Ricochet† Using Low Temperature Bolometers for Coherent Neutrino

Ricochet†

Using Low Temperature Bolometers for Coherent Neutrino Scattering

†An experiment in three courses

Joseph A. Formaggio Outline

(I) First Course: Science Motivation

(II) Second Course: The Ricochet Detector

(III) Last Course: Ricochet & Chooz Outline

(I) First Course: Science Motivation

(II) Second Course: The Ricochet Detector

(III) Last Course: Ricochet & Chooz Yet, some of their Make-and/or- properties may also Break eventually lead to it’s unraveling.

Neutrinos helped The Super-Kamiokande detector in Japan, which helped establish neutrino conﬁrm the validity of mass (a violation of the Standard Model) what we call “the Standard Model”

Image from Gargamelle neutrino experiment, which helped confirm the electroweak model Inverse Beta Decay: How Neutrinos + Interact ⌫¯e + p n + e ! Poltergeist Cd The fact that it took 20 years to Gd + detect the first neutrino is a Li } testament to the difficulty of the task.

We will focus on neutrino energies EMBARGOED UNTIL 2:00 PM US ET THURSDAY, 3 AUGUST 2017 associated with nuclear reactors Coherent Neutrino Scattering:

(2-20 MeV) for this discussion. ⌫ + A ⌫ + A !

There are two primary ways which one can use to detect neutrinos: inverse beta decay and coherent elastic neutrino nuclear scattering (CEνNS). COHERENT

Fig. 2. COHERENT detectors populating the “neutrino alley” at the SNS (34). Locations in this basement corridor profit from more than 19 m of continuous shielding against beam-related neutrons, and a modest 8 m.w.e. overburden able to reduce cosmic-ray induced backgrounds, while 11 2 sustaining an instantaneous neutrino flux as high as 1.7 × 10 νµ / cm s.

First release: 3 August 2017 www.sciencemag.org (Page numbers not final at time of first release) 8

What’s So Special About Coherence? ν ν

Incoming Neutrino Outgoing Neutrino

Neutrons Protons

Nucleon Size (<< 1 fm)

One can think of the energy (really, momentum) the neutrino imparts on its target as a way to probe the nucleus.

The higher the imparted momentum, the ﬁner the probe. Non-Coherent Interactions e- ν

Incoming Neutrino Outgoing Neutrino

Neutrons Protons

Nucleon Size (<< 1 fm)

In this case, the probe momentum is such that it singles out a single neutron or proton. Coherent Interactions

ν ν

Nucleus Size (> 1 fm)

However, if the exchange momentum is small enough, the probe just sees the entire nucleus, and no single proton or neutron is singled out.

Result: Interaction probability (cross-section) scales as the number of nucleons squared. Fundamental Coherent Interactions

Coherent scattering has been proposed and schemed as a means of detecting neutrinos for many decades.

Despite being proposed 40+ years ago, this process has yet to be observed. Fundamental Coherent Interactions

Coupling term (tiny)

Cross-section (probability of interacting) Coherence effect

Neutrino energy

Coherent scattering has been proposed and schemed as a means of detecting neutrinos for many decades.

Despite being proposed 40+ years ago, this process has yet to be observed. Fundamental Coherent Interactions

Coupling term (tiny)

Cross-section (probability of interacting) Coherence effect

Neutrino energy

Coherent scattering has been proposed and schemed as a means of detecting neutrinos for many decades.

Despite being proposed 40+ years ago, this process has yet to be observed.WAIT! That’s no longer true! EMBARGOED UNTIL 2:00 PM US ET THURSDAY, 3 AUGUST 2017

Discovery!

As of August 3rd, 2017, a ﬁrst detection of coherent neutrino scattering has been reported by COHERENT! The process does indeed take place.

Only 16 kg-years to get ~7 sigma! Fig. 3. Observation of Coherent Elastic Neutrino-Nucleus Scattering. Shown are residual differences (datapoints) between CsI[Na] signals in the 12 µs following POT triggers, and those in a 12-µs window before, as a function of their (A) energy (number of photoelectrons detected), and of (B) event arrival time (onset of scintillation). Steady-state environmental backgrounds contribute to both groups of signals equally, vanishing in the subtraction. Error bars are statistical. These residuals are shown for 153.5 live-days of SNS Coherent neutrino detection from inactivity (“Beam OFF”) and 308.1 live-days of neutrino production (“Beam ON”), over which 7.48 GWhr of 23 EMBARGOED UNTIL 2:00 PM US ET THURSDAY, 3 AUGUST 2017 energy (~1.76 × 10 protons) was delivered to the mercury target. Approximately 1.17 photoelectrons are reactors remains a goal for future expected per keV of cesium or iodine nuclear recoil energy (34). Characteristic excesses closely following the experiments. Standard Model CEνNS prediction (histograms) are observed for periods of neutrino production only, with a rate correlated to instantaneous beam power (fig. S14).

Fig. 2. COHERENT detectors populating the “neutrino alley” at the SNS (34). Locations in this basement corridor profit from more than 19 m of continuous shielding against beam-related neutrons, and a modest 8 m.w.e. overburden able to reduce cosmic-ray induced backgrounds, whileFirst release: 3 August 2017 www.sciencemag.org (Page numbers not final at time of first release) 9 11 2 sustaining an instantaneous neutrino flux as high as 1.7 × 10 νµ / cm s.

First release: 3 August 2017 www.sciencemag.org (Page numbers not final at time of first release) 8

ν ν Opening New Doors ν ν Opening New Doors

Step 1:

Conﬁrm Standard Model prediction

Statistics needed: ~ O(1) ν ν Opening New Doors

Step 1: Step n+1:

Conﬁrm Standard Model Measure Weinberg angle at prediction low Q2

Non-standard interactions

Sterile neutrinos

Neutrino magnetic moments

Reactor monitoring Statistics needed: Statistics needed: ~ O(1) >>O(100) Possible Science Reach *J. Billard Ricochet: sensitivity to µ J. Billard and B. Kavanagh, in preparation • Scientific Drivers: Threshold: 0.1 keV ×10−9 ] 0.3 B 90% C.L. contours µ Preliminary

• Better understanding of fundamental [ ν neutrino interactions. µ

• Understanding the physics of 0.2 supernova explosions. Back. Tot. = 1.4 /kg/day 5% signal syst. TEXONO (data?) 20% (betas) back. syst. (arXiv:1506.08377) • Direct probe into nuclear (neutron) 1000 kg.year Ge exposure

structure of nuclei. 0.1 APV (q = 2.4 MeV)

Ricochet • New physics (sterile neutrinos, magnetic moment, weak mixing Ricochet + APV 0 angle). 0.05 0.1 0.15 0.2 0.25 0.3 2 6 sin θw • New possibilities looking for axion and dark forces (arXiv:1504.07237).

• Technology Drivers:

• Leveraging technology of neutrino and dark matter detectors. Possible Science Reach *J. Billard Ricochet:Ricochet: sensitivitysensitivity toto µµ J.J. Billard and B. Kavanagh, inin preparationpreparation Threshold: 0.01 keV • Scientific Drivers: −9 Threshold: 0.1 keV ×10 −9 ] ×10 ] 0.3 B 0.3 B 90% C.L. contours µ Preliminary 90% C.L. contours µ Preliminary [ • Better understanding of fundamental [ ν ν µ neutrino interactions. µ

• Understanding the physics of 0.20.2 supernova explosions. Back.Back. Tot. Tot. = = 1.4 1.4 /kg/day /kg/day 5%5% signal signal syst. syst. TEXONO (data?) 20%20% (betas) (betas) back. back. syst. syst. (arXiv:1506.08377) • Direct probe into nuclear (neutron) 10001000 kg.year kg.year Ge Ge exposure exposure

structure of nuclei. 0.10.1 APVAPV (q (q = = 2.4 2.4 MeV)MeV)

Ricochet • New physics (sterile neutrinos, Ricochet magnetic moment, weak mixing RicochetRicochet + + APV APV 0 angle). 00.05 0.1 0.15 0.2 0.25 0.3 0.05 0.1 0.15 0.2 0.25 2 0.3 sin2 67 sin θθww • New possibilities looking for axion and dark forces (arXiv:1504.07237).

• Technology Drivers:

• Leveraging technology of neutrino and dark matter detectors. The Case for Sterile Neutrinos

•A number of recent (and not so recent) • All suggestive, but no “smoking results seem to indicate the possibility gun” accepted by the of sterile neutrinosThe†. Reactor Anomaly (RAA) community at the moment.

? ? ? Atm. oscillation Solar oscillation

Phys. Rev. D 83, 073006 (2011)

3σ anomaly

Reactor Anomaly

† The initial inception of RicochetTh. Lasserre was –toHEP search 2017 for sterile neutrinos, discussed over coffee. Genuine Neutrino “Application”

Monitoring of Heavy Water • When applied to nuclear Reactors reactors, it is possible to monitor arXiv:1403.7065v1 the activity of the reactor core activity and fuel composition with a relatively compact device.

Heavy Water Reactor in Arak, Iran

• Possibility of first genuine applied technology for neutrino physics.

• There have been spin-offs from the technology developed for neutrinos. But here we discuss actually using neutrinos as part of that technology. Monitoring of Spent Fuel arXiv:1606.06309v1 Outline

(I) First Course: Science Motivation

(II) Second Course: The Ricochet Detector

(III) Last Course: Ricochet & Chooz Different Approaches to Detection “Heat”

Phonons (meV/ph) 100% energy

Ionization Scintillation (10 eV/e-) (1 keV/Ɣ) 10% of energy few % of energy

“Charge” “Light” Where Phonon Ricochet Technology is Used

CUORE

CMB, Infrared detection Dark matter 0νββ

• To go to lower neutrino energies, lower threshold are required. Phonon readout is a promising technology already used in many other experiments.

• Ricochet uses phonon readout to reach low threshold, with eventual goal of reaching 10 eV recoil threshold. Cryogenic Bolometers Cryogenic Bolometers Technology for large NumberCalorimetry of Pixels no filter for events far(ECHO, from endpointHOLMES & NuMECS) needs large number of counts needs large number of pixels

needs

~ 1010 counts for m ~ 10 eV

~ 1013 counts for m ~ sub 1 eV Temperature rise in cryogenic bolometers proportional to energy deposition & capacitance.

3 Since capacitance drops as T in insulators/ superconductors, one can achieve high energy resolution.

Calorimetric Approach ALL energy is absorbed. No issues with 163 163 backscattering, final states, etc. Ho + e Dy⇤ + ⌫ technology to read ! e out such large pixel numbers is available and proven

needs

~ 1010 counts for m ~ 10 eV

13 The absorber allows conversion from ~ 10 counts for m ~ sub 1 eV energy to heat (phonons) Temperature rise in cryogenic bolometers For semi-conductors and proportional to energy deposition & superconductors, only lattice vibrations contribute tocapacitance. thermal capacitance (C ~ T3)

3 Since capacitanceSmall detectors drops & lowas Ttemperatures in insulators/ superconductors, one= can achieve high energylower resolution. thresholds Calorimetric Approach ALL energy is absorbed. No issues with 163 163 backscattering, final states, etc. Ho + e Dy⇤ + ⌫ technology to read ! e out such large pixel numbers is available and proven

Cryogenic Bolometers Cryogenic Bolometers Technology for large NumberCalorimetry of Pixels no filter for events far(ECHO, from endpointHOLMES & NuMECS) needs large number of counts needs large number of pixels needs Small changes in temperature can be captured by Transition Edge Sensors (TES), which allow great sensitivity to ~ 1010 counts for m ~ 10 eV small temperature depositions.

