- Email: [email protected]
Status of the CALDER project: Cryogenic light detectors for background suppression
Nuclear Inst. and Methods in Physics Research, A (
)
–
Contents lists available at ScienceDirect
Nuclear Inst. and Methods in Physics Research, A journal homepage: www.elsevier.com/locate/nima
Status of the CALDER project: Cryogenic light detectors for background suppression N. Casali a ,∗, F. Bellini b , M. Calvo c , L. Cardani a , M.G. Castellano d , C. Cosmelli b , A. Cruciani a , S. Di Domizio e,f , P. Fresch a , J. Goupy g , M. Martinez h , A. Monfardini c , G. Pettinari d , H. le Sueur g , M. Vignati a a
INFN Sezione di Roma, Piazzale Aldo Moro 2, I-00185, Roma, Italy Dipartimento di Fisica Sapienza Università di Roma, Piazzale Aldo Moro 2, I-00185, Roma, Italy c Institut Neel, CNRS/UJF, 25 rue des Martyrs, BP 166, F-38042 Grenoble, France d Istituto di Fotonica e Nanotecnologie CNR, Via Cineto Romano 42, I-00156, Roma, Italy e Dipartimento di Fisica Università degli Studi di Genova, Via Dodecaneso 33, I-16146, Genova, Italy f INFN Sezione di Genova, Via Dodecaneso 33, I-16146, Genova, Italy g CSNSM, Univ. Paris-Sud, CNRS/IN2P3, Université Paris-Saclay, F-91405 Orsay, France h Fundacion ARAID and U. Zaragoza, C/ Pedro Cerbuna 12, 50009 Zaragoza, Spain b
ARTICLE
INFO
Keywords: Kinetic inductance detectors Phonon-mediated light detectors Rare events searches Superconducting device
ABSTRACT The development of large area cryogenic light detectors is one of the priorities of next generation bolometric experiments searching for neutrinoless double beta decay. The simultaneous read-out of the heat and light signals enables particle identification, provided that the energy resolution and the light collection are sufficiently high. CALDER (Cryogenic wide-Area Light Detectors with Excellent Resolution) is developing phonon-mediated silicon light detectors using KIDs, with the goal of sensing an area of 5 × 5 cm2 with a resolution of 20 eV RMS. We present the latest results obtained with aluminum chips and with newly developed multi-layer titanium–aluminum chips featuring a remarkable sensitivity.
Contents 1. 2. 3. 4. 5. 6.
Introduction ....................................................................................................................................................................................................... First phase: Al resonator optimization ................................................................................................................................................................... Second phase: AlTiAl resonator optimization.......................................................................................................................................................... Third phase: 5 × 5 cm2 substrate.......................................................................................................................................................................... The substrate choice ............................................................................................................................................................................................ Conclusion and perspectives ................................................................................................................................................................................. Acknowledgments ............................................................................................................................................................................................... References..........................................................................................................................................................................................................
1. Introduction The CUORE experiment [1] is searching for the neutrino-less double beta decay (0𝜈DBD) of 130 Te operating 988 cubic crystals (edge = 5 cm) of TeO2 as cryogenic calorimeters. The signal produced by this reaction consists of two electrons with a total kinetic energy of about 2.5 MeV. Unfortunately the 𝛼 background events, produced by the radioactive contaminations located on the surfaces of the detector components limit the sensitivity of the experiment. The CUPID (CUORE Upgrade with
1 2 2 2 3 3 3 3
Particle IDentification) interest group [2] aims to develop a CUOREsized bolometric detector able to reach a sensitivity 50–100 times better than CUORE. This ambitious goal can be reached by increasing the source mass and reducing the background in the region of interest. To increase the number of 0𝜈DBD emitters, crystals grown with enriched material are needed. The background suppression can be achieved by discriminating 𝛽∕𝛾 against 𝛼 events by means of the Cherenkov radiation [3,4], produced in the TeO2 crystal only by electrons. The expected Cherenkov light signal detectable at the 0𝜈DBD of 130 Te is about 100 eV. To detect such a small light output, the light detector must
∗ Corresponding author. E-mail address: [email protected] (N. Casali).
https://doi.org/10.1016/j.nima.2018.10.079 Received 30 June 2018; Received in revised form 9 October 2018; Accepted 11 October 2018 Available online xxxx 0168-9002/© 2018 Elsevier B.V. All rights reserved.
