Calorimetric readout of Superfluid 4He for sensitivity to dark matter of keV-MeV mass

Scott Hertel (U. of Massachusetts Amherst) Dan McKinsey, Junsong Lin, Vetri Velan, Andreas Biekert (UC Berkeley)

Workshop: Table-Top Experiments with Skyscraper Reach MIT, Aug 10th, 2017 understanding the WIMP as one model among many

1020 10-25 The basic WIMP hypothesis electroweak 1018 mass scale 10-27

(neutrino-like interactions) ] 1016 10-29 2 ) 1014 10-31 2 Z-mediated 1012 10-33

1010 10-35

108 10-37

106 10-39

104 10-41

2 -43

WIMP-nucleon cross section ( zb ) 10 10 WIMP-nucleon cross section ( cm 100 10-45 WIMP-nucleon cross section [ zb ] 10-2 10-47 WIMP-nucleon cross section [ cm

10-4 keV MeV GeV TeV understanding the WIMP as one model among many

1020 10-25 The basic WIMP hypothesis electroweak 1018 mass scale 10-27

(neutrino-like interactions) ] -ruled out by ~1991 1016 10-29 2 ) 1014 10-31 2 Z-mediated 1012 Gotthard 1990 10-33

1010 10-35

108 10-37

106 10-39

104 10-41

2 -43

WIMP-nucleon cross section ( zb ) 10 10 WIMP-nucleon cross section ( cm 100 10-45 WIMP-nucleon cross section [ zb ] 10-2 10-47 WIMP-nucleon cross section [ cm

10-4 keV MeV GeV TeV understanding the WIMP as one model among many

1020 10-25 The basic WIMP hypothesis electroweak 1018 mass scale 10-27

(neutrino-like interactions) ] -ruled out by ~1991 1016 10-29 2 ) 1014 10-31 2 We keep digging down at this Z-mediated 12 Gotthard 1990 -33 same mass range... 10 10 1010 10-35

108 10-37

106 10-39

104 10-41

2 -43

WIMP-nucleon cross section ( zb ) 10 10 WIMP-nucleon cross section ( cm 100 LUX 2016 10-45 WIMP-nucleon cross section [ zb ] 10-2 10-47 WIMP-nucleon cross section [ cm

10-4 keV MeV GeV TeV understanding the WIMP as one model among many

1020 10-25 The basic WIMP hypothesis electroweak 1018 mass scale 10-27

(neutrino-like interactions) ] -ruled out by ~1991 1016 10-29 2 ) 1014 10-31 2 We keep digging down at this Z-mediated 12 Gotthard 1990 -33 same mass range... 10 10 1010 10-35 ...meaning, we keep 108 10-37 broadening the interaction type prior to increasing un-natural 106 10-39 versions of weak interaction. 104 10-41

2 -43

WIMP-nucleon cross section ( zb ) 10 H-mediated 10 WIMP-nucleon cross section ( cm 100 LUX 2016 10-45 WIMP-nucleon cross section [ zb ] 10-2 10-47 WIMP-nucleon cross section [ cm loop-mediated 10-4 keV MeV GeV TeV understanding the WIMP as one model among many

1020 10-25 The basic WIMP hypothesis electroweak 1018 mass scale 10-27

(neutrino-like interactions) ] -ruled out by ~1991 1016 10-29 2 ) 1014 10-31 2 We keep digging down at this Z-mediated 12 Gotthard 1990 -33 same mass range... 10 10 1010 10-35 ...meaning, we keep 108 10-37 broadening the interaction type prior to increasing un-natural 106 10-39 versions of weak interaction. 104 10-41

2 -43

WIMP-nucleon cross section ( zb ) 10 H-mediated 10 WIMP-nucleon cross section ( cm 0 LUX 2016 -45 Not sustainable for ever: 10 solar-ν 10 WIMP-nucleon cross section [ zb ]

1. increasingly un-natural WIMP-nucleon cross section [ cm 10-2 10-47 2. unavoidable backgrounds loop-mediated atm-ν 10-4 keV MeV GeV TeV Time to broaden other priors? understanding the WIMP as one model among many

