DAEδALUS. A Novel Approach to CP violation in the ν sector. Janet Conrad, INT Workshop August 5,6 2010

Transcription

1 DAEδALUS A Novel Approach to CP violation in the ν sector Janet Conrad, INT Workshop August 5,6 2010

2 Decay At rest Experiment for δ cp studies At the Laboratory for Underground Science Use decay-at-rest neutrino beams, and the planned 300 kton H 2 O detector (Gd doped) at the Deep Underground Science & Engineering Laboratory to search for CP violation in the neutrino sector

3 Outline: Some observations about CP violation The Conventional Beam Approach An unconventional approach: DAEδALUS Practicalities -- especially the accelerators Mike s talk will show combined-with-lbne sensitivity.

4 c ij =cosθ ij s ij =sinθ ij The CP Violation Parameter From Atmospheric and Long Baseline Disappearance Measurements From Reactor Disappearance Measurements From Appearance Measurements From Solar Neutrino Measurements

5 The oscillation of muon-flavor to electron-flavor at the atmospheric Δm 2 may show CP-violation dependence! in a vacuum We want to see if δ is nonzero } terms depending on mixing angles } terms depending on mass splittings

6 The oscillation of muon-flavor to electron-flavor at the atmospheric Δm 2 may show CP-violation dependence! in a vacuum We want to see if δ is nonzero CP violation is all about interference. The δ-dependent terms arise from interference between the Δm 13 2 and Δm 12 2 oscillations

7 Most parameters are well known N/A N/A Except for that pesky θ 13! We will end up having to quote our sensitivity as allowed regions in both θ 13 and δ

8 So what do we know about δ vs θ 13??? The actual values could be anywhere in this region! including at low sin 2 2θ 13 This region ruled out by Chooz and Palo Verde

9 The experimental goal is to minimize the Jelly Bean across this plane Imagine the real values are: δ = 80 sin 2 2θ 13 = sigma error 2 sigma error

10 Part II: The Conventional Wisdom on how to search for CP violation in the neutrino sector

11 Our equation flips sign between ν µ ν e & ν µ ν e in a vacuum what we want to measure } terms depending on mixing angles } terms depending on mass splittings

12 The classic idea for how to see CP violation: P osc (ν µ ν e ) P osc (ν µ ν e ) P osc (ν µ ν e ) δ 0 π CP This is in a vacuum (or air). CP parameter P osc (ν µ ν e )

13 But the proposed experiments to search for CP violation shoot the neutrinos through a lot of matter Here s why The easiest way to make a high-flux beam which switches from ν to ν: p target magnetic field region for π and K decay dump ν or ν Conventional neutrino beam s of MeV to a few GeV

14 P is maximized when Δm 2 (L/E) ~ 1 The atmospheric Δm 2 ~0.001 ev 2 E from a convention beam is ~ 1 GeV So L = 1000 km!!!

15 E.g., LBNE -- starting in 2020 Beam from Fermilab Shoots to detectors in South Dakota 1300 km And there is lots and lots of matter along a 1300 km path! also true for MEMPHYS and HyperK designs

16 And the ground is made of matter (electrons) not antimatter (positrons) Forward scattering affects neutrinos differently than antineutrinos. P osc (ν α ν β ) δ 0 CP parameter π CP CP + matter, This slides the allowed ring off the diagonal This a type of CP violation, but not what we are looking for! P osc (ν α ν β )

17 Worse, we actually don t know which direction CP + matter, Δm 2 <0 P osc (ν α ν β ) δ 0 π CP CP + matter, Δm 2 >0 CP parameter P osc (ν α ν β )

18 Worse, we actually don t know which direction CP + matter, Δm 2 <0 P osc (ν α ν β ) δ 0 π CP All long-baseline experiments will need to introduce a model for matter effects, before they can study CP-violation!!! CP + matter, Δm 2 >0 CP parameter P osc (ν α ν β )

