hep-ph/0011217 LMU-00-15 Democratic Neutrino Mixing and Radiative Corrections Zhi-zhong Xing Sektion Physik, Universit¨at Mu¨nchen, Theresienstrasse 37A, 80333 Mu¨nchen, Germany and 1 Theory Division, Institute of High Energy Physics, P.O. Box 918, Beijing 100039, China 0 0 (Electronic address: [email protected]) 2 n a J Abstract 4 1 2 The renormalization effect on a specific ansatz of lepton mass matrices, aris- v ing naturally from the breaking of flavor democracy for charged leptons and 7 that of mass degeneracy for light neutrinos, is studied from a superhigh en- 1 2 ergy scale M0 ∼ 1013 GeV to the electroweak scale in the framework of the 1 minimal supersymmetric standard model. We find that the democratic neu- 1 0 trino mixing pattern obtained from this ansatz may in general be instable 0 against radiative corrections. With the help of similar flavor symmetries we / h prescribe a slightly different scheme of lepton mass matrices at the scale M0, p - from which the democratic mixing pattern of lepton flavors can be achieved, p after radiative corrections, at the experimentally accessible scales. e h PACS number(s): 14.60.Pq, 12.15.Ff, 12.60.-i, 11.10.Hi : v i X r a Typeset using REVTEX 1 Recently a number of models of lepton mass matrices have been proposed at low energy scales [1], at which their consequences on the spectrum of neutrino masses and the mixing of lepton flavors can directly be confronted with the robust Super-Kamiokande data on atmospheric and solar neutrino oscillations [2]. From the theoretical point of view, however, a phenomenologically-favored scheme of lepton mass matrices might only serve as the low- scaleapproximationofamorefundamentaltheoryresponsible fortheleptonmassgeneration and flavor mixing at superhigh energy scales. It is therefore desirable to investigate the scale dependence of lepton mass matrices with the help of the renormalization-group equations. So far some attempts have been made in this direction [3–6]. In this Brief Report we aim to study whether the democratic neutrino mixing pattern, which is indeed a nearly bi-maximal mixing pattern of lepton flavors arising from the break- ing of flavor democracy for charged leptons and that of mass degeneracy for light neutrinos, can be stable or not against the effect of radiative corrections from a superhigh energy scale 13 M0 ∼ 10 GeV to the electroweak scale MZ in the framework of the minimal supersym- metric standard model. We find that the democratic neutrino mixing pattern at M0 is no longer of the same form at M . But it can still be obtained at low energy scales, if an Z additional term preserving the symmetry of flavor democracy is introduced to the original neutrino mass matrix at the scale M0. Therefore the democratic neutrino mixing pattern at the experimentally accessible scales might hint at certain lepton flavor symmetries at a superhigh scale, at which viable models of lepton mass matrices can naturally be built. Let us begin with a brief retrospection of the specific model of democratic neutrino mixing proposed first in Ref. [7] at low energy scales. The essential idea of this model is that the realistic textures of charged lepton and neutrino mass matrices might arise respectively from the breaking of S(3)L ×S(3)R and S(3) flavor symmetries: 1 1 1 c M = l 1 1 1 +∆M , l   l 3 1 1 1   1 0 0  M = c 0 1 0 +∆M , (1) ν ν   ν 0 0 1     where c and c measure the corresponding mass scales of charged leptons and neutrinos. l ν The explicit symmetry-breaking term ∆M is responsible for the generation of muon and l electron masses, and ∆M