One way to loosen the priors: 1) retain assumption of thermal production 2) stop assuming we already know all the force mediators

First order effects: - more unknowns (more theories) - thermal production works down to ~keV mass scale (note: new ‘freeze-in’ modes) - light mediators avoid standard collider searches understanding the WIMP as one model among many

1020 10-25 thermal production electroweak 18 -27 10 new mediator mass scale 10 ] 1016 10-29 2 ) 1014 10-31 2

1012 10-33

1010 10-35

108 10-37

106 10-39

104 10-41

2 -43

WIMP-nucleon cross section ( zb ) 10 10 WIMP-nucleon cross section ( cm 100 10-45 WIMP-nucleon cross section [ zb ] 10-2 10-47 WIMP-nucleon cross section [ cm

10-4 keV MeV GeV TeV DM kinetic energies

DM particle velocities cut off by (local) escape velocity: vmax ≃ 540 km/s

2 KEmax ≃ 1/2 mDM vmax

107 106 105 proportionality of 104 mDM to KEmax 103 [eV] 2 2 10

MeV/c ➔ eV max 2 101

GeV/c ➔ keV KE TeV/c2 ➔ MeV 100 -1 10 at cutoff DM KE 10-2 10-3 103 104 105 106 107 108 109 1010 1011 1012 1013 M [eV/c2] DM nuclear recoil energies when mDM ≃ mtarget , efficient coupling of KEDM into target order-GeV mass ➞ order-keV recoil endpoint energies order-MeV mass ➞ order-meV recoil endpoint energies

107 106 Xe 105 Si He 104 103 102 101 100 -1 10 at cutoff DM

Elastic Recoil Endpoint [eV] KE 10-2 10-3 103 104 105 106 107 108 109 1010 1011 1012 1013 M [eV/c2] DM nuclear recoil energies when mDM ≃ mtarget , efficient coupling of KEDM into target order-GeV mass ➞ order-keV recoil endpoint energies order-MeV mass ➞ order-meV recoil endpoint energies

107 106 Xe 105 Si He 104 103 102 101 100 -1 10 at cutoff 100meV He endpoint DM

Elastic Recoil Endpoint [eV] -2 KE 10 5meV Ge endpoint 10-3 103 104 105 106 107 108 109 1010 1011 1012 1013 M [eV/c2] DM nuclear recoil energies

when mDM ≃ mtarget , efficient coupling of KEDM into target

order-GeV mass ➞ order-keV recoil endpoint energies order-MeV mass ➞ order-meV recoil endpoint energies

107 106 Xe 105 Si He two punchlines: 104 103 1) light target particle desirable 102 2) meV-scale excitations desirable 101 100 -1 10 at cutoff 100meV He endpoint DM

Elastic Recoil Endpoint [eV] -2 KE 10 5meV Ge endpoint 10-3 103 104 105 106 107 108 109 1010 1011 1012 1013 M [eV/c2] DM the recent cresst/ν-cleus example arXiv:1707.06749v2

Al2O3 bronze balls clamp wire bonds copper Al2O3 + TES holder 5x5x5mm (0.49g)

20 eV threshold (on oxygen ➞ 140 MeV mass threshold) 2.27h exposure above ground, no shielding, with ~0.2Bq Fe55 source

my two cents: we are entering new regime - sensitivity: energy threshold matters more than target mass - backgrounds: dark counts & noise matter more than radiogenics counts [5eV binning] counts eV ie, we are entering the tabletop regime of this workshop Excitations in superfluid 4He

eV-scale excitations:

Excimers (He2*) He He* singlet: ~ns halflife (observable as scintillation) triplet: 13s halflife (observable as ballistic molecules) (+ a little IR from excitations to higher atomic states)

meV-scale excitations: phonons, R- rotons, R+ rotons (observable as athermal evaporation) Excitations in superfluid 4He : partitioning Nuclear Recoil 1

0.8

0.6 4He: atomic cross sections well measured, well understood. 0.4 Recoil energy partitioning can be estimated from the ground up. Energy Fraction 0.2

