GSI Effect Aspen Center for Physics June 19, 2009 Stuart Freedman University of California, Berkeley Lawrence Berkeley National Laboratory Stuart Freedman University of California, Berkeley Lawrence Berkeley National Laboratory

GSI

GSI Complex

Recording the Schottky-noise

time SMS 4 particles with different m/q Yu. A. Litvinov

Sin(  1 ) Sin(  2 ) Sin(  3 ) Sin(  4 ) 11 22 33 44 time Fast Fourier Transform SMS Yu. A. Litvinov

SMS: Single-ion sensitivity

Promethium Praseodymium

Also 122 I --

Fermi

6 Li W. Pauli Pauli’s Theory of Beta Decay

m2C2m2C2 m1C2m1C2 Neutrino Oscillations

6 He 6 Li e Pauli’s Theory of Beta Decay

E. Fermi e np Chadwick discovers the neutron Fermi invents a new theory of  decay We now understand this as an example of a Quantum (Gauge) Field Theory

Nucleus1/λ sec T Period sec α Amplitude Φ Phase Rad Χ 2 /DOF 140 Pr666(444)7.06(9)0.18(3)0.4(4)67.18/ Pm44.6(84)7.10(25)0.23(4)-1.6(5)31.82/35

Fermi Is the effect a consequence of neutrino mass and mixing? A  m 2 naively inferred from GSI data is within a factor of 3 of the most recent measurement from KamLAND.

Solar, KamLAND, EC Results on Δm²-tan²θ EC-Decay P. Kienle

Was the effect seen at the GSI missed in each of the numerous experiments which have measured electron capture lifetimes? Can the effect seen at the GSI be reproduced in a “simpler” experiment measuring electron capture decay? Experimental Approach

No previous experiment observed hydrogenic initial ion and/or “free decay” either in a trap or a storage ring. Many other experiments would not have observed the effect because of the production method employed. In most experiments designed to measure lifetimes the production time of the isotope in question is chosen to be comparable to or longer than, the expected lifetime. A literature search was made for experiments carried out with a short bombardment time and a long counting time that might have seen the same modulation reported at the GSI. Was it missed?

/ Sm 142 Eu min sec 81% 82% 11% “Structure of 142 Sm from the decay of 142 Eu” G.G. Kennedy, S.C. Gurathi, and S.K. Mark, Phys. Rev. C 12, 553 (1975). Study of Isomeric States in 142 Eu

Exponential / Sm 142 Eu min sec 81% 82% 11% Data of Kennedy et al

Υ-Ray Line keV 1/λ sec T Period sec Α Amplitude Φ Phase Rad Χ 2 /DOF (53)0.0136(55)5.6(8) (73)0.0133(84)1.13(55) Eu

It not surprising to find an amplitude at ~ 3σ -- if the phase is not fixed. A + δA

Fermi

Berkeley Gas Filled Separator 124 Sn( 23 Na,5n) 142 Pm Catcher foil Clover Detector

Nd Kα and Kβ X-rays0.511 MeV Spectrum in Clover Detector – Beam on and Beam off

Decay time spectrum of detected Kα X-rays from 142 Pm decay 142 Pm Kα

Decay time spectrum of detected 511 keV γ-rays from 142 Pm decay

Berkeley Experiment T Period sec α AmplitudeΦ Phase Rad Χ 2 /DOF Kα3.178(36) (74)-1.93(76)614.6/ keV0.8129(8)0.0173(70)2.15(59)597.2/593 Kα Beta decay

arXiv:nucl-ex/ Thick tantalum target/source 181 Ta( 3 He,4n) 180 Re Bombardment sec Detect 903 keV gamma

Decay time spectrumFrequency spectrum of decay

T sec α AmplitudeΦ phase Rad Χ 2 /DOF MeV (8)446.2/486

142 Pm Laboratory T Period sec α AmplitudeΦ Phase Rad Χ 2 /DOF GSI in flight7.10(25)0.23(4)-1.6(5)31.82/35 GSI at rest4.97(18)0.23(4)-1.6(5) LBNL3.178(36) (74)-1.93(76)614.6/ Pm

EC-decay vs. Beta-decay for 142Pm Single analysis only! Checks are to be done -!- Preliminary -!- F. Bosch

EC-decay vs. Beta-decay for 142Pm Single analysis only! Checks are to be done F. Bosch

Experiments observing electron capture disagree about the existence of time modulations in the decay probability. What is the explanation of the modulations from the GSI experiments? Fundamental physics? Experimental artifact? Is the explanation related to neutrino mass and mixing? Status

Fermi Is this surprising result another consequence of neutrino mass and mixing like neutrino oscillations? B. Kayser poll: The answer “NO!” is favored by 13 out of the 18 theorists who have expressed a strong opinion. Papers relating to GSI oscillations While experimentalists usually remain agnostic some have come down in favor of “YES!”. I have been in favor of an experimental resolution but “theory” has gotten in the way … I will come back to this point. Blue -- Neutrino Osc. Yes Red -- Neutrino Osc. No P. Vetter

Decay scheme of 122 I Experiment: F. Bosch

Decay Statistics Correlations: injections ~1080 EC-decays Many ions: 5718 injections ~5000 EC-decays F. Bosch

Fit by a Pure Exponential: dN EC /dt = N 0 λ EC exp(-λt) Single analysis only ! Checks are to be done -!- Preliminary -!- F. Bosch

