Scintillation Detector Development for LArTPC Experiments

ICATPP, Villa Olmo, Como, Thursday, 26th of September 2013

Ben Jones, MIT

Liquid Argon Scintillation Light

Liquid argon produces scintillation light at a wavelength of 128 nm.

Llight yield ~ few 10,000’s of photons per MeV (dependences on E field, particle type and purity)

Argon is transparent at 128nm, which makes LAr scintillation detectors very scalable.

Coupling scintillation detection with charge detection (eg in a TPC) offers many benefits

J Chem Phys vol 91 (1989) 1469 E Morikawa et al

Our Motivation: MicroBooNE OpticalSystem

32 cryogenic Hamamatsu R5912-02mod PMTs (with platinum undercoating)

Mounted on a rack behind the TPC wireplanes

Each PMT has a magnetic shield and wavelength shifting plate

A 14m optical fiber runs to each PMT, couped to an LED outside the cryostat

On paper…

For more info on MicroBooNE, see M. Webers plenary talk

2.5m Drift

Installation of final PMT rack

Left to right: Matt Toups

BJPJ Eric James

Janet Conrad

not shown:Teppei Katori

Photograph as installed, illuminated with amber

lights

And in real life!

Our Motivation: R&D towards large detectors

PMT-and-plate strategy not scalable to an N-kiloton scale, multi-TPC detector like LBNE

We are also working on lightguide based detectors to slide between TPC units

MicroBooNE will contains 4 prototypes as a long term R&D exercise

We have a dedicated lightguide test stand at MIT, and work is performed in collaboration with Indiana University

(in collaboration with Indiana University)

Matt Toups, MIT

Wavelength shifting plate (TPB)

One MicroBooNE PMT assembly

Mu-metal shield + mount

Tetraphenyl Butadiene

128 nm light will not penetrate, glass, air, acrylic, etc.

This is a problem for the design of optical liquid argon detectors.

Common solution is to use a fluorescent chemical like tetraphenyl butadiene (TPB)

TPB absorbs 128nm light and emits it in the visible.

In MicroBooNE we use a coating of 50% TPB in polystyrene, dissolved in toluene brush coated on acrylic

Environmental sensitivity As part of the MicroBooNE

system development we optimized coatings for performance, robustness and stability

During these investigations we found that TPB is very sensitive to UV light and degrades in performance

We also sometimes observe a yellowing of the coating, after ~days of lab light exposure

Photodegradation Mechanism

Working with GCMS we also identified the degradation mechanism – radical mediated photo-oxidation to benzophenone

Radical Mediator StudiesSome stabilization of the coating is possible using a radial mediator

Here we find a 20% admixture of 4-tert butylcatechol improves performance + somewhat slows degradation rate

But certainly, there is much room for improvement + further work here

4-tert butylcatechol

Cryogenic PMT

Testing MicroBooNE PMTs

Every PMT for MicroBooNE has been characterized both warm and in liquid nitrogen

Measurements include gains, dark rates, and stability over few days of operation

Largely the work of Teppei Katori, MIT. Full report published in JINST.

Full assembly characterization

Bo Vertical Slice Test

A long term, high purity liquid argon test stand

Used to make detailed characterization of a few PMTs and supporting hardware:

Cryogenic PMTs Base electronics Wavelength shifting plate High voltage system +

interlocks Cables and splitters Readout electronics Cryostat feedthrough Trace impurity monitors Etc…

uB style PMT assembly

Full assembly characterization :

Lots of results, but no time to tell you about them…

Understanding light yields in scintillation detectors UV photon

But also a LAr scintillation R&D detector

The Effects of Nitrogen in LAr Unlike oxygen and water, nitrogen

does not disturb charge drift in LArTPCs, and is difficult to remove from argon.

Nitrogen is an expected contaminant in any present or future large LArTPC detector (especially with vacuum-free purge)

Nitrogen at the ppm level leads to :

1) Scintillation Quenchingmeasured in a detailed study by the WArP collaboration in small test cells (R Acciarri et al 2010 JINST 5 P06003)

2) Absorption of Scintillation LightAbsorption effects of N2 in LAr have not previously been measured. Very important to know for big detectors!

