Metallic Magnetic Calorimeters

for high resolution particle detection

Yong-Hamb Kim

2

Equilibrium thermal detectors at Low Temp.

Absorber

Thermometer

Thermal link

Heat sink < 100 mK

, , ,

WIMP etc.

Energy absorption Heat (Temperature)

Measurement of the energy E of individual particles

as a temperature rise ΔT = E/C

Temperature pulse with decay time τ = C/G

Very low temperature(10–100 mK)

• Small C Large ΔT

• Low noise

3

Sensor technology

Choice of detector sensors (superconducting detectors)

• Thermistors (doped Ge, Si)

• TES (Transition Edge Sensor)

• MMC (Metallic Magnetic Calorimeter)

• STJ (Superconducting Tunneling Junction)

• KID (Kinetic Inductance device)

• etc.

Currently used as sensors

for equilibrium thermal detectors

Have great potentials useds

for equilibrium thermal detectors

4

Metallic Magnetic Calorimeter (MMC)

• Paramagnetic alloy in a magnetic field

Au:Er(300-1000 ppm), Ag:Er(300-1000 ppm)

Magnetization variation with temperature

• Readout: SQUID

• High energy+time resolution, Good linearity, Large dynamic

range, No bias heating, Absorber friendly, Fast

• Mux being developed, dipl.167Er needed,

5

Metallic Magnetic Calorimeters

Magnetic Material in loop of a dc SQUID (in early stage)

Choice of Material (Metal host Metallic)

(Au:Er, Ag:Er, Au:Yb, Bi2Te3:Er, etc)

• weakly-interacting paramagnetic system

(Curie system with small heat capacity)

• fast thermalization ( 10-9 /T [s] << 1ms)

Mag

net

izat

ion

140mA

100mA

60mA

30mA

10mK20mK50mK

Inverse Temperature, 1/T (K-1)

B

<KRISS>

<NASA/GSFC>

6

Er3+ ions in Au

4 Å

gold (fcc)

Er3+

Rare earth ions in metals

• Incomplete 4f shell, Er3+ (4f 11)

magnetic moment

J = 15/2 g = 1.2

• Cubic crystal field of fcc gold

group multiplets

ground state: Kramer’s doublet

(g = 6.8)16 K

E

B

100K

7

7 –Kramer’s doublet

5 mT e = 1.5 meV

1 keV 109 spin flips

B = 50 G

8

More physics of Au:Er and Ag:Er

Spin-Spin Interactions:

- Magnetic dipole-dipole interaction

- RKKY (indirect exchange interaction) : Stronger in Ag:Er

Hyperfine Interactions:

- 167Er nuclear spin(I=7/2), 23% anundance in natural Er

- Enrichement for 166Er or 168Er (commercially available from ORNL)

Nuclear quarupoles of 197Au(I=3/2, 100% anundance) (not in Ag:Er)

- Electric field graident due to Er impurities in gold fcc lattice

- Extra heat capcity ~1/T2

- Less concentaration preferable for lower temperatures

well understood now !

9

Sensor geometry in early setups

Au:Er sensors

<Ketchen’s susceptometer>

<Heidelberg, 2009><KRISS, 2006>

field coil

40 um Au:ErSi

SQUID loop

10

MMC on SQUID

largest coupling

for spins near loop

Coupling depends on spin position and width of loop

External B field

Better coupling !

Problems

- SQUID junctions under B fields

- Magnetic cross talks

- SQUID bias heating

- Fab. on SQUIDs

- Not optimal geometry

11

Meander geometry

A. Fleischmann et. al,

2005

• No heat dissipation

• Smaller magnetic cross-talk

• Easier to fabricate

• Reduced pickup of magnetic Johnson noise

• Higher filling-factor(F) than cylinder geometry

B field is not homogenous, F ≈ 0.4

“Field generation” & “Signal pickup”

12

Sandwich Geometry

Meander, F ≈ 0.4

Spiral Sandwich F > 0.6~0.85

<Pies et. al, LTD14>

- Sandwiched planar setup

- Au:Er sensor between spiral and ground plane

- More homogeneous B field

Heidelberg

13

Noise contributionLTD-13

A. Fleischmann

100e

Cabsfluctuations of energy between sub-systems

(optimum for Cabs = Cspins)

flux noise of SQUID-magnetometer

SF = 2 ε L, required:

magnetic Johnson noise

- thermal currents in the metallic components

- marginal in all present detectors

magnetic 1/f noise SF ~ NEr

SF ~ 1/f, Sm|1Hz ≈ 0.023 mEr2/Hz

temperature independent (20mK – 4K)

Cspins

14

Applications

• MMCs have many active applications in several institutions

• Here, only a few of them are introduced.

• Institutions working on magnetic calorimeters

- Brown University

- CEA Saclay

- Heidelberg Universisty

- IBS

- KRISS

- LLNL

- NASA/GSFC

- PTB

- University of New Mexico

15

Alpha spectrometer using MMC

Heat-pulse switchTemperature switch

Poster by SoRa Kim

Gold foil

absorberAu:Er

MMC fabrication: KRISS

16

241Am Alpha spectrum

emissions

(from nndc)

5388 1.66%

5416.5 0.0100%

5442.8 13.1%

5469 0.020 %

5485.56 84.8%

5511.5 0.225%

5544.5 0.37%

17

MMC vs SBD

SBDMMC

18

Full spectrum of 241Am

19

Low energy spectrum

Low E. and high E. spectrum can be measured in the same time with high resolution.

Significant tool for radionuclide analysis.