Please cite this article in press as: N. Casali, et al., Status of the CALDER project: Cryogenic light detectors for background suppression, Nuclear Inst. and Methods in Physics Research, A (2018), https://doi.org/10.1016/j.nima.2018.10.079.
N. Casali et al.
Nuclear Inst. and Methods in Physics Research, A (
satisfy very stringent requirements; the most important one is the energy resolution, that must be lower that 20 eV RMS in order to reject event by event the 𝛼 background. CALDER [5] will take advantage from the superb energy resolution and natural multiplexed read-out [6] provided by Kinetic Inductance Detectors (KIDs). These sensors, that have been successfully applied in astro-physics searches, are limited only by their poor active surface, of a few mm2 . For this reason, we are exploiting the phonon-mediated approach, proposed by Swenson et al. [7] and Moore et al. [8]: the KIDs are deposited on an insulating substrate featuring a surface of several cm2 . Photons emitted by the bolometer interact in the substrate and produce phonons, which can travel until they are absorbed by a KID. The first phase of the CALDER project was devoted to the optimization of the KIDs design, and to the understanding/suppression of the noise sources. For this phase we chose a 2 × 2 cm2 Si substrate 300 μm thick, and a well-known material for KIDs application, aluminum, which according to our detector model would have provided a noise resolution of about 80 eV RMS. In the second phase we investigated more sensitive superconducting materials like AlTiAl multi-layer which allowed to reach an energy resolution of 25 eV RMS. In the last phase, the optimized light detectors will be scaled to the final detector size (5 × 5 cm2 ) and will be coupled to an array of TeO2 bolometers to prove the potential of this technology. In this paper we show the status of the CALDER project and discuss the results obtained from the first two phases.
)
–
Fig. 1. Design of a CALDER sensor. The inductor (30 strips of 62.5 μm×2 mm) features an active area of 3.75 mm2 (4.0 mm2 including the active region that connects the inductor to the capacitor). To contain the geometrical inductance, we used a gap of 5 μm between the meanders and we closed the circuit using a capacitor made by only 2 fingers.
2. First phase: Al resonator optimization The detector shown in Fig. 1 is the result of the design optimization performed in the first phase of the project. We used a 60 nm thick Al film, as this thickness is a good compromise between detector sensitivity and quality of the superconductor. The active area of the KID (4.0 mm2 ) was enlarged with respect to the first prototypes [9] in order to increase the fraction of phonons that can be collected before being lost in the substrate as described in Ref. [10]. The detector was cooled below the critical temperature using a 3 He/4 He dilution refrigerator with base temperature of about 10 mK. The output signal was fed into a CITLF4 SiGe low noise amplifier [11], which was thermally anchored to the 4 K plate of the cryostat. The rest of the electronics was located at room temperature and its features, together with the description of the cryogenic facility and the acquisition software, can be found in Refs. [5]. The detector response to energy deposition on the Si substrate was evaluated exploiting a 400 nm LED light source located at room temperature and an optical fiber that drives the light signals down to the detector plate at 10 mK. The energy of this light pulses can be tuned from few hundreds of eV up to hundreds of keV. Furthermore, the detector was permanently exposed to X-rays from 55 Fe to cross check the energy calibration performed with optical pulses. The detector was operated in the most sensitive point, where the signal-to-noise ratio (S/N) in the phase direction is maximum (see Ref. [12]). The noise energy resolution obtained was 80 eV RMS with a phonon collection efficiency of about 10%. More details about the performance of the phase-1 prototypes can be found in Ref. [9,13].
Fig. 2. Sketch of the KID deposited on the Si substrate (lateral view). The KID is made by a multi-layer of Al+Ti+Al. The dimensions of the Si substrate are not in scale.