13 The absorber allows conversion from ~ 10 counts for m ~ sub 1 eV energy to heat (phonons) TES3Resistance3@3Tc Temperature rise in cryogenic bolometers For semi-conductors and proportional to energy deposition & superconductors, only lattice vibrations contribute tocapacitance. thermal capacitance (C ~ T3)

Since capacitance drops as T3 in insulators/ Resistance Small detectors & low temperatures Normalized superconductors, one= can achieve high energylower resolution. thresholds Calorimetric Approach ALL energy is absorbed. No issues with 163 163 backscattering, final states, etc. Ho + e Dy⇤ + ⌫ technology to read ! e out such large pixel numbers is available and proven

Cryogenic Bolometers Cryogenic Bolometers Technology for large NumberCalorimetry of Pixels

Readout of TES done using SQUID (ECHO, HOLMES & NuMECS) amplifiers, quantum-limited no filter for events far from endpoint magnetometers, ideal for small currents. needs large number of counts needs large number of pixels needs Small changes in temperature can be captured by Transition Edge Sensors (TES), which allow great sensitivity to ~ 1010 counts for m ~ 10 eV small temperature depositions.

Event rates for phonon versus ionization kg(✏) Ionization readout requires much lower f = thresholds for the same rates n 1+kg(✏)

D Ge-Phonon L 10.00 Ge-Ionization day 5.00 kg H ê

evt 1.00 @ 0.50 Ionization Phonons Rate 0.10 0.05 Integrated

10 20 50 100 200 500 1000 2000 Threshold Energy eVnr or eVee *from Tali Figueroa and Adam Anderson

@ D What Kind of Detectors to Use?

The goal is to reach recoil thresholds below 100 eV so as to see the onset of reactor neutrinos from CEvNS reactions.

Two different technologies discussed for deployment.

Germanium Superconducting Detectors Metals (EDELWEISS) (MIT)

We are also discussing potentially adding CaWO4 crystals to the list. See the next talk! 90% Gammas 90% Neutrons Repurposing DM 1.4 Ionisation Threshold EDELWEISS-II simulations Detectors 1.2 Ionisation Yield 1

0.8

0.6

0.4

0.2

0 0 20 40 60 80 100 120 140 160 180 200 ER, keV

Figure 5: IonisationEDELWEISS yield (ratio of ionisation Detectors: to phonons normalised to this ratio for electron recoils) versus recoil energy for simulated nuclear recoils in EDELWEISS-II from neutronsoriginatedintheuraniumdecaychainfromcontamination in the steel support structure around the main copper vessels. Blue curves show the average and the edges of the band which contains 90%Separation of nuclear recoilsof recoil in one from of the electromagnetic crystals ascalculatedfromtheexperimentalresolutions,thatwere events using light also included in the simulations [1]. Green curves show the band which contains 90% of electron recoils. They appear on the plot becauseand of neutron charge inelastic signatures. scattering and captureresultingingamma-rayproduction.Thepinkcurveshows the 3 keV software threshold for ionisation, applied as in real data. Statistics corresponds to about 4.5 × 104 years of live time for the uranium decay rate of 5 mBq/kg. New 25 gram-scale detectors projected to reach 50-100 eV threshold. pb. Materials and components which could contribute significantly to the gamma-ray or neutron background rate in EDELWEISS-II are being replaced by their counterparts with better radiopurity, for instance theRadioactivity cryostat screens levels 7 to reduced 11 and other down copper to ~ part1 event/keV/kg//days at 10 mK (disks supporting the Ge detectors, vertical bars and 10 mK chamber) are made of ultra radiopure NOSV copper [20]. Radioactivity measurements of most new components were done at LSM using low-background gamma- ray spectrometry. Extensive simulations of gamma-rays and neutrons were carried out for a geometry of EDELWEISS-III with additional neutron shielding and the results were normalised to the measured concentrations of radioactive isotopes. The results of the measurements and simulations are shown in Table 4. For some components, such as cables and connectors, various parts were screened separately using a HPGe detector. We present in Table 4 the data for the parts which contribute the most to the background rate. The uncertainties in the radioactivity levels are given at 90% C. L.. Some measurements gave only upper limits leading to large uncertainties in the expected background rates. Neutron event rates were calculated assuming secular equilibrium in the U/Th decay chains except for 210Pb sub-chain. The neutron rate is also affected by a large uncertainty in the chemical composition of the component or its part which may contribute to the background. Since a significant fraction of neutrons may come from (α,n) reactions, exact knowledge of the chemical composition of the material is crucial in the estimate of the neutron event rate. However, in some cases, for instance electronics parts, it is not known precisely which particular part is contaminated the most and hence, it is difficult to predict the expected rate of events with high accuracy. We emphasise that we try to avoid placing materials containing elements with high cross-section of (α,n) reactions (low energy threshold), for example fluorine, close to the crystals. As can be seen from Table 4, a large contamination of printed

11 to TES sensor gold pad (+wire) Why Si (or SiO2) insulating layer Superconducting ν ν Metals?

qp’s ɸ’s

Zn Absorber Tc ~ 0.85 K ϴD ~ 327 K

Metallic Superconductors as Detectors:

Zinc crystals become superconducting below 850 mK. If operating at 15 mK, this is well below Tc. Implies that capacitance dominated by lattice contributions (scale as T3).

High Debye temperature implies low capacitance.

Target atomic number very similar to germanium.

Energy breaks Cooper pairs; turning into either quasi-particles or phonons. PHOTOINDUCED TIME-RESOLVED ELECTRODYNAMICS… PHYSICAL REVIEW B 72, 024510 ͑2005͒

TABLE I. Parameters for Mattis-Bardeen fits of the far infrared superconducting to normal transmission ratio / . In all cases we used the dirty limit ͑1/␶ ϱ͒ approximation. The time ratio obtained from the Ts Tn → nonequilibrium fits, the laser fluences ͑filling the 5 mm sample aperture͒ utilized and the excess QP fraction NQP/NT are also shown. For the latter the number in parenthesis is the lowest sample temperature reached.

−1 −1 −2 0 0 Material Rᮀ͑⍀͒ Tc͑K͒ 2⌬0͑cm ͒ 2⌬0 /kBTc ␦S͑⍀ cm ͒ ␶R /␶B I͑nJ/pulse͒ NQP/NT Pb 70 7.2 22.5 4.5 −0.5 2.53 0.5 0.03 ͑3.7 K͒

Pb0.75Bi0.25 110 8.0 28 4.9 −0.7 2.2 0.5 0.24 ͑3.0 K͒ Nb 61 5.4 14 3.7 0 13.3 2 0.03 ͑3.5 K͒

Nb0.5Ti0.5N 110 9.8 26.5 3.9 −0.2 6 2 5 ͑2.0 K͒ NbN 70 13.5 35.0 3.7 0 1.53 2 14 ͑2.2 K͒

IV. THE EXCESS QUASIPARTICLE STATE to 2⌬ whereas phonons of energy less than 2⌬ do not par- ticipate in the pair breaking process. Therefore, the system A. Theory eventually reaches the state depicted in Fig. 3͑b͒. This state Figure 3͑a͒ shows the density of states at 0 K of a BCS is characterized by an excess ͑with respect to thermodynamic superconductor. Light having a photon energy h␯Ͼ2⌬ is equilibrium͒ of unpaired QPs of energy ⌬ and, as shown by 0 14 absorbed and breaks a Cooper pair, producing two QPs Owen and Scalapino, by a reduced energy gap. above the Fermi energy. In our experiments the photon en- The effective relaxation process must take into account a ergy is much higher than the superconducting gap and thus phonon bottleneck effect, shown schematically in Fig. 3͑c͒. the QPs are created far above the Fermi level with an energy Two QPs having energy ⌬ take a time ␶R to recombine into a ϳh␯/2. Very quickly ͑in subpicosecond times͒ these high- Cooper pair. At recombination, a 2⌬ phonon is emitted to energy QPs relax via electron-electron and, eventually, carry the excitation energy of the QPs. The phonons with electron-phonon scattering. These relaxation processes break energy 2⌬ can break a Cooper pair, in characteristic time ␶B, more pairs, and the QPs quickly settle to energies close to ⌬. creating two QPs at the gap edge. A steady-state dynamic In such a cascading process, each QP initially created by the balance is established between the phonons and QPs. Even- laser pulse will generate ϳh␯/2⌬ QPs close to the Fermi tually, however, the 2⌬ phonons become depleted; either level. they relax anharmonically to lower energies or they leave the A similar branching process occurs in the phonon sector. ﬁlm. ͑The phonons can escape into the substrate or into their Phonons of energy higher than 2⌬ can break pairs, relaxing surrounding environment, such as the helium bath.͒ The phonon escape time is denoted ␶␥. The Rothwarf and Taylor ͑RT͒ equations5 describing nonequilibrium superconductivity at any density of excess quasiparticles can be written as

