E ψII − − · 0 0 t ~τ ~σ τ ~t ~σ ψI for spinors and = 0− · 0− · , (59) τ + ~t ~σ t + ~τ ~σ ψII − · − · ˜ ¯ µ ¯ λ ˘λ ¯ i~(Dµψ)γ + mψ = α1T + β2T ψγλ which highlights the fact that the matrix on the right- − hand side contains the geometrical effects due to tor- α T˘λ + β T λ ψγ¯ γ5 , (53) − 2 1 λ sion, including spin-torsion interactions of both parity- breaking and parity-preserving types. for the adjoint spinors. The eigenvalue problem above is a system of two cou- To estimate the new physics involved in this model we pled equations for the 2-spinors ψI and ψII . To solve it will take again the zero-curvature limit in order to iden- we use the general form of the spinor ψ(~r)= χei~p·~r/~ and tify the effects of torsion and have a qualitative notion of the properties of Pauli matrices, so that the first of these its consequences in the context of beyond the standard equations can be written as model of particle physics interactions. In this limit, the 3 0 1 2 II Dirac equation (52) becomes p3 + t τ p1 + t i(p2 + t ) χ1 1 − 2 − 3 0 II p1 + t + i(p2 + t ) p3 t τ χ − − − 2 ~ µ λ ˘λ 0 3 1 2 I i γ ∂µψ mψ = α1T + β2T γλψ E m + t τ (τ iτ ) χ1 − − = − 1 −2 − − 0 3 I . (60) (τ + iτ ) E m + t + τ χ2 α T˘λ + β T λ γ γ5ψ , (54) − − − 2 1 λ Now let us consider the two orthogonal spin up/down As in the minimally coupled case, one can in principle 1 0 solutions for the particle: χI = and χI = , and consider different ansatze for the background torsion de- 0 1 pending on whether one is interested, for instance, in obtain the corresponding 4-spinor solutions. In the first gravitational wave astronomy (setting a dynamic, har- (spin up) case, we get the system of equations monic torsion), or in simple models of the RC geometry (p + t3 τ 0)χII + (p + t1 i(p + t2))χII around spherical compact objects (setting a spherically 3 − 1 1 − 2 2 symmetric ansatz). For free fermionic spinors we try = E m + t0 τ 3, (61) − − 8

1 2 II 3 0 II (p1 + t + i(p2 + t ))χ1 + ( p3 t τ )χ2 1 1 2 − − − 0 = (τ + iτ ), (62) and therefore we ﬁnd the solution χ =  II , with − χ1 χII   2    (E m + t0 τ 3)( p t3 τ 0) + (p + t1 i(p + t2))(τ 1 + iτ 2) χII = − − − 3 − − 1 − 2 , (63) 1 (p + t3 τ 0)( p t3 τ 0) (p + t1 i(p + t2))(p + t1 + i(p + t2)) 3 − − 3 − − − 1 − 2 1 2

(τ 1 + iτ 2)(p + t3 τ 0) (p + t1 + i(p + t2))(E m + t0 τ 3) χII = − 3 − − 1 2 − − . (64) 2 (p + t3 τ 0)( p t3 τ 0) (p + t1 i(p + t2))(p + t1 + i(p + t2)) 3 − − 3 − − − 1 − 2 1 2

1 2 II 3 0 II Note that, in the vanishing-torsion (Minkowski) limit (p1 + t + i(p2 + t ))χ1 + ( p3 t τ )χ2 II p3 II p1+ip2 0 − 3− − we obtain χ1 = E+m , χ2 = E+m , which is exactly the = E m + t + τ , (66) 4-spinor solution corresponding to the free fermion, spin − up state, with E2 = p2 + m2. As for the second (spin down) case, we obtain the system 3 0 II 1 2 II 0 (p3 + t τ )χ1 + (p1 + t i(p2 + t ))χ2 − 1 2 − 1 = (τ iτ ), (65) and therefore we ﬁnd the solution χ =  II , with − − χ1 χII   2    (τ 1 iτ 2)( p t3 τ 0) (p + t1 i(p + t2))(E m + t0 + τ 3) χII = − − − 3 − − − 1 − 2 − (67) 1 (p + t3 τ 0)( p t3 τ 0) (p + t1 i(p + t2))(p + t1 + i(p + t2)) 3 − − 3 − − − 1 − 2 1 2

(E m + t0 + τ 3)(p + t3 τ 0) + (p + t1 + i(p + t2))(τ 1 iτ 2) χII = − 3 − 1 2 − . (68) 2 (p + t3 τ 0)( p t3 τ 0) (p + t1 i(p + t2))(p + t1 + i(p + t2)) 3 − − 3 − − − 1 − 2 1 2