titanium nitride (TiN), or composite superconductors such as Ti+TiN or Ti+Al. We performed several tests on TiN substrates without obtaining satisfactory results: the reproducibility of the superconductor substrate as well as the performances of the resonators were not found to be sufficient. On the contrary, the first test on Ti+Al multi-layer showed very encouraging results. We started with a bilayer made of Ti+Al, with the Ti layer in contact with the Si substrate. Even if the detector performances were competitive with the Al film ones, we measured a decreasing in the phonons collection efficiency, that could be ascribed to a mismatching phonon transmission between Si and Ti. Since the Ti film cannot be exposed to the air to avoid its oxidation, we moved to a trilayer made of Al+Ti+Al: we used the Al layer as interface between Ti layer and Si substrate to increase the phonon absorption probability, followed by the Ti layer in order to decrease the 𝑇𝐶 and increase 𝐿𝑘 and finally, the last Al layer to prevent the oxidation of Ti. We tested a few resonators made with different thickness of both Al and Ti layers. The best results was obtained with the pattern shown in Fig. 2 and using the detector design optimized in the phase one (Fig. 1). We obtained a phonon collection efficiency comparable we the one measured on Al resonator while the noise energy resolution increase up to 25 eV RMS. More details about the AlTiAl substrate performances can be found in Ref. [14].
3. Second phase: AlTiAl resonator optimization The second phase of the project aimed at improving the noise energy resolution down to ∼ 20 eV RMS. To this purpose, we moved to the production of chips based on other, more sensitive superconductors. The resolution (𝛿𝐸), indeed, scales as: 𝑇 𝛿𝐸 ∝ √ 𝐶 𝜖 𝑄𝐿𝑘
(1) 4. Third phase: 5 × 5 cm𝟐 substrate
where 𝑇𝐶 in the critical temperature of the superconductor, 𝑄 the total quality factor of the resonator, 𝐿𝑘 the kinetic inductance fraction and 𝜖 the phonon collection efficiency. Therefore, superconductors with lower critical temperature and higher inductance would allow to further enhance the sensitivity. Possible candidates are sub-stoichiometric
With the results described in Refs. [13,14] we accomplished the first and second phase of the CALDER project. Then we started the final project phase increasing the surface of the substrate up to the final size of 5 × 5 cm2 . Fig. 3 shows the first prototype that we developed. 2
Please cite this article in press as: N. Casali, et al., Status of the CALDER project: Cryogenic light detectors for background suppression, Nuclear Inst. and Methods in Physics Research, A (2018), https://doi.org/10.1016/j.nima.2018.10.079.
N. Casali et al.
Nuclear Inst. and Methods in Physics Research, A (
)
–
Fig. 4. Noise power spectra measured for the same Al resonator evaporated on Si (left) and SOS (right) substrates with pulse tube off (red line) and on (black line). The noise increase at low frequency caused by the vibrations induced by the pulse tube (black line) is visible only for the Si substrate . (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
model suggests that we can reach this results sampling the 5 × 5 cm2 substrate with four AlTiAl resonators like the one described in Section 3. Furthermore, we discovered that detectors based on sapphire substrate (like SOS or GeOS) are not sensitive to the mechanical vibration induced, for example, by the pulse tube. This result is of the utmost importance for the future application of this type of light detectors in the CUPID experiment, that will be host in the same dry cryostats used now by the CUORE experiment.
Fig. 3. First prototype of 5 × 5 cm2 Si substrate 300 μm thick sampled with one Al KID.
In the first test we used just one resonator made of aluminum to study the detector response in the absence of cross-talk or competition among pixels in the absorption of the propagating phonons. The tests on the 5 × 5 cm2 prototypes are ongoing and the results obtained are still very preliminary, therefore they will not be presented in this work. Nevertheless, exploiting our phonons propagation and absorption model [15] we estimated that sampling the 5 × 5 cm2 substrate with four resonators as the one measured in Ref. [14] we will be able to achieve our final resolution target of 20 eV RMS.
Acknowledgments This work was supported by the European Research Council (FP7/2007-2013) under Contract No. CALDER No. 335359 and by the Italian Ministry of Research, Italy under the FIRB Contract No. RBFR1269SL. The authors thanks the personnel of INFN Sezione di Roma for the technical support, in particular M. Iannone, F. Pellegrino, L. Recchia and D. Ruggeri.
