Axion Decay of a 17-Kev Neutrino

VOLUME 67, NUMBER 7 PHYSICAL REVIEW LETTERS 12 AUGUST 1991

Axion Decay of a 17-keV Neutrino

Myron Bander ' Department of Physics, University of Californiai, rvine, California 92717 (Received 25 March 1991; revised manuscript received 15 May 1991) A 17-keV particle tentatively observed in P decays is assumed to be a new Dirac-type neutrino. Cou- pling to the Higgs sector of the Dine-Fischler-Srednicki-Zhitnitsky (DFSZ) axion model provides a mass for this particle, its mixing with the electron neutrino, and a channel for a reasonably fast decay into an electron neutrino and the light DFSZ axion. Results are consistent with most laboratory, astrophysical, and cosmological limits on the properties of this particle and of the axion.

PACS numbers: 12.15.Ff, 13.35.+s, 14.60.6h, 14.80.Gt

Over the past several years tentative evidence has ap- a light neutrino and a majoron, while Babu, Mohapatra, peared in nuclear P decay for the existence of a 17-keV and Rothstein [6] considered it to be a Dirac neutrino neutrino [1-3]. Approximately 1% of the decays produce whose right-handed component has new couplings, per- this new particle [2,3], which we shall call VH. Although mitting a fast decay into three light neutrinos. it is simple to envisage schemes where the ordinary elec- In this Letter we propose that VH is a new, SU(2)L- tron neutrino mixes with the vH, where vH is either one of singlet, Dirac neutrino that obtains its mass and mixes the known neutrino species or a new type, laboratory ex- with v, through the Higgs mechanism of the Dine- periments and astrophysical and cosmological considera- Fischler-Srednicki-Zhitnitsky (DFSZ) [7,8] "invisible" tions put constraints on the properties of this particle. axion scheme. The primary decay of v& is into v, and an One of the tightest is the requirement that the lifetime of axion; the lifetime for this decay is —10' s, within the VH be less than 10' s [4], in order to prevent an "over- allowed cosmological limit. closure" of the Universe. This limit was used as a guide The Higgs sector of the DFSZ model consists of two in the formulation of this model; constraints from primor- doublets h] and h2 with vacuum expectation values v] dial nucleosynthesis and from observations on the super- and U2, respectively, and a singlet g with a much larger nova 1987A will be discussed, as will be those from ter- expectation value u. The Yukawa Lagrangian has two restrial observations. Recently two proposals have ap- parts, XY1, the usual couplings of the Higgs particles to peared to account for this particle; Glashow [5] took the quarks and leptons, and an additional part involving the vH to be a Majorana particle whose primary decay is into VH, +Y2'

( ) (M) ' m"' qL;h1dR;+ qL;h»R;+ l«h1eR; +H.c. , I V] V2 V(

m pH g, m,„ XY2 VHLgVttR+ leLJl2VHR+ H.c. v2 u;, d;, e;, and v; denote the upper and lower components of the quarks and leptons in generation i. g, is the L The spectrum contains, of course, an axion a with decay amount of mixing of vH with v„according to the experi- constant ments [2,3] g, =0.1. The light neutrinos remain mass- (4 2 2+ 2 2) 1/2 less and the linear combination of states that couples to (3) both charged and neutral currents is ~v, )+(, ~VH). In 2U]U2 the above Lagrangian the new heavy neutrino is coupled where v =(vt+v2)'I . In addition to its usual couplings only to the electron family. Extra heavy neutrinos cou- to quarks and leptons the axion couples to vH and v„ pling to the other families could be introduced, or we could allow v~ to couple to the other lepton families. We U] +vHv a gv mva VeL VHRa+ H. C. (4) shall return to these topics towards the end of this Letter. a This Lagrangian is invariant under a U(1) symmetry with Peccei-Quinn charges This leads to a lifetime for the decay vH v, a of