2 dNQP 2NQP 2N⍀ = I0 − + ͑4͒ dt ␶R ␶B LOBO et al. PHYSICAL REVIEW B 72, 024510 ͑2005͒ Lobo et al. and Phys. Rev. B 72, 024510 (2005) Why 2 dN⍀ N⍀ N⍀ N gion covered by most of our= − experiments+ + andQP . we will use͑5͒ Superconducting this expression for somedt qualitativeͩ ␶B analysis.␶␥ ͪ ␶R Metals? The phononI0 is the escape density time of QPs depends injected on in several the system, factorsNQP is such the as the samplenumber quality, of excess the QPs film at thickness, energy ⌬ and acousticN⍀ is the mismatch number of between filmnon-equilibrium and substrate, phonons and with the energy presence 2⌬. We or are absence interested of in the case with without injection45 of QPs ͑I0 =0͒ in a system liquid heliumin the surrounding low-fluence limit, the i.e., film. the numberThere of is QPs no created detailed by calculationthe for laser the light temperature is much smaller dependence than the of thermally␶␥. However, broken at temperaturespairs. In much this smallercase, the than RT equations the Debye can temperature, be linearized the Kapitzagiving resistance10 in a material interface behaves as T3.46 FIG. 3. ͑Color online.͒ Ratio of the superconducting ͑NS͒ and normal ͑NN͒ densities of states at 0 K for a BCS superconductorAs a͑a higher͒ Kaptiza resistancedNQP makes2NQP it2 harderN⍀ for phonons in equilibrium and ͑b͒ after photoexcitation and initial cascading = − + ͑6͒ Metallic Superconductors as Detectors: to leave the material, in a firstdt approximation␶R ␶B we can assume effects but before QP recombination. The broken pairs in the pho- 0 3 FIG. 4. ͑Color online.͒͑a͒ Effective relaxation time ͑␶eff͒ calcu- that ␶␥ =␶␥ +aT . Several scattering rates for metallic super- toexcited state produce the empty states below EF and occupied and 47 lated from the intrinsic QPstates recombination above. Consequently, time ͑␶ theR͒, superconducting the pair break- gap isconductors depleted. were calculated by Kaplan et al. In particular, However, quasi-particles and phonons do not evolve in the same way. dN N N N ing time by phonons ͑␶BPanel͒, and͑c͒ theillustrates phonon the relaxation escape time process͑␶ with␥͒ using competition␶ betweenR and ␶B can be obtained from⍀ = − intrinsic⍀ + ⍀ materials+ QP . parameters7 0 0 ͩ ͪ ͑ ͒ Eqs. 8 , 10 , and 11 withQPs and␶ = phonons.␶ =100 ps. ␶ =300 ps is taken to dt ␶B ␶␥ ␶R ͑ ͒ ͑ ͒ ͑ ͒ R B ␥ 3 ϱ Recombination times for quasi-particles become extremely long at low temperatures (~ 1 1 ⌬0 1 ⍀ − ␻ be temperature independent. ͑b͒ Temperature dependence of the ex- = d⍀⍀2 seconds), while (a)thermal phononscess number operate of at QPs much from different Eq. ͑13 ͒(faster). The dashed time scales. line was calculated 024510-5 ͩ ͪ 0 ͵ 2 2 ␶R͑␻͒ 1 − f͑␻͒ kBTc ␶R ␻+␦ ͱ͑⍀ − ␻͒ − ␦ using the same parameters as those used in panel ͑a͒. For the other 2 curves, the number next to each line indicates the ratio ␶0 /␶0 uti- ␦ Separation of recoil from electromagnetic events using quasi-particle versus athermalR B ϫ 1+ ͓n͑⍀͒ +1͔f͑⍀ − ␻͒͑10͒ phonon timing signatures. lized. The error bars indicate the maximum uncertainty introduced ͫ ␻͑⍀ − ␻͒ͬ by a ±1 cm−1 variation in the gap value 2⌬. and The solution for this set of coupled differential equations 1 1 ␻−␦ d⍀ ⍀͑␻ − ⍀͒ + ␦2 = 1 − f ⍀ is a linear combination of two exponential relaxations with 0 2 2 2 2 ͓ ͑ ͒ † ͵ The result of a conversation with Flavio Gatti over lunch. ␶B͑␻͒ ␶ ␦ ͱ⍀ − ␦ ͱ͑␻ − ⍀͒ − ␦ characteristic decay times given by B − f͑␻ − ⍀͔͒. ͑11͒ −1 −1 −1 −1 −1 1 2␶R + ␶B + ␶␥ 8␶␥ ␶R 0 3 0 2 2 = 1± 1 − . Here, ␶ =បZ1͑0͒/2␲b͑kbTc͒ and ␶ =បN/4␲ N͑0͒͗␣ ͘⌬0 ␶± 2 ͫ ͱ ͑2␶−1 + ␶−1 + ␶−1͒2 ͬ R B R B ␥ are respectively the intrinsic pair recombination time close to ͑8͒ Tc /2 and the phonon pair-breaking time at 0 K. All energies in Eqs. ͑10͒ and ͑11͒ are measured in units of ⌬ . ␦=⌬/⌬ is ␶± represent the two time scales expected in the QP and 0 0 the reduced gap, Z1͑0͒ is the QP renormalization factor, N͑0͒ phonon populations. As discussed previously, just after is the electron single particle DOS at E , N is the density of pumping with the laser the system has mostly excited QPs F ions, and ͗␣͘2 is the electron-phonon coupling ␣2F͑⍀͒ func- but not nonequilibrium phonons. These phonons are created tion averaged over the whole phonon spectrum. In Eq. ͑10͒ by the cascading of high energy QPs to the gap edge. Even- ␣2F͑⍀͒ is approximated by b⍀2, but the correct function can tully, excess QP and phonon populations equilibrate into a be used if it is known. ␻ is the energy of QPs just before steady state dynamic balance ͑dN /dtϷdN /dt͒. A second QP ⍀ recombination or the energy of phonons available for pair process then follows as the system fully relaxes to the state breaking. n͑⍀͒ and f͑⍀͒ are the Bose and Fermi factors, without excess QPs and phonons. In order to assign the two respectively. characteristic times in Eq. ͑8͒ to the two processes above, it Figure 4 a shows ␶ and ␶ calculated from Eqs. 10 and is useful to look at the limiting situation where ␶ ϱ. With- ͑ ͒ R B ͑ ͒ ␥ 11 , assuming a weak coupling BCS temperature depen- out phonon escape, the system the system should→ reach the ͑ ͒ dence for the gap, QPs with energy ⌬ and phonons with steady state dynamic balance between QP and phonon popu- energy 2.1⌬. We note that ␶ diverges at 0 K while ␶ re- lations but should not relax to the fundamental state. In this R B mains finite. Therefore, at low temperatures, all the photon case, one obtains 1/␶+ =2/␶ +1/␶ and 1/␶− =0. Hence ␶+ is R B energy absorbed in the pair breaking process will appear as the time for excess QPs and phonons to reach the steady state excess QPs. At higher temperatures the photon energies will dynamic balance. The solution with the negative sign is the end up distributed between QPs and phonons so that one of interest for our work and represents the effective time − ͑␶eff=␶ ͒ for the system to fully relax. NQP͑0͒⌬0 = NQP͑T͒⌬͑T͒ + N⍀͑T͒2⌬͑T͒. ͑12͒ Figure 4͑a͒ shows that for temperatures above ϳ0.4T but c When the phonon and QP populations are in equilibrium, not very close to Tc the phonon escape time ␶␥ is much larger NQP/␶R ϷN⍀ /␶B and, thus, than ␶R or ␶B. Hence ␶eff can be approximated by NQP͑T͒ ⌬0 1 ␶ = . ͑13͒ R N ͑0͒ ⌬͑T͒ 1+2␶ /␶ ␶eff = ␶␥ͩ1+ ͪ. ͑9͒ QP B R 2␶B With Eqs. ͑8͓͒or ͑9͔͒ and ͑13͒ we can describe both the In our numerical analysis we utilized the full expression for excess number of QPs and the effective relaxation time as a ␶eff given by the negative root in Eq. ͑8͒. However, the tem- function of the temperature. Simulations of the effective re- perature range where Eq. ͑9͒ is valid corresponds to the relaxation time and number of excess QPs are shown in Fig. 4.

024510-6 PHOTOINDUCED TIME-RESOLVED ELECTRODYNAMICS… PHYSICAL REVIEW B 72, 024510 ͑2005͒

−1 −1 −2 0 0 Material Rᮀ͑⍀͒ Tc͑K͒ 2⌬0͑cm ͒ 2⌬0 /kBTc ␦S͑⍀ cm ͒ ␶R /␶B I͑nJ/pulse͒ NQP/NT Pb 70 7.2 22.5 4.5 −0.5 2.53 0.5 0.03 ͑3.7 K͒

Pb0.75Bi0.25 110 8.0 28 4.9 −0.7 2.2 0.5 0.24 ͑3.0 K͒ Nb 61 5.4 14 3.7 0 13.3 2 0.03 ͑3.5 K͒

Nb0.5Ti0.5N 110 9.8 26.5 3.9 −0.2 6 2 5 ͑2.0 K͒ NbN 70 13.5 35.0 3.7 0 1.53 2 14 ͑2.2 K͒