5. The substrate choice During the aforementioned R&D we performed several tests also on the detectors substrate: in parallel with silicon we studied also the silicon on sapphire (SOS) and the germanium on sapphire (GeOS) substrates. Both the substrate were 2 × 2 cm2 and 300 μm thick sapphire with a deposition respectively of Si or Ge to increase the optical photons absorption probability. The Al KID was evaporated on the opposite face with respect to the one with the Si or Ge deposition. We noticed a fundamental feature of the KID evaporated on the SOS or GeOS substrate: they were barely affected by the noise induced by mechanical vibrations. Indeed almost all the modern dilution refrigerators pre-cool the 3 He/4 He mixture with a cryocooler in place of liquid nitrogen, liquid helium, and a 1 K bath. No external supply of cryogenic liquids is needed in these ‘‘dry cryostats’’ and the operations can be highly automated. This is the case for the CUORE dilution refrigerator [16]. However, dry cryostats are subject to mechanical vibrations, such as those produced by pulse tube refrigerators, that can spoil the detectors performances. We observed this effect on KIDs evaporated on Si substrate as shown in Fig. 4, where an increase of the detector noise at low frequency is visible when the pulse tube is operating. The increase of the noise at low frequency spoiled the energy resolution of the detector evaporated on Si substrates. On the contrary, with the SOS substrate we obtained a comparable energy resolution despite the pulse tube operation. Even if this effect was measured in all the sapphire substrates that we characterized, a clear explanation of this experimental evidence has not been yet formulated.
References [1] D.R. Artusa, et al., Searching for neutrinoless double-beta decay of 130 Te with CUORE, Adv. High Energy Phys. 2015 (2015) 879871. [2] D.R. Artusa, et al., [CUORE Collaboration], Exploring the neutrinoless double beta decay in the inverted neutrino hierarchy with bolometric detectors, Eur. Phys. J. C 74 (2014) 3096. [3] N. Casali, Model for the Cherenkov light emission of TeO2 cryogenic calorimeters, Astropart. Phys. 91 (2017) 44. [4] N. Casali, et al., TeO2 bolometers with Cherenkov signal tagging: towards nextgeneration neutrinoless double beta decay experiments, Eur. Phys. J. C 75 (1) (2015) 12. [5] E. Battistelli, et al., CALDER: Neutrinoless double-beta decay identification in TeO2 bolometers with kinetic inductance detectors, Eur. Phys. J. C 75 (8) (2015) 353. [6] P.K. Day, et al., A broadband superconducting detector suitable for use in large arrays, Nature 425 (6960) (2003) 817–821. [7] L.J. Swenson, et al., High-speed phonon imaging using frequency-multiplexed kinetic inductance detectors, J. Appl. Phys. 96 (2010) 263511. [8] D.C. Moore, et al., Position and energy-resolved particle detection using phononmediated microwave kinetic inductance detectors, J. Appl. Phys. 100 (2012) 232601. [9] L. Cardani, et al., Energy resolution and efficiency of phonon-mediated Kinetic Inductance Detectors for light detection, Appl. Phys. Lett. 107 (2015) 093508. [10] L. Cardani, et al., New application of superconductors: High sensitivity cryogenic light detectors, Nucl. Instrum. Methods 845 (2017) 338–341. [11] http://radiometer.caltech.edu/datasheets/amplifiers/CITLF4.pdf. [12] N. Casali, et al., Characterization of the KID-based light detectors of CALDER, J. Low Temp. Phys. 184 (1–2) (2015) 142–147. [13] F. Bellini, et al., High sensitivity phonon-mediated kinetic inductance detector with combined amplitude and phase read-out, Appl. Phys. Lett. 110 (2017) 033504. [14] L. Cardani, et al., Al/Ti/Al phonon-mediated KIDs for UV-vis light detection over large areas, Supercond. Sci. Technol. 31 (2018) 075002. [15] M. Martinez, L. Cardani, N. Casali, A. Cruciani, G. Pettinari, M. Vignati, Measurements and simulations of athermal phonon transmission from silicon absorbers to aluminium sensors, arXiv:1805.02495, 2018, submitted for publication. [16] C. Ligi, et al., The CUORE cryostat: A 1-Ton scale setup for bolometric detectors, J. Low. Temp. Phys. 184 (2016) 590.
6. Conclusion and perspectives The CALDER project successfully completed the second phase reaching a noise energy resolution of 25 eV RMS through the operation of KIDs based on sensitive superconductor (such as AlTiAl multi-layer). The next step of the project is to scale up the dimensions of the detector substrate to 5 × 5 cm2 conserving the ∼20 eV RMS noise resolution. Our detector 3
Please cite this article in press as: N. Casali, et al., Status of the CALDER project: Cryogenic light detectors for background suppression, Nuclear Inst. and Methods in Physics Research, A (2018), https://doi.org/10.1016/j.nima.2018.10.079.