II p1−ip2 Again we have the correct Minkowski limit, χ1 = E+m , Analogously, for the spin down particle we get the solu- II −p3 tion and χ2 = E+m , describing the free particle, spin down state. Note that, proceeding in a similar manner, we 0 could derive the corresponding expressions for the 4- 1 spinor solutions associated to the anti-fermion in the spin eff eff p + ip  ~ up/down states. ψ = 1 2 ei(~p·~r−Et)/ , (74)  E + meff  To simplify further our analysis let us consider the  eff  µ 1 2 3  p3  ansatz for the torsion components t = (0,t ,t ,t ), and  −  µ  E + meff  τ = (0, 0, 0, τ) in Eq. (58). The spin up particle solution   is then given by where Eqs. (70)–(72) still hold but now the effective 1 mass in Eq. (73) becomes meff m τ. Therefore, ≡ − 0 two different energy levels are obtained for the spin up  eff  and spin down states. The energy of the anti-aligned p3 i(~p·~r−Et)/~ ψ = e , (69) state with respect to the direction of ~τ is lower than the  E + meff  eff eff 2 p + ip  aligned state. These two possible energy states, E2 =  1 2  2 2 2 2 2   p + (m + τ) and E1 = p + (m τ) , correspond to  E + meff  eff eff −   the energy transition where 2 2 2 4mτ E = peff + meff , (70) hν = E2 E1 = , (75) − E1 + E2 with the definitions which in the reference frame of the particle reads peff p + tk (71) k ≡ k 2 1 2 2 2 3 2 4mτ peff (p1 + t ) + (p2 + t ) + (p3 + t ) (72) ≡ hν = 1/2 1/2 , (76) 2 2 2 2 meff m + τ . (73) ~t + (m + τ) + ~t + (m τ) ≡ − 9 where ~t2 (t1)2 + (t2)2 + (t3)2 can be written simply as instead the Lagrangian in (51), then the Cartan equa- t2 assuming≡ that ~t is aligned in any of the spatial axis di- tions are rections of the reference system of coordinates. We recall µ µ µ 2 µ µ µ 2 µ µ that the torsion functions t and τ are constructed from T κ (ζ1j + ζ2a ) , T˘ κ (θ1j + θ2a ) , (82) ∼ ∼ the torsion trace vector and axial vectors and depend on the (parity-preserving) (α1, α2) and (parity-breaking) where ζi,θi (i = 1, 2) are constants, and the result- (β1,β2) coupling parameters. ing Dirac equation, after substitution in (52), includes The bottom line of this section is that parity-breaking parity-breaking and C-breaking cubic terms. effects can arise in a RC space-time from the interac- The EC theory can also be extended to include the tion of fermions with a background torsion field via non- parity-breaking Holst term [36], encapsulating additional minimal couplings. These effects include the prediction vector-axial (contact) self-interactions. The extended ac- of well-defined frequencies that a free fermion can absorb tion is thus or emit in order to make transitions between the pre- 1 S = d4x√ gR(Γ) + d4x√ g dicted two energy levels that arise depending on the spin EC 2κ2 − − Lfermions orientation with respect to external torsion quantities. Z Z 1 The signature of parity breaking might also be present + d4x√ gǫαβµν R , (83) 2γκ2 − αβµν in the radiated field itself. Z where γ is the Barbero-Immirzi parameter and the αβµν parity-breaking Holst term ǫ Rαβµν can be expressed V. NON-MINIMAL COUPLINGS IN THE as EINSTEIN-CARTAN THEORY 1 1 ǫαβµν R ˜ T˘α T˘λT + ǫαβµν T¯λ T¯ . This framework could be further specified by consider- αβµν ∼−∇α − 3 λ 2 αβ λµν ing the EC theory. The corresponding Lagrangian then (84) reads The generalized Cartan equations become

1 µ 2 3γ µ µ S = d4x√ gR(Γ) + d4x√ g , T κ (α1γj + a ) , (85) EC 2κ2 − − Lfermions ∼ 1+ γ2 Z Z (77) 3γ T˘µ κ2 (α jµ α γaµ) . (86) where the Ricci scalar of the independent connection, ∼ 1+ γ2 1 − 2 λ R(Γ), with Γ Γµν , can be related to the one con- ≡ Then, by choosing the fermionic Lagrangian in Eq. structed with the metric-compatible connection, R˜(Γ),˜ (49), one obtains a generalized Dirac equation and via an expression of the form Lagrangian with vector-vector, axial-axial and parity- 1 1 1 breaking vector-axial (contact) self-interactions. As in R R˜ 4 ˜ T α T λT + T˘λT˘ + T¯ T¯µνρ. (78) ∼ − ∇α − 3 λ 24 λ 2 µνρ the model-independent approach of the previous section, all these cases of non-minimal couplings in specific grav- Inserting this in the action (77) and taking the fermionic itational models (EC and its Holst extension) yield in- Lagrangian defined in Eq. (49), the corresponding Car- teresting and quite relevant