is responsible for the breaking of neutrino mass degeneracy. The ν lepton flavor mixing matrix results from the mismatch between the diagonalization of M l and that of M , therefore its pattern depends crucially on the forms of ∆M and ∆M . It ν l ν has been shown in Ref. [7] that current data on solar and atmospheric neutrino oscillations seem to favor the following forms of ∆M and ∆M : l ν −iδ 0 0 c l ∆M = l 0 iδ 0 , l  l  3 0 0 ε  l  −δ 0 0  ν ∆M = c 0 δ 0 , (2) ν ν  ν  0 0 ε ν     2 where (δ ,ε ) and (δ ,ε ) are dimensionless perturbation parameters of small magnitude. It l l ν ν 2 2 is easy to obtain m ≈ c , m ≈ 2|ε |m /9, and m ≈ |δ | m /(27m ) in the lowest order τ l µ l τ e l τ µ approximation. As for neutrino masses, we have m1 = cν(1 − δν), m2 = cν(1 + δν), and m3 = cν(1+εν). The simultaneous diagonalization of Ml and Mν leads to the lepton flavor mixing matrix V, which links the neutrino flavor eigenstates (ν ,ν ,ν ) to the neutrino mass e µ τ eigenstates (ν1,ν2,ν3): 1 1 0 √2 √−2 1 1 2 V =  √6 √6 √−6 +∆V , (3) 1 1 1  √3 √3 √3    where m m e µ ∆V = i ξ + ζ (4) V m V m s µ τ holds in the next-to-leading order approximation with 1 1 2 √6 √6 √−6 ξ = 1 1 0 , V  √2 √−2  0 0 0    0 0 0  1 1 1 ζ = . (5) V  √6 √6 √6  1 1 1  √−12 √−12 √3    Thisisjustanearlybi-maximalleptonmixingpatternwithlargeCPviolation[7]. Neglecting the term ∆V, which is remarkably suppressed by the small quantities m /m ≈ 0.07 and e µ m /m ≈ 0.06 [8], one often refers to V as the democratic neutrino miqxing pattern. µ τ Now we prescribe the same ansatz of lepton mass matrices, as that introduced above, at superhigh energy scales. To be specific, we only consider the simple possibility that the typical mass scale of light Majorana neutrinos is determined via the conventional seesaw 13 mechanism [9] by the mass of a heavy right-handed neutrino M0 ∼ 10 GeV; namely 2 cν ∼ v /M0 [10], where v is theelectroweak vacuum expectation value. The mass degeneracy of three active neutrinos is broken by the perturbative term ∆Mν in Eq. (1) at the scale M0. Belowthistypicalscale,theneutrinomassmatrixM andthechargedleptonmassmatrixM ν l have quite simple running behaviors in the framework of the standard electroweak model or its minimal supersymmetric extension. The relevant renormalization-group equations, which describe the radiative corrections to lepton mass matrices from the superhigh scale M0 to the electroweak scale MZ, have been derived by a number of authors in Refs. [3–6]. Subsequently we investigate whether the democratic neutrino mixing pattern V in Eq. (3) is stable or not against radiative corrections in the framework of the minimal supersymmetric standard model (MSSM). For simplicity we choose a specific flavor basis, in which the lepton mass matrix M is l diagonal (namely, M is transformed into the diagonal form Mˆ = Diag{m ,m ,m }) at l l e µ τ the scale M0. In this basis and at the same scale, the corresponding neutrino mass matrix takes the form Mˆ = V M V . Running Mˆ and Mˆ down to the scale M by use of the ν ∗ ν † l ν Z 3 renormalization-group equations in the framework of MSSM, one obtains the new lepton mass matrices Mˆ and Mˆ . Obviously Mˆ remains diagonal, but its mass eigenvalues are in l ν l general different from those of Mˆ due to radiative corrections [3,5]. At the scale M , the l Z form of Mˆ reads explicitly as [4] ν ˆ 6 ˆ M = I I T M T (6) ν g t l ν l (cid:16) (cid:17) with I 0 0 e T = 0 I 0 , (7) l  µ  0 0 I τ     in which I , I , and I (for α = e,µ,τ) denote the corresponding evolution functions of g t α the gauge couplings g and g , the