Paris - 14/11/20051 MIBETA 2 Semiconductor microbolometers MIBETA 2 Semiconductor microbolometers for a direct neutrino mass search Alessandro MONFARDINI – ITC-IRST and INFN Trento On behalf of the MIBETA group: INFN sez. Milano and Università di Milano-Bicocca, Dipartimento di Fisica, ITALY Università dell’Insubria, Dipartimento di Fisica e Matematica, ITALY ITC-IRST and INFN, Trento, ITALY Goddard Space Flight Center, NASA, Greenbelt, Maryland, USA University of Wisconsin, Madison, Wisconsin, USA

Paris - 14/11/20052 Science case: (m  0)  (m = ?) Neutrino oscillations (Δm 2 only): atmospheric atmospheric Δm 23 2  2  eV 2 (SK evidence + CHOOZ constrains) solar solar Δm 12 2  7  eV 2 (SNO + KAMland) Neutrinoless double beta decay - 0  (model-dependent):..but direct insights on the neutrino nature (Majorana ?) and access to Majorana phases Effective Majorana mass: m ee < 0.35 eV (Heidelberg-Moscow 76 Ge) m ee < 0.2 ÷ 1.1 eV (CUORICINO 130 Te) m ee = 0.1 ÷ 0.9 eV (Klapdor: 76 Ge reanalysis) m ee = 0.1 ÷ 0.9 eV (Klapdor: 76 Ge reanalysis) Cosmology (indirect): U. Seljak, Physics Review D 71 (2005)  m i < 0.42 eV (CMB+SDSS+SN) Direct (  decay) SAFE determination NEEDED !! … deserves more than just KATRIN (to start data taking in 2010) … deserves more than just KATRIN (to start data taking in 2010)

Paris - 14/11/20053 Calorimetric technique: status Lowest-Q (2.5keV) beta decay (most sensitive to small m ): 187 Re GOALS:- eliminate as much systematics as possible (sub-eV!!) GOALS: - eliminate as much systematics as possible (sub-eV!!) - scaling (in principle) possible up to the n th generation  Source  Detector (neutrino is the only allowed to escape from the bulk)  Published results:< 15 eV (90% C.L.) Milano MIBETA (AgReO 4 ) < 26 eV (95% C.L.) Genova MANU (metallic Re) STATUS:at present about one order of magnitude worse than spectrometers (best published result: 95% C.L.). STATUS: at present about one order of magnitude worse than spectrometers (best published result: 95% C.L.). Some pros in principle: micro vs. macro approach (scalability) and systematic effects (but to be studied carefully..).

Paris - 14/11/20054 MIBETA2 and MARE An ambitious experiment has been proposed by an international collaboration to directly compete with KATRIN using Cryodetectors (Ref. D. Pergolesi presentation). MARE (Microcalorimeters Arrays for a Rhenium Experiment) will be a two phases effort. MIBETA2 is part of MARE phase I MIBETA2 is part of MARE phase I MARE Phase I: - Present technology detectors ( ) - Scaling up to hundreds of devices  MIBETA2Semiconductor bolometers  MANU2TES Scientific Goal: - m < 2eV before KATRIN - phase II preliminary (systematics, technology..)

Paris - 14/11/20055 MIBETA 2 brief description STARTING POINT: MIBETA GOAL: significantly increase the statistics the statistics SOLUTION: scaling up to 200 detectors the MIBETA concept ABSORBER: AgReO 4 THERMISTOR: semiconductor (Si or Ge) (Si or Ge) Single absorber mass  500  g  ~ beta decays required   E and f pup achievable f pup ~ A   R

Paris - 14/11/20056 MARE phase I: MIBETA2 options ITC-IRST TMAH micromachined arrays with SU8 supports for the absorbers. Implanted silicon with the technology developed for the MIBETA single devices. STATUS: 10 devices arrays fabrication ongoing. IRST process BL12 NASA 6  6 silicon array (XRS2). STATUS: encouraging first results with 450  g AgReO 4. Coupling and electronics to be optimized. NTD Ge array (LBL + Bonn). STATUS: preferred for the larger e-ph thermal coupling. Reproducibility to be demonstrated.

Paris - 14/11/20057 NASA array results - 6  6 array (XRS2) - T 0  7 K - AgReO 4 (450  g) either mounted on Si platforms glued to the four SU8 spacers or directly to the thermistor (with the help of a silicon spacer) Best results obtained with the spacers: -  E FWHM 1.5keV) = 35eV -  RISE = 220  s Problems: - MIBETA electronics not well matched yet - coupling of AgReO 4 to be optimised. But…… good baseline choice in any case.

