Calibra'on*and*Modeling*of*Nuclear*and*
Electron*Recoils*in*Liquid*Argon*
Workshop(on(Calibra/on(of(
Low(Energy(Par/cle(Detectors(
Samuele(Sangiorgio(
Rare(Event(Detec/on(Group.(LLNL(
Chicago, Sep 24, 2015
LLNL-PRES-677393
This work was performed under the auspices of the U.S. Department of Energy by Lawrence Livermore
National Laboratory under contract DE-AC52-07NA27344. Lawrence Livermore National Security, LLC
LLNL’s*Noble*Liquid*R&D*Program*
Physics(Mo/va/ons(
Liquid(Xenon(and(Argon(Detectors(
• Two(small(dualIphase(
detectors(
• Dark(MaRer(
• Coherent(Elas/c(
NeutrinoINucleus(
ScaRering(
• Measure(electron(and(
nuclear(recoils(<(few(
keV(
• Understand(and((
control(lowIenergy(
backgrounds(
• HV(stability(in(noble(
liquids(
Dedicate(lowIenergy(neutron(beam(
• OnIsite(at(LLNL(
• QuasiImonoenerge/c(
filtered(neutron(
beam(
LLNL-PRES-677393
S.(Sangiorgio(
2(
LLNL*Dual?Phase*LAr*Detector*
• Ac/ve(volume:(~(100(g(Lar(
• TPB(as(wavelength(shiYer(
• HomeIbuilt(HV(feedIthroughs(
• Very(good(LAr(purity(
4x Hamamatsu
R8520 1” PMTs
(1 atm @ 87K)
Up to 11kV/cm
liquid Ar
Up to
3kV/cm
Egain
3.5 cm
gas Ar
2.5 cm
liquid
level
Edrift
liquid
level
field
rings
5 cm
rings
support
HV feedthroughs
LLNL-PRES-677393
S.(Sangiorgio(
3(
High*Gain*Detec'on*of*Ioniza'on*Signal*
• Interest(in(the(lowest(energy(
possible(
• Emphasis(on(detec/on(of(
ioniza/on(by(means(of(S2(only(
• Operate(close(to(electron(
mul/plica/on(in(gas(
Monteiro et al, IEEE Trans. Nucl. Sci., 48 (2001)
~10 kV/cm
~2-4 kV/cm
~11 kV/cm at
1atm at 87K
LLNL-PRES-677393
S.(Sangiorgio(
4(
Ar?37*as*a*Diffuse*Low?E*Calibra'on*Source*
natAr
Decay(scheme(
100%(electron(capture(
t1/2(=(35.04(d(
(
Decay(radia/on(
isotopes
Mass
number
Natural
Abundance
40
99.6%
36
0.34%
38
0.06%
KI(capture(2.82*keV*(90.2%)(
LI(capture(0.27*keV*(8.9%)(
MI(capture(0.02(keV((0.9%)(
(
Isotope(produc/on(
Produced(by(neutron(irradia/on(
of(natAr(at(a(nuclear(reactor(
Aalseth,(C.(E.(et#al.(NIM#A652,(58–61((2011).(
Barsanov,(V.(I.(et#al.(Phys.#Atomic#Nucl#70((2007).(
LLNL-PRES-677393
S.(Sangiorgio(
5(
Sub?keV*Calibra'on*for*Electron*Recoils*
S2
LLNL-PRES-677393
S.(Sangiorgio(
6(
Single*Electrons*
100
single ioniz electrons
µ = 8.2 ± 0.1 p.e.
! = 3.4 ± 0.1 p.e.
EventWidthT [µs]
Counts
• Typical(S.E.(event(as(seen(on(the(
scope(
80
• Experimental(spectrum(of(single(and(
double(ioniza/on(electrons(
25
h2
Entries
Mean
RMS
!2 / ndf
Prob
µ
"
SE amp
2E amp
20
15
10
60
5
2807
10.52
5.398
115.4 / 94
0.06611
8.2 ± 0.1
3.418 ± 0.128
67.25 ± 2.25
12.72 ± 1.38
45
40
35
30
25
20
15
10
5
40
0
0
5
10
15
20
25
30
35
40
0
Integral [p.e.]
• Provides(absolute(calibra/on(of(the(
number(of(detected(electrons(
20
0
0
LLNL-PRES-677393
S.(Sangiorgio(
5
10
15
20
25
30
35
40
Collected Light [p.e.]
