the PVLAS anomaly
the PVLAS anomaly 9/07/005-3/6/007 >10 papers
L anomalia di PVLAS: un miraggio* ad polosa infn roma `la sapienza` m bergantino, r faccini, l maiani, a melchiorri, a strumia. *arxiv:0706.3419 3 Jun 007
glossary birefringence = generation of an ellipticity in linearly polarized light in the presence of H. induction of ξ. ρ αβ = 1 ( ) 1 + ξ3 ξ 1 iξ ; ξ ξ 1 + iξ 1 ξ 1,,3 [ 1, 1] 3 ξ 3 = 1 lin y ξ 1 = 1 lin y(45 ) ξ 3 = 1 lin y ξ 1 = 1 lin y( 45 ) ξ = i A 1A + A A 1 A 1 + A dichroism = rotation of the plane of linear polarization of light in the presence of H; this has to do with loss of power in a certain direction of propagation (the imaginary part of the refraction index).
dielectric vacuum L Maiani, R Petronzio, E Zavattini, (MPZ) Phys. Lett. B175, 359 (1987) the original idea of e. zavattini was to measure vacuum dielectric properties; in particular birefringence due to photon-photon interactions in presence of an external magnetic field but in mpz, physics beyond qed is proposed: (dichroism+birefringence) See S Adler, Photon Splitting and Photon Dispersion in a Strong Magnetic Field, Ann. Phys. 67, 599 (1971)
mpz-dichroism 1 LI = φf F 4M 1 φf F 4M consider e.g. the pseudoscalar LI φ Eγ Hext cos λ H λ unaffected loose power along H see also G Raffelt and L Stodolsky, Phys. Rev. D37, 137 (1988)
the apparatus E Zavattini et al. (PVLAS collaboration), Phys. Rev. Lett. 96, 110406 (006) Lint = 1 m λlaser = 1064 nm 1 ev H = 5.5 T P 10 8 mbar νm 0.3 Hz νsom = 506 Hz!SOM = 10 3 rad the QWP transforms apparent rotations in ellipticities which then beat with the SOM (carrier ellipticity signal) and are detected
measuring dichroism p p1 superimpose the som carrier F Brandi et al., Meas. Sci. Technol. 1, 1503 (001) beyond QWP :: λ/4 e.g. fast axis p1 φ induced by H beyond polaroid p1 laser λ
the effect (κ! κ ) α= D sin λ I = I0 (before P ){σ + [α(t) + η(t) + Γ(t)] } α(t) = α0 cos(4πνm t + Θm ) η(t) = η0 cos(πνsom t + ΘSOM ) α (the rotation to be detected) is expected to be small In the Fourier amplitude spectra of the detection photodiode signal the largest intensity comes from the term αη (η gives an help) beating with νsom ± νm
νm H 3 H θ H 1 λ 3 1 rotation 1 L, 1 R 1 0-1 0 Π 4 cos Θ Π 3 Π 4 Π
α = (3.9 ± 0.5) 10 1 rad/pass results som som
results after QWP π/ inversion 6 0 90 a Y a systematic error occurs b
but life is hard... pvlas arxiv:0706.3419 3 Jun 007 No Rotation after 45000 passes < 1. 10 8 rad @ 5.5 T < 1.0 10 8 rad @.3 T No Ellipticity after 45000 passes (< 1.4 10 8 @.3 T ) while at 5.5 T still present... 10 7 @ 5.5 T these results exclude particle interpretation of pvlas
B = 5.5 T B =.3 T B=0 no QWP QWP 0 QWP 90
FORMER EXPERIMENTAL CHALLENGES
axion interpretation the dichroism amplitude was propto H** as it should in an axion model; measured to be ~10**4 than expected by QED Pγ φ! sin ql = g H L! " ql! = sin λ # HL 4M $ " Npass q is the transfered momentum % sin(mφ L/4ω) mφ L/4ω & pvlas claimed to observe a scalar (from rot. sign) with 105 GeV! M! 6 105 GeV 1 mev! mφ! 1.5 mev which strongly conflicts with the observation by cast
cast S Andriamonje et al. (CAST collaboration), see review [hep-ex/070006] ) -1 (GeV -7 10 pvlas a! g 10-8 Lazarus et al. -9 10 SOLAX, COSME DAMA Tokyo helioscope -10 10-11 10-1 10 10-5 HB stars CAST phase I -4-3 Axion models - KSVZ [E/N = 0] QCD 10 10 10 10 1 10 m axion (ev) imaging x-ray optical system -1 HDM
gravity A Dupays et al., Phys. Rev. Lett. 98, 13180 (007) L = yφψ ψ take the leading radiative to y!contribution " α mp y ln π M Λ mp then exp. < 10**-! " # " # $ m1 m 1 y Z Z 1 V (r)! G 1+ exp( m r) m φ φ = 0. mm r Gmp 4π A 1 A M~4.*10**16 GeV
cont d inconsistent with pvlas by a factor of ~10**11 Vab (r) = ga gb exp( r/λ); λ =!/mc 4πr!α " g 4π gφγγ π mp V (r) = G m1 m [1 + α exp( r/λ)] r Adelberger et al., Phys. Rev. Lett. 98, 131104 (007) Kapner et al., Phys. Rev. Lett. 98, 01101 (007) 1033 = 103 GeV gpvlas
regeneration R Rabadan, A Ringwald, K. Sigurdson, Phys. Rev. Lett. 96, 11407 (006) shining-through-wall experiment Pγ φ ql"1!" #" #$ #" g L H 10 9 10 T 10 m 10 6 GeV 1 (Nγreg /s) = (NγFEL /s) P (NγFEL /s) 1017 /s regeneration plans: pvlas+b; lipss(jlab); alps(desy); apfel(desy); bmv(luli/f);?(cern)