19 Other problems Long Baseline experiments are usually low in antineutrino statistics a combination of style of beam and cross section and the backgrounds are larger compared to signal

20 Expectation for inverted hierarchy:

21 If we know the mass hierarchy, then this is how well LBNE can do in 10 years of running (e.g. without Project X)

22 Decay At rest Experiment for δ cp studies At the Laboratory for Underground Science Part III An unconventional approach to the problem

23 A π + decay at rest beam: p+c Shape driven by nature! Only the normalization varies from beam to beam use νe νe to normalize flux No intrinsic ν e Perfect for a ν µ ν e search

24 The plan: Use ν µ ν e and use the L/E dependence to extract δ in a vacuum We want to see if δ is nonzero } terms depending on mixing angles } terms depending on mass splittings

25 For E=50 MeV L=1.5 km 8 km 20 km km - normal 1.5km - inverted km - normal 8km - inverted km - normal 20km - inverted Nuebar Events Nuebar Events Nuebar Events δ 0 CP δ CP δ CP Close to source Osc mid (π/4) Osc max (π/2) Find δ by comparing the number of events at multiple locations

26 For E=50 MeV L=1.5 km 8 km 20 km km - normal 1.5km - inverted km - normal 8km - inverted km - normal 20km - inverted Nuebar Events Nuebar Events Nuebar Events δ 0 CP δ CP δ CP There will be a degeneracy with hierarchy: We cannot tell the difference between δ=45 and normal hierarchy and δ=135 and inverted DAEδALUS cannot differentiate the hierarchy

27 How do you observe ~50 MeV ν e events? The signal: inverse beta decay, IBD ν e e + p n ν e +p e + + n You need a lot of free protons! Use the same ultra-large water Cerenkov detector system as LBNE at DUSEL! 300 kton fiducial volume

28 We want to observe a 2-fold signature in time The signal: inverse beta decay, IBD ν e e + p n 1st signal 2nd-- from capture We need to reject: ν e e - O F Lower xsec than IBD by 10 because of binding But even if the xsec is small there are a lot of ν e s in the beam!

29 To enhance the signal from n-capture, add gadolinium! thermal neutron capture xsec (barns) Gd Enormous xsec and produces Series1 ~8 MeV upon capture! Z of element Adding Gd is technically difficult But others need it too: Supernova Relic Neutrino Search Non-proliferation studies good progress is being made

30 Energy Dependence of IBD events The signal: inverse beta decay, IBD ν e e + p n ν e +p e + + n Event range is 20 < E ν < 55 MeV 20 MeV 55 MeV

31 Neutrino-electron scattering is also very important! ν e e - ν ν e - ν e e - e - Provides the normalization of the flux since the xsec is known to 1% Mostly from ν e s about 20% from muon flavor

32 We need 3 distances and we cannot have kton detectors! osc max (π/2) at 40 MeV off max (π/4) at 40 MeV Constrains flux 20km 8km 1.5km A multiple-baseline, single-detector experiment H2O w/ Gd An advantage: Nature assures decay-at-rest beams will be identical in flavor and energy

33 1.5 km Accelerator Beam Off Beam Off 100µs 400µs 400µs 100µs 100µs 8 km Accelerators 100µs 400µs 400µs 100µs 100µs 20 km Accelerators 100µs 400µs 100µs 400µs 100µs 20% DF 20% DF 20% DF 20km 8km 1.5km We can know the distance for an event by the timing H2O w/ Gd

34 What proton energy is required? There is a Delta plateau where you can trade energy for current to get the same rate of ν/mw <600 MeV too little π + production Delta Plateau >1500 MeV energy goes into producing other particles besides π + at a significant level

35 Wanted: ~1 MW sources of protons, w/ energy > 600 MeV and <1500 MeV for a reasonable price What isn t needed: 1. No need for a fancy beam structure -- CW is fine. (run 100 µs on and 400 µs off) 2. No need for injection into another accelerator or need to make clean secondary beams. 3. No need for flexibility on beam energy.

