NeutrinoOscillations.pdf

Transcription

1 Particle Physics Michaelmas Term 2010 Prof Mark Thomson Handout 11 : Neutrino Oscillations Prof. M.A. Thomson Michaelmas Neutrino Flavours Revisited Never directly observe neutrinos can only detect them by their weak interactions. Hence by definition is the neutrino state produced along with an electron. Similarly, charged current weak interactions of the state produce an electron = weak eigenstates For many years, assumed that were massless fundamental particles Experimental evidence: neutrinos produced along with an electron always produced an electron in CC Weak interactions, etc. e? e e - W e + W p u u d u d d n n d u d u u d p Experimental evidence: absence Suggests that and are distinct particles otherwise decay could go via: Prof. M.A. Thomson Michaelmas

2 Mass Eigenstates and Weak Eigenstates The essential feature in understanding the physics of neutrino oscillations is to understand what is meant by weak eigenstates and mass eigenstates Suppose the process below proceeds via two fundamental particle states? e - W e + W p u u d u d d n n d u d u u d p i.e. and Can t know which mass eigenstate (fundamental particle ) was involved In Quantum mechanics treat as a coherent state represents the wave-function of the coherent state produced along with an electron in the weak interaction, i.e. the weak eigenstate Prof. M.A. Thomson Michaelmas Neutrino Oscillations for Two Flavours Neutrinos are produced and interact as weak eigenstates, The weak eigenstates as coherent linear combinations of the fundamental mass eigenstates The mass eigenstates are the free particle solutions to the wave-equation and will be taken to propagate as plane waves The weak and mass eigenstates are related by the unitary 2x2 matrix (1) Equation (1) can be inverted to give (2) Prof. M.A. Thomson Michaelmas

3 Suppose at time a neutrino is produced in a pure state, e.g. in a decay Take the z-axis to be along the neutrino direction The wave-function evolves according to the time-evolution of the mass eigenstates (free particle solutions to the wave equation) where Suppose make an observation at a distance z from the production point. Making the (very good) approximation that gives For giving Prof. M.A. Thomson Michaelmas which is the phase of the wave for mass eigenstate with at a distance from the point of production Expressing the mass eigenstates,, in terms of weak eigenstates (eq 2): If the masses of are the same, the mass eigenstates remain in phase,, and the state remains the linear combination corresponding to and in a weak interaction will produce an electron If the masses are different, the wave-function no longer remains a pure with Prof. M.A. Thomson Michaelmas

4 Hence the two-flavour oscillation probability is: with The corresponding two-flavour survival probability is: e.g. wavelength e e e e e e e e e e e e e e e e e e Prof. M.A. Thomson Michaelmas Neutrino Oscillations for Three Flavours It is simple to extend this treatment to three generations of neutrinos. In this case we have: The 3x3 Unitary matrix is known as the Pontecorvo-Maki-Nakagawa-Sakata matrix, usually abbreviated PMNS Note : has to be unitary to conserve probability Using gives Prof. M.A. Thomson Michaelmas

5 Unitarity Relations The Unitarity of the PMNS matrix gives several useful relations: gives: (U1) (U2) (U3) (U4) (U5) (U6) To calculate the oscillation probability proceed as before Consider a state which is produced at as a (i.e. with an electron) Prof. M.A. Thomson Michaelmas The wave-function evolves as: where After a travelling a distance z axis in direction of propagation where As before we can approximate Expressing the mass eigenstates in terms of the weak eigenstates Which can be rearranged to give (3) Prof. M.A. Thomson Michaelmas

6 From which The terms in this expression can be represented as: Because of the unitarity of the PMNS matrix we have (U4): and, consequently, unless the phases of the different components are different, the sum of these three diagrams is zero, i.e., require different neutrino masses for osc. Prof. M.A. Thomson Michaelmas Evaluate using which gives: (4) (5) This can be simplified by applying identity (4) to (U4) 2 Substituting into equation (5) gives (6) Prof. M.A. Thomson Michaelmas

