the low/very low mass particle dark matter frontier

B.SadouletMoorea Pacific 2018.5 1 September 2018 1

The Low/Very Low Mass Particle Dark Matter Frontier

Bernard Sadoulet Dept. of Physics /LBNL UC Berkeley

UC Institute for Nuclear and Particle Astrophysics and Cosmology (INPAC)

Scope of this talk Situation September 2018 High mass region>6 GeV/c2 0.5-6 GeV/c2 intermediate mass region

G2 program. How to get to solar neutrino floor The low mass region: 1 MeV to 0.5 GeV/c2

Reachable with current technology?

The ultra low mass frontier: 1 keV-1 MeV/c2 Also 1 meV to 1 keV ultra light boson absorption

Thanks to Matt Pyle Kathryn Zurek and many colleagues but misunderstandings are mine

B.SadouletMoorea Pacific 2018.5 1 September 2018

Scope of this talkWhat do I mean by Particle Dark Matter?

We tend to speak about WIMPs: Weakly Interacting Massive Particles with mass and interaction at the weak scale, typically SUSY

=> appropriate dark matter density But lack of SUSY at LHC has obliged us to expand our exploration space!

• Theoretically: Dark Sector, Heavy or Light carriers, Dark Matter asymmetry etc.

Plenty of viable models, but much less constrained than WIMPs • Experimentally: New techniques

I would like still to speak about DM particles which act more as individual particles scattering on a target than waves!

≠ Ultra light bosons: axions, dark photons, moduli Because they are very light, long wave length, large occupation number:

best described as coherent fields =>We will hear a number of talks ot this conference

However, such bosons can be absorbed by either nuclei or electrons. Same amount of deposited energy as scattering of halo

particle of mass 106 more massive: same technology!

≠ Sterile neutrinos Different techniques: x ray detection

2


Basic KinematicsKinematics

Purely elastic

ultimately due to the conservation of energy and momentum. Bad news for low mass

But at low mass

Putative Dark Matter event rate goes as

Electron recoil background from Compton, betas ≈ flat at low energy Proportionally less important!

Main problem, spurious pulses, not enough redundancy to be sure of a signal

3

Ed =!q 2

2mT

=m

r

2

mT

v2 1− cosθ *( ) = mχ2mT

mx +mT( )2 v2 1− cosθ *( )

where mT is the mass of the nucleon or electron

For mχ << mT , Ed ≈mχ2

mT

v2 1− cosθ *( )

1mχ

match mT to mχ as much as possible


Design Drivers for Dark Matter Scattering/Absorption

Maximize potential signal Kinematics Matrix elements

Maximize sensor energy sensitivity Lower gap materials, lower temperature or Amplification

Minimize background/dark current/shot noise Radioactivity, neutrino background External influence (RF, IR, Vibration,Crystal Cracking) Gap inhomogeneity (hot spots)

Maximize discrimination Energy spectrum shape Time and pulse shape Nuclear recoil recognition (phonon/ionization or scientillation, pulse

shape) Directionality? Velocity modulation?

Maximize target mass/unit cost, leverage R&D 4


Particle DM Situation in September 2018At High Mass >10GeV/c2

Nothing so far Broadly consistent with the absence of Super Sym. observation at

LHC Focus point solution in CMSSM ≈10-45 is mostly excluded

5

Intermediate Mass ≈10GeV/c2

A number of closed contours, and strong limits

What is going on? See Reina’s talk

� �� -��

��-��

��-��

��-��

��-��

��-��

��-��

��-��

��-��

��-�

��-�

��-�

��-�

�� [��/��]

��-��σ �

�[�� ]

��-��σ �

�[��]

��

��-��

��

��

��-��

��

��

��

��

��

��-��

��-��

��

�

��

��-��

��

��

��

��

Below 1 GeV/c2 Nearly totally open


High Mass Region M>6 GeV/c2

6

5’



Favored region of WIMP mechanism Noble liquids are the technologies of choice

Large target masses at reasonable cost Xe excellent self shielding, moderate Nuclear Recoil discrimination (39Ar depleted) Ar excellent Nuclear Recoil discrimination, but higher thresholds

7

� � �� -��

��-��

��-��

��-��

��-��

��-��

��-��

��-��

��-��

��-�

��-�

��-�

��-�

�� [��/��]

��-��σ �

�[�� ]

��-��σ �

�[��]

��

��-��

��

��

��-��

��

��

�� -��

��-��

��

��-��

��

��

��



Complementarity of other technologies Diversity of nuclei: More general than spin independent/dependent

Velocity dependent effects (including Fermi)Haxton, Zurek Different systematics (but of course need to be in the same sensitivity

region)

Large target mass candidates Scintillators (NaI):

challenge of radioactivity but need to check DAMA See Reina’s talk

PICO Mostly for “spin dependent” Discrimination against alphas

8


The Intermediate Mass Region 500 MeV/c2<M<6 GeV/c2

9

10’’


Amplification MechanismsDAMIC and CRESST have currently the energy sensitivity to cover this

region, but not yet the mass But you can improve energy sensitivity by amplification. 3 incarnations

Gas multiplication Multiwire Proportional chambers, NEWS-G, Drift

Gas field induced scintillation Noble gas S2 signal

Neganov-Trofimov-Luke (NTL): CDMSLite

But lost Nuclear Recoil Discrimination 10


CDMSLite Run 3 Just out of the press ArXiv 1808.09098 Neganov Trofimov Luke + Profile likelihood analysis (about factor 3 compared to Optimal Interval)

Recent Results 1: CDMSLite Run 3

11

Is this believable?