top-quark Yukawa coupling f , and the charged lepton 1 2 t Yukawa couplings f , f and f : e µ τ 1 lnM0 6 2 2 I = exp + g (χ)+6g (χ) dχ , g " 16π2 ZlnMZ (cid:18)5 1 2 (cid:19) # 1 lnM0 2 I = exp − f (χ)dχ , t " 16π2 ZlnMZ t # 1 lnM0 2 I = exp − f (χ)dχ . (8) α " 16π2 ZlnMZ α # Note that the power of I in the expression of Mˆ depends on the definition of I in Eq. (8). t ν t 6 The overall factor (I I ) in Eq. (6) does not affect the relative magnitudes of the matrix g t elements of Mˆ . Only the matrix T , which amounts to the unity matrix at the energy scale ν l M0, can modify the texture of the neutrino mass matrix from M0 to MZ. The magnitude of I may somehow deviate from unity, if tanβ (the ratio of Higgs vacuum expectation values τ in the MSSM) takes large values. In contrast, I ≈ I ≈ 1 is an excellent approximation. e µ Denoting κ ≡ I /I −1 ≈ I /I −1, one arrives from Eq. (8) at e τ µ τ 2 κ ≈ mτ ln M0 . (9) 16π2v2cos2β M Z 13 It turns out that κ ≈ 0.03 for M0 ∼ 10 GeV and tanβ = 60. With the help of Eqs. (3) and (6), one can straightforwardly figure out the explicit expression of Mˆ at the scale M . To a good degree of accuracy, the small term ∆V of ν Z V (namely, the small corrections from the ratios of charged lepton masses) is negligible in ˆ T ˆ the calculation. We then make the transformations V M V ≡ M and V M V ≡ M at † l l ν ν the scale M . Of course M should have a quasi-democratic texture in the lowest order Z l approximation, just as M . The neutrino mass matrix M takes the following form: l ν 6 2 M ≈ I I I [M +2κ(Ω −Λ )] , (10) ν g t τ ν ν ν (cid:16) (cid:17) where the κ-induced term signifies the renormalization effect, and the constant matrices Ω ν and Λ read as ν 4 1 0 0 Ω ≈ c 0 1 0 , ν ν   0 0 1    1 1 1 c ν Λ ≈ 1 1 1 . (11) ν   3 1 1 1     2 In obtaining Eqs. (10) and (11), we have neglected the small contributions of O(κ ), O(κε ) ν 2 and O(κδ ). Note that the factor I in M comes from the product of two T matrices on ν τ ν l the right-hand side of Mˆ . Comparing M with M , we see that the diagonal texture of ν ν ν M is not affected by the radiative correction term Ω , which is also diagonal. The latter ν ν modifies three neutrino mass eigenvalues of M with the same magnitude (proportional to ν 2κc ). In contrast, the diagonal texture of M is spoiled by the other radiative correction ν ν term Λ , which has a democratic form. As a consequence of the appearance of Λ in M , ν ν ν the corresponding lepton flavor mixing matrix V, which arises from the mismatch between the diagonalization of M and that of M at the electroweak scale M , may substantially l ν Z deviate from the original flavor mixing matrix V at the superhigh scale M0. Unless κ is negligibly small, the democratic neutrino mixing pattern is expected to be instable against radiative corrections. The sensitivity of a nearly bi-maximal neutrino mixing pattern to the renormalization effect is of course not a big surprise [3,5]. However, it is not impossible to find out the appropriate textures of lepton mass matrices, which are essentially stable against radiative corrections [5,6]. As the democratic neutrino mixing pattern is only favored at the experi- mentally accessible energy scales, the question turns out to be whether there is a scheme of lepton mass matrices at the superhigh scale M0, from which the democratic mixing pattern of lepton flavors can be obtained at the electroweak scale M . We find that such a scheme Z of lepton mass matrices does exist and it is very similar to that discussed above. To be specific, let us prescribe the new ansatz of lepton