From(R Acciarri et al 2010 JINST 5 P06003)

Add nitrogen, monitor light yield at 2 source positions

Light loss due to N2 in 8” source configuration

27ppb N23.7ppm N27.4ppm N215.5 ppm N2

Measure intensity of polonium alpha peak

Divergence of 2 curves shows absorption effect

8”

14

.5”

PMT

Nitrogen Results:

Attenuation strength :

Some characteristic LAr samples :

Underground Argon for DM Experiments Dark matter LAr experiments

suffer from pervasive 39Ar background

39Ar is a beta emitter with endpoinr 565 keV and a half-life of 269 years

Produced by cosmic ray interactions in air

Industrial argon distilled from air contains significant 39Ar

Underground argon extracted from carbon dioxide wells has a much lower 39Ar concentration.

Underground argon distillation column at

FermilabFor more information: arXiv 1204.6024, 1204.6061, 1204.6011

Methane as a contaminant

Unlike industrial argon, UAr contains methane as a contaminant

Concentration of argon through distillation also concentrates methane

Can be removed using hot getters – but very expensive

Methane has been shown not to harm charge collection

No spec exists on the allowed methane concentration in a LAr scintillation detector

Gas composition from CO2 well

Distillation concentrates both methane and argon

+ submitted to JINST

We made a study of absorption, quenching and visible re-emissions of methane / argon mixtures.

Key conclusions:

Purity spec seemsto be about 10ppb

We see no signs of visible re-emission features

At higher concentrations (50-100ppb) some quenching is observed

Most losses are due to UV absorption

Lig

ht

yie

ld f

rom

alp

ha s

ou

rce

(PE)

Methane was discovered by local Como hero Alessandro Volta!

On vacation in 1776, Volta collected gas he noticed bubbling from mud in Lake Maggiore

Interest piqued by a recent paper from Benjamin Franklin on “flammable air”, he discovered the gas was flammable

By 1778, he had isolated methane from the marsh gas.

Statue of Volta in Piazza Volta, Como Volta’s summer holiday activities

Just for fun:

Summary At MIT we have been developing scintillation detectors for

current and future liquid argon TPC experiments

This includes development and installation the MicroBooNE optical systems, and work on prototype systems for LBNE

We have made studies of the performance and photochemistry of wavelength shifting coatings

Using high purity test stands we have both characterized detector elements and made R&D measurements

Nitrogen absorption is important for large LArTPCs - purity spec of 2 ppm is sufficient for MicroBooNE

Methane absorption is important for DM detectors - purity spec of 10 ppb is likely appropriate.

Thank you for your attention!

Backups and / or no time

Some technical achievements of Bo VST…

Measurement of global collection efficiency of PMT assembly

Linearity of PMT / base / splitter system up to 300 PE

Development of PMT gain and timing calibration methods

Successful operation of MicroBooNE PMT readout and trigger electronics

And more

LED pulses read through uB electronics

Scintillation spectrum from sourceto extract collection efficiency

(more on this in next slides)

LArTPC Detectors

MicroBooNE at FNAL

Ability to build massive detectors with long drift distances make LArTPCs appealing for neutrino detection

A LArTPC also offers bubble-chamber level position resolution and excellent calorimetric resolution with a large active volume and electronic readout

Simulated MicroBooNE event, reconstructed in 3D

ICARUS at LNGS

Why do TPCs need Optical Systems?

Typical LArTPC has a finite drift time (~ms). A priori we don’t know the interaction position.

So a LArTPC in a beam integrates milliseconds of cosmics around the beam gate

A correlated flash from the optical system allows timing of subevents to be specified to the few nanosecond level

This timing information can be used to reject cosmic rays (+other uncorrelated BGs)

Simulated MicroBooNE event with on-beam reco flash position from optical system overlaid

MicroBooNE Optical Calibration System

Feedthrough

LED

Fiber

(One fiber per PMT)

Pulser

An LED driven optical fiber calibration system will be used to:

1. Calibrate gains,2. Time in the PMTs 3. Test PMT functionality during commissioning.

In situ MicroBooNE PMT illuminated using calibration system LEDs

Trig pulse

PMT waveform

Light spot from fiber on PMT

Fiber installation was completed 1 week ago

We have already used parts of the system to verify the connection and functionality of all 32 PMTs

35

Experimental Configuration for This Study

Prompt peak window

Light in Liquid Argon The scintillation light in liquid argon is produced copiously

alongside all ionization charge deposits. There are two scintillation pathways, with different time

constants – a fast component with t=6ns and a slow time constant with t=1500ns.