)
–
Contents lists available at ScienceDirect
Nuclear Inst. and Methods in Physics Research, A journal homepage: www.elsevier.com/locate/nima
Status of the CALDER project: Cryogenic light detectors for background suppression N. Casali a ,∗, F. Bellini b , M. Calvo c , L. Cardani a , M.G. Castellano d , C. Cosmelli b , A. Cruciani a , S. Di Domizio e,f , P. Fresch a , J. Goupy g , M. Martinez h , A. Monfardini c , G. Pettinari d , H. le Sueur g , M. Vignati a a
INFN Sezione di Roma, Piazzale Aldo Moro 2, I-00185, Roma, Italy Dipartimento di Fisica Sapienza Università di Roma, Piazzale Aldo Moro 2, I-00185, Roma, Italy c Institut Neel, CNRS/UJF, 25 rue des Martyrs, BP 166, F-38042 Grenoble, France d Istituto di Fotonica e Nanotecnologie CNR, Via Cineto Romano 42, I-00156, Roma, Italy e Dipartimento di Fisica Università degli Studi di Genova, Via Dodecaneso 33, I-16146, Genova, Italy f INFN Sezione di Genova, Via Dodecaneso 33, I-16146, Genova, Italy g CSNSM, Univ. Paris-Sud, CNRS/IN2P3, Université Paris-Saclay, F-91405 Orsay, France h Fundacion ARAID and U. Zaragoza, C/ Pedro Cerbuna 12, 50009 Zaragoza, Spain b
ARTICLE
INFO
Keywords: Kinetic inductance detectors Phonon-mediated light detectors Rare events searches Superconducting device
ABSTRACT The development of large area cryogenic light detectors is one of the priorities of next generation bolometric experiments searching for neutrinoless double beta decay. The simultaneous read-out of the heat and light signals enables particle identification, provided that the energy resolution and the light collection are sufficiently high. CALDER (Cryogenic wide-Area Light Detectors with Excellent Resolution) is developing phonon-mediated silicon light detectors using KIDs, with the goal of sensing an area of 5 × 5 cm2 with a resolution of 20 eV RMS. We present the latest results obtained with aluminum chips and with newly developed multi-layer titanium–aluminum chips featuring a remarkable sensitivity.
Contents 1. 2. 3. 4. 5. 6.
Introduction ....................................................................................................................................................................................................... First phase: Al resonator optimization ................................................................................................................................................................... Second phase: AlTiAl resonator optimization.......................................................................................................................................................... Third phase: 5 × 5 cm2 substrate.......................................................................................................................................................................... The substrate choice ............................................................................................................................................................................................ Conclusion and perspectives ................................................................................................................................................................................. Acknowledgments ............................................................................................................................................................................................... References..........................................................................................................................................................................................................
1. Introduction The CUORE experiment [1] is searching for the neutrino-less double beta decay (0𝜈DBD) of 130 Te operating 988 cubic crystals (edge = 5 cm) of TeO2 as cryogenic calorimeters. The signal produced by this reaction consists of two electrons with a total kinetic energy of about 2.5 MeV. Unfortunately the 𝛼 background events, produced by the radioactive contaminations located on the surfaces of the detector components limit the sensitivity of the experiment. The CUPID (CUORE Upgrade with
1 2 2 2 3 3 3 3
Particle IDentification) interest group [2] aims to develop a CUOREsized bolometric detector able to reach a sensitivity 50–100 times better than CUORE. This ambitious goal can be reached by increasing the source mass and reducing the background in the region of interest. To increase the number of 0𝜈DBD emitters, crystals grown with enriched material are needed. The background suppression can be achieved by discriminating 𝛽∕𝛾 against 𝛼 events by means of the Cherenkov radiation [3,4], produced in the TeO2 crystal only by electrons. The expected Cherenkov light signal detectable at the 0𝜈DBD of 130 Te is about 100 eV. To detect such a small light output, the light detector must
∗ Corresponding author. E-mail address: [email protected] (N. Casali).
https://doi.org/10.1016/j.nima.2018.10.079 Received 30 June 2018; Received in revised form 9 October 2018; Accepted 11 October 2018 Available online xxxx 0168-9002/© 2018 Elsevier B.V. All rights reserved.
