1/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 MARE Microcalorimenter Arrays for a Rhenium Experiment Proposal of a next generation neutrino mass experiment based on calorimetric study of the 187 Re beta spectrum Samuele Sangiorgio on behalf of the MARE collaboration University of Insubria Como - Italy INFN Milano - Italy

2/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/ eV 0.5 eV 2.2 eV 0.1 eV 0.05 eV 0.2 eV Present sensitivity Future sensitivity (a few year scale) Cosmology (CMB + LSS) Neutrinoless Double Beta Decay Single Beta Decay Tools Model dependent Direct determination Tools for the investigation of the mass scale

3/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 Complementarity of cosmology, single and double β decay Cosmology, single and double  decay measure different combinations of the neutrino mass eigenvalues, constraining the neutrino mass scale In a standard three active neutrino scenario:  Mi Mi i=1 3  cosmology simple sum pure kinematical effect  M i 2 |U ei | 2 i=1 3 1/2  M    beta decay incoherent sum real neutrino  M i |U ei | 2 e i   i i=1 3  M    double beta decay coherent sum virtual neutrino Majorana phases

4/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 Neutrinoless Double Beta Decay  it works only if neutrino is a Majorana particle (  c )  uncertainties from nuclear physics  other mechanisms (not only massive neutrinos) can mediate the process necessity of direct measurement and cross checks at this scale Cosmology (Cosmic Microwave Background + Large Scale Structure) very sensitive, but considerable spread in recently published results  parameter degeneracy  dependence on priors cosmological parameters  sensitivity to even small changes of input data Model dependent tools

5/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 finite neutrino mass only a small spectral region very close to Q is affected electron kinetic energy distribution (A,Z)  (A,Z+1) + e - + e Q = M at (A,Z) – M at (A,Z+1)  E e + E Single Beta Decay processes involving neutrinos in the final state E 2 = M 2 c 4 + p 2 c 2 basic idea: use only kinematics Model independent tool: the kinematics of β decay phase space term coulombian correction term form factor contains NME radiative electromagnetic correction

6/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 The count fraction laying in this range is  (M  Q) 3 low Q are preferred The modified part of the beta spectrum is over range of the order of [Q – M c 2, Q] E – Q [eV] Tritium as an example Effects of a finite neutrino mass on the β decay

7/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 Effects of a finite neutrino mass on the Kurie plot The Kurie plot K(E e ) is a convenient linearization of the beta spectrum Q Q–M c 2 Q K(E) zero neutrino massfinite neutrino masseffect of:  background  energy resolution  excited final states  Q-  E Q (dN/dE) dE  2(  E/Q) 3 fraction of decays below endpoint

8/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 Experimental searches based on nuclear beta decay Requests :  high energy resolution  a tiny spectral distortion must be observed  high statistics in a very narrow region of the beta spectrum  well known response of the detector  control of any systematic effect that could distort the spectral shape Approximate approach to evaluate sensitivity to neutrino mass  M  Require that the deficit of counts close to the end point due to neutrino mass be equal to the Poissonian fluctuation of number of counts in the massless spectrum  M  1.6 Q 3  E A T M  4 energy resolution total source activity live time

9/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 Two complementary experimental approaches  determine all the “visible” energy of the decay with a high resolution low energy “nuclear” detector  cryogenic microcalorimeters  present achieved sensitivity:  15 eV  future planned sensitivity: see later  measurement of the neutrino energy source  detector ( calorimetric approach ) (the source is 187 Re - Q=2.5 keV)   determine electron energy by means of a selection on the beta electrons operated by proper electric and magnetic fields  measurement of the electron energy out of the source  present achieved sensitivity:  2 eV  future planned sensitivity:  0.2 eV  magnetic and electrostatic spectrometers source separate from detector (the source is T - Q=18.6 keV)  completely different systematic uncertainties

10/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 The calorimetric approach to the measurement of the mass 187 Re  187 Os + e - + e 5/2 +  1/2 – unique first forbidden (computable S(E e )) Advantages of calorimetry  no backscattering  no energy loss in the source  no excited final state problem  no solid state excitation Drawbacks of calorimetry  systematic induced by pile-up effects  energy dependent background Calorimeters measure the entire spectrum at once  use low Q beta decaying isotopes to achieve enough statistic close to Q  best choice: 187 Re – Q = 2.47 keV but T 1/2 = 43.2 Gy vs. 3x for T beta spectrum event frac. in the last 10 eV: 1.3x10 -7

11/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 Bolometric detectors of particles: basic concepts Energy absorber crystal containing Re M ~ 0.25 mg basic parameter: C Thermal coupling read-out wires in the future,  machined legs basic parameter: G G  0.02 pW / mK  Temperature signal:  T = E/C  1 mK for E = 2.5 keV  Bias: I  0.5 nA  Joule power  0.4 pW  Temperature rise  20 mK  Voltage signal:  V = I  dR/dT   T   V  30  V for E = 2.5 keV  Noise over signal bandwidth (  1 kHz): V rms = 0.2  V  Signal recovery time:  = C/G  20 ms Advantages over conventional techniques:  high energy resolution  wide choice for absorber material Thermometer Si-implanted thermistor basic parameters: R  1.5 M   100 mK dR/dT  50 k  mK Variable Range Hopping conduction regime exponential increase of R with decreasing T Heat sink T ~ 80 mK dilution refrigerator MicroBolometer:

12/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 Calorimetric tecniques: status MIBETA Milano/Como (AgReO4)  M β  < 15.6 eV (90% C.L.) MANU Genova (Metallic Re)  M β  < 26 eV (95% C.L.) Presently, two groups are dealing with Rhenium single beta decay measurements by means of microbolometers:

13/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 MIBETA (Milano/Como) experiment: the detectors Energy absorbers  AgReO 4 single crystals  187 Re activity  0.54 Hz/mg  M  0.25 mg  A  0.13 mHz Thermistors  Si-implanted thermistors  high sensitivity  many parameters to play with  high reproducibility  array  possibility of  -machining typically, array of 10 detectors lower pile up & higher statistics ~ 1 mm single detector

14/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 MIBETA experiment: the Kurie - plot E (keV) K(E) total Kurie – plot 5 x Re decays above 700 eV indipendent from fit interval » good fit func from statistical distribution of time interval between two consecutive events

15/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 MIBETA experiment: the neutrino mass  M   2 = -141  211 stat  90 sys eV 2 (preliminary)  M     15.6 eV (90% c.l.) Fit parameters single gaussian:  E FWHM = 27.8 eV fitting interval: 0.8 – 3.5 keV free constant background: 6 x c/keV/h free pile-up fraction: 1.7 x 10 -4

16/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 spectrum detector scheme Similar technique as MIBETA  One detector only  Metallic Rhenium (1.5 mg)   E FWHM = 96 eV  Q = 2470  1  4 eV   ½ = 41.2  0.02  0.11 Gy  M   < 26 eV (95 % c.l.) MANU experiment (Genova)

17/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 The Future still open! To reach a sub-eV sensitivity on mass (like Katrin) we have to improve our sensitivity by two orders of magnitude too challenging task for a single step! two stages effort

18/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 simulations I phase GOAL:  reach 2 eV sensitivity  improve understanding on systematics  R&D for phase II Time Schedule: MARE - phase I Present technology detectors Single channel optimization Scaling up to hundreds devices Theoretical spectral shape of decay Solid state BEFS effect Detector response function Unidentified pile-up Data reduction …

19/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 MARE - phase I MIBETA 2 MANU 2 Transition Edge Sensor (TES) instead of NTD thermistors » faster risetime and better S/N ITC-irst micromachined array. Implanted silicon with the technology developed for the MIBETA single devices. Status: ongoing production & tests NASA 6x6 silicon array. Status: encouraging first results. Coupling and electronics to be optimized. LBL+Bonn NTD Ge array. Status: excess noise observed; reproducibility to be demostrated.

20/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 MARE – phase II the full MARE phase I dataset is required to drawn a definitive conclusion. The kick-off of the phase II will be subordinated to: safe reduction of all the known sources of systematic uncertainties; verification that no new sources come up to impair the sensitivity; understanding of the 187 Re decay spectrum with the required precision; demonstration that the estimated sensitivity can be maintained though the experiment is segmented in a large number of channels GOAL:  reach 0.2 eV sensitivity Time Schedule: Need substancial improvements :  sensors: TES or MMC  electronics: multiplexed SQUID  methods: modularity technology already under study in several other experiments

21/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 MARE – phase II simulations II phase  E(eV),  R (  s), A(Hz) channels in 5 y Approach: design a kind of modular pixel array kit which can be relatively easily installed in any available refrigerator

22/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 Conclusions and summary DRAFT PROPOSAL  Importance of another β decay experiment complementary to KATRIN  Calorimetric approach looks very promising  Joining together the know-how and technology of the involved groups MARE phase I could be implemented with a relatively straightforward optimization and scaling of the current MIBETA and MANU detectors  MARE phase II is challenging but there’s no real technological limit » large margin of improvements exists né dolcezza di figlio, né la pieta del vecchio padre, né 'l debito amore lo qual dovea Penelopè far lieta, vincer potero dentro a me l'ardore ch'i' ebbi a divenir del mondo esperto e de li vizi umani e del valore; ma misi me per l'alto MARE aperto [Dante, Inferno, XXVI]

23/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005

24/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 Mass hierarchy In case of mass hierarchy:  the Kurie plot  superposition of three different sub - Kurie plots  each sub - Kurie plot corresponds to one of the three different mass eigenvalues The weight of each sub – Kurie plot will be given by |U ej | 2, where | e  =  U ei | Mi  i=1 3 Q – M 3 Q – M 2 Q – M 1 Q E e K(E e ) This detailed structure will not be resolved with present and planned experimental sensitivities (~ 0.2 eV) K(E e ) EeEe

25/22 Samuele Sangiorgio, Universita’ dell’Insubria, Como – INFN MilanoInternational School of Nuclear Physics – ERICE – 23/09/2005 Mass degeneracy Q – M  K(E e ) Q E e If the 3 mass components cannot be resolved or degeneracy holds: the Kurie-plot can be described in terms of a single mass parameter, a mean value of the three mass eigenstates M=M=  M i 2 |U ei | 2 1/2