Results obtained with large mass surface sensitive detectors

1. Results obtained withlarge mass surfacesensitive detectors University of Insubria - Como, Italy INFN - Milano, Italy Zaragoza -November 7, 2005

2. BKG = 0.18±0.01 c/(keV kg y) T1/20n > 2×1024y @ 90% C.L. bb (0n) 130Te Surface background in CUORE Qbb0n= 2528.8 keV CUORE – bb0n of 130Te Cuoricino background analysis and Monte Carlo simulations allow to make predictions on the future background expected for CUORE… From what we know until now, experimental data and simulations suggest that the main contribute expected for CUORE background in the DBD region could be: a and b degraded particles emitted by 238U and 232Th surface contaminations on the Cu frame

3. Surface background reduction There are three possible ways to reduce the background due to surface Cu contaminations: Improving the quality of surface treatment for Cu components Minimizing the amount of Cu surface facing the detectors Creating new types of detectors able to recognize surface events This can be done by shielding a bolometer with layers of different materials. Each layer is an auxiliary bolometer… …thermally coupled to a “classical” bolometer. This detector is a Surface Sensitive Bolometer (SSB)

4. Event originating inside the main bolometer (DBD event) Event originating outside the main bolometer (degraded a) Classic pulse Classic pulse Classic pulse Fast and high pulse Dynamic behaviour of SSBs Particles releasing energy in the detector heat up each element of this composite bolometer: Active shield TeO2 main bolometer The origin of events can be determined. The addition of active shields alters the dynamic behaviour of the bolometer, leading to pulses different in both amplitude and shape. An a-surface event gives distinct pulses on the thermistors of the SSB.

5. Amplitude on active shield sensor [mV] Amplitude on main absorber sensor [mV] Identification of surface events – Amplitude (1) Surface events may be identified comparing pulse amplitudes read simultaneously by the thermistor on the main absorber and on the active shield. The main tool is a graph reporting the pulse height from the shield sensor vs the pulse height from the main sensor. A DBD event occurring inside the absorber gives an equal amplitude pulse on both sensors SCATTER PLOT An a particle coming from outside the detector gives a higher amplitude pulse on the shield sensor than on the main sensor

6. Pros • Low cost • Pros • Thermal contractions with main absorber • Known material • Cons • Low purity Surface events Surface events • Cons • Frailty • Thermal coupling with thermistors • Notes • IET Si slabs (Polish) gave good results • IRST Si slabs (Italian) gave some unexplained results Surface events SSB prototypes @ Como Various materials for active shields were tested on small size detectors (main absorbers 2×2×2 - 2×2×0.5 cm3) Ge shields Si shields TeO2 shields • Pros • Excellent results • High purity • Cons • High cost Parallel readout!

7. Counts Rise time on active shield thermistor [ms] Identification of surface events – Rise time (2) But also pulse shape discrimination Not just pulse amplitude discrimination Two classes of events Surface events correspond to pulses with shorter rise times on the shield sensor Rise time discrimination is a powerful tool to isolate outer events

8. Preparation for the large mass test After testing a great deal of SSB prototypes, a test with CUORE real size crystals was organized. Main absorber TeO2 crystals, size 5×5×5 cm3, m=750 g Initial idea for the active shields: TeO2. However, the producer could not deliver slabs in time for the beginning of assembly. Therefore… Active shields Si slabs produced by IRST, size 5×5×0.3 cm3, resistivity 20kΩ (floating zone crystal with < 1012 atoms/cm3 of P) Moreover, in order to reduce the number of necessary channels… Signal readout Slab sensors connected in parallel configuration Four SSBs with these characteristics were assembled.

9. Test @ LNGS – Single SSB Slabs thermally coupled to absorbers through Ge stand-offs; teflon holders keeping crystals attached to the Cu frame. Two configurations - 3 detectors type A, 1 detector type B Type A Type B Very different from both a thermal and a mechanical point of view

10. Test @ LNGS – Four detectors module The four SSB module (Cu frame and PTFE holders)… …and the whole structure mounted in the cryostat (along with other eight nude crystals for an assembly test).

11. Test @ LNGS - Problems Crystals with unequal heights Only two crystals were firmly held. The fully covered crystal (even if its height was increased by PTFE holding it through the slabs) was not securely fixed. The last one was not held at all. Slabs not fully realiable Si slabs by IRST had previously lead to results whose physical origin has yet to be determined. Difficult assembly Because of very different thermal contractions between Si and TeO2, slabs must be glued only close to the center of the TeO2 crystal face. Many slabs detached during assembly and had to be provisionally attached with vacuum grease. Lost signals During the test, we lost contact with a set of slab sensors in parallel: signals from one SSB could not be acquired. A lot of problems, but in spite of them… The two firmly held crystals worked properly and one of them had readable slabs, from which we gathered data.

12. Test @ LNGS – Scatter plot The scatter plot is coherent with the expected behaviour. However, there are some problems. For example… • There is a considerable contamination (a) in the slabs or close to them. Surface events Mixed events • This is probably a slab with a bad thermal coupling to the absorber. Bulk events • While surface events may be identified, the parallel configuration seems to confuse the physical understanding of each element in the plot.

13. Test @ LNGS – Rise time distribution on shields Again as expected, we can recognize various classes of events from the rise time distribution on the parallel slabs channel. Rise time vs amplitude plot on slab sensors:

14. Surface events & decay time on main crystal Surprise: there is another way to distinguish surface events from bulk events. Decay time – amplitude relation on main absorber sensor Detector with no slabs Detector with slabs

15. Identification of surface events – Decay time (3) Bulk events are well isolated by a cut in the decay time – amplitude plot on the main sensor

16. Shields as pulse shape modifiers Shields act as pulse shape modifiers – behaviour to be investigated in the future. We found a similar feature in the previous tests at Como. Important feature… Is it possible to get rid of the slab sensors? In this way, slabs could be a very promising option for CUORE.

17. Background reduction Cutting on the DT distribution allows to isolate a great deal of unwanted events. Quantitatively – background obtained (a region): E [MeV] 2.9-3.2 3.2-3.4 3.4-3.9 Bkg (no slabs) [c/kg keV y] 0.44 ± 0.06 0.58 ± 0.08 0.51 ± 0.04 Bkg (slabs) [c/kg keV y] 0.18 ± 0.08 0.51 ± 0.16 0.29 ± 0.08 The a region shows a considerable improvement: the obtained value is very similar to the Cuoricino value. Important result: even though the test was not performed in extremely clean conditions, the background is comparable to that obtained with accurately cleaned detectors… Rejection by slabs works

18. Conclusions & future investigations Topics to be cleared by analysis & future tests Not complete efficiency of the SSB rejection (high contamination, not working slabs) Origin of the decay time distribution on the main absorber sensor and reproducibility of surface events identification with this parameter Motivation for the background value (~ 0.2 c/kg keV y): SSBs are proving themselves to be very helpful in understanding the nature of the background. Next test at LNGS Four detectors with TeO2 slabs (problem of thermistor coupling to be solved) Careful cleaning of all the detectors Single readout of the slab sensors of two detectors (instead of parallel) “Passive” slabs for the other two detectors (can we use just the main absorber sensor?)