7(
37Ar*Electron*Recoils*vs*Electric*Field*
2.82 keV – electron recoil
• Electric(field(reduces(recombina/on(of(electron(with(ions(
• Measurements(of(the(0.27(keV(peak(vs(E(field(are(ongoing(
• Need(to(deal(with(lowIenergy(background(
LLNL-PRES-677393
S.(Sangiorgio(
8(
Recombina'on*in*LAr*
Consider(electron(recoils(first(
For(electron(recoils(the(amount(of(ini/al(
ioniza/on(Ni(is(calculable:(
• Nex(/(Ni(=(0.21(
• E(=(2.82(keV(for(37Ar(KIshell(
• W(=(19.5(eV(
• q(E)(=(1(
S2 ∝ ne = rNi
Ni + Nex
E
· q(E)
=
W
ThomasIImel(parameteriza/on(of(
recombina/on(
r = ln(1 + ξ)/ξ
Introduce(
phenomenological(
scaling(for(field(
dependence:(
2.82 keV – electron recoil
ξ = CNi · E −b
Extract(field(dependence(
parameter(C,(b(from(fit(
Cfr.(Sorensen,(P.(and(Dahl,(C.(E.,(Phys.#Rev.#D.#83*(2011)(
LLNL-PRES-677393
S.(Sangiorgio(
9(
Modeling*recombina'on*in*Liquid*Argon*
At(low(energy,(empirical(ThomasIImel(box(model(seems(successful(but(
I Empirical(field(dependence(
I All(electronIion(pairs(recombine(for(zero(electric(field(
I LiRle(insight(on(physical(processes(involved(
M.#Foxe,#C.#Hagmann,#et#al,#NIM#A#771#(2015)#
Simula'on*Approach*
1. Ini/al(interac/on(
•
Simulate(ini/al(emission(of(photoelectrons(and/or(auger(electrons((
2. Follow(electrons(using(electron(transport(algorithm(
•
•
•
based(on(prior(work(by(Wojcik(et(al(for(thermal(electrons(
Solves(equa/on(of(mo/on(for(electrons(under(external(fields(and(ions(field(
Posi/ons(and(velocity(of(electrons(are(forward(propagated(
3. Compute(interac/ons(as(electrons(slow(down(
•
electronsIinduced((excita/on,(ioniza/on(and(elas/c(
scaRering(
• secondary(electron(generated(and(followed(as(well(
• Thermal(model(validated(against(measurements(
(driY(velocity,(escape(probability,…)(
Elas/c(energy(
transfer(
Momentum(
transfer(
ioniza/on(
excita/on(
4. Recombina/on(criteria:(
•
•
Electron(energy(<(1(eV(
ElectronIion(distance(<(1.3(nm(
No tunable parameter!
LLNL-PRES-677393
S.(Sangiorgio(
10(
M.#Foxe,#C.#Hagmann,#et#al,#NIM#A#771#(2015)#
Modeling*37Ar*Decays*
Average(ioniza/on(track(
length(<<(electron(
thermaliza/on(length(
(2.6(µm)(
Compare with 0.21
from Doke from 217Bi
conversion electrons
5.9(keV((55Fe)((
2.8(keV((37Ar(K)((
DATA:(2.8(keV((37Ar(K)((
1.0(keV(eI((
Model(limita/ons:(
• Ioniza/on(crossIsec/on(
uses(binding(energies(of(
gas(
0.27(keV((37Ar(L)((
LLNL-PRES-677393
S.(Sangiorgio(
11(
Nuclear Recoil
Neutron?induced*Nuclear*Recoils*in*LAr*
! Elas/c(neutron(scaRering(
Neutron
rvable signal than electron recoi
! Two(complementary(approaches:(
NARRLI
SCENE
Neutron Argon Recoils Resulting
in Liquid Ionization
SCintillation (and ionization)
Efficiency Noble Elements
• Recoils from tagged neutron scatter
• Energy 11 – 57 keV " DarkMatter
• Scintillation & Ionization
LLNL-PRES-677393
S.(Sangiorgio(
• End-point measurement
• Low energy < 10 keV " CENNS
• Ionization signal only
12(
T.#Joshi,#S.#Sangiorgio,#et#al,#NIM#B#333#(2014)#
Crea'ng*a*low?E*neutron*beam*
7Li(p,n)7Be
NearIthreshold(kinema/cs(of(
7Li(p,n)(allow(control(of(maximum(
neutron(energy(
Lab Neutron Energy (keV)
Proton Energy Countours for a Thick Lithium Target
from Lee and Zhou NIMB 152 (1999)
45º - 1.93MeV
Requirements:
- Continuous p beam
- No ToF in detector (no S1)
Lab Neutron Emission Angle (degrees)
LLNL-PRES-677393
S.(Sangiorgio(
13(
T.#Joshi,#S.#Sangiorgio,#et#al,#NIM#B#333#(2014)#
The*Li*target*
Same total Li content
Ep = 1.93 MeV
Li metal target
LLNL-PRES-677393
Li2O target
LiF target