7 This expression for the electron survival probability is obtained from the coefficient for in eqn. (3): which using the unitarity relation (U1) can be written (7) This expression can simplified using with Phase of mass eigenstate i at z = L Prof. M.A. Thomson Michaelmas Define: NOTE: with is a phase difference (i.e. dimensionless) Which gives the electron neutrino survival probability Similar expressions can be obtained for the muon and tau neutrino survival probabilities for muon and tau neutrinos. Note that since we only have three neutrino generations there are only two independent mass-squared differences, i.e. and in the above equation only two of the are independent All expressions are in Natural Units, conversion to more useful units here gives: and Prof. M.A. Thomson Michaelmas

8 CP and CPT in the Weak Interaction In addition to parity there are two other important discrete symmetries: Parity Time Reversal Charge Conjugation Particle Anti-particle The weak interaction violates parity conservation, but what about C? Consider pion decay remembering that the neutrino is ultra-relativistic and only left-handed neutrinos and right-handed anti-neutrinos participate in WI RH Not Allowed LH Not Allowed RH Prof. M.A. Thomson Michaelmas LH Hence weak interaction also violates charge conjugation symmetry but appears to be invariant under combined effect of C and P CP transforms: RH Particles LH Particles LH Anti-particles RH Anti-particles If the weak interaction were invariant under CP expect All Lorentz invariant Quantum Field Theories can be shown to be invariant under CPT (charge conjugation + parity + time reversal) Particles/anti-particles have identical mass, lifetime, magnetic moments, Best current experimental test: Believe CPT has to hold: if CP invariance holds time reversal symmetry if CP is violated time reversal symmetry violated To account for the small excess of matter over anti-matter that must have existed early in the universe require CP violation in particle physics! CP violation can arise in the weak interaction (see also handout 12). Prof. M.A. Thomson Michaelmas

9 CP and T Violation in Neutrino Oscillations Previously derived the oscillation probability for The oscillation probability for by simply exchanging the labels can be obtained in the same manner or (8) Unless the elements of the PMNS matrix are real (see note below) (9) If any of the elements of the PMNS matrix are complex, neutrino oscillations are not invariant under time reversal NOTE: can multiply entire PMNS matrix by a complex phase without changing the oscillation prob. T is violated if one of the elements has a different complex phase than the others Prof. M.A. Thomson Michaelmas Consider the effects of T, CP and CPT on neutrino oscillations T CP CPT If the weak interactions is invariant under CPT Note C alone is not sufficient as it transforms LH neutrinos into LH anti-neutrinos (not involved in Weak Interaction) and similarly (10) If the PMNS matrix is not purely real, then (9) and from (10) Hence unless the PMNS matrix is real, CP is violated in neutrino oscillations! Future experiments, e.g. a neutrino factory, are being considered as a way to investigate CP violation in neutrino oscillations. However, CP violating effects are well below the current experimental sensitivity. In the following discussion we will take the PMNS matrix to be real. (question 22) Prof. M.A. Thomson Michaelmas

10 Neutrino Mass Hierarchy To date, results on neutrino oscillations only determine Two distinct and very different mass scales: Atmospheric neutrino oscillations : Solar neutrino oscillations: Two possible assignments of mass hierarchy: Normal Inverted In both cases: (solar) (atmospheric) Hence we can approximate Prof. M.A. Thomson Michaelmas Three Flavour Oscillations Neglecting CP Violation Neglecting CP violation considerably simplifies the algebra of three flavour neutrino oscillations. Taking the PMNS matrix to be real, equation (6) becomes: with Using: (see p. 365) Which can be simplified using (U4) Can apply to the expression for electron neutrino survival probability Which can be simplified using (U1) Prof. M.A. Thomson Michaelmas

11 Neglecting CP violation (i.e. taking the PMNS matrix to be real) and making the approximation that obtain the following expressions for neutrino oscillation probabilities: (11) (12) (13) (14) (15) (16) The wavelengths associated with and are: SOLAR and ATMOSPHERIC Long -Wavelength Short -Wavelength Prof. M.A. Thomson Michaelmas PMNS Matrix The PMNS matrix is usually expressed in terms of 3 rotation angles and a complex phase, using the notation Dominates: Writing this out in full: Atmospheric Solar There are six SM parameters that can be measured in oscillation experiments Solar and reactor neutrino experiments Atmospheric and beam neutrino experiments Reactor neutrino experiments + future beam Future beam experiments Prof. M.A. Thomson Michaelmas