36 Wanted: ~1 MW sources of protons, w/ energy > 600 MeV and <1500 MeV for a reasonable price Luckily there are others looking for this too! ADS -- accelerator driven systems for subcritical reactors. Also DTRA -- Defense Threat Reduction Agency We can gain a lot from what is learned in these efforts!

37 Approaches using cyclotrons: The compact cyclotron with self-extraction An H2+ accelerator for ADS applications Under dev. by INFN, Catania The stacked cyclotron: under development for DTRA at MIT 7 cyclotrons in one flux return Under dev. for ADS at TAMU (I will come back to these later in the talk)

38 Strategy: Phase I -- find it In 5 years, a 3σ measurement -- δ =( 90 ± 2 2θ 13 =0.05 The canonical comparison point 3 MW 2 MW 1 MW 20km 8km 1.5km Likely more than one cyclotron at some locations H2O w/ Gd

39 Strategy: Phase II -- measure it In an additional 5 years: LBNE-level sensitivity δ= sin 2 2θ 13 = MW 2 MW 1 MW 20km 8km 1.5km H2O w/ Gd

40 Measurement strategy: Using near accelerator measure absolute flux normalization with ν-e events to ~1%, Also, measure the ν e O event rate. At far and mid accelerator, Compare predicted to measured ν e O event rates to get the relative flux normalizations between 3 accelerators In all three accelerators, given the known flux, fit for the ν µ ν e signal with free parameters: θ 13 and δ

41 Non-beam backgrounds Atmospheric ν µ Invisible muons : ν µ + p µ + + n where µ + is below Cherenkov threshold, stops and decays. ν µ µ + e+ p n Atmospheric ν e IBD events: ν e + p e + + n Measured in beam-off! Diffuse supernova neutrinos ν e e + p n

42 Beam-related Background Intrinsic ν e in beam From π µ events which failed to capture in the beam stop ~ ν e rate (low) Beam ν e in coincidence with random neutron capture signal Estimated to be very small from Super-K rates ν e -Oxygen CC scatters producing an electron+ n signal Subsequent n from nuclear de-excitation should be very small. All fall as 1/r 2 from the 3 accelerators, near accelerator provides a measurement

43 Total Events and background level, for sin 2 2θ 13 = km - normal 1.5km - inverted km - normal 8km - inverted km - normal 20km - inverted Nuebar Events Bkgnd Nuebar Events Bkgnd Nuebar Events Bkgnd δ 0 CP δ CP δ CP For the signal accelerators signal-to-background is excellent!

44 1.5km Daedalus Event Energy Distributions (Signal & Background) (sin 2 2θ 13 = 0.04) MeV beam off beam on 8km Blue: Intrinsic ν e bkgnd Red: Beam off bkgnd Black: δ CP = 0 0 Violet: δ CP = 45 0 Green: δ CP = km MeV MeV

45 Compare signal to-background 8km Blue: Intrinsic ν e bkgnd Red: Beam off bkgnd Black: δ CP = 0 0 Violet: δ CP = 45 0 Green: δ CP = km MeV MeV With LBNE LBNE ν 5yr LBNE ν 5yr

46 Systematic Uncertainties before fits By comparing measurements in the 3 accelerators, several of these systematics effectively cancel.

47 With either hierarchy, we can clearly observe CP violation!

48 How well do we do? By construction our capability is equal to LBNE, But our measurement has completely different issues! Daedalus Phase LBNE 5 yrs nu + 5 yrs nubar

49 Part IV Practicalities

50 Our schedule is tied to when the large H 2 O detector exists

51 Our schedule is not tied to proton economics for outside programs. These are single-purpose machines.

52 Requests for funding comes in reasonable chunks. We could work on smoothing this schedule (install 1 cyclotron every N years?) No phase requires complex construction. No Tunneling! The proposed near hall is based on the MiniBooNE target hall.