physics (C and P symmetry- tan equations become breaking, beyond standard model interactions, etc) that in principle can be observationally probed, upon the T µ κ2α jµ , T˘µ κ2α aµ . (79) ∼ 1 ∼ 2 computation of the energy levels of fermionic systems and searching for its signatures using advanced spectro- Re-inserting these expressions in the (49) we obtain effec- graphs. tive vector-vector contact interactions besides the usual well-known axial-axial (spin-sin) interaction (Hehl-Data term) as VI. DISCUSSION AND CONCLUSION κ2 κ2 ˜ + (α )2j j (α )2a a . (80) Lfermions ∼ LDirac 3 1 · − 24 2 · Experimental constraints on the minimal and non- minimal coupling of spinors with torsion using high- The corresponding Dirac equation can be written as precision data has been a topic of interest in the liter- κ2α2 ature since quite a long time ago [37–39]. In this sense, i~γµD˜ ψ mψ = 2 (ψγ¯ λγ5ψ)γ γ5ψ the results derived in the present paper may have a phys- µ 12 λ − ical impact at several levels. First, by generalizing these 2 2α1 λ κ (ψγ¯ ψ)γλψ . (81) results to the case of bound states of electrons within − 3 atoms and molecules, and also to bound states of nu- As in the usual Dirac-Hehl-Data equation, under charge cleons within atomic nuclei, one opens up the possibil- conjugation operation ψ ψch one obtains different dy- ity of detecting the effects of torsion in the strong grav- namics for the ψch representing→ anti-fermions. If we use itational regime via the measurement of spectral lines 10 and searching for new fine structures, for instance, us- slightly deviated from the corresponding predicted fre- ing ultra-sensitive spectrographs. Therefore, there could quency when torsion is absent. be astrophysical spectral signatures of torsion waiting to Beyond astrophysics and cosmology, torsion-induced be discovered around intense gravitational fields of neu- effects of the kind studied in this work might be found in tron stars, or even in X-ray binaries, where one of the the field of condensed matter physics. Such is the case of objects is a black hole candidate surrounded by an ac- the interaction of a Dirac fermion with the torsion of a sea cretion disk. For a given bound system, a specific initial of vacuum fermion-condensates, provided that the latter energy level could be chosen such that the transitions has a non-zero expectation value. This way, fermions from this level into the two (Zeeman-like) lower levels in vacuum would have a different effective mass accord- due to the torsion-spin interaction could be searched for. ing to the relative orientation of the fermionic spin with Indeed, this could be carried out for different values of respect to the background vacuum axial torsion, which the predicted torsion in the emission regions according could be tested in laboratories, putting bounds on the to different gravitational theories accommodating torsion predicted effects. While these tests do not require strong effects. gravitational fields the challenge lies on reproducing the There are further examples of astrophysical interest conditions of fermionic vacuum condensates in the labo- where torsion effects into the physics of fermions/anti- ratory. fermions could be observationally detected. A particular Another possibility would be the existence of a continu- case is that of polarized fermion/anti-fermion pairs pro- ous and smooth phase transition for a Bose-Einstein sys- duced within quantum field theory, for instance in the tem as a superconducting fluid/material in a space-time space-time torsion around black holes. After their sub- background with torsion. Cooper pairs of anti-aligned sequent annihilation, the generated photons would have fermions in bound states are required in the BCS model different energies