UC Santa Barbara

Numerous astrophysical observations point to the existence of dark matter,

making up about a quarter of the mass-energy budget of the Universe.

Although its exact nature is unknown,

the quest for dark matter has been a long-standing journey

in physics.

The most popular dark matter model is the

weakly massive interacting particle (WIMP).

The majority of the direct detection dark matter experiments

searches for rare nuclear collisions between the WIMPs

and the target nuclei

in low background detectors located deep underground.

No conclusive detection of nuclear recoil (NR) dark matter interactions

has been reported up to date.

A compelling feature predicted by various dark matter models

is the annual and diurnal modulations in the interaction rates measured by the dark matter detectors

as a result of the Earth’s motion in the halo.

The DAMA experiment has observed an annual

event rate modulation in the low energy (2-6 keV) regime of the

electron recoil (ER) events

for two decades. Motivated by that,

a search for both annual and diurnal rate modulations in the low energy

ER events was performed by the

Large Underground Xenon (LUX) experiment.

LUX is a liquid xenon (LXe)-based dark matter detector located

at the 4850 ft level of the Sanford Underground Research Facility (SURF)

in Lead, South Dakota, USA. LUX operated in WIMP search mode

between 2013 and 2016, finding no evidence for WIMP NR interactions.

The analysis steps and results of the search for

annual and diurnal modulations in the LUX experiment are presented

in great detail.

LUX-ZEPLIN (LZ)

experiment is a second generation dark matter direct detection experiment, which is

currently being constructed at SURF and on track for commissioning in 2020.

LZ will contain

about 10 tonnes of LXe, which is more than that of LUX by nearly a factor of 30,

in order to reach an unprecedented WIMP detection sensitivity.

Another distinctive characteristic of LZ is

the large gadolinium-loaded liquid scintillator (GdLS) volume ($\sim$17 tonnes)

held by acrylic tanks surrounding the LXe volume.

The main goal of this outer detector (OD)

is to veto the neutrons that escape the LXe after single-scattering.

They are mostly ($\sim$88\% of the time) captured by the gadolinium isotopes ($^{155}$Gd and $^{157}$Gd)

in the medium that have large capture cross sections. The liquid scintillator converts the

energy released following the neutron capture into detectable light. Thus, the NR collision from a neutron

in the LXe

mimicking a WIMP interaction is rejected thanks to the coincident OD signal.

During the final design review for the OD acrylic vessels,

the reviewers noted that GEANT4 has a shortcoming

of depicting the Gd deexcitation cascade

and that GEANT4-based

evaluations of the LZ OD neutron efficiency

were most likely inaccurate, and they encouraged the OD group to

find and implement an improved simulation.

We adopted the DICEBOX (a nuclear physics software)

simulation of the deexcitation

cascade after neutron capture

on $^{155}$Gd and $^{157}$Gd.

The DICEBOX model addresses various

issues in the default GEANT4 final state

neutron capture model.

The implementation of the

improved

$\gamma$-cascade model after neutron capture on $^{155}$Gd and $^{157}$Gd is reported.

GEANT4 simulations indicate that neutrons tarry in the acrylic tanks

prior to the capture in the OD. In the LZ experiment

searching for rare WIMP interactions,

this can create the peril of confusing the

NRs from neutrons that are followed by a late capture signal (outside the veto window)

with those from WIMPs. In order to attack this problem,

LZ will carry out a tagged neutron source calibration.

In a common neutron source, such as AmBe,

the neutron emission

is usually followed by a high energy $\gamma$, which is prompt.

This 4.4 MeV $\gamma$ detected by a fast, dense and inorganic scintillator crystal

will serve as a tag for the neutrons within 5-10 ns of their birth.

The objective is to accurately measure

the neutron detection efficiency and the OD neutron capture time

in comparison with the GEANT4 predictions

in a way that is independent from the LXe detector.

The design and the simulation of the in-situ tagged AmBe neutron source for the OD

are presented. Furthermore,

the results of the neutron tagging experiments performed

at UCSB with the tagging detector built for this calibration are

discussed.