2 dNQP 2NQP 2N⍀ = I0 − + ͑4͒ dt ␶R ␶B LOBO et al. PHYSICAL REVIEW B 72, 024510 ͑2005͒ Lobo et al. and Phys. Rev. B 72, 024510 (2005) Why 2 dN⍀ N⍀ N⍀ N gion covered by most of our= − experiments+ + andQP . we will use͑5͒ Superconducting this expression for somedt qualitativeͩ ␶B analysis.␶␥ ͪ ␶R Metals? The phononI0 is the escape density time of QPs depends injected on in several the system, factorsNQP is such the as the samplenumber quality, of excess the QPs film at thickness, energy ⌬ and acousticN⍀ is the mismatch number of between filmnon-equilibrium and substrate, phonons and with the energy presence 2⌬. We or are absence interested of in the case with without injection45 of QPs ͑I0 =0͒ in a system liquid heliumin the surrounding low-fluence limit, the i.e., film. the numberThere of is QPs no created detailed by calculationthe for laser the light temperature is much smaller dependence than the of thermally␶␥. However, broken at temperaturespairs. In much this smallercase, the than RT equations the Debye can temperature, be linearized the Kapitzagiving resistance10 in a material interface behaves as T3.46 FIG. 3. ͑Color online.͒ Ratio of the superconducting ͑NS͒ and normal ͑NN͒ densities of states at 0 K for a BCS superconductorAs a͑a higher͒ Kaptiza resistancedNQP makes2NQP it2 harderN⍀ for phonons in equilibrium and ͑b͒ after photoexcitation and initial cascading = − + ͑6͒ Metallic Superconductors as Detectors: to leave the material, in a firstdt approximation␶R ␶B we can assume effects but before QP recombination. The broken pairs in the pho- 0 3 FIG. 4. ͑Color online.͒͑a͒ Effective relaxation time ͑␶eff͒ calcu- that ␶␥ =␶␥ +aT . Several scattering rates for metallic super- toexcited state produce the empty states below EF and occupied and 47 lated from the intrinsic QPstates recombination above. Consequently, time ͑␶ theR͒, superconducting the pair break- gap isconductors depleted. were calculated by Kaplan et al. In particular, However, quasi-particles and phonons do not evolve in the same way. dN N N N ing time by phonons ͑␶BPanel͒, and͑c͒ theillustrates phonon the relaxation escape time process͑␶ with␥͒ using competition␶ betweenR and ␶B can be obtained from⍀ = − intrinsic⍀ + ⍀ materials+ QP . parameters7 0 0 ͩ ͪ ͑ ͒ Eqs. 8 , 10 , and 11 withQPs and␶ = phonons.␶ =100 ps. ␶ =300 ps is taken to dt ␶B ␶␥ ␶R ͑ ͒ ͑ ͒ ͑ ͒ R B ␥ 3 ϱ Recombination times for quasi-particles become extremely long at low temperatures (~ 1 1 ⌬0 1 ⍀ − ␻ be temperature independent. ͑b͒ Temperature dependence of the ex- = d⍀⍀2 seconds), while (a)thermal phononscess number operate of at QPs much from different Eq. ͑13 ͒(faster). The dashed time scales. line was calculated 024510-5 ͩ ͪ 0 ͵ 2 2 ␶R͑␻͒ 1 − f͑␻͒ kBTc ␶R ␻+␦ ͱ͑⍀ − ␻͒ − ␦ using the same parameters as those used in panel ͑a͒. For the other 2 curves, the number next to each line indicates the ratio ␶0 /␶0 uti- ␦ Separation of recoil from electromagnetic events using quasi-particle versus athermalR B ϫ 1+ ͓n͑⍀͒ +1͔f͑⍀ − ␻͒͑10͒ phonon timing signatures. lized. The error bars indicate the maximum uncertainty introduced ͫ ␻͑⍀ − ␻͒ͬ by a ±1 cm−1 variation in the gap value 2⌬. Potential recoil discrimination! and The solution(maybe) for this set of coupled differential equations 1 1 ␻−␦ d⍀ ⍀͑␻ − ⍀͒ + ␦2 = 1 − f ⍀ is a linear combination of two exponential relaxations with 0 2 2 2 2 ͓ ͑ ͒ † ͵ The result of a conversation with Flavio Gatti over lunch. ␶B͑␻͒ ␶ ␦ ͱ⍀ − ␦ ͱ͑␻ − ⍀͒ − ␦ characteristic decay times given by B − f͑␻ − ⍀͔͒. ͑11͒ −1 −1 −1 −1 −1 1 2␶R + ␶B + ␶␥ 8␶␥ ␶R 0 3 0 2 2 = 1± 1 − . Here, ␶ =បZ1͑0͒/2␲b͑kbTc͒ and ␶ =បN/4␲ N͑0͒͗␣ ͘⌬0 ␶± 2 ͫ ͱ ͑2␶−1 + ␶−1 + ␶−1͒2 ͬ R B R B ␥ are respectively the intrinsic pair recombination time close to ͑8͒ Tc /2 and the phonon pair-breaking time at 0 K. All energies in Eqs. ͑10͒ and ͑11͒ are measured in units of ⌬ . ␦=⌬/⌬ is ␶± represent the two time scales expected in the QP and 0 0 the reduced gap, Z1͑0͒ is the QP renormalization factor, N͑0͒ phonon populations. As discussed previously, just after is the electron single particle DOS at E , N is the density of pumping with the laser the system has mostly excited QPs F ions, and ͗␣͘2 is the electron-phonon coupling ␣2F͑⍀͒ func- but not nonequilibrium phonons. These phonons are created tion averaged over the whole phonon spectrum. In Eq. ͑10͒ by the cascading of high energy QPs to the gap edge. Even- ␣2F͑⍀͒ is approximated by b⍀2, but the correct function can tully, excess QP and phonon populations equilibrate into a be used if it is known. ␻ is the energy of QPs just before steady state dynamic balance ͑dN /dtϷdN /dt͒. A second QP ⍀ recombination or the energy of phonons available for pair process then follows as the system fully relaxes to the state breaking. n͑⍀͒ and f͑⍀͒ are the Bose and Fermi factors, without excess QPs and phonons. In order to assign the two respectively. characteristic times in Eq. ͑8͒ to the two processes above, it Figure 4 a shows ␶ and ␶ calculated from Eqs. 10 and is useful to look at the limiting situation where ␶ ϱ. With- ͑ ͒ R B ͑ ͒ ␥ 11 , assuming a weak coupling BCS temperature depen- out phonon escape, the system the system should→ reach the ͑ ͒ dence for the gap, QPs with energy ⌬ and phonons with steady state dynamic balance between QP and phonon popu- energy 2.1⌬. We note that ␶ diverges at 0 K while ␶ re- lations but should not relax to the fundamental state. In this R B mains finite. Therefore, at low temperatures, all the photon case, one obtains 1/␶+ =2/␶ +1/␶ and 1/␶− =0. Hence ␶+ is R B energy absorbed in the pair breaking process will appear as the time for excess QPs and phonons to reach the steady state excess QPs. At higher temperatures the photon energies will dynamic balance. The solution with the negative sign is the end up distributed between QPs and phonons so that one of interest for our work and represents the effective time − ͑␶eff=␶ ͒ for the system to fully relax. NQP͑0͒⌬0 = NQP͑T͒⌬͑T͒ + N⍀͑T͒2⌬͑T͒. ͑12͒ Figure 4͑a͒ shows that for temperatures above ϳ0.4T but c When the phonon and QP populations are in equilibrium, not very close to Tc the phonon escape time ␶␥ is much larger NQP/␶R ϷN⍀ /␶B and, thus, than ␶R or ␶B. Hence ␶eff can be approximated by NQP͑T͒ ⌬0 1 ␶ = . ͑13͒ R N ͑0͒ ⌬͑T͒ 1+2␶ /␶ ␶eff = ␶␥ͩ1+ ͪ. ͑9͒ QP B R 2␶B With Eqs. ͑8͓͒or ͑9͔͒ and ͑13͒ we can describe both the In our numerical analysis we utilized the full expression for excess number of QPs and the effective relaxation time as a ␶eff given by the negative root in Eq. ͑8͒. However, the tem- function of the temperature. Simulations of the effective re- perature range where Eq. ͑9͒ is valid corresponds to the relaxation time and number of excess QPs are shown in Fig. 4.

024510-6 5

Hidden photon dark matter

8 10 Stellar constraints How Else Can We Use (Stuckelberg case) 10 10 Resonant Superconductors? LC Xenon10 12 10 HB stars

(Higgs case, e0=0.1) Ronit Hochberg, Tongan Lin, and Kathryn 14 Al superconductor 10 Zurek recently proposed also using metallic 16 superconductors to probe bosonic dark 10 1 kg-day matter at very low mass scales. 1 kg-yr 2 18 10 4 3 2 1 0 1 2 10 10 10 10 10 10 10 same way that superconductors and metals are excellent X X mV [eV] absorbers of electromagnetic fields. For instance, we find q Q q R.Hochberg,Q T. Lin, and K. Zurek that a kg-day exposure on a superconducting target is FIG. 3. Estimated sensitivity of an aluminum superconductor target for 1-kg-year (thick solid black) and 1-kg-day (thin solid black) exposures, for absorptionarXiv:1604.06800v1 of hidden photon relic dark [hep-ph] matter. For comparison, we show solar and horizontal branch sucient to exceed the stellar constraints for a hidden constraints for the Stuckelberg (shaded orange) and Higgs cases (dashed purple) [32]; Xenon10 bounds (shaded red) [33]; and the projected reach for an LC circuit experiment (solid gray curve) [34]. photon whose mass is obtained via the Stuckelberg mech- k k0 k k0 anism. e e In what follows,e we use the results of this sectione to where F u⌫ (V µ⌫ ) are the field strengths for the photon The outline of this paper is as follows. In Section II A relate theSuperconducting DM absorption Substrate (Al) rate to that of a photon, and (hiddenSuperconducting photon). Substrate (Ta) For the parameter space considered