Recent Results 2 Dark SidearXiv 1802.06994

Decent resolution

Better job than LUX, Xe on e emission from grid

12


Recent Results 2 Dark Side

My main problem: assumption about behavior of the yield at low energy

13


Low Mass Region Projections≈2020

DAMIC/SENSEI CCD NEWS-G (Spherical Proportional Counter) Low temperature calorimeters: SuperCDMS, CRESST, Edelweiss

14

�� -��

��-��

��-��

��-��

��-��

��-��

��-��

��-��

��-��

��-�

��-�

��-�

��-�

�� [��/��]

��-��σ �

�[�� ]

��-��σ �

�[��]

��

��

��

��

��-��

��

��

��

��

��

��-�� (��)

��

��

��

��

��

��-�


SuperCDMS: Current Limits on resolution

15

In principle TES r.m.s. ≈Tc3 (Pyle) Best so far 20eV baseline r.m.s. for 1.3 kg Ge 2 x our goal, <project threshold But need to control

Vibrations √ IR √ RF <= seems to be a major problem (+ correlated noise in dry fridges )

Compact TES 50µx200µ 2 10-18 W/√Hz

Distributed TES Collect over 4” 2 10-17 W/√Hz


Radioactive BackgroundsTritium, 32Si

16

CDMSLite Run 2 arXiv 1806.07043


Low E-field Dark Current

Use of NTL amplification Need to hold voltage Delicate but our contact seem OK

However large dark current

≈ independent of field IR ? Metastable states Appears as shot noise but will give individual pulses

Maybe due to being at the surface Muons/gammas->metastable states?

Plan to test at CUTE, a new underground test facility at SNOLAB

17

an early R&D detector significant improvements since


Reaching The Solar Neutrino Floor

Neganov Trofimov Luke amplification mixes the primary phonons with the ionization generated phonons.

We need Nuclear Recoil discrimination in 1-6 GeV/c2 to reach solar neutrino floor

How can we restore? 1) Displacement of NR peaks

no very far for what we already measure!

2) 2 types phonon sensors One close to region where Neganov Trofimov Luke phonons are created One further away more sensitive to original phonons => Separate E0 from ENeganovTrofimovLuke some preliminary demonstration in existing detectors

3) Also progress in ionization charge amplifier (91 eV rms for 300pf detector capacitance)

18

• σ=5eVt

• Singlee-/h+Sensitivity• ER/NRDiscrimination

TotalPhononEnergy[keVt]

1e-/h+ER1x(100+3)

2e-/h+ER2x(100+3)

3e-/h+ER3x(100+3)

1e-/h+NR1x(100+30)


The low mass region 1 MeV/c2-0.5 GeV/c2

19


Our Low Mass DreamsEasiest:e scattering!

in condensed matter, everything is inelastic

excitations, phonons

Liquid He Lightest nucleus

HERON D. McKinsey

Can go further than naive estimates

Interesting calculations e.g., by the Zurek group of multiple excitations

e.g., back to back phonons (≠ large momentum compared to energy carried by phonons)

Penalty due to off shell process, but some sensitivity!

20

US Cosmic Visions: New Ideas in Dark Matter 2017 : Community ReportarXiv:1707.04591v1

��-� �� -��

��-��

��-��

��-��

��-��

��-��

��-��

�� [��/��]

��-��[�� ]

��-��

��

��

��

��-��

��

��

��-��

� ��

��

��-��

��

��

��-��

��

��

��

-��

�� (�

)

��(��)

��(��)

��

�� (�

)

��

��

��

��-��

��(�-��)

��(��)� ��

��+

��

��

��

��

��

��

�-�

��

-��


Jeremy Mardon, SITP, Stanford

DIRECT DETECTION VS OTHER BOUNDS

10 100 1000 10410-44

10-41

10-38

10-35

10-32

10-29

mDM @MeVD

se@cm

2 D

FDMµ1êqLEP

PlanckN eff

XENON10 ER

XENON10 NR

CRESST

XQC

EDM-DM

10 100 1000 10410-42

10-40

10-38

10-36

10-34

mDM @MeVD

se@cm

2 D

FDM=1LEP

CMB

PlanckN eff

XENON10 ER

XENON10 NR

CRESST

MDM-DM

10 102 10310-4310-4210-4110-4010-3910-3810-3710-3610-35

mDM @MeVD

se@cm

2 D

FDM=1

XENON10 ER

CRESST

HeXeArGe

PlanckN eff

Massive Dark Photon

1 10 102 10310-4310-4210-4110-4010-3910-3810-3710-3610-3510-3410-3310-32

mDM @MeVD

se@cm

2 DFDM µ 1êq2

XENON10 ERHeArXeGe

Freeze-In

Ultra-light Dark Photon

e scatter:Smaller Band Gaps are better

21

(Ultra-lightMediator)