12 Neutrino Experiments Before discussing current experimental data, need to consider how neutrinos interact in matter (i.e. our detectors) Two processes: Charged current (CC) interactions (via a W-boson) charged lepton Neutral current (NC) interactions (via a Z-boson) Two possible targets : can have neutrino interactions with atomic electrons nucleons within the nucleus CHARGED CURRENT NEUTRAL CURRENT Prof. M.A. Thomson Michaelmas Neutrino Interaction Thresholds Neutrino detection method depends on the neutrino energy and (weak) flavour Neutrinos from the sun and nuclear reactions have Atmospheric neutrinos have These energies are relatively low and not all interactions are kinematically allowed, i.e. there is a threshold energy before an interaction can occur. Require sufficient energy in the centre-of-mass frame to produce the final state particles Charged current interactions on atomic electrons (in laboratory frame) Require: Putting in the numbers, for CC interactions with atomic electrons require High energy thresholds compared to typical energies considered here Prof. M.A. Thomson Michaelmas

13 charged current interactions on nucleons (in lab. frame) Require: For CC interactions from neutrons require Electron neutrinos from the sun and nuclear reactors which oscillate into muon or tau neutrinos cannot interact via charged current interactions they effectively disappear Atmospheric muon neutrinos which oscillate into tau neutrinos cannot interact via charged current interactions disappear To date, most experimental signatures for neutrino oscillation are a deficit of neutrino interactions (with the exception of SNO) because below threshold for produce lepton of different flavour from original neutrino Prof. M.A. Thomson Michaelmas In Handout 10 derived expressions for CC neutrino-quark cross sections in ultra-relativistic limit (neglecting masses of neutrinos/quarks) For high energy muon neutrinos can directly use the results from page 316 with Cross section increases linearly with lab. frame neutrino energy For electron neutrinos there is another lowest order diagram with the same final state It turns out that the cross section is lower than the pure CC cross section due to negative interference when summing matrix elements In the high energy limit the CC neutrino-nucleon cross sections are larger due to the higher centre-of-mass energy: Prof. M.A. Thomson Michaelmas

14 Neutrino Detection The detector technology/interaction process depends on type of neutrino and energy REACTOR SOLAR ATMOSPHERIC/BEAM Deep Inelastic Scattering (p. 332) CC only Threshold 11 GeV NC only (see handout 13 ) NC + CC Prof. M.A. Thomson Michaelmas Atmospheric/Beam Neutrinos Water erenkov: e.g. Super Kamiokande Iron Calorimeters: e.g. MINOS, CDHS (see handout 10) Produce high energy charged lepton relatively easy to detect Solar Neutrinos Water erenkov: e.g. Super Kamiokande Detect erenkov light from electron produced in Because of background from natural radioactivity limited to Because Oxygen is a doubly magic nucleus don t get Radio-Chemical: e.g. Homestake, SAGE, GALLEX Use inverse beta decay process, e.g. Chemically extract produced isotope and count decays (only gives a rate) Reactor Neutrinos Liquid Scintillator: e.g. KamLAND Low energies large radioactive background Dominant interaction: Prompt positron annihilation signal + delayed signal from n (space/time correlation reduces background) ~100 s electrons produced by photons excite scintillator which produces light Prof. M.A. Thomson Michaelmas

15 1) Long Baseline Neutrino Experiments Initial studies of neutrino oscillations from atmospheric and solar neutrinos atmospheric neutrinos discussed in examinable appendix Emphasis of neutrino research now on neutrino beam experiments Allows the physicist to take control design experiment with specific goals In the last few years, long baseline neutrino oscillation experiments have started taking data: K2K, MINOS, CNGS, T2K Basic Idea: Intense neutrino beam Two detectors: one close to beam the other hundreds of km away Measure ratio of the neutrino energy spectrum in far detector (oscillated) to that in the near detector (unoscillated) Partial cancellation of systematic biases Near Detector (unoscillated) Far Detector (oscillated) Depth of minimum sin2 2 Position of min. m 2 Prof. M.A. Thomson Michaelmas MINOS 120 GeV protons extracted from the MAIN INJECTOR at Fermilab (see p. 271) 2.5x10 13 protons per pulse hit target very intense beam MW on target Soudan Mine, Minnesota 735 km Fermilab Two detectors: 1000 ton, NEAR Detector at Fermilab : 1 km from beam 5400 ton FAR Detector, 720m underground in Soudan mine, N. Minnesota: 735 km from beam Prof. M.A. Thomson Michaelmas