Please cite this article in press as: N. Casali, et al., Status of the CALDER project: Cryogenic light detectors for background suppression, Nuclear Inst. and Methods in Physics Research, A (2018), https://doi.org/10.1016/j.nima.2018.10.079.
N. Casali et al.
Nuclear Inst. and Methods in Physics Research, A (
satisfy very stringent requirements; the most important one is the energy resolution, that must be lower that 20 eV RMS in order to reject event by event the 𝛼 background. CALDER [5] will take advantage from the superb energy resolution and natural multiplexed read-out [6] provided by Kinetic Inductance Detectors (KIDs). These sensors, that have been successfully applied in astro-physics searches, are limited only by their poor active surface, of a few mm2 . For this reason, we are exploiting the phonon-mediated approach, proposed by Swenson et al. [7] and Moore et al. [8]: the KIDs are deposited on an insulating substrate featuring a surface of several cm2 . Photons emitted by the bolometer interact in the substrate and produce phonons, which can travel until they are absorbed by a KID. The first phase of the CALDER project was devoted to the optimization of the KIDs design, and to the understanding/suppression of the noise sources. For this phase we chose a 2 × 2 cm2 Si substrate 300 μm thick, and a well-known material for KIDs application, aluminum, which according to our detector model would have provided a noise resolution of about 80 eV RMS. In the second phase we investigated more sensitive superconducting materials like AlTiAl multi-layer which allowed to reach an energy resolution of 25 eV RMS. In the last phase, the optimized light detectors will be scaled to the final detector size (5 × 5 cm2 ) and will be coupled to an array of TeO2 bolometers to prove the potential of this technology. In this paper we show the status of the CALDER project and discuss the results obtained from the first two phases.
)
–
Fig. 1. Design of a CALDER sensor. The inductor (30 strips of 62.5 μm×2 mm) features an active area of 3.75 mm2 (4.0 mm2 including the active region that connects the inductor to the capacitor). To contain the geometrical inductance, we used a gap of 5 μm between the meanders and we closed the circuit using a capacitor made by only 2 fingers.
2. First phase: Al resonator optimization The detector shown in Fig. 1 is the result of the design optimization performed in the first phase of the project. We used a 60 nm thick Al film, as this thickness is a good compromise between detector sensitivity and quality of the superconductor. The active area of the KID (4.0 mm2 ) was enlarged with respect to the first prototypes [9] in order to increase the fraction of phonons that can be collected before being lost in the substrate as described in Ref. [10]. The detector was cooled below the critical temperature using a 3 He/4 He dilution refrigerator with base temperature of about 10 mK. The output signal was fed into a CITLF4 SiGe low noise amplifier [11], which was thermally anchored to the 4 K plate of the cryostat. The rest of the electronics was located at room temperature and its features, together with the description of the cryogenic facility and the acquisition software, can be found in Refs. [5]. The detector response to energy deposition on the Si substrate was evaluated exploiting a 400 nm LED light source located at room temperature and an optical fiber that drives the light signals down to the detector plate at 10 mK. The energy of this light pulses can be tuned from few hundreds of eV up to hundreds of keV. Furthermore, the detector was permanently exposed to X-rays from 55 Fe to cross check the energy calibration performed with optical pulses. The detector was operated in the most sensitive point, where the signal-to-noise ratio (S/N) in the phase direction is maximum (see Ref. [12]). The noise energy resolution obtained was 80 eV RMS with a phonon collection efficiency of about 10%. More details about the performance of the phase-1 prototypes can be found in Ref. [9,13].
Fig. 2. Sketch of the KID deposited on the Si substrate (lateral view). The KID is made by a multi-layer of Al+Ti+Al. The dimensions of the Si substrate are not in scale.
