Barium Ion Tagging :

Ion Acquisition in LXe &Laser Fluorescence Identification

P.C. RowsonSLAC

As has been outlined in the plenary talk, the tonne-scaleEXO experiment can reach the 10 meV mass scale

exploiting the dramatic background reduction provided bycoincidence tagging of the barium daughter

of xenon double beta decay.

The ongoing R&D program will be summarized here.

Background reduction by coincidence measurement

It was recognized early on that coincident detection of the two decay electrons and the daughter decay species

can dramatically reduce bkgrd.

X (Y++)* + e– + e–

Y++ +

One possibility would be the Observation of a from an excited daughter ion, but the rates compared to ground state decays are generally very small(best chance might be 150Nd, but E is only 30 keV.)

136Xe 136Ba++ + e– + e–

Identify event-by-event

A more promising approach : Barium detection from 136Xe decay

Described in 1991 by M. Moe (PRC, 44, R931,(1991)).The method exploits the well-studied spectroscopy of Ba and the demonstrated sensitivity to a single Ba+ ion in an

ion trap.

Our decision to proceed with a LXe TPC (as opposed to gXe)led us to investigate ion retrieval, or “ion to laser”, schemes

for barium tagging.

So far, this work has proceeded in parallel mainly at Stanford and SLAC :

• Laser ion trap program - trap design & operation• Ion capture program - electrostatic probe designs

• Interface program* - ion-to-trap transfer

* (recently begun)

In addition, at CSU, W. Fairbank, is investigating :

• “in situ” tagging - laser tagging in LXe

Barium Tagging R&D

Xe136(p,n)Cs136 : Cs136 production by cosmics (Cs→Ba via decay)

Xe136(,)Cs136 : Cs136 production by solar neutrinos

Xe136(n,γ)Xe137 : Xe137 production by cosmics (Xe→Cs→Ba via )

Correlated sources of barium ions have been investigatedand appear to be negligible. Rates are low and in addition,

event topologies should be distinctive. Detailed MC simulation has

not yet been deemed high priority but will be done.

Comment on Barium backgrounds

Barium atoms hypothetically present in the xenon would notnormally constitute a background, as we only collect barium ions.

Barium ions from 2 decay are produced in the xenon at a ratenot yet determined, but limited to ~300,000/tonne-year, or roughly1 per 100 seconds per tonne. These are continually swept out of

the liquid by the TPC E-field in < 30 seconds for our nominal~3 kV/cm field strength.

(The ion mobility is known - more on this later).

... some preliminary studies …

Liquid Xenon TPC conceptual design

The basic concept, shown here for a LXe option, is :

• Use ionization and scintillation light in the TPC to determine the event location, and to do precise calorimetry.

• Extract the Barium ion from the event location (electrostatic probe eg.)

• Deliver the Barium to a laser system for Ba136 identification.

Compact and scalable (3 m3 for 10 tons).

Ion capture in LXe TPC

175 nm scintillation

e-

Issues to be addressed (R&D progress where indicated) :

Ba+ lifetimes in LXe (expected to be long - data exists)

Ba ion drift velocities (should be a few mm/sec - confirmed)

Ba capture and release – various probe designs

Ba transport to the laser spectroscopy station

cathode

TPC charge & UV detection

LXe level

electrostatic probe

Basic electrostatic capture procedure

Probe motion (3 d.o.f.) triggered by event E threshold.

The probe moves above the TPC, and then vertically down to the event location.

The Ba+ is collected electrostatically (doesn’t move far from the event location),

and the probe is withdrawn.

Ba++ lines in the UV – convert ion to Ba+ or Ba.

“Intermodulation” “Shelving” into metastable D state allows for modulation of 650nm light

to induce modulated 493nm emission out of synch. with excitation (493nm) light – improves S/N

Laser fluorescence barium identification

A well-studied techniquepioneered by atomic

physicists in the 1980’sfor the detection of single

atoms and ions, in particular,alkali and alkaline-earth

metals.

RF appliedBa oven

laser, Ba ionizerand detectionline-of-sightthrough these

gaps

blue laser

red laser

reference cavities

Laser Spectroscopy Lab at Stanford

Stable and reliable laser system

Ion trap : hyperbolic Paul type

The trap is loaded with multiple ions:We observe the signal intensity as ions

are dropped one by one…

Glare fromelectrodes

Single Ba+

signal

The effects of buffer gas on trap performance

The operating environment of the EXO ion trap will likelyinclude some level of background xenon gas, and the effects

of this “buffer gas” have been studied.

It has been found that the addition of helium canimprove trapping times (which are essentially indefinite

for UHV conditions for modest xenon pressures.

Differential pumping can/will be used to maintain a low ion trap buffer gas pressure.