16 The MINOS Detectors: Dealing with high energy neutrinos The muons produced by interactions travel several metres Steel-Scintillator sampling calorimeter Each plane: 2.54 cm steel +1 cm scintillator A charged particle crossing the scintillator produces light detect with PMTs Steel Plastic scintillator Alternate layers have strips in x/y directions FAR NEAR Prof. M.A. Thomson Michaelmas Neutrino detection via CC interactions on nucleon Example event: Signal from hadronic shower The main feature of the MINOS detector is the very good neutrino energy resolution Muon energy from range/curvature in B-field Hadronic energy from amount of light observed Prof. M.A. Thomson Michaelmas

17 MINOS Results For the MINOS experiment L is fixed and observe oscillations as function of For first oscillation minimum at To a very good approximation can use two flavour formula as oscillations corresponding to occur at, beam contains very few neutrinos at this energy + well below detection threshold MINOS Collaboration, Phys. Rev. Lett. 101, , 2008 no oscillations best fit osc Prof. M.A. Thomson Michaelmas ) Solar Neutrinos The Sun is powered by the weak interaction producing a very large flux of electron neutrinos Several different nuclear reactions in the sun complex neutrino energy spectrum All experiments saw a deficit of electron neutrinos compared to experimental prediction the SOLAR NEUTRINO PROBLEM e.g. Super Kamiokande Prof. M.A. Thomson Michaelmas

18 Solar Neutrinos I: Super Kamiokande ton water erenkov detector Water viewed by Photo-multiplier tubes Deep underground to filter out cosmic rays otherwise difficult to detect rare neutrino interactions Mt. Ikenoyama, Japan 36 m 34 m Prof. M.A. Thomson Michaelmas Detect neutrinos by observing erenkov radiation from charged particles which travel faster than speed of light in water c/n e Can distinguish electrons from muons from pattern of light muons produce clean rings whereas electrons produce more diffuse fuzzy rings Prof. M.A. Thomson Michaelmas

19 Sensitive to solar neutrinos with For lower energies too much background from natural radioactivity ( -decays) Hence detect mostly neutrinos from Detect electron erenkov rings from In LAB frame the electron is produced preferentially along the direction Event/day/kton/bin S.Fukada et al., Phys. Rev. Lett , background due to natural radioactivity ( -decay ) e from the sun Results: Clear signal of neutrinos from the sun However too few neutrinos DATA/SSM = 0.45±0.02 SSM = Standard Solar Model Prediction The Solar Neutrino Problem cosθ sun Prof. M.A. Thomson Michaelmas Solar Neutrinos II: SNO Sudbury Neutrino Observatory located in a deep mine in Ontario, Canada 1000 ton heavy water (D 2 O) erenkov detector D 2 O inside a 12m diameter acrylic vessel Surrounded by 3000 tons of normal water Main experimental challenge is the need for very low background from radioactivity Ultra-pure H 2 O and D 2 O Surrounded by 9546 PMTs H 2 O Transparent acrylic vessel D 2 O PMTs Ultra-pure H 2 0 and D 2 0 Prof. M.A. Thomson Michaelmas