mass matrices at the scale M0. We take the charged lepton mass matrix M to have the same form as M in Eq. (1); namely, l′ l Ml′ = Ml has the S(3)L ×S(3)R flavor symmetry in the limit ∆Ml = 0. The corresponding neutrino mass matrix M is a linear combination of M in Eq. (1) and an additional term, ν′ ν which preserves the symmetry of flavor democracy: 1 1 1 M = M +c 1 1 1 . (12) ν′ ν ′ν   1 1 1     Obviously M has the same S(3) flavor symmetry as M in the limit ∆M = 0. One can ν′ ν ν therefore see much similarity between the new ansatz and the old one. The coefficient c is ′ν a free parameter [6], but its value may be physically nontrivial, as we shall see below. Following the same procedure as outlined above, we calculate the counterpart of M at ν′ the electroweak scale M . We obtain Z 6 2 M ≈ I I I [M +2κ(Ω −Λ )] , (13) ′ν g t τ ν′ ν ν (cid:16) (cid:17) where κ, Ω , and Λ have been given in Eqs. (9) and (11). Note that the texture of Ω is ν ν ν essentially the same as that of M (the first term of M ), and the texture of Λ is essentially ν ν′ ν 5 the same as that of the second term of M . Therefore the basic texture of M is the same ν′ ′ν as that of M . In other words, the structure of M is stable against radiative corrections. ν′ ν′ It is particularly interesting that the second term of M and the term Λ in M are ν′ ν ′ν possible to cancel eath other. Indeed the cancellation between these two terms takes place, if the coefficient c satisfies the condition c /c = 2κ/3. In this case, the resultant neutrino ′ν ′ν ν mass matrix reads as 6 2 M ≈ I I I (M +2κΩ ) , (14) ′ν′ g t τ ν ν (cid:16) (cid:17) which is diagonal at the scale M . Since the corresponding charged lepton mass matrix M Z ′l is of a quasi-democratic format the same scale, just asM discussed above, thesimultaneous l diagonalization of M and M must lead to the democratic flavor mixing pattern V in the ′l ′ν′ lowest order approximation (namely, in the approximation of neglecting the ∆V term). One might question whether the condition c /c = 2κ/3 ≪ 1 suffers from fine-tuning or ′ν ν not. Indeed it is rather natural to expect c ≪ c in the model under consideration, because ′ν ν this hierarchy assures the near degeneracy of three neutrino masses. On the other hand, a relation between c and c allows us to reduce the number of free parameters in M from ′ν ν ν′ four to three, which can fully be determined by three neutrino mass eigenvalues m (for ′i i = 1,2,3) at low energy scales. Indeed we obtain 6 2 m ≈ I I I (1+2κ−δ )c , ′1 g t τ ν ν (cid:16) 6 2(cid:17) m ≈ I I I (1+2κ+δ )c , ′2 g t τ ν ν (cid:16) 6 2(cid:17) m ≈ I I I (1+2κ+ε )c . (15) ′3 g t τ ν ν (cid:16) (cid:17) 6 2 Note that the overall factor (I I I ) takes approximate values 0.80 and 0.63, respectively, g t τ for tanβ = 10 and 60 [4]. The present Super-Kamiokande [2] and CHOOZ [11] data favor the approximate decoupling between solar and atmospheric neutrino oscillations, which are respectively attributed to ν → ν and ν → ν transitions in the framework of three active e µ µ τ 2 2 2 2 2 2 neutrinos.Thus one may take ∆m = |(m ) − (m ) | and ∆m = |(m ) − (m ) |. sun ′2 ′1 atm ′3 ′1 2 2 Taking ∆m ≪ ∆m into account [2], we arrive at sun atm 2 ∆m 2|δ | |δ | sun ν ν ≈ ≈ 2 . (16) 2 ∆m |ε +δ | |ε | atm ν ν ν 2 2 This result implies that the observables ∆m and ∆m are completely insensitive to the sun atm small parameter κ. Given the energy scale at which the proposed textures of M and M l′ ν′ hold, κ is a well-defined quantity in respect to the fixed value of tanβ within the MSSM or other extensions of the standard electroweak