53 What about the θ 13 question??? The mid and far sites do not require funding until after the results of the reactor searches are available. Near-site has a broad physics program see Working Groups II and III -- a good investment no matter θ 13 & it lets us learn to run a machine, & understand the real costs, etc.

54 Thoughts about schedule and cost The schedule gives plenty of room for R&D, since the accelerators are small and quick to build. Repeating the same design at different locations during the different phases saves money. Ordinarily stretching a schedule adds cost, but not in the case where market forces are likely to bring the costs down.

55 An example: Nb 3 Sn Number of producers of coil doubled last year. In the next 5-7 years, it is reasonable to expect the price of superconducting coil to drop Driven by industrial applications.

56 Among all of the types of accelerators out there Very interesting R&D ongoing, but these machines are not yet proven Cyclotrons Synchrotrons Linacs FFAGs etc. Can do what we need right now, but are expensive. Use linacs if you want a nice beam for transfer to another line and flexibility on energy (We don t) Why cyclotrons? Inexpensive, Only practical below ~1 GeV (ok for us!) Only good if you don t need timing structure (ok!) Typically single-energy (ok!) Taps into existing industry We do not rule out other options, but cyclotrons seem like a good fit.

57 Today s cyclotron industry is based on: Making isotopes -- low energy (30 MeV), high intensity (few ma) Treating cancers -- higher energy (250 MeV), low intensity! Countries with cyclotron industries: From the latest List of Cyclotrons -- Norway, countries Sweden w/ major initiatives Canada US Belgium, Netherlands, Poland Germany, Switzerland, Austria France, Italy, Hungary, Serbia Ukraine, Russia Japan, China, Taiwan South Africa

58 Why do I say that Cyclotrons are inexpensive? Consider what is on market right now The Monarch250 (250 MeV, 9 T, low intensity, p) is $2M (accel only, gantry, patient table, etc) The cost of a TR30 (30 MeV, 1 T, high intensity, H-) was quoted to us as: A 30 MeV cyclotron (protons) will run $6-7M USD. Beamlines $ K/each, target stations $750K-$1M. Facility costs are extra and can range widely depending on circumstance and location, but will approximate $5-10M USD. treating people making isotopes Costs are dropping as market becomes more competitive.

59 Tomorrow s industry is at high energy & high intensity 1. This is the next step in the arc of the research. 2. There are suddenly market drivers. Energy issues & homeland security

60 The best place to learn about the latest news: This is for cyclotron-physicists what Neutrino 2010 was for us. There are 4 invited talks related to DAEδALUS An overview talk on the experiment by Jose Alonso And 3 talks on our designs*

61 * Truth in advertising Our Designs are really already-existing designs! These are not our invention! They exist because of Homeland Security -- The Compact Cyclotrons ADS -- The H2+ and Stacked Cyclotrons Luckily the ADS machines are a perfect match to us, and the Homeland Security design should be close.

62 Two comments on finding the right people 1) The Cyclotron Physicists come from a different world... Not the usual HEP/NUC suspects (FNAL, BNL, etc.) 2) They specialize, just like us -- like 0νββ is a niche of our field high energy, high intensity cyclotrons are a niche of their field. So once you have found the community, you still need to look further! Our contacts are growing!

63 Who are the accelerator physicists who are authors on the EOI?

64 current needed for 1 MW delivered w/ 20% DF We need high current and high energy

65 What are the issues to achieving this? These are covered in detail in our EOI. The biggest issue by far is space charge effects at injection which lead to difficulties in extraction. Extraction issues are exacerbated by compact size. But compact size leads to lower cost. Advertisement: MIT has an open postdoc position in beams simulation, to work on space charge simulation under guidance of Bill Barletta (location: PSI)

66 Introducing the designs Design #1: A Compact Cyclotron The cheapest by far if it works. This work is supported by DTRA for active interrogation. The research is fully funded, but must pass milestones. The next milestone review is in October, and is being organized by LANL.