Takes advantage of long quasi-particlethen apply Insulating the layer combined solid 1 curve of Fig. 2 to derive here,Insulating this layer hidden photon may be all of the DM, where we discuss how metals can be ecient absorbers of low the sensitivity of a superconducting aluminum target to the origin of the relic abundance is set by a misalignment lifetimes and thermal phonon emissions (and TES and QP collection antennas (W) TES and QP collection antennas (W) mass particles. The process we consider involves ab- various DM candidates. mechanism during or before inflation [37–39]. SuperConducting Bias Rails (Al) Athermal Phonon Collection Fins (Al) FIG. 1. Absorption processFor on the electrons hidden photon for model an described incoming next, relic we will Performing a field redefinition of the photon Aµ very small s.c. gap energies) to see low ! sorbing all the mass-energy of the DM particle via an also require knowledge of 2 at low temperatures; here we Aµ Vµ leads to the canonical basis, where the electro- particle X, where a phonon is emitted in the final state: µ µ energy interactions. simply use the result in the Drude theory, Eq. (14), over magnetic current JEM picks up a dark charge, eVµJEM electron recoil, with emission of an athermal phonon to X(q)+e(k) e(k0)+(theQ). whole energy range. We have verified the validity of in vacuum. However, this mixing angle can vary sub- conserve momentum. We then describe in Sections II B ! this approximation by comparing with measurements of stantially from  due to in-medium e↵ects, which a↵ect Mechanism for detection very similar2 atto room Zn temperature [35, 36], finding at most 50% the polarization tensor ⇧ (related to the conductivity ˆ and II C our method to determine the DM absorption di↵erence with the Drude theory. ⇠ via Eq. (10)). In a metallic target such as aluminum, crystals; our crystals can also be explored for the e↵ective mixing angle is suppressed by powers of the rate from the optical properties of a metal. In Section III vation of the target material. Consider an electron with plasma frequency, this purpose. 2 we present the reach of superconducting detectors for ul- initial momentum ~k and energyIII. RATESE = AND~k / CONSTRAINTS(2m ). Assum- 2m4 2m4 i i i e 2 = V V , (19) e↵ 2 2 2 4 tralight DM that couples to electrons, including hidden Figure 1. Schematic designs for superconducting detectors that are sensitive to DM-electron[m Re scattering.⇧(!)] +[Im⇧(!)] ' !p ing the electron absorbs aUtilizing single the particle results of the of previous energy section,!,the we now V Left: Quasiparticles produced by a recoiling e in a large aluminum arbsorber are collected by tungsten photons, pseudoscalars, and scalars. We conclude in Sec- turn to ultralight~ bosonic~ DM — hidden photons, pseu- where we used Eqs. (10), (13), and (14). Since Re ⇧ final momentumquasiparticle of the collection electron fins and is thenkf their= energyki + is~q sensedand by energy a TES. Right: Athermal phonons produced 2 2 ⇡ doscalars, and scalars — in each case assuming that the !2 is larger than both Im ⇧ !1 and mV ! in tion IV. by a recoil e in a large tantalum absorber are collected by aluminum collection fins and then their energy ⇡ ' conservation gives candidate composes all the DM. our region of interest, we then have e↵ . Note that is sensed by a TES. the suppression by the plasma frequency⌧ is di↵erent than the electron-scattering case explored in Refs. [23, 24], ~ 2 ~ 2 (ki + ~q) ki A. Dark Photons where the Thomas-Fermi screening length was relevant superconducting gap= is not important+ !. for the scattering process(1) itself,for its determining existence meanse↵ . Thethat reason is that the absorption II. DARK MATTER ABSORPTION WITH process occurs when the momentum transfer is much 2me Consider2 am hiddene photon which is kinetically mixed athermal phonons and quasiparticles have very long lifetimes, and assmaller such can than potentially the absorbed be energy, ~q 10 3m !, SUPERCONDUCTORS with the hypercharge gauge boson, leading to kinetic collected before they thermalize. Thus in the systems we consider, detectionwhereas of DM scattering operates in viathe non-relativistic| | ⇠ limit occurs⌧ mixing with the photon, (Note that momentum on the lattice is conserved up to an when ~q !. (See Sec. 5.2 of Ref. [24] for a discus- the breaking of Cooper pairs in a superconducting target. We consider this| idea| in more detail  sion of the (q, !)-dependence of the screening mass.) additive reciprocalnext. lattice vector, K~ . ForF V electrons,µ⌫ , the(18) We begin by describing the DM absorption process, be- L 2 µ⌫ For the absorption of the kinetically mixed hidden pho- fore computing its rate in a superconductor. We compare typical energy2.2 scale Detector associated design with with milli-eV transitions sensitivity involving our results for consistency against the standard Drude K~ is K2/2m 10 eV, which is above the energies con- e Our⇠ detector concept is based on collecting and concentrating long lived athermal excitations theory for low-energy photon absorption in metals. Then, sidered here.)from Then DM interactions the required in a superconducting momentum target transfer absorber onto to a small volume (and thus highly sensitive) sensor. The collection and concentration of long lived excitations is a general concept in order to obtain accurate predictions at higher (& 0.1 the electron is ~q !(m / ~k ) !/v 100 !,where that| | has⇠ been ae core| principlei| ⇠ of detectorF ⇠ physics, from ionization in semiconductor CCDs to eV) energies, we relate the DM absorption rate to mea- vF is the Fermiathermal velocity. phonon collectionThis cannot in CDMS. be Here satisfied we propose that for this an general detection philosophy be sured photon absorption rates. on-shell DM particleapplied in large in the volume halo, (very pure, which single carries crystal) superconductors momen- to search for DM with mass tum 10 3!.as However, low as the warm energy DM limit and of a momentum keV using standardcan superconductingstill sensor technology that has been pushed to its ultimate theoretical sensitivity. A schematic of two proposed detector be conserved⇠ if a phonon with momentum ~q is emit- concepts for light dark matter, that we describe⇠ in greater detail through the remainder of this A. General Principle: Phonon emission ted by the electronsection, inis shown the in final Fig. 1. state; in other words, the electron recoils againstDetection ofthe dark lattice. matter in such The detectors emitted is comprised phonon of a three part process: Dark Matter Scattering on Target Absorber and Subsequent Excitation Production. ADM Absorption of low energy particles in a superconductor carries away a fraction• of the excitation energy, but can particle scatters o↵ an e in the target metal or superconducting absorber. In subse- can proceed when the energy of the absorbed radiation balance the large recoilquent interactions, momentum the recoil of energy the is electron. converted into long lived athermal phonons and (in this case the mass of the DM particle) exceeds the su- In the Debye model,quasiparticles. the dispersion relation of a phonon ~ perconducting gap. In the absorption process, a Cooper with 4-momentumCollection (⌦, Q) of is Excitations. given byThe resulting excitations must be collected and concentrated • pair is broken, and a pair of excitations is created. These onto a small volume (and thus very sensitive) sensor; this is typically done via ‘collection ⌦ = c Q~ (2) excitations have a long recombination and thermalization s| | time (of order a few milliseconds in aluminum), which al- where the speed of sound in aluminum is c lows for their collection and measurement, as described –6– s 6320 m/sec 2 10 5 in natural units. There is' a in Refs. [23, 24]. Once the energy of the absorbed par- maximum frequency⇠ ⇥ ! = c k for phonons, where the ticle significantly exceeds the superconducting gap, the D s D maximum wavevector for lattice vibrations k 1/a absorption process is identical in the superconducting D is set by the lattice spacing a. For aluminum, ⇠! and normal phases of a metal. There are several ways D 0.037 eV; therefore the maximum phonon energy is rel-⇡ to absorb a particle (be it a photon or DM) in a metal. atively low, but the maximum momentum can be much One way is via impurities, where an o↵-shell electron pro- higher, ! /c keV. duced in the absorption process becomes on-shell through D s ⇡ interaction with an impurity. In the case of interest here, however, the target superconductor must be ultrapure in B. Dark Matter Absorption order to enable the collection and measurement of the created athermal excitations, and so this possibility is not viable. We now turn to computing the rate of DM absorption Instead, we make use of another process – that of par- in a material. The total DM absorption rate per unit ticle absorption on electrons through the emission of an mass per unit time R is athermal phonon in the final state, as shown in Fig. 1. 1 ⇢X R = neabsvrel , (3) The emitted phonon is required for momentum conser- ⇢ mX h i Outline

(I) First Course: Science Motivation

(II) Second Course: The Ricochet Detector

(III) Last Course: Ricochet & Chooz Neutrino Sources

• The variety of sources trade off ﬂux, energy and knowledge of spectrum.

Sources Pros Cons

Radioactive Mono-energetic, can < 1 MeV energies require place detector < 1m from very low (~10 eVnr) Sources (Electron source, ideal for sterile thresholds, limited half- Capture) neutrino search life, costly

Spectrum not well known below 1.8 MeV, site Nuclear Reactors Free*, highest ﬂux access can be difﬁcult, potential neutron background

Higher energies can use Spallation/Decay higher detector Prompt neutron ﬂux; thresholds, timing can large shielding or at Rest cut down backgrounds distances needed signiﬁcantly Neutrino Sources Neutrino Sources

• The variety of sources trade off ﬂux, energy and • The variety of sources trade off ﬂux, energy and knowledge of spectrum. knowledgeRelevant sources for of spectrum. CEνNS detec:on

§ The variety of sources trade oﬀ ﬂux, energy and knowledge of spectrum

1021 SNS: Reactors n 19 m 10 nm D

L ne s 17 10 Reactors: Electron capture source MIT reactor

MeV 5 MW

H 15

ê 10 Advanced Test

n SNS @ reactor 110 MW 1013 SanH OnofreL reactor 3.4 GW Flux H L 1011 EC Sources: 37Ar 5 MCi H L 109 0.5 1.0 5.0 10.0 50.0 H L En MeV

Ma0hieu Vivier - GdR Neutrino 2017 8

@ D The Early Ricochet Program

• Ricochet aims to make a first detection of coherent neutrino scattering using neutrinos from a nuclear reactor as its source. MIT Research • Hoping to demonstrate pulse shape Reactor discrimination from metallic superconductors for background suppression.

• Original plan to use the MIT Research Reactor (~5 MW power, 7 meters from core).

• At such proximity, neutron backgrounds could mimic a signal. Modest shielding can reduce this background to < 3 events/kg/day in ROI. Ricochet @ Chooz

• The option of moving to the Chooz near site complex was mentioned as a possibility, and gathered immediate attention†.

• Chooz two core reactors (8.25 GW power combined) in France is an ideal location.

• Almost zero neutron background from reactor. Allows for clean determination of a signal correlated with the reactor power.

• Infrastructure already exists. Existing good relation with power company.

The Chooz Reactors † The result of a conversation between Lindley Winslow and Jaime Dawson at the wine reception at Neutrino 2016 Ricochet Collaboration

Massachusetts Institute of Technology Rachel Carr Joseph Formaggio Joseph Johnson Alexander Leder Valerian Sibille Sarah Trowbridge Lindley Winslow IPNL Lyon University of Wisconsin, Madison Corinne Augier Julien Billard Jules Gascon Maryvonne De Jesus Kimberly Palladino Alexandre Julliard Romain Maisonobe Northwestern University Centre de Saclay, DRF/Irfu Enectali Figueroa-Feliciano Hong Ziqing Thierry Lasserre Matthieu Vivier Claudia Nones CSNSM Orsay Technical University of Munich Emiliano Olivieri Louis Dumoulin Xavier Defay Alex Langenkaemper Stefanos Marnieros Elizabeth Mondragon Université Paris Diderot Lothar Oberauer Stefan Schoenert Michael Willers Jaime Dawson A possible detector conﬁgura:on… Near Site Ricochet @ Chooz Low CEνNS rate @ near lab but…

§ No reactor induced backgrounds • An estimate of the potential sensitivity of § 120 mwe è strong reduc:on of Ricochet @ Chooz was quickly developed cosmogenicMakes induced use backgrounds of (see arXiv:1612.09035). Integral count rates at near lab existing lab § 1 m ultra pure infrastructurecopper for pit for an • Backgrounds: easier cryostat/detector handling 8400 t Pth background NCE⌫NS(T Tth)= • 2 2700 Internal backgrounds taken from 4⇡D ⇥ k ↵kEf,k § 3.5 m of water reduction.shielding to reduced measurements taken with the externalTmax(m background i,E⌫ ) Mdet P 1 d EDELWEISS Ge-detectors. Ai dE (E⌫ ) dT (mi, T, E⌫ ) ⇥ m1715 i ⇥ 0 T dT - Fission rate i Z § ZPossibilityth to install ac:ve shielding • X Muons and neutrons, we normalize- Isotope natural abundance (muon veto, etc…) GEANT 4 simulations to measured- Reactor ν spectrum 910 - CEνNS recoil spectrum rates at Chooz’s near site. Ma0hieu Vivier - GdR Neutrino 2017 23 -1 -1 Rates [kg d ] above energy threshold with B1,2 opera:ng at full power • Signal:

Tth Zn Ge CaWO4 • We assume a correlated signal with 10 eV 0.87 1.03 3.79 the power reactors (baseline of 20% 20 eV 0.78 0.92 3.19 ON, 80% OFF, but variations also studied). 50 eV 0.60 0.68 2.16 100 eV 0.41 0.46 1.38 • Systematics estimated at roughly 5%. 200 eV 0.22 0.23 0.80

Need a few kilograms of material to get a couple of events per day

Ma0hieu Vivier - GdR Neutrino 2017 20 Ricochet at Chooz 6

Detector Rate (per kg per day) Ricochet at Chooz Threshold Si Zn Ge Os 13 50 eV 0.32 0.66 0.76 1.25 Background and Signal 300 100 eV 0.26 0.46 0.51 0.55 Expected Total Counts Generated Data 200 eV 0.19 0.25Extracted 0.26 Fit 0.14