0 101 102 103 104 105 Recoil Energy [eV] NR and ER have quite different partitioning in a three-way Electron Recoil George Seidel, unpub. partition (kinetic + triplet + singlet). 1 Singlet Excimers 0.8 Triplet Excimers IR photons Phonons/R-/R+ 0.6 Beauty of calorimetric sensors: All recoil energy appears as observable excitations. 0.4

Energy Fraction 0.2

0 101 102 103 104 105 Recoil Energy [eV] Reading out Singlet Excitations (16eV photons)

Detecting photons is a standard calorimetry application.

simple detector: box with calorimetry inside Operating calorimetry in LHe: less standard. Possible, thanks to 1) huge LHe-solid Kapitza resistance 2) fast conversion of photon energy in calorimeter to trapped excitations (eg, Al quasiparticles) 16eV photon

4He Reading Out Triplet Excitations (ballistic molecules)

Superfluid ➔ friction-free ballistic propagation simple detector: box with calorimetry inside Touching a solid supplies mechanism for decay

Some fraction of energy appears in surface -energy transferred through electron exchange (not phonons) -fraction dependent on material’s electron density of states ballistic molecule Reading Out Triplet Excitations (ballistic molecules)

Superfluid ➔ friction-free ballistic propagation simple detector: box with calorimetry inside Touching a solid supplies mechanism for decay

Some fraction of energy appears in surface -energy transferred through electron exchange (not phonons) -fraction dependent on material’s electron density of states ballistic molecule

1 triplet 0.8 molecules

0.6 singlet photons 0.4 counts (scaled) counts (arbitrary) 0.2

Journal of Low Temperature Physics 0 0 2 4 6 8 10 12 14 16 18 20 February 2017, Volume 186, Issue 3, pp 183–196 [eV] Energy in TES [eV] https://arxiv.org/abs/1605.00694 Excitations in superfluid 4He

eV-scale excitations:

Excimers (He2*) ✔ detection: easy He He* singlet: ~ns halflife (observable as scintillation) triplet: 13s halflife (observable as ballistic molecules) ✔ detection: easy (+ a little IR from excitations to higher atomic states)

meV-scale excitations: phonons, R- rotons, R+ rotons (observable as athermal evaporation) 4He Quasiparticles

1.5 phonons R- R+ Things to know:

Emaxon meV-scale (hear ‘MeV-scale DM’...) 1 Not on a crystal lattice (isotropic dispersion)

Ballistic propagation Eroton

Most downconversions forbidden

energy [meV] 0.5 Multiple ‘flavors’ with distinguishing characteristics: - slope is velocity - R- propagation opposite to momentum

Below atomic excitation energy, all recoil energy appears in these kinetic modes 0 0 1 2 3 4 momentum [keV/c] 4He Quasiparticles

Don’t let the word ‘roton’ distract you.

Illustration is incorrect in two ways: -rotons are few-atom-scale kinetic excitations -rotons do not carry angular momentum.

sufficient shorthand: rotons are “high-momentum phonons” Reading Out 4He Quasiparticles

crossing into solid extremely suppressed (Kapitza resistance)

(etc. etc...)

roton P R- R+ incident angle (rad.)

p [eV/c]

Probability Reading Out 4He Quasiparticles (quantum evaporation)

crossing into solid extremely suppressed (Kapitza resistance) atom

...saved by significant probability vacuum interface of quantum evaporation at vacuum

roton P R- R+ incident angle (rad.)

p [eV/c] p [eV/c]

Probability Reading Out 4He Quasiparticles (quantum evaporation)

crossing into solid extremely suppressed (Kapitza resistance) atom

...saved by significant probability vacuum interface of quantum evaporation at vacuum P R- P R- R+

R- R+ incident angle (rad.)

p [eV/c] p [eV/c] 1.5 phonons R- R+

Probability by the way: 1 (etc.) Energy [meV]

energy [meV] 0.5

0 0 1 2 3 4 momentum [keV/c] Reading Out 4He Quasiparticles (quantum evaporation) → van der Waals gain

van der Waals free atom quasiparticle binding Typical helium-solid binding energy: ~10meV