titanium nitride (TiN), or composite superconductors such as Ti+TiN or Ti+Al. We performed several tests on TiN substrates without obtaining satisfactory results: the reproducibility of the superconductor substrate as well as the performances of the resonators were not found to be sufficient. On the contrary, the first test on Ti+Al multi-layer showed very encouraging results. We started with a bilayer made of Ti+Al, with the Ti layer in contact with the Si substrate. Even if the detector performances were competitive with the Al film ones, we measured a decreasing in the phonons collection efficiency, that could be ascribed to a mismatching phonon transmission between Si and Ti. Since the Ti film cannot be exposed to the air to avoid its oxidation, we moved to a trilayer made of Al+Ti+Al: we used the Al layer as interface between Ti layer and Si substrate to increase the phonon absorption probability, followed by the Ti layer in order to decrease the 𝑇𝐶 and increase 𝐿𝑘 and finally, the last Al layer to prevent the oxidation of Ti. We tested a few resonators made with different thickness of both Al and Ti layers. The best results was obtained with the pattern shown in Fig. 2 and using the detector design optimized in the phase one (Fig. 1). We obtained a phonon collection efficiency comparable we the one measured on Al resonator while the noise energy resolution increase up to 25 eV RMS. More details about the AlTiAl substrate performances can be found in Ref. [14].
3. Second phase: AlTiAl resonator optimization The second phase of the project aimed at improving the noise energy resolution down to ∼ 20 eV RMS. To this purpose, we moved to the production of chips based on other, more sensitive superconductors. The resolution (𝛿𝐸), indeed, scales as: 𝑇 𝛿𝐸 ∝ √ 𝐶 𝜖 𝑄𝐿𝑘
(1) 4. Third phase: 5 × 5 cm𝟐 substrate
where 𝑇𝐶 in the critical temperature of the superconductor, 𝑄 the total quality factor of the resonator, 𝐿𝑘 the kinetic inductance fraction and 𝜖 the phonon collection efficiency. Therefore, superconductors with lower critical temperature and higher inductance would allow to further enhance the sensitivity. Possible candidates are sub-stoichiometric
With the results described in Refs. [13,14] we accomplished the first and second phase of the CALDER project. Then we started the final project phase increasing the surface of the substrate up to the final size of 5 × 5 cm2 . Fig. 3 shows the first prototype that we developed. 2
Please cite this article in press as: N. Casali, et al., Status of the CALDER project: Cryogenic light detectors for background suppression, Nuclear Inst. and Methods in Physics Research, A (2018), https://doi.org/10.1016/j.nima.2018.10.079.
N. Casali et al.
Nuclear Inst. and Methods in Physics Research, A (
)
–
Fig. 4. Noise power spectra measured for the same Al resonator evaporated on Si (left) and SOS (right) substrates with pulse tube off (red line) and on (black line). The noise increase at low frequency caused by the vibrations induced by the pulse tube (black line) is visible only for the Si substrate . (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)
model suggests that we can reach this results sampling the 5 × 5 cm2 substrate with four AlTiAl resonators like the one described in Section 3. Furthermore, we discovered that detectors based on sapphire substrate (like SOS or GeOS) are not sensitive to the mechanical vibration induced, for example, by the pulse tube. This result is of the utmost importance for the future application of this type of light detectors in the CUPID experiment, that will be host in the same dry cryostats used now by the CUORE experiment.
Fig. 3. First prototype of 5 × 5 cm2 Si substrate 300 μm thick sampled with one Al KID.
In the first test we used just one resonator made of aluminum to study the detector response in the absence of cross-talk or competition among pixels in the absorption of the propagating phonons. The tests on the 5 × 5 cm2 prototypes are ongoing and the results obtained are still very preliminary, therefore they will not be presented in this work. Nevertheless, exploiting our phonons propagation and absorption model [15] we estimated that sampling the 5 × 5 cm2 substrate with four resonators as the one measured in Ref. [14] we will be able to achieve our final resolution target of 20 eV RMS.
Acknowledgments This work was supported by the European Research Council (FP7/2007-2013) under Contract No. CALDER No. 335359 and by the Italian Ministry of Research, Italy under the FIRB Contract No. RBFR1269SL. The authors thanks the personnel of INFN Sezione di Roma for the technical support, in particular M. Iannone, F. Pellegrino, L. Recchia and D. Ruggeri.