Q(u;R) =b, Q(d;R) =c, Q(l;R) =c, Q(VHR) =b, r Q(h1) = —c, Q(h2) =b, Q(g) = —(b+c)l2, (2) v 2fa 32tr Q (vHL ) = (b —c)/2 . m„'„ 1991 The American Physical Society 801 VOLUME 67, NUMBER 7 PHYSICAL REVIEW LETTERS 12 AUGUST 1991 or acceptable/unacceptable; modifications, such as decreas- t ' 2 ing slightly the rise in temperature as one approaches the z = 1.35x108 10 GeV center of the collapsed core, would lower this rate considerably. —3 2 f1' pH Laboratory constraints arise from experiments on v, s. (6) disappearance. The predicted sin (28) for v, disappear- 17 keV 0. 1 ance is 10, while the experimental bound [16] is 7 Astrophysical considerations put constraints on the ax- x10 ion mass [9,10] and in turn on f,. The evolution of red As discussed earlier, these ideas may be extended tp in- giants provided the most stringent stellar limit, rn, (0.02 volve the other light neutrinos. Whether there is a single eV or f, & 3 x 10 GeV. The observed cooling rate of su- vz mixing with neutrinos from each family or each fami- pernova 1987A tightened this limit to m, & 3 x 10 eV ly has its own vht, all the previous discussions and lim- or f, & 2&&10 GeV. With Uz/vl in Eq. (6) of order uni- its apply. In addition, the v„-disappearance results [17] ty, and f, =10 GeV we obtain r = 10' s, within this place a limit on the mixing of this neutrino with v~ of cosmological limit. (Theories of structure formation in (g, ) (0.02. Should vH couple both to the electron and the Universe place a more stringent limit on the particles to the muon neutrinos, it would induce a p ey transi- decaying into relativistic daughters, namely, z ~ 6x10 s tion. We may estimate the branching ratio for this pro- [11]. Should these ideas on structure formation hold, cess, they would rule out this and other models for this particle. ) Q 2 —l2g2 8(p ey) = (g g ) 10 (7) Primordial nucleosynthesis is used to place a limit on mu the number of neutrino types. A recent analysis [12] limits the number of species (neutrino plus antineutrino of which is well within present bounds [18]. one handedness) to be less than 3.4. Neutrinos that de- We have presented a mechanism for a "rapid" decay of couple very much earlier than the nucleosynthesis epoch the hinted-at 17-keV neutrino that involves no hypotheti- contribute significantly less than 1 to this bound; this is cal particles other than those already postulated for other the situation with the right-handed component of vII. reasons. The DFSZ axion scheme involves a singlet The left-handed component of vH decouples only shortly Higgs meson, the g, whose only purpose was to make f, prior to ordinary neutrino decoupling and thus contrib- large; we have provided it with another raison d' etre, utes fully to the expansion rate of the Universe at the namely, that of giving v~ its mass. time of nucleosynthesis, seriously violating the above Useful discussions with Dr. S. Barwick, Dr. Y. Nir, bound. However, a recent reanalysis of nucleosynthesis and Dr. H. Rubinstein are gratefully acknowledged. This [13], which allowed for the possibility of a neutrino work was supported in part by the National Science chemical potential, weakened the limit on the number of Foundation under Grant No. PHY-89-06641. neutrinos considerably. Astrophysical constraints come from a study of the cooling rates of the supernova 1987A. Cooling due to '" new quanta should not be so fast as to interfere with the Electronic address: mbanderucivmsa. bitnet; mbander observed neutrino fluxes. In order to prevent the "sterile" funth. ps. uci.edu. [1] J. J. Simpson, Phys. Rev. Lett. 54, 1891 right-handed component of ordinary neutrinos from car- (1985); J. J. Simpson and A. Hime, Phys. Rev. D 39, 1825 (1989); A. rying away energy too rapidly, an upper limit of 14-17 Hime and J. J. Simpson, ibid 39, 1837 (1.989). for keV [14] the mass of such a particle was obtained. As [2] A. Hime and N. A. Jelley, Phys. Lett. B 257, 441 (1991). vH interacts with a strength of (, GF, the limit on its [3] B. Sur et al. , Phys. Rev. Lett. 66, 2444 (1991). mass from such considerations is 140-170 keV. The [4] E. W. Kolb and M. S. Turner, The Early Universe right-handed component of our 17-keV vH will not be (Addison-Wesley, Redwood City, CA, 1990), p. 141. produced at a rate sufficient to aff'ect the evolution of a [5] S. L. Glashow, Phys. Lett. B 256, 255 (1991). supernova. The left-handed component of vH has a mean [6] K. S. Babu, R. N. Mohapatra, and I. Z. Rothstein, Phys. free path I/g, or 100 times that of an ordinary neutrino. Rev. Lett. 67, 545 (1991). It will set up a vH-sphere and thermally radiate these [7] A. R. Zhitnitsky, Yad. Fiz. 31, 497 (1980) [Sov. J. Nucl. Phys. 260 particles. Details depend strongly on the density profile 31, (1980)]; M. Dine, W. Fischler, and M. Srednicki, Phys. Lett. 104B, 199 (1981). and equation of state of a supernova core. Following Gri- [8] H-Y. Cheng, Phys. Rep. 128, 1 (1988). fols and Masso [15],we find that vHt radiates at a rate of [9] Kolb and Turner, Ref. [4], p. 401. 100 =24 times that pf an prdinary neutrinp. Fpr the [10] G. G. RaA'elt, Mod. Phys. Lett. A 5, 2581 (1990). parameters used in this analysis, the rate of ordinary neu- [11]G. Steigman and M. S. Turner, Nucl. Phys. B253, 375 trino radiation is 10 ergs/s. Thus 2.4&&10 ergs/s are (1985). radiated into vH s. Such a radiation rate is marginally [12] K. A. Olive, D. N. Schramm, G. Steigman, and T. P.

802 VOLUME 67, NUMBER 7 PHYSICAL REVIEW LETTERS 12 AUGUS+ 1991

Walker, Phys. Lett. B 236, 454 (1990). we use n=7 in the parametrization of the density profile [13] K. A. Olive, D. N. Schramm, D. Thomas, and T. P. p(r) =(r/ro) ", which is the value preferred by these au- Walker, University of Minnesota Report No. UMN-TH- thors. 929l91 (to be published). [16] Particle Data Group, J. Hernandez et al. , Phys. Lett. B [14] G. G. Raffelt and D. Seckel, Phys. Rev. Lett. 60, 1793 239, 1 (1990), p. vi. 31. (1988); R. Gandhi and A. Burrows, Phys. Lett. B 246, [17] Hernandez et al. , Ref. [16],p. vi. 30. 149 (1990). [18] Hernandez er al. , Ref. [16],p. vi. 1 l. [15] J. J. Grifols and E. Masso, Nucl. Phys. B331, 244 (1990);

803