Higher binding energies exist (graphene-fluorine: 42.9meV) ~1 meV 0.62 meV

most recoil energy is in roton modes (DOS ~ p2) LHe vacuum Cal. each 1 meV roton energy becomes ~40 meV observation

10s of meV → x40 gain ‘Shovel Ready’ Technology Years Ago

3.3. The Experimental Cell R&D for the proposed HERON pp neutrino observatory

Wires running from film free region to Mixing Chamber SQUID sensor

Film Flow Constriction

Thermal Link to 200 mK Evaporation 364 keV Plate Thermal Link to 100 mK Condensing Plate electron recoil Condensing Baffles He film burner Heater roton time of flight

Film Free Mounting wafer calorimeter Platform single TES channel Calorimeter Vacuum

Helium Surface

230mm Phonons/Rotons

kg-scale LHe mass Radioactive Source Linear Drive now: technology newly motivated by light dark matter + improved calorimetry steadily improving

Figure 3.1: The experimental cell and the film burner.

70 ER/NR discrimination using excimer production

Toy MC Result: Production Statistics Detection Efficiencies Assumed Poissonian singlet UV photons : 0.95 (4pi coverage by calorimetry) extreme discrimination (biggest first, remaineder in QP) triplet excimers : 5/6 (only solid interfaces) (in `high energy’ >100eV range) IR photons : 0.95 (similar to UV photons)

Nuclear Recoil Nuclear Recoil 1 1 100 0.8 0.8

0.6 0.6 10-1

0.4 0.4

Energy Fraction Energy Fraction -2 0.2 0.2 10

0 0 1 2 3 4 5 1 2 3 4 5 10 10 10 10 10 10 10 10 10 10 -3 Recoil Energy [eV] Recoil Energy [eV] (reconstructed) 10 Electron Recoil Electron Recoil 1 1 Singlet Excimers IR photons 10-4 0.8 Triplet Excimers 0.8 Triplet Excimers

Singlet Excimers Singlet Excimers ER Leakage Fraction Quasiparticles Quasiparticles 0.6 0.6 10-5 0.4 0.4

Energy Fraction 0.2 Energy Fraction 0.2 10-6 1 2 3 0 0 10 10 10 1 2 3 4 5 1 2 3 4 5 10 10 10 10 10 10 10 10 10 10 Recoil Energy [eV] Recoil Energy [eV] Recoil Energy [eV] (reconstructed) Simulated Backgrounds

gamma backgrounds included: gamma neutrino electron scattering -neutrino nuclear coherent scattering nuclear nuclear -gamma backgrounds copy scattering scattering SuperCDMS & DAMIC projections Delbrück https://arxiv.org/abs/1610.00006 Rayleigh neutrons -note: LHe is naturally itself radiopure Thomson two details: after ER discrimination -excimers allow ER discrimination (>20eV) -newly-discussed gamma-NR background Robinson Phys. Rev. D 95, 021301 (2017) arguments for low dark count rate:

-calorimetry, no applied potential energies -low-mass calorimeter: low-energy clamps -superfluid target: highly isolated from environment Nuclear Recoil Sensitivity

1020 10-25

1018 10-27 ] 1016 10-29 2 ) gen1: “shovel ready” 2 1014 10-31 10 eVr threshold, 1 kg-y assuming 40meV per evaporated atom (graphene-fluorine) 1012 10-33 20 eV calorimeter threshold w/ 5% evap. efficiency 1010 10-35 gen2: “feasible after R&D” 8 -37 100 meVr threshold, 10 kg-y 10 10 gen1 gen3 gen2 assuming 40meV per evaporated atom (graphene-fluorine) 6 -39 1 eV calorimeter threshold w/ 25% evap. efficiency 10 10 104 10-41 gen3: “theoretically possible” 2 -43

1 meVr threshold, 100 kg-y WIMP-nucleon cross section ( zb ) 10 10 WIMP-nucleon cross section ( cm limit of single-atom counting (~40meV calorimeter threshold) 100 10-45 WIMP-nucleon cross section [ zb ] 10-2 10-47 WIMP-nucleon cross section [ cm

10-4 keV MeV GeV TeV Nuclear Recoil Sensitivity

1020 10-25 100m depth 1018 10-27 ]