laser-laser M Bergantino, R Faccini, ADP L = 10 40 cm sec 1 via axions # evts s -7-8 -9-30 -31-3 -33 0 0.0000.00040.00060.0008 0.001 Γ energy ev allowing multiple collision regions in the FP
crystals M Bergantino, R Faccini, ADP W Buchmuller and F Hoogeveen, Phys Lett B37, 78 (1990) X-ray source crystal wall detector crystal what about primakoff in a crystal? H-field wall crystal
other interpretations
a partial list of models paraphotons and millicharged particles bounds from cmb :: cmb ellipticities? mohapatra-nasri model chern-simons coupled vectors...
a microscopic point of view E Masso and J Redondo Phys. Rev. Lett. 97, 15180 (006) axion parafermion 1 α # 1 v =! # # 10 M π v ev if v is a low energy scale, we need a very tiny charge for the parafermion the `millicharge`. qed with extra u(1) fields can accomodate this B Holdom, Phys. Lett. B166, 196 (1986) consider a multiplication of u(1) s! 1 T 1 T L = F Mmix F + A Mmass A + e ji Ai 4 i=0
diagonalize Mmix hyp :: 1 =!!!! 1 0 0 1 mixings are assumed to be small making the hyp that they are induced by ultramassive fermions circulating in loops; 1 0 1 m 0 U=! m1 m0 m0! m m 0 m1! m m 0 1 m! m m 0 0 1 1 0 we can rotate, by U, the (para)photons fields in such a way to obtain the kinetic part in the standard F.F form -- keep up to first order in ε hyp :: e = e1 = e
reconciling with stars m0 q m1 0 m µ masso-redondo qeff = e! (vacuum) µ qeff = e! small povided ωp ( KeV) " µ q µ! < 4 10 8 HBstars ev gravity exclusion plot
gravity again A Dupays et al., Phys. Rev. Lett. 98, 13180 (007) -alp does not have a direct coupling to photons -alp-photons vertex arises because of a photonparaphoton mixing ε -a paraphoton mass μ induces an effective photon form factor such that the coupling is reduced for q>> μ 1 1 µ effective propagator! k k µ k α µ! y = finite [LO(µ/mp )] 4M include e para-γ parafermion! $ " # " # m1 m 4 y! Z Z 1 V (r)! G 1+ exp( m r) m φ φ = 0. mm r Gmp 4π A 1 A since y is smaller than y, by μ/mp, a smaller value of M is allowed! M~10**5 GeV γ
millicharges & dichroism H Gies, J Jaeckel, A Ringwald, Phys. Rev. Lett. 97, 14040 (006) γ(lin pol) e+ e with ω > me in H ext dichroism never observed in lab because of the threshold in pvlas ω > mε 1 λ! (κ! κ )L sin(λ) 4 NB. the landau levels are very dense here :: no absorption peaks expected review in W Dittrich and H Gies, Springer Tracts Mod. Phys. 166, 1 (000)
milli-parameters birefringence φ = (n! n )ωl The propagation speed of the laser photons is slightly changed in the magnetic field owing to the coupling to virtual charged pairs M Ahlers, H Gies, J Jaeckel, A Ringwald, Phys. Rev. D75, 035011 (007) See also SN Gninenko, NV Krasnikov and A Rubbia, Phys. Rev. D75, 075014 (007)
cosmic bounds A Melchiorri, AD Polosa, A Strumia, Phys. Lett. B (007) ::hyp:: -cosmology after decoupling -only SM particles at beginning start producing millicharged by photons ε ε e ε e ne! 1 after decoupling nγ Γ "nγ σγγ " " v#t T nγ = (ζ(3)/π )T 3 H = a /a T 3/ Γ/H maximal at low T :: T max(t0, m" )
fitting firas we expect an energy dependent depletion of the CMB spectrum f (E) (small) deviation from fbe (E) r(e) f (E)/fBE (E) x E/T! f (E) d ndev d γ!! r(x) = dp f (E )σ (s) # nγ H = γt fbe (E)H d ln z 4E d ln z nγ expect small spectral distorsions! d T H r(x) = dcθ dx! x! fbe (x! )σ (s = xx! T (1 cθ )) d ln z 3π x 0.1 0 frequency in 1/cm 10 15 5 0 0-0.05 I"E,T# % deficit m! = 0.1 ev 1-r(x) 0.05! = 10 7 e -0.1 E!T 0 4 6 E! /T 8 10 1
fitting firas as a result one finds that n!! 6 10 5 at 3σ c.l. for m! = 0.1 ev Y nγ nγ H d Y = γt d ln z for general masses and charges isocurves of Y
model dependent exclusion once millicharges have been produced they can start disintegrating into paraphotons pairs -- going rapidly to thermal equilibrium -- but still they can keep on depleting CMB via γ CMB ɛ γ ɛ other bounds can be studied Y depletion γ min σ(γɛ γ ɛ) ɛ ɛ e 4 ɛ O(1)? [ 1, σ(γɛ γ ɛ) Y ɛn γ (T ) H(T ) 4πT ] depending on the paraphoton model...