Table 2. The predicted CE⌫NS event rate for various detector materials and recoil threshold250 energies. Rates are calculated for a reactor power of 8.54 GW, a detector 235 239 241 distanceCounts per 12.2 days of 400 m, and ﬁssion fractions U = 55.6%, Pu= 32.6%, Pu = 7.1%, • An estimate of the potential sensitivity of and 238U = 4.7%. Ricochet @ Chooz was quickly developed (see arXiv:1612.09035). 200 we would expect 4.85 events/day with a 100 eV recoil threshold and full reactor • Backgrounds: ⇠ power. This rate increases by almost 50% if the threshold drops to 50 eV, placing a 150 • Internal backgrounds taken from high premium on the0 detector100 threshold one200 can achieve.300 measurements taken with the Time (days) EDELWEISS Ge-detectors. 2 10Figure 5. The blue envelope is the expected number of counts for the true reactor • CEvNS Muons and neutrons, we normalize power and background rate, with a 5% variation allowed in the signal. The black and 10 kg target ν-e GEANT 4 simulations to measured red points are respectively simulated data and the extracted ﬁt. The background is Compton 14.0 events/day and the signal is 5.0 evts/day. The reactor is modeled at full power Tritium rates at Chooz’s near site. 60% of the time and half power for 40%. 10 Pb-206 Betas • Signal: Neutrons Acknowledgments • We assume a correlated signal with 1 the power reactors (baseline of 20% We are grateful to the Heising-Simons Foundation for their generous support of the work presented here. We wish to thank the MIT Electronics Research Society for access to ON, 80% OFF, but variations also their computingEvent rate [evts/kg/keV/day] facilities. studied). 10−1 References • Systematics estimated at roughly 5%. [1] Freedman D Z 1974 Phys. Rev. D9 1389–1392 [2] Scholberg10 K 2006−2 Phys. Rev. D73 033005 (Preprint hep-ex/0511042) −3 2 1 2 [3] Formaggio J A,10 Figueroa-Feliciano10− E and10 Anderson− A J 20121 Phys. Rev.10D85 01300910 (Preprint 1107.3512) Recoil energy [keV] [4] Patton K, Engel J, McLaughlin G C and Schunck N 2012 Phys. Rev. C86 024612 (Preprint 1207.0693Figure) 1. The di↵erential CE⌫NS rate and projected background rates versus [5] McLaughlin G 2015 AIP Conf. Proc. 1666 160001 206 [6] Wilsondetector J R 1974 recoilPhys. Rev. energy. Lett. 32 The849–852 internal backgrounds, such as tritium and Pb, are based [7] Horowitzon observed C J, Perez-Garcia rates M as A, measured Carriere J, Berry by theD K and EDELWEISS-III Piekarewicz J 2004 Phys. experiment Rev. C70 [23] 065806 (Preprint astro-ph/0409296) [8] Scholberg K 2012 Ann. Rev. Nucl. Part. Sci. 62 81–103. (Preprint 1205.6003) [9] Billard J, Strigari L and Figueroa-Feliciano E 2014 Phys. Rev. D89 023524 (Preprint 1307.5458)

4. Background Estimates

Placement of the Ricochet experiment at the Chooz near reactor site –over 400 meters of rock and soil separating the detector from the reactor cores– has the unique advantage in that it reduces the reactor-correlated neutron and gamma backgrounds to undetectable levels. This allows for a clean separation of “neutrino-induced events”, which are correlated with the reactor thermal power, and “background events”, which ProjectedDiscovery Sensitivitypotential: Summary

Minimum exposure required to get a 95% probability to reach at least a 5 Minimum exposuresigma CEvNS required detection*. to get a 95% probability to reach at least a 5 sigma CEvNS detection

Background 100 [eVnr] 200 [eVnr] 300 [eVnr] 400 [eVnr] 500 [eVnr] rate

1.4 [/kg/day] 266 1123 4115 14148 33328 (nominal)

3.6 592 2676 10688 30750 56660 [/kg/day]

5.0 785 3660 16000 42100 76300 [/kg/day] *J. Billard

* Achievable with only 1 kg target material and exposition time less than 3 years * Achievable with only 10 kg target material and exposition time less than 3 years * Extrapolated This study includes all systematics: 5% on signal efﬁciency, 10% on all backgrounds except for the betas that are at 20% Confirmed detection with just 1 kg of target mass seems achievable! 15 ProjectedDiscovery Sensitivitypotential: Summary Discovery potential: Summary Minimum exposure required to get a 95% probability to reach at least a 5 MinimumMinimum exposuresigma CEvNS exposure required detection*. required to get a to 95% get aprobability 95%Systematics: probability to reach to all reachat least at leasta 5 sigma a 5 sigma CEvNS CEvNS detection detection 105 Background 100 [eVnr] 200 [eVnr] 300 [eVnr] 400 [eVnr] 500 [eVnr] rate 10 kg x 3 years 101.44 [/kg/day] 266 1123 4115 14148 33328 (nominal)

Exposure [kg.day] 1 kg x 3 years 3.63 10 592 2676 10688 30750 56660 [/kg/day] Back. Rate: 1.4 evt/kg/day

Back. Rate: 3.6 evt/kg/day 5.0 785 3660 16000 42100 76300 [/kg/day]2 10 Back. Rate: 5.0 evt/kg/day *J. Billard

* Achievable with only 1 kg target material and exposition time less than 3 years * Achievable with only 10 kg target material and exposition time less than 3 years * Extrapolated 0 50 100 150 200 250 300 350 400 450 500 550

This study includes all systematics: 5% on signal efﬁciency, 10% onThreshold all backgrounds [eVnr] except for 8 the betas that are at 20% Confirmed detection with just 1 kg of target mass seems achievable! 15 Projected Sensitivity

DiscoveryMinimum exposure potential: required to get signiﬁcance a 95% probability to reach at least a 5 sigma CEvNS detection*.

median value and 95% contours

*J. Billard

Confirmed detection with just 1 kg of target mass seems achievable! “Very” Near Site Ricochet @ Chooz CEνNS at a very near site (< 100m)

Near lab

• A new location is being explored. Half-way bewteen B1 and B2 Integral count rates at a very near site Flux > 1012 cm-2 s-1

• The Near site (400 m) t P N (T T )= th CE⌫NS th 4⇡D2 ⇥ ↵ E • The “Very Near” site (80 m)! k k f,k 160 m Tmax(mi,E⌫ ) Mdet P 1 d Ai dE (E⌫ ) dT (mi, T, E⌫ ) ⇥ mi ⇥ 0 T dT - Fission rate i Z Z th - Isotope natural abundance X • Would imply a x30 increase in flux. - Reactor ν spectrum Ma0hieu Vivier - GdR Neutrino 2017 25 Allows one to deploy smaller - CEνNS recoil spectrum detectors and/or loosen requirements Rates [kg-1 d-1] above energy threshold at 80 m from B opera:ng at full power on background levels. 1,2

Tth Zn Ge CaWO4 10 eV 20.68 24.46 89.67 • Could provide a signal in far less than 20 eV 18.50 21.68 75.42 a year. 50 eV 14.12 16.18 51.05 100 eV 9.72 10.80 32.66

200 eV 5.19 5.49 18.86

Rates x 30 with respect to Double Chooz near lab Need a few 10 to 100 grams of material to get a couple of events per day Ma0hieu Vivier - GdR Neutrino 2017 26 Sensi:vity at a very near site

§ A simple (but unrealis:c) baseline scenario: o Single detector with mass 0.1 kg

o Region of interest for signal detec:on: recoil events from Tth= 50 eV up to 2 keV o Two kinds of background “modeling” in the RoI: flat or exponen/al o Background rate is assigned a 20% uncertainty o Detec:on efficiency is assigned a 5% uncertainty “Very” Near Site Ricochet @ Chooz o Reactor and νe spectra uncertainty embedded in an addi:onal 5% signal normaliza:on uncertainty Differen/al recoil spectra 3 10 Ge Rate + shape significance Zn CaWO4 2 12 10 Flat +exponential background ]

Ge 1 −

kg -1 -1 Zn 1 1 − 10 RBck[RoI]= 500 kg d

10 CaWO4 keV 1 Assumes:− 0 10 Detectors placed 80 m • A new location is being explored. from the cores

8 d2N/dtdE [d ) −1 σ Integral count rates at a very near site 10010 gram payload

−2 10 6 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 • Recoil energy [keV] The Near site (400 m) t Pth 50 eV threshold NCE⌫NS(T Tth)= Significance ( 4⇡D2 ⇥ ↵ E • The “Very Near” site (80 m)! 4 k k f,k Tmax500(mi ,bkg/day/kgE⌫ ) Mdet P 1 d Ai dE (E⌫ ) dT (mi, T, E⌫ ) 2 ⇥ mi ⇥ 0 T dT - Fission rate i Z Z th - Isotope natural abundance X Can relax background • Would imply a x30 increase in flux.- Reactor ν spectrum 0 speciﬁca/ons by a factor 50 Allows one to deploy smaller - CEνNS recoil spectrum 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 … Time [years]

detectors and/or loosen requirements -1 -1 Rates [kg d ] above energy threshold at 80 m from B1,2 opera:ng at full power on background levels. Ma0hieu Vivier - GdR Neutrino 2017 27 Tth Zn Ge CaWO4 10 eV 20.68 24.46 89.67 • Could provide a signal in far less than 20 eV 18.50 21.68 75.42 a year. 50 eV 14.12 16.18 51.05 100 eV 9.72 10.80 32.66

200 eV 5.19 5.49 18.86

Rates x 30 with respect to Double Chooz near lab Need a few 10 to 100 grams of material to get a couple of events per day Ma0hieu Vivier - GdR Neutrino 2017 26 Current Status

Dilution Refrigerator:

• MIT currently has R&D funding from the Heising-Simons Foundation. This pays for the dilution refrigerator + personnel and some detector work.

Detectors (Zn Crystals):

• Working with RMD (Watertown, MA) for development of low background, high purity Zn crystals. Northwestern working on TES chips.

• Working with Lincoln Labs to fabricate large SQUID arrays. Making use of high quality fabrication facilities for SQUID design.

Operations:

• Could perform a run at MITR to determine recoil discrimination and ability to extract reactor ON/OFF data.

• Timeline for entering Chooz (~2019). Our new fridge, Desperaux! Current Status

Dilution Refrigerator:

• MIT currently has R&D funding from the Heising-Simons Foundation. This pays for the dilution refrigerator + personnel and some detector work.

Detectors (Zn Crystals):

• Working with RMD (Watertown, MA) for development of low background, high purity Zn crystals. Northwestern working on TES chips.

• Working with Lincoln Labs to fabricate large SQUID arrays. Making use of high quality fabrication facilities for SQUID design.

Operations:

• Could perform a run at MITR to determine recoil discrimination and ability to extract reactor ON/OFF data.

• Timeline for entering Chooz (~2019). Our new fridge, Desperaux! First Pulses!

• Single zinc crystal instrumented in Lyon with NTD now tested. Pulses created by particles clearly seen with system completely un-optimized!

• Tests will continue with new RMD crystals (higher quality and purification). Will also begin testing with TES+SQUID setup Particle pulse this fall. with Zn bolometer

• Begin testing energy resolution and pulse shape discrimination in coming months. Summary Summary

“Never skip a meal!” Summary

“Never skip a meal!”

After forty years, we are finally at the point where coherent neutrino scattering is detectable. This opens a myriad of doors in the ability to explore new physics and even in applications.

Ricochet is quickly building as an experiment with fast sensitivity to first CEνNS detection once installed at the Chooz near site using promising and proven bolometric technologies.

Could open the the door for a wide range of physics beyond the Standard Model. Buon Appétit!

Thank you for your attention. Science Goal: First Detection

• Why the delay in detection? E⌫ • Simple Answer: It’s hard to detect. Tmax = 1+ MA • Signature is the recoil (thermal) energy of the neutrino 2E⌫ impacting on the nucleus.