ßß Decay then Ba++

Ba+

CCD/APD

Concept for Ba+ tagging in the Liquid in a LXe Double Beta Decay Experiment :

“laser-to-ion” schemes.

FiltersSlit

Laser

Fluorescence

Focus

Ba+ cloud image in liquid xenon

Liquid surface

Grid

8 mm

CCD camera image of Ba+ fluorescence in LXe

At CSU, fluorescencedata in LXe has taken, andstudies are continuing. The

issues here are :

• Line broadening/loss of specificity.

• S/N improvement for in situ ion detection.

230Pa (17.4d)

230U (20.8d)

222Ra (38s)

226Th (30.5min)

8.4%

5.99MeV

6.45MeV

Pa produced in a cyclotronPa produced in a cyclotron230230Th + p Th + p 230230Pa + 3nPa + 3n

3-steps of decay

Ion capture test simulates

Ba ions by using a 230U source to recoil 222Ra into the Xenon – Ba and Ra are

chemically similar (ionization potentials 5.2 eV

and 5.3 eV respectively).

Barium ion extraction R&D at SLAC

1st Prototype electrostaticprobe – W tipped.

Variations have been tried (diamond coated),but ions not released by

reversed HV in these cases(required E field too high)

Xenon cell

Probe lowered for ion collection (1)

Electrode (source)

PMT

3-position

pneumatic actuator

probe up position

for release (3).

Xenon cell

outer vac. vessel

detector flange

counting (2) station

Probe test cell

230230U source U source αα spectrum spectrumas delivered by LLNLas delivered by LLNL

(measured in vacuum)(measured in vacuum)

αα spectrum fromspectrum fromwhatever is grabbedwhatever is grabbed

by the tipby the tip(in Xe atmosphere)(in Xe atmosphere)

An additional An additional signature from signature from

the observed Th the observed Th and Ra and Ra

lifetimes.lifetimes.

Ion extraction from Xe and LXe

Ion mobility studies in LXe

We use the probe test cell to measure ion drift speed

Observed mobility of 0.24±0.02 cm2/kVs for Thorium ions compares with result for Thallium ions 0.133 cm2/kVs. (A.J. Walters et al. J. Phys. D: Appl. Phys.) and with Fairbank etal. for EXO (Ba,Sr,Ca,Mg).

Our work submitted to Phys. Rev. B.

Modulate the electrode voltage,and measure ion collection rate.

Data taken for various separationdistances and voltage differences.

LXe level“Paddle” probe

U230 source electrode

forward bias

reverse bias

Ion Capture “Cryo Probe” prototype

incoming gas(inner tube)

gas return(outer tube)

small apertureat tube end

In order to release a captured ion, the electrostatic probe can be cooled such that Xe ice coats the tip. The captured ion can then be released by thawing.

Joule-Thompson cooling is usedfor cooling (argon gas).

An additional benefit : the Ba+

charge state may be stable in solid and liquid xenon.

Expected gas cooling from calculated J-T coefficient and our data with cryoprobe.

Argon

Probe tip detail

Remarkably, surgicalcryoprobes seem to be ideally suited to our application. Wehave adapted 2.4 mmdiameter probes for use in our probe test cell.

Testing the ion extraction probe

U230 sources were installed, xenon was liquefied in the cell, ion capture and release from Xe ice has been demonstrated.

First cryo-probe was not equipped for acceptably “graceful” Xe ice release. New version is under test.

Refinement of ion release procedure (rapid ice sublimation is best).

Issue for cold probe method - Xe gas release

X-ray imageof new cryoprobe

TC J-Tnozzle

Vacuumjacket

2.4mm

test version of “thaw” heater

Saha-Langmuir effect :Ion emission from heatedhigh-work function surfaces(shown here for alkaline earth metals)known from ion beam experiments

May be possible to release Ba+ ions by heating Pt probe.This procedure would be simpler than the cold probe.

Method requires the Pt surface is heated to a high enoughtemperature to efficiently liberate barium, but not so high that neutral atoms become a significant fraction w.r.t. ions.

Surface ionization or “hot probe” R&D

10eV5.2eV ~5.9eV

Ba Pt (111)

First (eV) Second (eV)Tl 6.1 20.4Th 6.1 11.5Ra 5.3 10.1

kT

We

n

n )(exp2

work function ionization E

It is well known that heatedmetal surfaces can release

captured metal atoms in boththe neutral or ionized state :

“impact ionization”

Movable Pt Foil

source (Th228)

Alpha counter

source collimatorplate

stopperplate

HV

HV

heater PS

Vessel is filled with 1 atm Xe.This limits the diffusion of the

ions. The α’s range out in ~5 cm

Test apparatus for thermal ion release experiments

Th228 (1.9 yr) source produces Ra224 (3.6 d) daughters

• Source can be forward or backward biased (±500 V typ.)