20 Detect erenkov light from three different reactions: CHARGE CURRENT Detect erenkov light from electron Only sensitive to (thresholds) Gives a measure of flux NEUTRAL CURRENT Neutron capture on a deuteron gives 6.25 MeV Detect erenkov light from electrons scattered by Measures total neutrino flux ELASTIC SCATTERING Detect erenkov light from electron Sensitive to all neutrinos (NC part) but larger cross section for Prof. M.A. Thomson Michaelmas Experimentally can determine rates for different interactions from: angle with respect to sun: electrons from ES point back to sun energy: NC events have lower energy 6.25 MeV photon from neutron capture radius from centre of detector: gives a measure of background from neutrons SNO Collaboration, Q.R. Ahmad et al., Phys. Rev. Lett. 89:011301, 2002 Using different distributions obtain a measure of numbers of events of each type: CC : 1968 ± 61 ES : 264 ± 26 NC : 576 ± 50 Measure of electron neutrino flux + total flux! Prof. M.A. Thomson Michaelmas

21 Using known cross sections can convert observed numbers of events into fluxes The different processes impose different constraints Where constraints meet gives separate measurements of and fluxes SNO Results: ( e only) (NC constrains total flux) SNO Collaboration, Q.R. Ahmad et al., Phys. Rev. Lett. 89:011301, 2002 SSM Prediction: Clear evidence for a flux of and/or from the sun Total neutrino flux is consistent with expectation from SSM Clear evidence of and/or neutrino transitions Prof. M.A. Thomson Michaelmas Interpretation of Solar Neutrino Data The interpretation of the solar neutrino data is complicated by MATTER EFFECTS The quantitative treatment is non-trivial and is not given here Basic idea is that as a neutrino leaves the sun it crosses a region of high electron density The coherent forward scattering process ( ) for an electron neutrino CC + NC is different to that for a muon or tau neutrino NC Can enhance oscillations MSW effect A combined analysis of all solar neutrino data gives: Prof. M.A. Thomson Michaelmas

22 3) Reactor Experiments To explain reactor neutrino experiments we need the full three neutrino expression for the electron neutrino survival probability (11) which depends on Substituting these PMNS matrix elements in Equation (11): Contributions with short wavelength (atmospheric) and long wavelength (solar) For a 1 MeV neutrino Amplitude of short wavelength oscillations given by Prof. M.A. Thomson Michaelmas Reactor Experiments I : CHOOZ Two nuclear reactors, each producing 4.2 GW Place detector 1km from reactor cores Reactors produce intense flux of Detector France Detector reactors 150m underground Anti-neutrinos interact via inverse beta decay Detector is liquid scintillator loaded with Gadolinium (large n capture cross section) Detect photons from positron annihilation and a delayed signal from photons from neutron capture on Gadolinium Prof. M.A. Thomson Michaelmas

23 At 1km and energies > 1 MeV, only the short wavelength component matters CHOOZ Raw Data Background subtracted Compare to effect of oscillations Data agree with unoscillated prediction both in terms of rate and energy spectrum Hence must be small! From atmospheric neutrinos (see appendix) can exclude Hence the CHOOZ limit: can be interpreted as CHOOZ Collaboration, M.Apollonio et al., Phys. Lett. B420, , 1998 Exact limit depends on Prof. M.A. Thomson Michaelmas Reactor Experiments II : KamLAND Detector located in same mine as Super Kamiokande 18m 70 GW from nuclear power (7% of World total) from reactors within km Liquid scintillator detector, 1789 PMTs Detection via inverse beta decay: Followed by prompt delayed Prof. M.A. Thomson Michaelmas

24 For MeV neutrinos at a distance of km oscillations due to are very rapid Experimentally, only see average effect Here: Averaging over rapid oscillations neglect Obtain two-flavour oscillation formula multiplied by From CHOOZ (Try Question 21) Prof. M.A. Thomson Michaelmas KamLAND RESULTS: Observe: 1609 events Expect: 2179±89 events (if no oscillations) KamLAND Collaboration, Phys. Rev. Lett., , 2008 Clear evidence of electron anti-neutrino oscillations consistent with the results from solar neutrinos Oscillatory structure clearly visible Compare data with expectations for different osc. parameters and perform 2 fit to extract measurment Prof. M.A. Thomson Michaelmas