Paris - 14/11/20058 Germanium NTD results - 37 devices array originally built for astronomical purposes - Ge NTD bump bonded on SiN thin (0.8  m) membranes - Nb wiring to the pads - AgReO 4 (450  g) glued on the NTD (ST2850FT) Among 10 actually bonded NTDs, the best result is summarized here: -  E FWHM 1.5keV) = 50eV -  RISE = 270  s Problems: - NTD mechanical stress

Paris - 14/11/20059 Systematic effects for MIBETA2 Under investigation using “old” MIBETA data: - theoretical spectral shape of the 1 st forbidden 187 Re decay; - solid state BEFS effect; - internal detector response function calibration; - unidentified pile-up spectrum; - external radioactive background; - energy scale calibration; - surface electron escape; - data reduction. Briefly discussed here

Paris - 14/11/ BEFS and a Nuclear Physics result A recent re-analysis of the MIBETA fit residuals with the state-of-the-art EXAFS software (GNXAS) led to a substantial improvement of our understanding of the effect. The important result here is: F (l e =1) = 0.84  0.30 Fraction of electrons emitted with l=1 (p-wave electrons). To satisfy the overall angular momentum conservation the antineutrino is mainly emitted with null orbital momentum. Now extrapolation to the end-point is much safer, and we the effect is negligible for MARE phase I. know that the effect is negligible for MARE phase I. Crucial for MARE Phase II, on the other hand. Re-submitted to PRL

Paris - 14/11/ Ti gamma: “deeply excited” Re A 6keV x-ray photon can only penetrate for about 4  m in an AgReO 4 target, path to be compared with 300  m, linear dimensions of a typical MIBETA crystal. QUESTION: QUESTION: are the asymmetric calibration line profiles due to escape effects ? DEDICATED RUN: DEDICATED RUN: a 44 Ti gamma source (E  =78.4keV) has been used to uniformly excite Re K,L escape peaks throughout the absorbers. Preliminary result: Preliminary result: the symmetry seems recovered…but the large intrinsic width of the escape peaks (42eV) is annoying. Paper in preparation External 55 Fe photons impinging onto the crystals Internal Re Escape peak (K  2 ) determined by an external 44 Ti source

Paris - 14/11/ ITC-IRST capabilities applied to Cryogenics Detectors: - “usual” silicon technologies (lithography down to 2  m) for the thermistor implant and heater - surface micromachining for thermistor-to-absorber coupling - Bulk micromachining (chemical) for the realization of the suspended structure - Experience with thin membranes and films for other detectors concepts (e.g. MKID, IR and THz detectors) ITC-IRST technology for Cryodetectors Old single devices are shown here. New run: - 10 detectors modular arrays; - absorbers SU8 supports; - heaters for calibration.

Paris - 14/11/ IRST silicon arrays fabrication steps Present status: in production (multiple implantations for main doping and compensation ongoing)

Paris - 14/11/ Wiring and “old” ITC-IRST Cryoflats COLD END WIRING General problem: COLD END WIRING 1)detector  “cold” electronics plate 20mK  4K 2)plate  (e.g. JFET input) 4K  120K below In order to keep the power transmitted per wire below 10nW 10nW for usable wire lengths and sections the first Titanium connection has to be realised in Titanium. First RUN problems: First RUN problems:- single front mask - aluminium - cleaving - bonding - weak suspension points Achievement: Achievement:- Micromachining process tuning

Paris - 14/11/ New IRST cryoflats for MIBETA PROCESS SCHEME SINGLE MEANDERS LAYOUT Old problems: hopefully all understood and solved with the new layout/process Present Status: Ready for fabrication

Paris - 14/11/ Conclusions MIBETA2 is now part of the two steps experiment MARE MIBETA2 is now part of the two steps experiment MARE MIBETA2 is intended to reach a 2  4 eV limit, depending on the actual MIBETA2 is intended to reach a 2  4 eV limit, depending on the actual performances of the detectors, the number of available channels and so on A number of systematic effects are under investigation using MIBETA A number of systematic effects are under investigation using MIBETA data (e.g. BEFS, line profile) We do have a baseline design with acceptable performances We do have a baseline design with acceptable performances We are waiting for the first results from the new ITC-IRST run to improve We are waiting for the first results from the new ITC-IRST run to improve the baseline expected results.