• Detectors need to be able to detect very low energy recoils (100 eV - 1 keV) to ensure detection. Only Recoil Energy until recently has technology advanced sufficiently to (max) enable detection. Element 3 MeV 30 MeV • If realized, this opens a lot of doors for physics. Ar 500 eV 50 keV

Benefits: Ge / Zn 250 eV 25 keV • First detection of this predicted process. • Applications to other science goals (sterile neutrino searches, dark matter detection, nuclear monitoring). Os 100 eV 10 keV

Challenges: • Backgrounds (in particular neutron recoils). Energy range Energy range • Extremely low recoil energy deposited as its signature. for reactors for spallation sources 2

The elastic neutrino cross section di↵erential with respect to the outgoing neutrino solid angle can be written as

d 1 2 2 2 1 = G "0 cos (✓/2) f , (2) d⌦ 2⇡2 F ⌫ rec R ✓ ◆(⌫,⌫) e where stands for the square of the WNC matrix element of the scattering process, namely the contraction of the correspondingR leptonic and hadronic tensors (see later for the normalization chosen). It ise also useful to express the cross section in a form that is di↵erential with respect to the target recoil energy (equal to the energy transfer), related to the previous expression through a Jacobian,

d d = J(⌦, !) , (3) d! d⌦ ✓ ◆(⌫,⌫) ✓ ◆(⌫,⌫) which is given by

A unified approach to electron and neutrinod⌦ 2⇡ elastic(MA + scattering!) frec o↵ nuclei with an applicationJ(⌦ to, ! the)= study= of the axial structure. (4) d! k⌫ k⌫0 (1 + !/MA) O. Moreno1 and T. W. Donnelly1 In these expressions "⌫ and "⌫0 are the initial and final neutrino energies, respectively (k⌫ and k⌫0 the corresponding 1Center for Theoretical Physics, Laboratory for Nuclear Science and Department of Physics, momenta), ✓ is the neutrinoMassachusetts scattering angle, InstituteM ofA Technology,is the target Cambridge, mass MA and 02139,frec USAis a kinematic recoil factor. The di↵erential neutrino cross sections imply(Dated: the June detection 16, 2015) of the recoiling energy or momentum (magnitude or direction) of the targetWe show with a relationship reasonable between precision; elastic electron if, on scattering the contrary, observables the and the detectors elastic neutrino have a large energy acceptance from a minimum valuecross section (!m that, given provides by a straightforward the detector determination threshold), of the up latter to from a maximum experimental valuedata of (!M , given by the specific the former and relates their uncertainties. An illustration of this procedure is presented using a kinematic conditions),Hartree-Fock what mean is actually field for the measured nuclear structure is of a set of even-even nuclear targets, using the spectra of the neutrinos produced in pion decay at rest. We also analyze the prospects to measure the incoherent axial contribution to the neutrino elastic!M scatteringd in odd targets. (⌫,⌫)(!m)= d!. (5) PACS numbers: 12.15.Mm, 24.80.+y, 25.30.Bf, 25.30.Pt d! Z!m ✓ ◆(⌫,⌫)

The matrixIn lepton-nucleus element squared elastic scattering in Eq. (2) the particularizedincident and the outgoing to coherent lepton neutrino is the same scattering and its energy is loss= ! iscoh,with transformed entirely into kinetic energy of the recoiling nuclear target; we denote the process as (⌫, ⌫) forR neutrinosR (of any flavor) and (e, e) for charged leptons (again of any flavor, but electrons being of most experimental interest). = V (F VV,T=0)2 , (6) Coherent scattering is a particular case of elasticRcoh scatteringL whereCC, all J of=0 the nucleons in the target contributee to thee cross section through the vector Coulomb monopole isoscalar form factor of the nucleus, which is, unlike the rest of VV,Tincoherent=0 elastic form factors, proportional to the number of nucleons. Coherence applies for momentum transfers where FCC, J=0 is the WNC Coulomb monopolee vector1/3 isoscalare form factor, normalized so that in the long wavelength corresponding to nuclear-size wavelengths, q 160 A MeV, and below; for larger values the Coulomb form factor limit (LWL),decreasesi.e. and, as the the incoherent momentum elastic form transfer factors,⇠ goes when to possible zero, (see it becomes below), become comparable. e Elastically scattered charged leptons can be easily detected, but in the case of neutrinos the proposed observable is the recoil energy of the nuclear target throughVV,T the=0 ionization induced in the2 detector. Elastic neutrino scattering o↵ F (q 0) A sin ✓W , (7) nuclei can be exploited to determine electroweakCC, parameters J=0 ! at very! low momentum transfers, to test the universality of the weak interaction for charged and neutral leptons, or to estimate the escape rate of neutrinos created in a 2 where A varietyis the of target stages mass of star number evolution. and These✓We motivationsis the weak support mixing recent angle, experimental sin ✓W proposals0.23. to measure The same neutrino normalization for the elastic scattering, such as the neutrino program at SNS-ORNL [1] and the analysis of sensitivity⇡ to this process of p n full CoulombFun Fact! monopole form factor (isoscalar plus isovector) in LWL yields the nuclear weak charge W = ZV +N V , pseveral neutrino and2 dark matter detectors [2].n Q where V Parity-violating=0.5 2 sin (PV)✓W elastic0 electron.04 and scatteringV = is another0.5 are nuclear the electroweak proton and process neutron that has WNC drawn much vector atten- coupling constants, ⇡ ⌫ 2 2 2 respectivelytion recently. [20]. The The Rosenbluth usual observable factor is the parity-violatingVL in the extreme asymmetry, relativistic defined as the limit relative (ERL) di↵erence is VL between= ↵ the(1 ! /q ) where ↵⌫ = (across⌫ )2 +( sectionsa⌫ ) of2) electrons/2 is a with combination spin projection of parallel neutrino (same WNC direction, couplingh = +1) constants, and antiparallel with (opposite (a⌫ )2 direction,=(a⌫ )2=1intheSM. hA= 1) toV their momentum: A V ⇥ ⇤ h=+1 h= 1 d d d⌦ d⌦ I. RELATIONSHIP BETWEEN ELECTRON(e,e) = h=+1 AND h NEUTRINO= 1 COHERENT CROSS(1) SECTIONS • It is possible to relate the cross- A d d d⌦ + d⌦ section for coherent neutrino The elasticscatteringMeasurements electron directly to of crosselectron parity-violating section, elastic elastic the electron parity-violating scatteringParity o↵ asymmetrynuclei violating can be asymmetry used in for elastic precise electron tests of the scattering Standard and the elastic scatteringModel (SM), (Moreno including and Donnelly, the 2015) evaluation. of the weak mixing angle or of higher-order radiative corrections, as well as neutrinoto cross determine section the neutronfor even-even radii of nuclei nuclear [3], with targets implications fulfill the to the following neutron-rich relationship: matter equation of state and to the structure of neutron stars. Recent or planned experimental e↵orts such us PREX I and II, using 208Pb [4], and 48 arXiv:1506.04733v1 [nucl-th] 15 Jun 2015 CREX, using Ca [5], have focused on the extractiond of the neutron2 radiid of the target nuclei with precisions as good • Allows for a direct comparison to = , (8) as 1.2% in the PV observable. There has alsod⌦ been recent interestA(e,e) in relativelyd⌦ low-energy electron beams for studies electronof PV electrondata (typically scattering, known such to as the MESA✓ accelerator◆(⌫,⌫) at Mainz [6]✓ or an◆ upgraded(e,e) version of the FEL at Je↵erson betterLab [7],than aimed 1%). at Deviation tenths of would percent precision in the PV measurements. where theimplyThe cross new dominant physics sections electron-nucleusbetween and the electron asymmetry scattering process are evaluated is overwhelmingly at the an same electromagnetic kinematic (EM) conditions one and (incident therefore momentum and v-e Correspondence scatteringandparity-conserving angle)neutrino andcouplings the (PC). with ERL On the the for Z. other the hand, leptons the weak has neutral been assumed.current (WNC) An is additional responsible for factor the parity of WNC violation leptonic couplings, in⌫ electrone 2 scattering, since it contains vector and axial components that behave di↵erently under inversion of spatial namely ↵coordinates,/(aA) , and has it isbeen also responsible particularized for neutrino-nucleus to its SM scattering. value of The 1; probabilities we note in of PC passing electron, that PV electron the neutrino and scattering on ⌫ 4 2 ⌫ 10 4 which weneutrino are focused scatterings is insensitive follow approximately to the the values ratio of 1 :a 3V10and q a:3A 10independently, q ,withq the and characteristic therefore momentum to the possible Majorana transfer⌫ of the process in GeV. In what follows we consider· the exchange· of a single gauge boson for each of the nature (aV = 0) of the neutrinos. interactions involved: one Z0, one photon, or one of each; we also neglect the distortion of the electron wave functions The relationshipdue to the nuclear in Eq. Coulomb (8) is field, valid although for any in neutrino practice it flavor is usually and taken for anyinto account. charged These lepton two flavor conditions (as are long as the ERL still holds), andknown for as leptons plane wave as Born well approximation as for antileptons (PWBA). in any combination, always within PWBA. For non even-even, J =0 6 Going Cold…

• Given historical precedent, we focus on Transition Edge Sensors (TES) as the 4k T 2C +1 B tot technology to push down to the 10 eV E ⇡ s ↵ r 2 scale.

• Energy resolution dominated by the total heat capacitance of system (Ctot).

• At 15 mK, a 10 eV threshold could be achieved with a system capacitance of Ctot < 300 pJ/K.

Impurities

TES (α T)

Lattice (α T 3) Capacitance sources Capacitance 6 4 Energy Threshold for Bolometric Detectors Going Cold… 2 Mo/Au TES 100 8 Si 20 g 6 Si 30 g 4 Si 40 g Si 50 g • Given historical precedent, we focus on Si 75 g Transition Edge Sensors (TES) as the 2 Si 100 g Ge 20 g technology to push down to the 10 eV 10 Ge 30 g 8 Ge 40 g scale. 6 Ge 50 g 4 Ge 75 g Ge 100 g Experimental Energy Threshold [eV] • Energy resolution dominated by the 2 2 3 4 5 6 7 8 9 10 100 total heat capacitance of system (Ctot). TES Tc [mK]

• At 15 mK, a 10 eV threshold could be achieved with a system capacitance of Ctot < 300 pJ/K.

Impurities

TES (α T)

Lattice (α T 3) Capacitance sources Capacitance Phonon-Light Technique Repurposing DM Detectors light phonon CRESST-III Detector Prototype e-/γ

CaWO4 sticks (with holding clamps) light yield light reﬂective and scintillating housing

light detector (with TES) neutrons

block-shaped target crystal (with TES) LD phonon CRESST-III Detector Prototype First modules ready PD

Reduced light output for highly-ionizing parKcles Quenching Raimund Strauss, MPI Munich 18 CRESST (Type 3) Detectors: Raimund Strauss, MPI Munich 11 CaWO sticks 4 CaWO4 crystals with TES readout. (with holding clamps)

reﬂective and scintillating housing Separation of recoil from electromagnetic events using light and heat signatures. light detector (with TES)

New 24 gram detectors reach ~50 eV threshold. block-shaped target crystal (with TES) LD Low radioactivity / impurities.