depending on whether the spins of the of superconductivity and in general Bose-Einstein con- fermions in the pair are up or down with respect to the densates, and since the effective mass of the pair would background torsion. In the particle pair production in depend on the interaction with the torsion background, general, there are three possible scenarios: i) particle and the effective spin-zero bosonic field due to the ensemble antiparticle are anti-aligned with each other, ii) both are of Cooper pairs would have a differential effective mass with their spins up (i.e. aligned to torsion), iii) both are powered by the relative strength of torsion. If such were with their spins down. These correspond to the three en- the case, then this would have a non-negligible impact on ergy/mass levels that the (initial) photons can generate the superconducting and superfluid phases in the interi- via pair production, with the first case having an inter- ors of neutron stars and hypothetical quark and strange mediate energy, the second the highest, and the third stars, with consequences on the macroscopic predictions the lowest. In the very early Universe, such effects would of stellar models (mass-radius relations, moment of iner- depend on the temperature of the quark-gluon-lepton- tia, etc). Moreover, the stability of Cooper pairs might photon plasma, while in Hawking’s radiation the outgo- be strongly perturbed as the torsion increases above a ing and ingoing energy flux through the event horizon certain threshold, since the background torsion axial vec- would be spin state-dependent. In particular, the (out- tor along a well defined direction can act exactly as an flux) energy loss via particles (or antiparticles) aligned external magnetic field does in paramagnetic materials, with the background torsion would be more efficient. i.e., above a certain critical value of the external field a These effects could also be sought for in the emission significant number of large clusters of “aligned” spins are of gamma rays in high energy astrophysical environments developed (and percolating the whole system) and the driven by strong gravitational fields. An observable sig- material is magnetized. The spin-spin interaction that nature of the background torsion could be then obtained naturally exists in a system with spins is analogous to by comparing the measured flux spectra with the de- a thermal-like interaction (increasing temperature tends tailed theoretical prediction of the emission curves. In to rise the entropy, and generate a random distribution the most general case the theory suggests that the radi- of spins), while the external field tends to counteract ated flux should result from the superposition of the three the random distribution of spins, by establishing grad- possibilities above peaked at the characteristic nearby ually a more ordered state. Therefore, torsion can also frequencies, corresponding to the three possible values act as an external field driving a phase transition in a of the energy of the (annihilating) pair. Disregarding macroscopic system of microphysical components with complex environmental effects and significant changes in spin, magnetizing the material, with the emergence of a the background torsion (in the typical scales of the emis- macroscopic (intrinsic) spin. In that sense, the super- sion region), the emission curves would resemble emission fluid/superconducting phase of the BCS models could lines very close together (a kind of hyperfine structure). suffer a phase transition for sufficiently strong “external” In the more conventional case of pair production with space-time torsion, inside ultra-dense compact objects. the fermion and anti-fermion anti-aligned with respect These