turbolent magnetic fields A Mirizzi, G G Raffelt, P Serpico, arxiv:0704.3044 spectral modification of a gamma-tev source at the galactic center. photon-axion oscillations cause a downward shift of the high energy spectrum (a change of normalization of the typical power spectrum between low and high energies) dn de E Γ Γ =.5 solid line
more scalars RN Mohapatra and S Nasri Phys. Rev. Lett. 98, 05040 (007) Φ(S), σ(s), φ(p S) Φφ φ F F rather than F F M M but φ Φφ F F! F F if T < kev M MPVLAS hyp :: low temperature phase transition!φ" 1 = M MPVLAS in the sun it is ph.s. inhibited also the process γγ Φφ if mφ 10 50 MeV M**~10**5 GeV** is consistent with cosmological & astrophysical data
a vector I Antoniadis, A Boyarsky, O Ruchayskiy, hep-ph/0606306 κ κ 10 17 suppress axions from stars pvlas dichroism? κ κh PVLAS birefringence mφ mφ κvφ and mφ 0 as κ 0 κ m(h) γ in a c.s. lagrangian L = 1/4 FA 1/4 Fφ + mγ / A + mφ φ κ#(a, φ; FA ) what if a vector axion is produced with longitudinal polarization?!µ! =! "k ω"k, mφ mφ "k " maybe ok with pvlas dichroism but again at odds with stars?
conclusions I (before 3/6) experimentally driven field :: maybe we are close to a final answer to confirm/disprove the PVLAS result* all theoretical models here described seem to have troubles with data i have not mentioned other models like the chamaleon by brax et al. [...] 0.00001 o *M Fairbairn et al., Searching for Energetic Cosmic Axions in Laboratory Experiments, arxiv:0706.0108
conclusions II experimentally driven field :: pvlas disproves pvlas* all theoretical models here described seem to have troubles with data :: it seems that now we know why 0.0000 o *PVLAS arxiv:0706.3419
the mpz calculation
conversion probability!a + 1 M B φ (! + m )φ =0 pseudoscalar 1 M B A = 0 look for solutions A = A! i + A j A = A 0 e i k t+ikz mixing axion and a! "! " A! λ ψ= = e iωt+ikz u e iωt+ikz φ µ and solve the linear system Mu = 0 M= k ω B im ω B ω i M k + m ω L Maiani, R Petronzio, E Zavattini, Phys. Lett. B175, 359 (1987)
the secular determinant: #$ % 1 B k B B ω± = k + m + ± +4 m + M M M 1 iω± B M k ω±! " λ± u± = 1 λ± = initial condition :: 0-axions A! (t = z = 0) = cos λ φ(t = z = 0) = 0 A! = cos λ [" e iω+ t + (1 ") e iω t ] eikz φ = cos λ Φ [e iω+ t + e iω t ] eikz ω+ (k ω ) "= D(k) )(k ω ) im (k ω+ Φ= B D(k) D(k) = ω+ (k ω ) ω (k ω+ ) B! small M
expanding in the small parameter! " #$ ωl A! = cos λ 1 " sin & % " # BL sin x = cos λ 1 4M x x=l ω/<<1 proceed similarly for Φ dichroism Pφ γ φ A" (t = z = 0)
heterodine & som the som (stress optic modulator) is a new type of polarization modulator deviced within the pvlas collaboration: it induces a controllable birefringence on a glass window by means of an electrical stress applied to it F Brandi et al., Meas. Sci. Technol. 1, 1503 (001) the modulation of ellipticity to be measured is at very low frequency, where 1/f noise and other sources of low freq. noise are dangerous the ellipticity modulated through a magnetic field modulation beats with known ellipticity induced on light using a polarization modulator (som)