5. The substrate choice During the aforementioned R&D we performed several tests also on the detectors substrate: in parallel with silicon we studied also the silicon on sapphire (SOS) and the germanium on sapphire (GeOS) substrates. Both the substrate were 2 × 2 cm2 and 300 μm thick sapphire with a deposition respectively of Si or Ge to increase the optical photons absorption probability. The Al KID was evaporated on the opposite face with respect to the one with the Si or Ge deposition. We noticed a fundamental feature of the KID evaporated on the SOS or GeOS substrate: they were barely affected by the noise induced by mechanical vibrations. Indeed almost all the modern dilution refrigerators pre-cool the 3 He/4 He mixture with a cryocooler in place of liquid nitrogen, liquid helium, and a 1 K bath. No external supply of cryogenic liquids is needed in these ‘‘dry cryostats’’ and the operations can be highly automated. This is the case for the CUORE dilution refrigerator [16]. However, dry cryostats are subject to mechanical vibrations, such as those produced by pulse tube refrigerators, that can spoil the detectors performances. We observed this effect on KIDs evaporated on Si substrate as shown in Fig. 4, where an increase of the detector noise at low frequency is visible when the pulse tube is operating. The increase of the noise at low frequency spoiled the energy resolution of the detector evaporated on Si substrates. On the contrary, with the SOS substrate we obtained a comparable energy resolution despite the pulse tube operation. Even if this effect was measured in all the sapphire substrates that we characterized, a clear explanation of this experimental evidence has not been yet formulated.
References [1] D.R. Artusa, et al., Searching for neutrinoless double-beta decay of 130 Te with CUORE, Adv. High Energy Phys. 2015 (2015) 879871. [2] D.R. Artusa, et al., [CUORE Collaboration], Exploring the neutrinoless double beta decay in the inverted neutrino hierarchy with bolometric detectors, Eur. Phys. J. C 74 (2014) 3096. [3] N. Casali, Model for the Cherenkov light emission of TeO2 cryogenic calorimeters, Astropart. Phys. 91 (2017) 44. [4] N. Casali, et al., TeO2 bolometers with Cherenkov signal tagging: towards nextgeneration neutrinoless double beta decay experiments, Eur. Phys. J. C 75 (1) (2015) 12. [5] E. Battistelli, et al., CALDER: Neutrinoless double-beta decay identification in TeO2 bolometers with kinetic inductance detectors, Eur. Phys. J. C 75 (8) (2015) 353. [6] P.K. Day, et al., A broadband superconducting detector suitable for use in large arrays, Nature 425 (6960) (2003) 817–821. [7] L.J. Swenson, et al., High-speed phonon imaging using frequency-multiplexed kinetic inductance detectors, J. Appl. Phys. 96 (2010) 263511. [8] D.C. Moore, et al., Position and energy-resolved particle detection using phononmediated microwave kinetic inductance detectors, J. Appl. Phys. 100 (2012) 232601. [9] L. Cardani, et al., Energy resolution and efficiency of phonon-mediated Kinetic Inductance Detectors for light detection, Appl. Phys. Lett. 107 (2015) 093508. [10] L. Cardani, et al., New application of superconductors: High sensitivity cryogenic light detectors, Nucl. Instrum. Methods 845 (2017) 338–341. [11] http://radiometer.caltech.edu/datasheets/amplifiers/CITLF4.pdf. [12] N. Casali, et al., Characterization of the KID-based light detectors of CALDER, J. Low Temp. Phys. 184 (1–2) (2015) 142–147. [13] F. Bellini, et al., High sensitivity phonon-mediated kinetic inductance detector with combined amplitude and phase read-out, Appl. Phys. Lett. 110 (2017) 033504. [14] L. Cardani, et al., Al/Ti/Al phonon-mediated KIDs for UV-vis light detection over large areas, Supercond. Sci. Technol. 31 (2018) 075002. [15] M. Martinez, L. Cardani, N. Casali, A. Cruciani, G. Pettinari, M. Vignati, Measurements and simulations of athermal phonon transmission from silicon absorbers to aluminium sensors, arXiv:1805.02495, 2018, submitted for publication. [16] C. Ligi, et al., The CUORE cryostat: A 1-Ton scale setup for bolometric detectors, J. Low. Temp. Phys. 184 (2016) 590.
6. Conclusion and perspectives The CALDER project successfully completed the second phase reaching a noise energy resolution of 25 eV RMS through the operation of KIDs based on sensitive superconductor (such as AlTiAl multi-layer). The next step of the project is to scale up the dimensions of the detector substrate to 5 × 5 cm2 conserving the ∼20 eV RMS noise resolution. Our detector 3
Please cite this article in press as: N. Casali, et al., Status of the CALDER project: Cryogenic light detectors for background suppression, Nuclear Inst. and Methods in Physics Research, A (2018), https://doi.org/10.1016/j.nima.2018.10.079.