topics deserve a much more careful analysis, since to each other, the theory predicts a specific character- they evolve very complicated physics of the interiors of istic frequency, determined by the particle’s mass, the neutron stars and related objects. torsion field, and the temperature of the emission region, Further avenues of research are those involving labora- 11 tory tests of the space-time torsion near Earth. In per- torsion using splitting of energy levels in both minimally fect analogy with the magnetic spin resonance, one could and non-minimally coupled fermions to the background design a torsion-spin resonance. In this effect, an exter- torsion in a variety of astrophysical/cosmological envi- nal torsion field generates the splitting of energy levels ronments. Further work along these lines is currently (Zeeman-like effect) in an appropriate material sample, underway. while a time-varying current produces an electromagnetic wave which suffers a measurable absorption once the resonance frequency is achieved matching the energy gap. Therefore, the indirect detection of torsion would be ACKNOWLEDGMENTS achieved by the measurement of absorption (decrease in intensity) of the electromagnetic wave interacting with the material sample, once the resonance frequency is We thank A. Delhom for useful comments and achieved. For free fermions we saw that the predicted insights. FC is funded by the Funda¸cão para a frequency (in the particle’s frame) does not depend on Ciência e a Tecnologia (FCT, Portugal) doctoral grant the fermionic mass, only on the background torsion. If No.PD/BD/128017/2016. FSNL acknowledges support torsion has a magnitude of about 10−16 10−15 m−1 then from the FCT Scientific Employment Stimulus con- we get an estimated resonance frequency− around 1 10 tract with reference CEECIND/04057/2017. DRG is nHz, which corresponds to resonance in the radio band.− funded by the Atracción de Talento Investigador pro- Finally, regarding the non-minimal couplings within gramme of the Comunidad de Madrid (Spain) No. specific gravity models such as the EC one and its 2018-T1/TIC-10431, and acknowledges further support extension with the Holst term, one also obtains gen- from the Ministerio de Ciencia, Innovación y Uni- eralized Dirac equations and Lagrangians with vector- versidades (Spain) project No. PID2019-108485GB- vector, axial-axial and parity-breaking vector-axial (con- I00/AEI/10.13039/501100011033, the Spanish project tact) self-interactions. These might be relevant inside No. FIS2017-84440-C2-1-P (MINECO/FEDER, EU), compact objects like neutron, quark, and strange (quark) the project PROMETEO/2020/079 (Generalitat Valen- stars and also in the early Universe. If the coupling con- ciana), and the Edital 006/2018 PRONEX (FAPESQ- stants are taken to be dynamical scalar fields, then this PB/CNPQ, Brazil) Grant No. 0015/2019. The au- scenario leads naturally to the idea of parity-breaking thors also acknowledge funding from FCT Projects No. phase transitions for matter under extreme conditions, UID/FIS/04434/2020, No. CERN/FIS-PAR/0037/2019 induced by the torsion-fermion currents couplings. We and No. PTDC/FIS- OUT/29048/2017. FC thanks also see that EC gravity plus Holst with T j and T˘ a the hospitality of the Department of Theoretical Physics couplings can be made equivalent to the usual· EC theory· and IPARCOS of the Complutense University of Madrid, with T j and T˘ a plus (parity-breaking) T a and T˘ j where part of this work was carried out. This article couplings.· · · · is based upon work from COST Action CA18108, sup- To conclude, the results obtained in this paper open up ported by COST (European Cooperation in Science and new avenues for testing non-Riemannian geometries with Technology).