S.(Sangiorgio(
Li2CO3 target
14(
T.#Joshi,#S.#Sangiorgio,#et#al,#NIM#B#333#(2014)#
P.#Barbeau#et#al,#NIM#A#(2007)#
Neutron*Filtering*
• Take advantage of nuclear physics to selectively transmit
neutrons through interference dips in scattering x-sections
40
Ar, 56Fe, and 48Ti (n,el) cross-sections
1
82 keV
10-2
73 keV
24 keV
Cross Section (barns)
102
56
Fe
Ar
48
Ti
40
10-4
20
40
60
80
100
Incident Energy (keV)
LLNL-PRES-677393
S.(Sangiorgio(
15(
T.#Joshi,#S.#Sangiorgio,#et#al,#NIM#B#333#(2014)#
Pb(
LAr(
detector(
Fe(filter(
Li(target(
106 neutrons/250eV/sr/mC
Crea'ng*a*low?E*neutron*beam*
Ideal 45º collimation
p(
Backgrounds:
- gammas from 7Li(p,p’)
- neutron-capture gammas in shield
- 24 keV neutrons through the filter
"
LLNL-PRES-677393
Measure at different
collimation angle,
normalize and subtract
106 neutrons/250eV/sr/mC
Ideal 55º collimation
S.(Sangiorgio(
16(
Expected*Recoil*Spectrum*in*LAr*
detectors
MCNP calculation of neutron transport and interaction
using detailed geometry
Lead
n
LAr
detector
Fe filter
Li target
BPoly
All scatters in active volume (x0.5) - 45 deg
Counts (counts/keV/ѥC)
3
2.4 keVr
p
Single scatters fiducial volume - 45 deg
Single scatters fiducial volume - 55 deg
2.5
Background subtracted signal - model input
2
6.7 keVr
1.5
Endpoint Measurement
4mM
MAX
En
TAr
=
(m + M )2
1
6.7 keV endpoint
0.5
0
Multiple
scatter
Scatters from
24 keV neutrons
0
2
4
6
Deposited energy (keV)
8
10
Endpoint measurement at 6.7 keV nuclear recoils
LLNL-PRES-677393
S.(Sangiorgio(
17(
LLNL’s*on?site*dedicated*neutron*beam*
1.7 MeV Tandem accelerator at LLNL’s CAMS
The collimator
setup
7mm dia Li
1µm thick
Proton
beam
Unique neutron facility for detector calibration
to low-energy neutrons (< 150 keV)
LLNL-PRES-677393
S.(Sangiorgio(
Protons on target
Beam on target
18(
T.#Joshi,#S.#Sangiorgio,#et#al,#PRL#112#(2014)#
Ioniza'on*Yield*at*6.7*keVr*
Fit(using(the(MCNP(spectrum(convolved(with(measured(detector(resolu/on(and(
three(free(parameters:(
+0.1
+0.7
−
Q
=
4.9
(stat)
(syst)
e
/keV
• fixed(ioniza/on(yield,((
y
−0.2
−0.9
• rate(normaliza/on,((
at 640V/cm
19(
S.(Sangiorgio(
• fano(factor((((
LLNL-PRES-677393
T.#Joshi,#S.#Sangiorgio,#et#al,#PRL#112#(2014)#
Uncertainty*Es'ma'on*
LLNL-PRES-677393
S.(Sangiorgio(
20(
T.#Joshi,#S.#Sangiorgio,#et#al,#PRL#112#(2014)#
Nuclear recoils
at 6.7 keVr at
varying electric
field
Counts (counts/electron/ѥC)
Electric*Field*Dependence*of*Ioniza'on*Yield*
= ܦ640 V/cm
ѯ2/ndf = 8.1/10
= ܦ240 V/cm
ѯ2/ndf = 3.4/8
0.12
0.10
Signal - all cuts
Background - all cuts
Background subtracted
Best fit of model
0.08
0.06
0.04
0.02
0.00
20
30
40
50
1600 V/cm
= ܦElectrons
Detected
2
ѯ /ndf = 12.6/9
60 10
20
30
40
50
Detected
Electrons
= ܦ2130
V/cm
2
ѯ /ndf = 11.2/10
60
20
30
40
Detected Electrons
60 10
20
30
40
Detected Electrons
60
counts/electron/µC
ܦ
ѯ
10
60 10
LLNL-PRES-677393
S.(Sangiorgio(
50
50
21(
T.#Joshi,#S.#Sangiorgio,#et#al,#PRL#112#(2014)#
For nuclear recoils
S2 ∝ ne = rNi
Ni + Nex
E
· q(E)
=
W
Use modified Thomas-Imel to
account for recombination using
parameters obtained from 2.82
keV electron recoils
Detected number of electrons
Field*Dependence*
70
2.82 keV – electron recoil
60
50
40
6.7 keV – nuclear recoil
30
20
Fit with Ni as only free
parameter
LLNL-PRES-677393
1000
1500
2000
2500
3000
Drift Field (V/cm)
Same phenomenological model of
recombination holds in both cases
"
For nuclear recoils the amount of initial
ionization Ni is unknown:
• Nex / Ni = ??