ArAr

p+

Ar Ar* *1Σu excimer

Ar

Arγ

6ns

e

Ar Ar+

-

Ar

p+ Ar

e -

+

Ar Ar *3Σu excimer

1590ns

Special bonus – possible PID information

Ar Ar *Ar Ar *

Ar

Ar

ArAr

Ar Ar *Ar

Ar γ

•Utlized in dark matter searches (MiniCLEAN, DEAP), and we are investigating the applications of this technique to augment TPC based particle ID in MicroBooNE.

Scintillation process

Competing Excimer Dissociation Process

Pulse shape discrimination – a vital tool in dark matter detection, also useful to us!

Individual components (separated using PSD)

Fit function for alpha + background

General Idea: Source set in one of two possible

positions.

Controlled amounts of N2 injected into the liquid

Quenching affects both source positions equally

Absorption hinders the further more than the nearer source.

If fractional losses from each source deviate we see an N2 absorption length effect.

A future analysis will address the effects of quenching (more extensively studied by other groups) separately.

14

.5”

PPM amounts of nitrogen are injected into the liquid from a gas canister, charged to a known pressure.

From known volume of canister and known pressure we can calculate how many ppm we injected.

Nitrogen concentration monitored in both liquid and gas phases using LDetek8000 N2 monitor

We also monitor H20 and O2 to ~10ppb precision from the same sample lines.

Trace nitrogen monitor

Injection Canister

Kindly loaned by Jong Hee Yoo – Thanks!

Attenuation Data

PreliminaryDivergence of these two lines is clear evidence for the nitrogen absorption effect!

Stability of 1PE

- SPE scale stable to within 1% for each run

- This is similar to the precision of our SPE measurements

- Therefore we assume constant and fold in variations as a systematic error on each point

Just to be sure its really the nitrogen…Preliminary

No light loss during periods with no nitrogen injection – gives confidence in system stability, constrains outgassing effects, etc.

Getting to the Attenuation Strength

Measured Attenuation Strength:

Measured Absorption Cross Section:

Preliminary

Comparison to N2 gas absorption cross section world data

Preliminary

Nice result, but whats it gonna do for me?

$$$$ $

Preliminary

Summary + Prospects Bo VST has been constructed to test elements of MicroBooNE

optical system – also an R&D detector for LAr scintillation light.

Detailed studies of alpha source response have been made and area used in various Bo VST studies

We have measured the effects of nitrogen absorption of 128nm argon scintillation light in liquid argon. We find that the effect is on the order 0.015% / (ppm cm)

This means absorption is no problem for MicroBooNE, and could be useful information for the design of cryo systems for large LArTPCs

Backup Slides

Understanding the Geometrical EffectRay trace to understand

expected light yields per percent of absorption at each position

8”

14

.5”

Taking ratio, any quenching effect cancels

Ratio = Light loss at 8”

Light loss at 14.5”

Our region of interest

We will measure the nitrogen absorption effect as % light loss per ppm^-1 cm^-1.

First, measure the light loss ratio as a function of N2 concentration.

In our region of interest the relationship should be ~linear.

Absorption strength extracted by comparing the gradient of the measured line to the gradient of the line right, which gives proportionality factor for X axis scales.

This factor tells us the % light loss per ppm cm of nitrogen.

2) Measurement from liquid and gas capillaries in agreement with saturation pressure based equilibrium calculation

1) Amount of N2 in liquid agrees with amount injected to within our uncertainty of the injection volume.

How do we know we get N2 concentration right?

Inje

ctio

n vo

lum

e un

certai

nty

regi

on

Single exponent power law (cosmic background) + Poisson (alpha source)

Detected light spectrum – clean

argon, source at 8”

Check on functional form of fits:

Power law background is great. Alpha fit needs improvement (not exactly poissonian).

Why?“Shadowing” of outer source edges leads to reduced poisson mean light yield from edge area elements

This leads to an enhanced low tail of the source spectrum

Disc source kindly loaned by Adam Para – Thanks!

So we Measure the Shadowing Function…

Now we know how the source is shadowed, we know how to fit all points.

Improved fit from shadowing function

Major improvement with new fit function.

Note : no extra free parameters, since shadowing function was tuned on an independent dataset.

Aside: Pulse Shape Discrimination in Action

Alpha enriched

Cosmic only

Satu

ratio

n

PMT Characterizations for MicroBooNE Measured dark rat

63

128

nm

1.18 ± 0.1Visible photons out / UV photon in for evaporative TPB

Geh

man

et a

l