• Pt foil (@ ground) receives ions recoiled from source.

• Foil can be moved in front of detector, and down to the stopper plate.

• Foil heated >1000K, see if Ra released as neutral or charged.

(if the observed post-heating signal is modulated by the HV on the plates, ions were released)

Pt foil(power leads visible

as is the mounted TC)

At top, the Th228 source

Below, the Si SB α detector

Test apparatus : Source collimator not installed for this photo

E field calculation for collimator. Ra224 deposited near foil center

Ra-224

Bi-212Rn-220

Po-216

Po-212

Red histo : alpha spectrum from foil prepared with reversed biased source → Ra ions do not reach foil.

Black histo : … and when source at + potential → foil plated.

Experiments are underway in out lab to test the performanceof a Pt foil.

If promising, we will proceed to design a hot probe, and experiment with different metal tips

(iridium is a possibility - higher m.p. than platinum),and perhaps high-work-function dielectrics.

Recently, a third probe option is under study at Stanford - High field emission from “STM” tip,

or “sharp probe” R&D

Published data suggests that barium will desorb from tungsten needle tipsas a Ba+ ion at electric fields of ~150 MV/cm. These high fields can bereached with very sharp STM needle tips (radius of curvature of ~10 nm)

at moderate (10 kV) voltages .

Electric field calculations for ion capture are underway.

One of the issues here will be the robustness of these delicate sharp tips

SEM image of W needle

1. event energy & space location from TPC

2. “ion fetch” triggered by energy threshold & ~veto

3. TPC field switched off (prior ion drift very small).

4. move probe tip to (just above) ion location.

5. capture ion electrostatically with ~1 cm radius.

6. withdraw probe - TPC field back on - detector live

7. deliver ion to laser for identification.

Acceptable deadtime/Δt for steps 2-6 sets maximum “ion fetch” rate.

Our measurements of the mobility of ions (Th and Ba) in LXe indicate adrift speed of ~2 mm/s in a 1 kV/cm E field. For a 1 mm radius

probe tip, this translates into a 0.8 s collection time from 5 mm, 3.8 sfrom 10 mm. The deadtime will be dominated by probe motion and/or

high voltage ramping, if necessary - < 1 minute a reasonable target.

Backgrounds/trigger threshold sets “ion fetch” trigger rate.

While it is difficult to extrapolate from our prototype simulations to a largemulti-tonne detector, we can guess by scaling our bkgrd. simulations bya factor of 10 tonnes/200 kg = 50. For a low energy trigger threshold of

2.250 MeV (for an E resolution of 1%, this corresponds to 10σ), trigger rate would be < 1/hour. This is a plausible “ion fetch” rate.

(2 events not as important for the large detector - these and other low energyphenomena can be acquired using a scaled trigger).

Issues for Trigger rates

We have decided to focus our initial R&D efforts onan interface between an electrostatic probe and a linear

ion trap, including cryo- and differential pumping.

We have made progress studying electrostatic probes.A number of issues remain …

The ion-release procedure for the designs considered todate will have different challenges (assuming the basic

concepts are fully demonstrated in R&D).

• The cold probe will deliver a larger Xe load – Is effective pumping possible ?

• The hot probe may release Ba if it is present in the probe surface material –

Sensitive tests needed during R&D.

…and a bit further down the road …

• Significant engineering problems will need our attention –R&D for probe “robot” and interface to the TPC.

The probe-to-ion trap interface

probe unloading TMP/cryo pump ion trap/laser tag

Linear trap confinement : radially by RF quads, axially by DC fields

detect and counttrapped ions

conventional Ba+

loadingcapture ions on

probe tip

6 mm

600 mm

linear trap RF quadrupolessegmented (15 sections)to grade DC axial field.

linear trap vacuum chamber(excluding probe interface section)

There is considerable experience among nuclear/atomicphysicists with ion transport in linear traps.

Parts are on order for a linear ion trap to be built at Stanford.

R&D continues on trap/probe interface at SLAC/Stanford.

Linear Paul ion trap R&D

release ions to trap,detect and measure

efficiency

Progress to date

• We have developed the atomic physics and spectroscopy techniques to achieve good quality tagging in presence of some

Xe gas.

• Gained experience with grabbing on Xe-ice and on metal tips

Continuing R&D

• Building a linear trap that should be very close to final device and can be used to test loading efficiency.

• Ion release needs more R&D work, field emission from STM-tip, “impact” ionization and “cryoprobe”

all under development in parallel. Highest present priority/risk

Ba tagging R&D must continue in parallel with the construction of the 200 kgexperiment in order to move EXO

towards the 10 meV regime.