• E = 6.7 keV
• W = 19.5 eV
• q(E) = ??
500
Similarities in spatial distributions of
ions and electrons
S.(Sangiorgio(
22(
Comparison*with*SCENE*Measurements*
• Different(energies(and(electric(field(range.(Very(complementary(but(hard(
to(crossIcheck(directly(
• Agreement(on(recombina/on:(same(fit(result(for(the(electric(field(
parameter(‘b’(in(the(modified(ThomasIImel((b(=(0.61)(
• Combined(ioniza/on(yield(data:(
7"
6"
5"
4"
16.9"keVr"SCENE"
3"
25.4"keVr"SCENE"
2"
57.3"keVr"SCENE"
36.1"keVr"SCENE"
1"
6.7"keVr"LLNL"
Ionization Yield [e-/keV]
Ionization Yield [e-/keV]
8"
0"
30"
LLNL-PRES-677393
300"
5"
4.5"
4"
3.5"
3"
2.5"
2"
1.5"
1"
0.5"
0"
At 240 V/cm field
SCENE((interpolated)(
LLNL(
0"
3000"
Electric Field [V/cm]
20"
40"
60"
80"
Nuclear Recoil Energy [keVr]
S.(Sangiorgio(
23(
Modeling*Low?E*Nuclear*Recoils*in*Liquid*Argon*
ArIAr(Inelas/c(Cross(Sec/ons(
TRIMIbased(binary(collision(Monte(Carlo(Model(
• Elas/c(Coulomb(collisions(
• Inelas/c(collisions(producing(excita/on(
and(ioniza/on(
Ar + Ar " Ar* + Ar
Ar + Ar " Ar+ + Ar + e-
Recoiling
Ar
Ar+IAr(Inelas/c(Cross(Sec/ons(
Ar+ + Ar " Ar + Ar+
Ar+ + Ar " Ar + Ar+
Ar+ + Ar " Ar*
Notes:(
• Ioniza/on(energy(spectrum(is(not(well(known(and(
depends(on(collision(energy("(use(3(and(10(eV(
• ThreeIbody(collisions(are(neglected(
• BiIexcitonic(quenching(mechanism(not(included(
M.#Foxe,#C.#Hagmann,#et#al,#AstroparOcle#Physics#69#(2015)#
LLNL-PRES-677393
S.(Sangiorgio(
24(
M.#Foxe,#C.#Hagmann,#et#al,#AstroparOcle#Physics#69#(2015)#
Modeling*Results*
Track(Length((
Electron(thermaliza/on(length(
Electron(recoil(
Ioniza/on(Yield(vs(E(Field(
Ar(recoil(
Ioniza/on(Yield(at(1(kV/cm(
LLNL-PRES-677393
S.(Sangiorgio(
25(
Conclusions*and*outlook*
• Demonstrated use of 37Ar to calibrate down to sub-keV energies
• Measured the ionization yield at 6.7 keVr in liquid argon as a function of
electric field
• Developed atomic collision simulation for low-energy (< 10 keV)
interactions in liquid argon
• Appreciably good agreement
• Would be interesting to extend it to xenon
• Nuclear recoil measurements:
• Refurbishment of Li target for
higher neutron efficiency
• Access lower recoil energy using
different filters
• Xe target
• Things to consider:
• Liquid Argon vs Liquid Xenon
• Few-electrons backgrounds
• Single electron calibration
LLNL-PRES-677393
S.(Sangiorgio(
26(
Acknowledgements **
! A.(Bernstein,(C.(Hagmann,(K.(Kazkaz,(((
V.(Mozin,(S.(Pereverzev,(F.(Rebassoo,(
S.(Sangiorgio(
! T.(Joshi(
! P.(Sorensen(
! I.(Jovanovic(
! M.(Foxe(
LLNL-PRES-677393
S.(Sangiorgio(
27(
© Copyright 2025 Paperzz