67 250 MeV, high intensity, 7 T, small radius (<0.33 m), is under construction at MIT deliver to DTRA in $9M Unit cost once in production ~$3M 800 MeV needs 9T, ~3m radius (size of the above) cost est in production in about 2020 ~$15M, PHASE 0 to II = 10 machines (1,2,7) This new cyclotron will use resonant self-extraction This has worked at low energies, but at MeV is extracted beam under good control? Still River is also using resonant MeV Cheap cyclotron + expensive shielding = no net gain, so we need clean extraction!!!

68 Design #2: The H2+ machine Many cyclotron experts favor this design! This is work by Luciano Calabretta at INFN-Catania This machine gives 1.5 MW in the 20% DF (more power = fewer machines)

69 Elegant features: 1. Injector cyclotron -- today s technology, but running H2+ Get control of the beam at low energy where activation is low! This gives a clean injection into SCR w/ large radius of first bend.

70 Elegant features: 2. Extraction by stripping w/ foils (like H- machines) -- very clean! Some questions were raised about the lifetime of the foils, but experts in this area say this is not a problem.

71 Elegant features: 3. Multiple extraction lines make targeting feasible, given very high power. 2 MW (13 ma) is initial goal for phase I (1,1,1 or 2) We may be able to go to 3 MW or phase II (1,1,3 or 4) very prelim: 50M euros for 1st machine -- less thereafter

72 Design #3: The stacked cyclotron Many PSI beams in one flux return! This has been developed by Peter McIntyre, TAMU In principle we only need 3 machines, each with a different number of stacks Has both ADS and FRIB applications

73 There is a machine which runs at the energy we need right now: PSI But not enough intensity (2 ma). To produce many PSI machines would be cost prohibitive, mostly because of the cost of the iron in the giant machine. This proposal stacks many beams into one magnet. For example a machine that gives 2 MW in the 20% DF, has 7 beams at 2 ma 15 m Each beam hits a separate region on target, making targeting much simpler.

74 Some thoughts on targeting: We plan to use carbon (graphite). As long as we keep the intensity < ~ 1MW we can base our design on existing targets. 1 MW Low A -- minimize energy lost to n production, No stringers! (reduce decay-in-flight)

75 There is NO QUESTION that we need better cost estimates. But we are not far off! We have the great advantage of market forces to help. And we have time to understand these costs. We also suspect that other ideas may be out there. This is why we are so excited about Cyclotrons 2010

76 Summary DAEδALUS is a novel way to approach CP violation It is even more powerful when combined with LBNE! This idea is only a year old this month Our first major meeting was only 6 months ago! It is gaining traction fast In the HEP community and in the Cyclotron community. It builds on assets in both communities! Questions about DAEδALUS as a stand-alone project?

77 Backup

78 Some more examples for various sin 2 2θ 13 and δ Standard DAEδALUS (10 yrs) Short Combined Standard LBNE (10 yrs) Long Combined

79 Total of each type of event: Normalization Off Osc Max Osc Max sin 2 2θ 13 = 0.05 Phase 1 + Phase 2 Running

80

81

82 3 kinds of cyclotrons: 1. Classical operates at a fixed frequency ω 0 =qb/m accelerates protons up to 20 MeV Can continuously pump in beam -- so high intensity. Still used where very compact protable systems are desired. 2. Synchrocylotron (what we propose) lets the RF frequency decrease as energy increases ω RF =ω 0 /γ=qb/γm intrinsically low energy because it accelerates one batch per RF-sweep. Nice if you want low intensity in a compact package 3. Isochronous cyclotron Raise the B field with radius s.t. B=γB 0 So ω RF =qγb 0 /γm is a constant (γ s cancel) Permits high intensity but requires a complicated B field The most widely used and the best understood & simulated

83 Don t medical accelerators cost $15-20M? That s for the whole system for treating patients. The precision gantry & patient tables, in particular, are expensive. Still River Systems

84 modify an off-the-shelf cyclotron for isotope production Parameters for the H2+ machine

85