[1] M. Blagojevic, “Gravitation and gauge symmetries,” (In- [7] F. Cabral, F. S. N. Lobo and D. Rubiera-Garcia, “The stitute of Physics Publishing, 2002). cosmological principle in theories with torsion: The case [2] M. Blagojevic and F. W. Hehl, “Gauge Theories of Grav- of Einstein-Cartan-Dirac-Maxwell gravity,” JCAP 10 itation,” [arXiv:1210.3775 [gr-qc]]. (2020), 057 [arXiv:2004.13693 [gr-qc]]. [3] F. Cabral, F. S. N. Lobo and D. Rubiera-Garcia, “Funda- [8] C. G. Boehmer and J. Burnett, “Dark spinors with tor- mental Symmetries and Spacetime Geometries in Gauge sion in cosmology,” Phys. Rev. D 78 (2008), 104001 Theories of Gravity: Prospects for Unified Field Theo- [arXiv:0809.0469 [gr-qc]]. ries,” Universe 6 (2020), 238 [arXiv:2012.06356 [gr-qc]]. [9] N. J. Poplawski, “Nonsingular, big-bounce cosmology [4] F. W. Hehl and Y. N. Obukhov, “Elie Cartan’s torsion from spinor-torsion coupling,” Phys. Rev. D 85 (2012), in geometry and in field theory, an essay,” Annales Fond. 107502 [arXiv:1111.4595 [gr-qc]]. Broglie 32 (2007), 157-194 [arXiv:0711.1535 [gr-qc]]. [10] N. J. Poplawski, “Big bounce from spin and torsion,” [5] F. Cabral, F. S. N. Lobo and D. Rubiera-Garcia, Gen. Rel. Grav. 44 (2012), 1007-1014 [arXiv:1105.6127 “Einstein–Cartan–Dirac gravity with U(1) symmetry [astro-ph.CO]]. breaking,” Eur. Phys. J. C 79 (2019) no.12, 1023 [11] B. Vakili and S. Jalalzadeh, “Signature transition in [arXiv:1902.02222 [gr-qc]]. Einstein-Cartan cosmology,” Phys. Lett. B 726 (2013), [6] F. Cabral, F. S. N. Lobo and D. Rubiera-Garcia, “Cosmo- 28-32 [arXiv:1308.2877 [gr-qc]]. logical bounces, cyclic universes, and effective cosmolog- [12] K. A. Bronnikov and A. M. Galiakhmetov, “Wormholes ical constant in Einstein-Cartan-Dirac-Maxwell theory,” and black universes without phantom fields in Einstein- Phys. Rev. D 102 (2020) no.8, 083509 [arXiv:2003.07463 Cartan theory,” Phys. Rev. D 94 (2016) no.12, 124006 [gr-qc]]. [arXiv:1607.07791 [gr-qc]]. 12

[13] A. N. Ivanov and M. Wellenzohn, “Einstein-Cartan Grav- torsion fields”, Physics Letters A 247 (1998) 21-26. ity with Torsion Field Serving as an Origin for the Cos- [27] B. Goncalves, Y. N. Obukhov, I. L. Shapiro, “On mological Constant or Dark Energy Density,” Astrophys. the exact Foldy-Wouthuysen transformation for a Dirac J. 829 (2016) no.1, 47 [arXiv:1607.01128 [gr-qc]]. spinor in torsion and other CPT and Lorentz vio- [14] J. A. R. Cembranos, J. Gigante Valcarcel and F. J. Mal- lating backgrounds”, Phys. Rev. D 80 (2009) 125034 donado Torralba, “Singularities and n-dimensional black [arXiv:0908.0437]. holes in torsion theories,” JCAP 04 (2017), 021 [28] I. L. Shapiro, “Physical Aspects of the Space- [arXiv:1609.07814 [gr-qc]]. Time Torsion”, Physics Reports 357 (2002) 113-213 [15] J. A. R. Cembranos and J. Gigante Valcarcel, “Ex- [hep-th/0103093]. tended Reissner–Nordström solutions sourced by dy- [29] J. Audretsch, “Dirac Electron in Space-times With Tor- namical torsion,” Phys. Lett. B 779 (2018), 143-150 sion: Spinor Propagation, Spin Precession, and Non- [arXiv:1708.00374 [gr-qc]]. geodesic Orbits,” Phys. Rev. D 24 (1981) 1470. [16] G. Unger and N. Poplawski, “Big bounce and closed uni- [30] Y. N. Obukhov, A. J. Silenko and O. V. Teryaev, “Spin- verse from spin