First modules ready PD

Raimund Strauss, MPI Munich 18 First Results of CRESST-III Detector Phonon-Light Technique Repurposing DM Detectors light phonon Preliminary Promising results: CRESST-III Detector Prototype CRESST-III prototype Improvement by e-/γ factor E=5.9keV 6.2 (S/N) compared to Cryostat 1 Munich best CRESST-II detector

CaWO4 sticks (Eth = 307eV) (with holding clamps) light yield light reﬂective and CRESST-II “Lise” scintillating housing à E=5.9keV Baseline noise @GS light detector (with TES) Cryostat Gran Sasso 1.8-3.0mV RMS neutrons à Threshold: block-shaped target crystal (with TES) Eth= 45-60eV LD phonon CRESST-III Detector Prototype First modules ready PD

Design goal (EReduced light th=100eV) for output forCRESST-III highly-ionizingPhase 1 probably exceeded! parKcles Quenching Raimund Strauss, MPI Munich 18 CRESST (Type 3) Detectors:

Raimund StraussRaimund Strauss, MPI Munich, MPI Munich 20 11 CaWO sticks 4 CaWO4 crystals with TES readout. (with holding clamps)

reﬂective and scintillating housing Separation of recoil from electromagnetic events using light and heat signatures. light detector (with TES)

New 24 gram detectors reach ~50 eV threshold. block-shaped target crystal (with TES) LD Low radioactivity / impurities.

First modules ready PD

Raimund Strauss, MPI Munich 18 Exploring Metallic Super-conductors

Top: Image of • Metallic super-conductors (zinc, rhenium, metallic super- osmium) used at very cold (10-50 mK) conductor should quench responses to (MARE electromagnetic radiation (X-rays, collaboration) gammas) while retaining recoil sensitivity.†

Bottom: Benefits: Efficiency of • Having a “background-free” detector, electromagnetic which is solely sensitive (or preferentially signals (alphas) sensitive to) recoil events. as a function of • Certain metals (e.g. Al, Zn, Os) easier to temperature (in fabricate as thermal detectors. fractions of • Applications to other types of recoil Debye detectors. temperature) for various metals, Challenges: illustrating quenching. • Only quenching response observed so far. Sensitivity to recoils previously unexplored. Journal of Low Temperature Physics November 1993, Volume 93, Issue 3-4, pp 263-268 Further results on μ-calorimeters with superconducting absorber E. Cosulich, F. Gatti, S. Vitale Thermal Techniques Thermal Detectors

• Exploit that energy deposited in a solid will lead to thermalization (i.e. a change in the temperature).

• At ultra-low temperatures,the heat capacity is dominated by the lattice contribution. Allows essentially the counting of phonons, rather than photons or other particles. Final States • The available statistics for the number of thermalized phonons allows one to C(T)ΔT= ΔE e-t/τ access very low energies with exquisite energy resolution. Minimizing heat capacitance leads to greater energy sensitivity Thermal Techniques

Transition • Typical thermal detector readouts involve Edge Sensor either neutron-transmutation doped (NTD) devices, Microwave Kinetic Induction Detectors (MKIDs) or Transition Edge Sensors (TES). Small change in • TES devices work at the edge of the temperature superconducting-to-normal transition, leads to large change thus a small change in temperature in resistance becomes a large change in the resistance.

• Can measure the change in resistance Sample inductively (usually using SQUIDs, though resolution other techniques also in use). from low temperature bolometers (ECHO experiment) Target & Cryostat

2SARAHHEINEThe Ricochet CAD Model and cryostat THE RICOCHET COHERENT NEUTRINO SCATTERING EXPERIMENT 3 • We currently envision an expandable mass of 1 kg (expandable to 10 kg) payload of mixed detector technologies: metallic Zn (50%) and semi-conductor Ge (50%) crystals.

• Two targets have quite similar neutrino responses. The combination mitigates the risk associated with each technology and allows for growth and expandability.

Figure 1. CAD visualizationVibrational of the dilution profile refrigeratorFigure of Cryoconcept 2. forThe the Cryoconcepts Ricochet dil fridge, experiment. dilution from refrigerator EDELWEISS with dual-frame vibra- • Our refrigerator is designed by tion isolation for the Ricochet experiment. Delivery of the fridge to MIT for commissioning is expected in late February 2017 Cryoconcepts and utilizes a uniquebefore shipment to the US, is shown in Figure 2. The footprint is 1.15 meters in diameter (as shown in Figure 3 ) with a height of 1.9615 meters. space requirement is not terribly large compared to the space required for the fridge and double-framed design to isolate theThe fridge has several requirements to service its pumps and power needs. These have its pumps so it is not anticipated to be a problem. been defined by Cryoconcepts as follows: pulse tube from the rest of the (1) 200 V, 3 phase power with ground2.4. andSafety. neutralThere (5 wires) are relatively with wall few receptacle safety concerns in- for a cryogen-free system like the refrigerator and minimize the transferstalled of and ready dilution refrigerator we plan to use for Ricochet. The system will be securely mounted (2) Plug for PT compressor (3 phasesas with described ground-4 in Section wires) 2.2.with We wall will receptacle need a smallin- amount of liquid nitrogen (LN2) for vibration into the cold stage of the stalled and ready the cold trap of the dilution refrigerator. LN2 is extremely safe when used carefully. It (3) Water supply for cooling-flow 7L pershould minute be transferred at 20 degrees using C for vacuum PT compressor transfer lines to reduce spraying cryogens and cold experiment. (4) Water supply for cooling-flowE. 1LOlivieri perequipment, et minute al. NIM-A at and 20 (2017) personal degrees C protective for turbo equipment pump (PPE) like cryogenic gloves, closed shoes (5) Compressed air minimum 5 bars standardand long pants shop airand or sleeves equivalent should compressed be utilized while gas handling the nitrogen. Also, in a space source where any reasonably large amount (such as a standard 240 liter dewar) of cryogen is stored, an oxygen alarm should be installed in the room to notify personnel in the event 2.3. Requirements for Long Term Operation.of a cryogenSince leak the that dilution could refrigerator displace oxygen is cryogen- in a contained space and prove hazardous. free it should, in theory, be able to run for longThe periods only other of time safety without concerns warming for the up, experiment how- are standard to any lab environment: ever as in all cryogenic experiments, occasionallyeliminating/labeling there will be trip need hazards, to open proper the fridge use and to disposal of any chemicals used (likely perform maintenance or changes to the experimentjust isopropyl section, alcohol, wiring, acetone, or cooling and deionized apparatus. water that will be used for cleaning parts This will require cleanroom-style lab spacebefore in which they’re to installed open sensitive in the dewar), boxes, and and enough caution used when moving large parts of the ex- room in the experiment space to be able toperiments open the likefridge pumps. and store Grounding the outer will layers. also be This assured properly throughout the experiment Detector & Signal Chain

6SARAHHEINE

THE RICOCHET COHERENT NEUTRINO SCATTERING EXPERIMENT 9

• Detectors: single-Zn crystals, 25-50 g each (10-100 units needed, eventually).

• Has Tc of 0.85 K. Planning to SA IC SA initially operate at 50 mK but will push down to ~10 mK operation.

Figure 4. Schematic of a single crystal thermal readout. The current SQ2 IC design calls for the silicon detector chip to be epoxied directly to the crystal, • Readout: Phonon readout via THE RICOCHET COHERENT NEUTRINO SCATTERING EXPERIMENT 7 however in the future this chip may be moved to a separate location so that Transition Edge Sensor (TES), the Zn crystals can be kept at a lower temperature than the TES requires. with SQUID amplification. of the TES using Superconducting Quantum Interference Devices (SQUIDs). The temperature of the TES need not be held exactly at its transition. In fact, it is ideal to hold the temperature of the TES below its transition and use a voltage bias to drive it into the superconducting/normal transition. This allows stabilization of the TES temperature • TES “chip” currently under through electrothermal feedback (as the temperature rises, the resistance does as well, low-

SQ1 IC ering the current/thermal load on the TES and leading to cooling back to an equilibrium state in the transition) [6]. design by T. Figueroa-Feliciano SQUIDs are cryogenic devices made up of two parallel Josephson junctions in a superconducting loop [7]. When a SQUID is biased above a critical current level it becomes at Northwestern. resistive, causing a voltage drop to form across it. This voltage drop oscillates with in- creasing magnetic ﬂux, producing a sine-like signal that is nearly linear in between extrema. Typically a feedback loop will be used to ‘lock’ the SQUID in this region. We are planning on using a three-stage, multiplexed SQUID system to read out our roughly 100-200 detec- • SQUID array based on MicroX tors. We will be basing the design of this SQUID system on the design of that used to read design.

Figure 5. Schematic of the planned silicon detector chip. The meander will be designed to achieve a conductance from the crystal to the cold bath through the detector chip of at least 10-100 times that from the crystal to the cold bath through the mount so that a large fraction of the thermal signal travels through the TES.

Figure 7. Circuit diagram for the SQUID readout of the Micro-Xout the exper- 128 TESs in the Micro-X rocket (provided by NIST Boulder) [10], [5]. We may iment. The temperatures of the various components are labelledmake and modifications an to this design to follow more recent versions of the system described in [9] that are less sensitive to external magnetic field fluctuations. A circuit diagram from example of the bias signal used to multiplex are shown to the left.Micro-X system is shown in Figure 7. The detailed design and fabrication of the SQUIDs will likely be completed at Lincoln Labs. As previously mentioned, TESs are biased with a constant voltage, and the current running through them changes depending upon their resistance, the signal of interest. The change in current through the TES creates a change in flux through its corresponding first stage SQUID (SQ1), which is flux-coupled to a second stage SQUID (SQ2). Both the SQ1 and SQ2 are located on the cold stage along with the TESs in the Micro-X experiment. The signal from the SQ2 will run up cabling to a warmer stage (in Micro-X the stage is kept at 2 K) where it is further amplified through a SQUID array (a phase-coupled array of 100 SQUIDs that acts as a signal amplifier by inductive coupling and can be biased in the same way as an individual SQUID). CEνNS at a very near site (< 100m)

Near lab Ricochet @ Chooz

Half-way bewteen B1 and B2 Flux > 1012 cm-2 s-1

• The option of moving to the Chooz near site complex was mentioned as a possibility, and gathered immediate attention†. 160 m

• Chooz two core reactors (8.25 GW power combined) in France is an ideal location. Ma0hieu Vivier - GdR Neutrino 2017 25 • Almost zero neutron background from reactor. Allows for clean determination of a signal correlated with the reactor power.

• Infrastructure already exists. Existing Event Rate (per kg per day in ROI) good relation with power company. Chooz Very Threshold MITR Chooz Near Near • Two possible locations being explored: 50 eV 1.4 0.7 16.9 • The Near site (400 m) 100 eV 1.0 0.5 11.5 • The “Very Near” site (80 m)!

200 eV 0.5 0.3 6.1