16 -29 2 10 1400m depth 10 ) 1014 10-31 2

1012 10-33 rarely-considered but newly-relevant: 1010 10-35 earth shielding at large cross-sections 108 10-37 gen1 gen3 gen2 106 10-39

104 10-41

2 -43

WIMP-nucleon cross section ( zb ) 10 10 WIMP-nucleon cross section ( cm 100 10-45 WIMP-nucleon cross section [ zb ] 10-2 10-47 WIMP-nucleon cross section [ cm

10-4 keV MeV GeV TeV two-body diagram the nuclear mass (dispersion relation) this talk has so far been “we have a mismatch, but we is very different from have strategies to mitigate that mismatch.” the DM mass (dispersion relation) ~10x energy boost from light nucleus ~10x energy boost from van der Waals gain χ χ 107 106 Xe 105 Si He 104 - 103 e nucleus nucleus 102 101 100 1eV w/ binding gain 10-1 100meV He endpoint

Elastic Recoil Endpoint [eV] -2 10 5meV Ge endpoint 10-3 103 104 105 106 107 108 109 1010 1011 1012 1013 M [eV/c2] DM three-body option #1: nuclear bremsstrahlung

trick: outgoing [ nucleus+γ ] can have wide range of [E,p] χ χ no longer limited to nuclear dispersion relation γ

nucleus nucleus what you get: recoil E can be up to DM KE Kouvaris, Pradler: arXiv:1607.01789 (and E can be up to recoil E) McCabe: arXiv:1702.04730v1 γ

what you pay: large phase space suppression three-body option #2: multiple outgoing phonons

χ trick: χ outgoing multi-phonon states can have wide range of [E,p]

no longer limited to phonon outgoing dispersion relation virtual real phonons phonon what you get: recoil E can be up to DM KE

what you pay: Schutz, Zurek: arXiv:1604.08206 Knappen, Lin, Zurek arXiv:1611.06228 large phase space suppression three-body option #2: multiple outgoing phonons

as observed in neutron scattering data

0.5

TransferMomentum [A 1.0 1.5

2.0 -1 3.0 ] 2.5 2.0 1.0 Energy Transfer [meV] Schmidt-Wellenburg et al NIMA611 2009 p259 Nuclear Recoil Sensitivity

1020 10-25 100m depth 1018 10-27 ]

16 -29 2 10 1400m depth 10 )

14 -31 2 result of double-roton diagram: 10 10 1012 10-33

1010 10-35 bypass the He atom dispersion relation 108 10-37 gen1 gen3 gen2 sensitivity to keV-scale masses 106 10-39

104 10-41

2 -43

WIMP-nucleon cross section ( zb ) 10 10 WIMP-nucleon cross section ( cm 100 10-45 WIMP-nucleon cross section [ zb ] 10-2 10-47 WIMP-nucleon cross section [ cm

10-4 keV MeV GeV TeV starting up some new efforts

McKinsey Group at Berkeley Hertel Group at U. Mass. (no table yet)

4He scintillation yield at low energies designing, setting up 4He evaporation channel test bed

goals: evaporation channel R&D evaporation channel calibration (bring to n facility) early DM limits from on-campus lab (“ν-cleus style”)

LHe

PMT

wavelength shifter Summary

1020 10-25 100m depth 1018 10-27

Ideal technology for low-mass NR ] 16 -29 2 -meV-scale long-lived kinetic excitations 10 1400m depth 10 ) -light-element material 1014 10-31 2 -suppressed “dark counts” in superfluid 1012 10-33

1010 10-35 Shovel-ready technology 108 10-37 gen1 gen3 -4He evaporation: HERON gen2 -calorimeters: CDMS, CRESST, etc. 106 10-39

104 10-41

2 -43

Small and cheap, $1M-scale WIMP-nucleon cross section ( zb ) 10 10 WIMP-nucleon cross section ( cm - small target (grams-to-kg) 100 10-45

- few channels (6+) WIMP-nucleon cross section [ zb ] 10-2 10-47 WIMP-nucleon cross section [ cm

10-4 keV MeV GeV TeV