25 Combined Solar Neutrino and KamLAND Results KamLAND data provides strong constraints on Solar neutrino data (especially SNO) provides a strong constraint on Solar Neutrinos KamLAND Combined Prof. M.A. Thomson Michaelmas Summary of Current Knowledge SOLAR Neutrinos/KamLAND KamLAND + Solar: SNO + KamLAND + Solar: Atmospheric Neutrinos/Long Baseline experiments MINOS: Super Kamiokande: CHOOZ + (atmospheric) Currently no knowledge about CP violating phase Prof. M.A. Thomson Michaelmas

26 In the limit For the approximate values of the mixing angles on the previous page obtain: Have approximate expressions for mass eigenstates in terms of weak eigenstates: Prof. M.A. Thomson Michaelmas Final Words: Neutrino Masses Neutrino oscillations require non-zero neutrino masses But only determine mass-squared differences not the masses themselves No direct measure of neutrino mass only mass limits: Note the refer to charged lepton flavour in the experiment, e.g. refers to the limit from tritium beta-decay Also from cosmological evolution infer that the sum 10 years ago assumed massless neutrinos + hints that neutrinos might oscillate! Now, know a great deal about massive neutrinos But many unknowns:, mass hierarchy, absolute values of neutrino masses Measurements of these SM parameters is the focus of the next generation of expts. Prof. M.A. Thomson Michaelmas

27 Examinable Appendix: Atmospheric Neutrinos High energy cosmic rays (up to ev) interact in the upper part of the Earth s atmosphere The cosmic rays (~86% protons, 11% He Nuclei, ~1% heavier nuclei, 2% electrons ) mostly interact hadronically giving showers of mesons (mainly pions) Neutrinos produced by: Flux Typical energy : Expect Observe a lower ratio with deficit of coming from below the horizon, i.e. large distance from production point on other side of the Earth Prof. M.A. Thomson Michaelmas Super Kamiokande Atmospheric Results Typical energy: (much greater than solar neutrinos no confusion) Identify and interactions from nature of erenkov rings Measure rate as a function of angle with respect to local vertical Neutrinos coming from above travel ~20 km Neutrinos coming from below (i.e. other side of the Earth) travel ~12800 km from below from above Above expected Below Prediction for rate agrees with data Strong evidence for disappearance of for large distances Consistent with oscillations Don t detect the oscillated as typically below interaction threshold of 3.5 GeV Prof. M.A. Thomson Michaelmas

28 Interpretation of Atmospheric Neutrino Data Measure muon direction and energy not neutrino direction/energy Don t have E/ resolution to see oscillations Oscillations smeared out in data Compare data to predictions for Data consistent with: Prof. M.A. Thomson Michaelmas Flavour Treatment of Atmospheric Neutrinos non-examinable The energies of the detected atmospheric neutrinos are of order 1 GeV The wavelength of oscillations associated with is If we neglect the corresponding term in the expression for - equation (16) The Super-Kamiokande data are consistent with the possibility of being small Hence the CHOOZ limit: can be interpreted as which excludes NOTE: the three flavour treatment of atmospheric neutrinos is discussed below. The oscillation parameters in nature conspire in such a manner that the two flavour treatment provides a good approximation of the observable effects of atmospheric neutrino oscillations Prof. M.A. Thomson Michaelmas

29 3-Flavour Treatment of Atmospheric Neutrinos non-examinable Previously stated that the long-wavelength oscillations due to have little effect on atmospheric neutrino oscillations because for a the wavelength for a 1 GeV neutrino is approx km. However, maximum oscillation probability occurs at This is not small compared to diameter of Earth and cannot be neglected As an example, take the oscillation parameters to be Predict neutrino flux as function of Consider what happens to muon and electron neutrinos separately muon neutrinos only electron neutrinos only has a big effect at Prof. M.A. Thomson Michaelmas From previous page it is clear that the two neutrino treatment of oscillations of atmospheric muon neutrinos is a very poor approximation However, in atmosphere produce two muon neutrinos for every electron neutrino Need to consider the combined effect of oscillations on a mixed beam with both and 2/3 + 1/3 e At large distances the average muon neutrino flux is still approximately half the initial flux, but only because of the oscillations of the original electron neutrinos and the fact that Because the atmospheric neutrino experiments do not resolve fine structure, the observable effects of oscillations approximated by two flavour formula Prof. M.A. Thomson Michaelmas