and torsion,” Astrophys. J. 870 (2019) torsion coupling and gravitational moments of Dirac no.2, 78 [arXiv:1808.08327 [gr-qc]]. fermions: theory and experimental bounds, Phys. Rev. [17] M. R. Mehdizadeh and A. H. Ziaie, “Charged Wormhole D 90 (2014) no.12, 124068. Solutions in Einstein-Cartan gravity,” Phys. Rev. D 99 [31] Y. N. Obukhov, A. J. Silenko and O. V. Teryaev, “Dirac (2019) no.6, 064033 [arXiv:1811.03364 [gr-qc]]. particle in Riemann–Cartan spacetimes,” Phys. Part. [18] E. Barausse, E. Berti, T. Hertog, S. A. Hughes, P. Jet- Nucl. 49 (2018) no.1, 9-10. zer, P. Pani, T. P. Sotiriou, N. Tamanini, H. Witek [32] M. Adak, T. Dereli and L. H. Ryder, “Dirac equation in and K. Yagi, et al. “Prospects for Fundamental Physics space-times with nonmetricity and torsion,” Int. J. Mod. with LISA,” Gen. Rel. Grav. 52 (2020) no.8, 81 Phys. D 12 (2003), 145-156 [arXiv:gr-qc/0208042 [gr-qc]]. [arXiv:2001.09793 [gr-qc]]. [33] J. A. R. Cembranos, J. G. Valcarcel and F. J. Maldonado [19] J. B. Jiménez, L. Heisenberg and T. S. Koivisto, “The Torralba, “Fermion dynamics in torsion theories,” JCAP Geometrical Trinity of Gravity,” Universe 5 (2019) no.7, 04 (2019), 039 [arXiv:1805.09577 [gr-qc]]. 173 [arXiv:1903.06830 [hep-th]]. [34] S. Bahamonde and J. G. Valcarcel, “New models with [20] L. Parker, “One-Electron Atom in Curved Space-Time,” independent dynamical torsion and nonmetricity fields,” Phys. Rev. Lett. 44 (1980) no.23, 1559. JCAP 09 (2020), 057 [arXiv:2006.06749 [gr-qc]]. [21] L. Parker, “One electron atom as a probe of space-time [35] Y. N. Obukhov, “Parity violation in Poincarégauge grav- curvature,” Phys. Rev. D 22 (1980), 1922-1934. ity,” [arXiv:2010.16276 [gr-qc]]. [22] L. Parker and L. O. Pimentel, “Gravitational perturba- [36] S. Holst, “Barbero’s Hamiltonian derived from a gener- tion of the hydrogen spectrum,” Phys. Rev. D 25 (1982), alized Hilbert-Palatini action,” Phys. Rev. D 53 (1996), 3180-3190. 5966-5969 [arXiv:gr-qc/9511026 [gr-qc]]. [23] A. Delhom, V. Miralles and A. Peñuelas, “Effective [37] V. A. Kostelecky, N. Russell and J. Tasson, “New Con- interactions in Ricci-Based Gravity below the non- straints on Torsion from Lorentz Violation,” Phys. Rev. metricity scale”, Eur. Phys. J. C 80 (2020) no.4, 340 Lett. 100, 111102 (2008) [arXiv:0712.4393 [gr-qc]]. [arXiv:1907.05615 [hep-th]]. [38] B. R. Heckel, E. G. Adelberger, C. E. Cramer, [24] A. D. I. Latorre, G. J. Olmo and M. Ronco, “Ob- T. S. Cook, S. Schlamminger and U. Schmidt, “Preferred- servable traces of non-metricity: new constraints on Frame and CP-Violation Tests with Polarized Electrons,” metric-affine gravity,” Phys. Lett. B 780 (2018), 294-299 Phys. Rev. D 78 (2008), 092006 [arXiv:0808.2673 [hep- [arXiv:1709.04249 [hep-th]]. ex]]. [25] V. G. Bagrov, I. L. Buchbinder, I .L. Shapiro, “Possible [39] R. Lehnert, W. M. Snow and H. Yan, “A First Experi- experimental manifestations of the torsion field”, Sov. J. mental Limit on In-matter Torsion from Neutron Spin Phys., 35 (1992) n3, p.5-12. Rotation in Liquid 4He,” Phys. Lett. B 730 (2014), [26] L. H. Ryder, I.L. Shapiro, “On the interaction of mas- 353-356 [erratum: Phys. Lett. B 744 (2015), 415-415] sive spinor particles with external electromagnetic and [arXiv:1311.0467 [nucl-ex]].