20

alpha decay in 4 metal absorber

emitters

20~50 mm

gold foil

Alpha particle

Recoil

ce., , x, Auger electron

Heat generation

1. No energy loss in source and detector

2. No count loss

Alpha decay energy (Q)

“Low emission probabilities for high energy photons and electrons”

21

Full Q spectrum of 226Ra

3-5 keV FWHM

22

Energy linearity

23

Pu isotopes

239Pu 240Pu

T1/2 2.411e4 y 6561 y

Q

(keV)5244.5 5255.8

alpha

(keV,%)

5105.5, 11.9

5144.3, 17.1

5156.5, 70.8

5123.7, 27.1

5168.2, 72.8

From commercial

plutonium solutions

239Pu 240Pu

238Pu 241Am

Q spectrum

24

Isotopic ratio

087.0Pu)Activity(

Pu)Activity(239

240

99.0Pu)Activity(

Pu)Activity(239

240

239Pu240Pu

239Pu

240Pu

Measured for commercially available low-activity test samples.

25

Center for Underground Physics

KIMS (Korea Invisible Mass Search)

Now, 100-kg NaI: Cosine-100

CsI NaI LowTemp

AMoRE (Advanced Mo based Rare

process Experiment)

Now, AMoRE-pilot with 1.5 kg

(Ø 4cmx4cm crystal)

26

Neutrinoless double beta decay (0νββ)

Double Beta Decay with two neutrinos

Double Beta Decay with no neutrino

requires massive Majorana ν !

Key test proposed by Racah in 1937

It may answer

- Mass of neutrinos ( )

- Majorana ( ), or Dirac particles ( )

- Lepton number violation

20

2/11 mT

27

AMoRE detector technology

40Ca100MoO4 + MMC : Source = Detector

<Advanced Mo-based Rare process Experiment>

CaMoO4

- Scintillating crystal

- High Debye temperature: TD = 438 K, C ~ (T/TD)3

- 48Ca, 100Mo 0ν candidates

- AMoRE uses 40Ca100MoO4 w. enriched 100Mo and

depleted 48Ca

MMC (Metallic Magnetic Calorimeter)

- Magnetic temperature sensor (Au:Er) + SQUID

- Sensitive low temperature detector with highest

resolution

- Wide operating temperatures

- Relatively fast signals

- Adjustable parameters in design and operation

stages

Phonon(Heat)

sensor

Photon(Light)

sensor

28

Heat(Phonon) sensor for AMoRE

Gold film

We measure both thermal

and athermal phonons.

Phonon collector

Patterned gold film

Gold wires

(thermal connection)

MMC

rise-time: ~ 0.5ms

<Heat flow optimization>

216 g CaMoO4

SQUID

29

Photon sensor with MMC

MMC chip

SQUID sensor

Thermal connection

Phonon collector

(Gold films)

Light absorber

(Ge wafer)

Wafer holder

(Cu, heat bath)

rise-time:

~ 0.2 ms

Temperature independent

rise-time !

SeungYoon Oh’s talk

- Heat capacity optimization

- Inductance optimization

30

Energy spectrum (above-ground)

• Better than 9 keV energy resolution was obtained at 10 mK temperature.

• Internal alpha background levels of each isotopes were calculated successfully.

Electron and alpha events can be efficiently identified.

31

Alpha and electron events

PSD Light/Heat ratio

Clear separation in

PSD and Light/Heat ratios!

Poster by Inwook Kim and HeLim Kim

AMoRE@over-ground

AMoRE@over-ground

32

YangYang Pumped

Storage Power Plant

YangYang(Y2L) Underground Laboratory

Minimum depth : 700 m / Access to the lab by car (~2km)

(Upper Dam)

(Lower Dam)

(Power Plant)

COSINE (Dark Matter Search)

AMoRE (Double Beta Decay Experiment)

Y2L

700m

1000m

33

AMoRE pilot: 1.5 kg 40Ca100MoO4

NSB29

390 g

SS68

350 g

SB28

196 gS35

256 g

SE#1

354 g

5 Crystals (40Ca100MoO4) with total mass ~1.5 kg

34

AMoRE pilot : 5 phonon + 6 photons

All are installed in a dry dilution refrigerator

in YangYang Underground Lab.

35

We are taking measurement now.

with an external source at 20 mK

• All of the shields are mounted.

• The dil. fridge reaches 8 mK with 250kg lead & copper

attached to M.C.

• We are improving noise figures now.

- High friq.: reasonably low.

- Low friq.: recently improved.

Above ground

Y2L w.o. vib. damper

Y2L w. vib. damper

36

Schedule of the AMoRE project

Pilot Phase I Phase II

Mass 1.5 kg ~5 kg ~200 kg

Background (keV kg year)-1 0.01 0.001 0.0001

Sensitivity(T1/2) (year) ~1024 ~1025 ~5×1025

Sensitivity(mee) (meV)< 300-

90060-180 15-40

Location Y2L Y2L New Lab

Schedule 2016 2017-2019 2020-2022

<AMoRE Pilot, now> <AMoRE-1, 2017> <AMoRE-II, 2020>

About 500 crystals

w. 1000 MMCs + SQUIDs

2020-2022

37

Summary

• MMCs using superconducting sensor tech have great

performance and applications in many aspects of science.

• Excellence in energy resolution, timing resolution, linearity,

dynamic range.

IBS-KRISS

• High resolution alpha and Q spectroscopies for radionuclide

analysis

• Underground particle astrophysics

– Search for neutrinoless double beta decay

– Direct detection for dark matter (WIMPs)