model. Therefore we think that the condition c /c = 2κ/3 is plausible for our new ansatz of lepton mass matrices at a superhigh energy ′ν ν scale, from which the democratic neutrino mixing pattern can be obtained, after radiative corrections, at the experimentally accessible energy scales. In summary, we have investigated the renormalization effects on lepton mass matrices and flavor mixing from a superhigh energy scale to the electroweak scale in the framework of MSSM. We find that the democratic neutrino mixing pattern may in general be instable againstradiativecorrections. Anewansatzofleptonmassmatrices, basedonthebreaking of 6 flavor democracy for charged leptons and the mass degeneracy for light neutrinos, has been prescribed at superhigh scales. Taken the effect of radiative corrections into account, this ansatz can lead to the democratic mixing pattern of lepton flavors at low energy scales. We expect that the forthcoming neutrino oscillation experiments will provide a stringent test of the democratic neutrino mixing pattern and other nearly bi-maximal neutrino mixing schemes, from which one can get more hints to explore possible flavor symmetries and to build viable models of lepton mass matrices at appropriate superhigh energy scales. The author would like to thank H. Fritzsch, N. Haba, and M. Tanimoto for useful dis- cussions. He is also grateful to K.R.S. Balaji for helpful comments. 7 REFERENCES [1] For recent reviews with extensive references, see: R.N. Mohapatra, hep-ph/9910365; H. Fritzsch and Z.Z. Xing, hep-ph/9912358. [2] Y. Fukuda et al., Phys. Rev. Lett. 81, 1562 (1998) 1562; Phys. Rev. Lett. 81, 4279 (1998); http://www-sk.icrr.u-tokyo.ac.jp/dpc/sk/. [3] K.S. Babu, C.N. Leung, and J. Pantaleone, Phys. Lett. B 139, 191 (1993); P.H. Chankowski and Z. Pluciennik, Phys. Lett. B 316, 312 (1993); M. Tanimoto, Phys. Lett. B 360, 41 (1995); J. Ellis, G.K. Leontaris, S. Lola, and D.V. Nanopoulos, Eur. Phys. J. C 9, 389 (1999); E. Ma, J. Phys. G 25, L97 (1999); N. Haba, Y. Matsui, N. Okamura, and M. Sugiura, Prog. Theor. Phys. 103, 145 (2000); Y.L. Wu, J. Phys. G 26, 1131 (2000); E.J. Chun and S.K. Kang, Phys. Rev. D 61, 075012 (2000). [4] H. Fritzsch and Z.Z. Xing, Prog. Part. Nucl. Phys. 45, 1 (2000). [5] J.A. Casas, J.R. Espinosa, A. Ibarra, and I. Navarro, Nucl. Phys. B 573, 652 (2000); K.R.S.Balaji, A.S.Dighe, R.N.Mohapatra, andM.K.Parida, Phys. Rev.Lett.84, 5034 (2000); R.Adhikari, E. Ma, and G.Rajasekaran, Phys. Lett. B 486, 134 (2000); S. Lola, hep-ph/0005093; S.F. King and N.N. Singh, hep-ph/0006229; T. Miura, E. Takasugi, and M. Yoshimura, hep-ph/0007066; T. Kitabayashi and M. Yasu´e, hep-ph/0011153. [6] M. Tanimoto, Phys. Lett. B 483, 417 (2000); N. Haba, Y. Matsui, N. Okamura, and T. Suzuki, Phys. Lett. B 489, 184 (2000). [7] H.Fritzsch andZ.Z.Xing, Phys. Lett.B372, 265(1996); Phys. Lett.B413, 396(1997); Phys. Lett. B 440, 313 (1998); Phys. Rev. D 61, 073016 (2000); Z.Z. Xing, Nucl. Phys. B (Proc. Suppl.) 85, 187 (2000). [8] Particle Data Group, D.E. Groom et al., Eur. Phys. J. C 15, 1 (2000). [9] M. Gell-Mann, P. Ramond, and R. Slansky, in Supergravity, edited by D. Freedman and P. van Nieuwenhuizen (North Holland, Amsterdam, 1979), p. 315; T. Yanagida, in Proceedings of the Workshop on Unified Theory and Baryon Number in the Universe, edited by O. Sawada and A. Sugamoto (KEK, Tsukuba, Japan, 1979); R. Mohapatra and G. Senjanovic, Phys. Rev. Lett. 44, 912 (1980). [10] R. Barbieri, L.J. Hall, D. Smith, A. Strumia, and N. Weiner, JHEP 9812, 017 (1998). [11] M. Apollonio et al., Phys. Lett. B 420, 397 (1998). 8