Essigetal1607.01009

1kgGe~3tonofXe

https://arxiv.org/pdf/ 1108.5383.pdf


SuperCDMS 1g Si deviceGo to lower crystal mass: Single e-h pair sensitivity

22


Dark Matter Limit at the Surface

Use only events in peaks No Background Subtraction: Optimal Interval Method

23

Mean σ ! 14eV


Dark Matter LimitPhys. Rev. Lett. June 2018 p121.051301

Inelastic electron recoil Dark Photon limit

24

Light mediator

Heavy Mediator

As expected, factor ≈ 10-6 in mass scale

1 eV/c2!


Parasitic phenomenaExcellent diagnostic tool

+ leakage current indistinguishable from dark matter

25


The Ultra Low Mass Frontier <1 MeV/c2

26

16’


The Ultra Low Mass Frontier

How do the creative schemes proposed align to the major design drivers?

1. Maximize potential signal 2. Maximum sensor energy sensitivity 3. Minimize background/dark current/shot noise 4. Maximize discrimination 5.Maximize target mass/unit cost, leverage R&D

27


1. Maximize Potential Signal

Direct Coupling to Optical Phonons In many dark sector models, coupling to nuclei/electrons is through dark

photon and its small mixing with ordinary photons

Advantage of polar crystals, e.g., GaAs, Al2O3 arXiv:1712.06598v1 (Knapen, Lin, Pyle, Zurek) including coupling to L Optical phonons

28

B.SadouletMoorea Pacific 2018.5 1 September 2018 29

LO phonon resonant absorption needs 36 meV sensitivity

Scintillation Direct gap: Photon threshold = e threshold

Light Dark Photon mediator

LO phonon Production needs 36 meV sensitivity

Dark Photon absorption

Polar Crystal Example: Dark Photon couplings in GaAs


2. Maximum Sensor SensitivityLooking for small gap

Semiconductors ≈ 1eV (scintillation in Ga As is also 1 eV) Liquid helium rotons 0.75 eV Superconductors 1 meV

Creative exuberance (cf 1985) which is refreshing Graphene and carbon nanotubes where gap can be engineered Dirac materials (Semi metal with small gaps) Spins in a magnetic field Labelled DNA (Druikier) ≈eV binding energy Color centers ≈ eV binding energy?

Need to couple to a sensitive enough sensor CCDs with skipper readout (multiple sensing) SENSEI/DAMIC but 1 eV

Low temperature sensors TES ≈Tc3

MKID Amplification mechanisms

Luke Neganov but amplifies only ionization 4He Absorption on bare surface Spin flip avalanche (cf old granules & PICO)

30


Promising TES result: 70 meV50µx200µ

31

• Noisematchestheorytowithinx1.4– Sptot =2x10-18 W/rtHz

• !sensor ~8kHz• "E =55meV (withoutcollectionlosses

…~20%)


3. Minimize Dark Count 4. Maximize Discrimination

My main worry Fabrication defect=>zero gap flooding the sensors All parasitic phenomena will show up

Can we recognize nuclear recoil? probably not below a few 100MeV

Directionality inherent in crystal asymmetries e.g., Al2O3 LO phonons (Zurek, Pyle)

32

22’


5. Target Mass and CostRate goes at 1/mDM!

We will probably see soon results with very small detectors

DAMIC/SENSEI Low Temperature Detectors R&D (e.g., SuperCDMS)

Do not underestimate cost of R&D cf. history

Thermistors used in CMB SQUIDs

Balance enthusiasm with new direction and amount needed to get in the 100g region

33


ConclusionsThe current combination of experiments and

technologies will cover 300 MeV/c2 to TeV/c2. “G2”: part of the way towards neutrino floor 2020-2025 G2+: go to neutrino floor

New window below 500 MeV/c2 Nuclear Recoils: Liquid He, Polar crystals Electron scattering: -> keV DM absorption:-> meV

New experimental results soon R&D for G2/G2+ paves part the way toward those ambitious goals Table top experiments

Exciting new ideas (cf 1985 era) Use/manufacture low gap systems Of course many of these clever schemes will fail, but fun

We do not need large target masses My worries

(3.) Dark Count problem (we already see in 500 MeV/c2 to 6 GeV/c2, and learn from it)

(4.) Paucity of dark matter specific signatures 34

the low/very low mass particle dark matter frontier

Documents