Fit by : dN EC /dt = N 0 λ EC exp(-λt) [1 + a cos(ωt + φ)] Single analysis only! Checks are to be done -!- Preliminary -!- F. Bosch

Sum of All Evaluated EC Decays Single analysis only ! Checks are to be done -!- Preliminary -!- F. Bosch

Synopsis M (amu) ω (1/s) T lab (s) a φ (rad) 122* 1.038(6) 6.05(4) 0.21(2) - 0.2(2) (10) 7.06(8) 0.18(3) 0.4(4) (27) 7.10(22) 0.23(4) - 1.6(4) * preliminary → oscillation period T scales in proportion to the nuclear mass M of the parent nucleus F. Bosch

Cohen, Ligeti, and Glashow l for light, h for heavy Entangled states in electron capture

Fermi Neutrino-mass Induced lifetime modulation possible: Neutrinos existed in the initial state -- electron neutrinos at creation (a coherent mixture of mass eigenstates). A coherent mixture of mass eigenstates the “electron neutrino” is “measured” in the decay, which occurs later. Pauli vs Fermi Lifetime-mass Induced lifetime modulation not possible: No neutrinos existed in the initial state at the time of creation of the nucleus. Production and decay occur simultaneously. Fermi’s description of beta decay is correct and the evidence for modulation in decay probabilities does not have anything to do with neutrino oscillations.

Decay Rates Do Decay Rates O s c i l l a t e Because Of Neutrino Mass? Boris Kayser Aspen June 19, 2009 Photo courtesy of Y. Litvinov

Theoretical Opinion Neutrino mass can make decay rates oscillate: Faber, Ivanov, Kienle, Kleinert, Lipkin, Pitschmann, Reda No it cannot: Adler, Cohen, Gal, Frere, Giunti, Glashow, Kayser, Kienert, Kopp, Ligeti, Lindner, Merle, Peshkin

Our View Stephen Adler and B. K. With thanks to Jean-Marie Frere

Quantum Mechanical Rules If different intermediate states (or paths) lead to the same final state, and we don’t know which path is taken in each event, then the paths contribute to the event rate coherently. Source Screen Detector Total Amp = Amp(1) + Amp(2) 1 2

The rates to produce different final states that differ from one another in any way (particle content, kinematical properties, etc.) contribute to the total event rate incoherently. This is true whether or not we can actually distinguish the different final states in practice.

Neutrino Flavor Change (Oscillation)

oscillatescoherent interference Flavor change oscillates because of coherent interference between different neutrino mass eigensates i with different masses m i. Intermediate state ++ Detector e–e– ++ Source Amp ++ i Detector e–e– ++ Source Amp intermediate The neutrinos are an intermediate state.

++ Detector e–e– ++ Source Amp ++ i e–e– ++ Mixing matrix Mass of i Distance Energy Neutrino mass-splitting dependence is from interference.

Electron Capture Decay

in which a parent particle P decays to a daughter particle D plus a neutrino, there are actually 3 distinct decay modes: In electron-capture (EC) decays such as —, There are (at least) 3 neutrino mass eigenstates i, with unequal masses m i. P

In principle, we can capture the final-state neutrino and measure its mass. Then we will know which i it is. In principle, we can also measure the energy-momentum of the recoil D, and that of the neutrino. P None of these measurements affects the decay.

The 3 possible final states differ in particle content. For given P energy, they also differ in the energy-momenta of the individual particles. P The rates for decay to these 3 final states contribute incoherently to the total decay rate.

Mass eigenstate For example — An incoherent sum

Left-handed Taking mixing into account The Standard-Model Lagrangian for the leptonic couplings to the W boson is — Semi-weak coupling Mixing matrix comes from the term involving. This term doesn’t know about the other neutrino mass eigenstates or their masses.

does not depend on neutrino mass splittings.

coherent Q: Isn’t the neutrino in P  D + a e, and isn’t the e a coherent superposition of mass eigenstates? not A: No, actually the neutrino is not a e. But, for given P energy-momentum, the energy-momenta of the different i differ. Of course, the produced neutrino is almost a e.

But even if we approximate it as exactly a e, it is still true that — An incoherent sum Let’s recall — z

In, the pions are produced in a P wave: A coherent superposition of momentum eigenstates. Nevertheless — An incoherent sum

The Trap Of the Observed Period e is approximately a superposition of just two mass eigenstates, 1 and 2, with. In P D +, in the P rest frame, the mass eigenstates 1 and 2 are produced with energies differing by — If 1 and 2 were present in the initial state and contributed coherently, this could lead to an oscillation in the decay rate with period — T = 14 sec.

At GSI, the parent ions are moving with  = 1.4, so in the lab. the time-dilated oscillation period would be — T = 20 sec. Observed period: T = 7 sec. Some people have been suckered in by this coincidence.

But we can escape the trap by a sobering observation: The GSI anomaly can be considered equally well as an oscillation with time or with distance. Whereas, in the P rest frame, the energies of the mass eigenstates 1 and 2 differ by —, their momenta differ by — ;. The above an oscillation wavelength — = 110 km. Observed : = 1,500,000 km.

The oscillation seen at GSI is not due to neutrino mass splittings. Summary

Fermi F=3/2 F=1/2  EC  ,EC F=1/2 Hydrogenic Atoms in a Storage Ring First mentioned G. Lambiase, G. Papini, G. Scarpetta, arXiv: