2 of 46

LUX-ZEPLIN and “The Rn Problem”

CWODS 2026, UC Riverside

July 1, 2026

Introducing Crystal Xenon (CXe)

Preliminary studies @LBNL (700g)

UT Austin studies (7kg)

Future prospects & conclusions

3 of 46

CWODS 2026, UC Riverside

July 1, 2026

LUX-ZEPLIN (LZ)

A direct dark matter detection experiment

Located 4850ft under Sanford Underground Research Facility (SURF)

1.7t GdLS in outer detector, for rejecting external backgrounds

9t LXe Dual Phase Time Projection Chamber (TPC)

[1703.09144]

6t water tank, for shielding from external backgrounds

4 of 46

CWODS 2026, UC Riverside

July 1, 2026

Liquid Noble TPCs

PMT Hitmap Gives �(x,y) Position

Pulse Timing Gives (z) Position

Reconstruct

Energy,�Particle Type

from Pulse Size

Particle interaction with Xe produces immediate scintillation (S1) and delayed ionization (S2)

5 of 46

ER β

WIMP

¹²⁴Xe

Accidentals

CWODS 2026, UC Riverside

July 1, 2026

The Rn Problem

[2410.17036]

6 of 46

CWODS 2026, UC Riverside

July 1, 2026

The Rn Problem

Removable w/ distillation, transient from calibrations

ER β

WIMP

¹²⁴Xe

Accidentals

[2410.17036]

7 of 46

CWODS 2026, UC Riverside

July 1, 2026

The Rn Problem

10x Sensitivity improvement

requires Rn mitigation

8 of 46

CWODS 2026, UC Riverside

July 1, 2026

The Rn Problem

NR ν

How do we reduce our Rn levels?

Just purify the Xe better - [2009.06069], [2502.04209]

Best current methods reduce Rn 2-4x at 1tonne/day
At 100% efficiency, would need to purify full detector in < 3.8d Rn half life
Need to exceed 10x lower Rn than world best, 10-100x higher flow rate for XLZD

Tag Rn events during operation - [2508.19117], [1506.02562]

“At a cost of 9.0% of exposure, [...] (63 ± 6stat ± 7sys)% identification of 214Pb decays to ground state”
Still has its limits

Isolate TPC from surroundings - [2209.00362v1]

“Most likely, a combination of several methods will be required”

Stop Rn from entering in the first place...

10x

9 of 46

CWODS 2026, UC Riverside

July 1, 2026

The CXe TPC

A LXe TPC allows Rn flow during operation...

... But a frozen Crystal Xenon (CXe) TPC

effectively excludes Rn* from the bulk

*See Backup Slides

10 of 46

The CXe TPC

Rn surface emanation is excluded from bulk
Rn in system decays away in 𝓞(100) days
Internal decays now stay at same (x,y,z) as parent, allowing for easier tagging
Subdominant to neutrino backgrounds

CWODS 2026, UC Riverside

July 1, 2026

Table III, [1802.06039]

Requires studies of Rn diffusion to verify

[2312.15082]

11 of 46

Not just Rn - accidentals are likely also mitigated in a CXe TPC.

Reduced phonon scattering → drift time down 1/2x → 2x fewer accidentals
Improved electron extraction efficiency* → lower extraction field needed → Lower rate of spurious electron emission
Better extraction efficiency* → Less electron trapping at gas interface → Smaller electron-trains?

CWODS 2026, UC Riverside

July 1, 2026

Accidentals

[2312.15082]

Requires studies at UT to verify

*See Backup Slides

12 of 46

Benefits	Concerns
Effective Rn exclusion*	Can CXe maintain VUV transparency?
Improved e- extraction*	Can CXe be maintained over extended periods?
2x e- drift velocity	Can HV be achieved in CXe?
Increased density / mass	Can CXe be extended to tonne-scale?

CWODS 2026, UC Riverside

July 1, 2026

The CXe TPC

*See Backup Slides

13 of 46

CWODS 2026, UC Riverside

July 1, 2026

CrystaLiZe @ LBNL

² “Towards a Neutrino-Limited Dark Matter Search with Crystalline Xenon”

-LBNL Group, 2023, [2312.15082]

700 g-scale CXe TPC

Established similar S1 and S2 signals in CXe as in LXe¹
Confirmed expected faster drift and higher e- extraction¹
Confirmed expected Rn exclusion²

(See plots in backups)

¹ [2201.05740]

14 of 46

CWODS 2026, UC Riverside

July 1, 2026

UT Austin Studies

Can we scale to 7 tonnes? Let’s start with 7 kg

Coldhead

Inner Cryostat Vessel

GXe

SXe

Heater

Copper Thermal Straps

SiPMs

15 of 46

CWODS 2026, UC Riverside

July 1, 2026

UT Austin Studies

Can we scale to 7 tonnes? Let’s start with 7 kg

SiPM Cabling

Camera Viewport

Xe Inlet

HV Cabling

Coldhead

16 of 46

CWODS 2026, UC Riverside

July 1, 2026

Crystal Growth Process

How do we freeze transparently?

Cool bottom of system to Xe triple point: 161.5K
Utilizing Bridgeman method w/ constant gradient ~1K/cm (see also Yoo et al.*)

CXe

GXe

LXe

Temperature

Triple �point

t₀

Height

*[1508.05897]

17 of 46

CWODS 2026, UC Riverside

July 1, 2026

Crystal Growth Process

How do we freeze transparently?

Cool bottom of system to Xe triple point: 161.5K
Utilizing Bridgeman method w/ constant gradient ~1K/cm (see also Yoo et al.*)
Gradually lower heater power to freeze system while maintaining gradient
Goal: Develop method to freeze as fast as possible, without sacrificing high VUV transparency / scintillation signal (S1)

*[1508.05897]

CXe

GXe

Temperature

Height

Triple �point

Vessel cools

w/ time

t₁

t₀

18 of 46

CWODS 2026, UC Riverside

July 1, 2026

VUV Transparency in Crystal Xe

As in LBNL setup, achieve similar S1 signal in CXe and LXe with careful crystal growth

¹³⁷Cs gamma ~660 keV

19 of 46

CWODS 2026, UC Riverside

July 1, 2026

Crystal Growth Rate Studies

Faster growth �= more signal loss
Many latent variables …

20 of 46

CWODS 2026, UC Riverside

July 1, 2026

Crystal Growth Rate Studies - Mass

Most data at 6 kg
Good freezes �(w/in error of 0 signal loss) in ~36 hr
Ongoing work to explore mass dependence

21 of 46

CWODS 2026, UC Riverside

July 1, 2026

Crystal Growth Rate Studies – Surface Quality

Opaque surface sometimes forms at end of growth
Fixable but takes time
No impact on S1 signal �(SiPM array below surface)

22 of 46

CWODS 2026, UC Riverside

July 1, 2026

Crystal Growth Rate Studies – End of Freeze

S1 signal loss reduced by slowing down final stages of freeze

23 of 46

CWODS 2026, UC Riverside

July 1, 2026

Crystal Growth Rate Studies – Greatest Hits

Technique refined over time
Less dependence on growth rate with better control / established process

Last three freezes

24 of 46

CWODS 2026, UC Riverside

July 1, 2026

Crystal Growth Rate Studies – Projections

*[1508.05897]

25 of 46

CWODS 2026, UC Riverside

July 1, 2026

Crystal Growth Rate Studies – Projections

How long to freeze LZ?
For larger detector, want to allow cooling from the walls as well
Additional cooling clamps stacked at various heights can “reset” freezing t₀ => 3 clamps = freeze LZ in 3 months

*[1508.05897]

Additional Cooling Clamp

26 of 46

CWODS 2026, UC Riverside

July 1, 2026

Fast Change in Temperatures Hurts Scintillation

~1.1 K/hr�cooling rate

27 of 46

CWODS 2026, UC Riverside

July 1, 2026

Controlled Change in Temperature is Fine

~0.4 K/hr�cooling rate

28 of 46

CWODS 2026, UC Riverside

July 1, 2026

Stability Established on 2-Day Timescale

29 of 46

CWODS 2026, UC Riverside

July 1, 2026

Conclusions and Future Work

This year

2-3 years

Future

30 of 46

CWODS 2026, UC Riverside

July 1, 2026

Thank you!

https://www.youtube.com/shorts/AeHplHhBV2c

Learn more at kravitz-lab.ph.utexas.edu

31 of 46

CWODS 2026, UC Riverside

July 1, 2026

Backup Slides

32 of 46

CWODS 2026, UC Riverside

July 1, 2026

CrystaLiZe @ LBNL

How do we measure transparency?

Use single photons as gain calibration
Measure ⁵⁷Co source before and after freezing
Result: comparable S1s and S2s

33 of 46

CWODS 2026, UC Riverside

July 1, 2026

CrystaLiZe @ LBNL

How do we quantify Rn reduction?

Measure rate of Rn alphas vs time
Rate matches half life of 3.8 days
Alpha counts go from 2000->3

LXe: ~2000 ⍺s

CXe: 3 ⍺s

34 of 46

CWODS 2026, UC Riverside

July 1, 2026

Initial Freezing Studies with Ar

Preliminary tests using LAr demonstrate the ability to freeze on this length scale

with transparency to optical photons

https://utexas.box.com/v/CrystaLiZeArRun1

https://utexas.box.com/v/CrystaLiZeArRun2

35 of 46

A flat CXe surface gives a consistent S2 shape/size
Freezing process generally forms ‘iceberg’ - CXe-GXe boundary causes temperature of remaining LXe to drop, causing a ‘flash-freeze’
Slowing freezing process at last minute should mitigate this, as seen in most recent Xe run
Can be measured using SE pulse shapes and sizes

CWODS 2026, UC Riverside

July 1, 2026

CXe Meniscus

CXe

GXe

Anode

Gate

36 of 46

CWODS 2026, UC Riverside

July 1, 2026

Single-Phase

Single-phase detector - operate in entirely CXe
No lower bound on temperature
We know there’s a temperature dependence on the longitudinal diffusion coefficient in xenon [2303.13963]
Diffusion limits multiple scatter resolution, essential in multiple scatter searches like in 0vbb [2104.13374], [2410.19016]
Proposal: freeze LZ as a very cold single-phase CXe 0vbb observatory

Lower

Diffusion

0vbb

37 of 46

CWODS 2026, UC Riverside

July 1, 2026

Hydrogen doping - allows for a xenon detector to pursue more sensitive low-mass and spin-dependent results [2505.13402]
Technical challenges associated with hydrogen doping:

Requires constant flow of hydrogen and xenon, with two separate flow paths
Hydrogen in Xe vapor can quench S2

Solved by freezing:

Hydrogen is trapped in CXe
Pure Xe gas can be circulated in

Hydrogen Doping

38 of 46

CWODS 2026, UC Riverside

July 1, 2026

Exciton Analysis

People working on LXe TPCs are used to saying ‘the excimer’. This is incorrect!
Actually multiple excimer modes: most common are Self-Trapped (STE, 7.2eV) and Free Exciton (FE, 8.35eV)
STE is orders of magnitude more probable, as Xe atoms are constrained by jostling of neighbours... Until you reach CXe

DOI:10.1002/pssb.2221850126

39 of 46

CWODS 2026, UC Riverside

July 1, 2026

Exciton Analysis

We have a VUV optical filter which will reject only FE excitons
Plan to take first-ever data at this scale and temperature
Can we observe a difference from the absence of the exciton mode in CXe?

If this difference is considerable and differs between particles, this could improve ER/NR discrimination, as with recombination time affecting S1 shape in LAr

40 of 46

CWODS 2026, UC Riverside

July 1, 2026

Example Waveform (Bi-Po)

41 of 46

CWODS 2026, UC Riverside

July 1, 2026

Calibration

B10

A limitation of a CXe TPC we’re thinking about: you can’t circulate calibration sources if you’re frozen
External sources are fine, but e.g. Kr, Rn, CH3T injections become difficult
Suggestion: freeze a layer of calibration source into your detector
Allows for mid-run calibration as a sideband

42 of 46

CWODS 2026, UC Riverside

July 1, 2026

Electron Emission

B11

Electron extraction efficiency is slightly higher in solid (highlighted) compared to liquid

Lower conduction band
Lower dielectric constant

43 of 46

CWODS 2026, UC Riverside

July 1, 2026

ER/NR Discrimination

B12

Double electron mobility in CXe - does this influence more than just accidentals?
On paper, would lead to less recombination of thermal electrons → smaller fluctuations in S1-S2 contour → better ER-NR discrimination
UT group plan to test this with our ER sources and an NR source

Proportional to electron drift?

44 of 46

CWODS 2026, UC Riverside

July 1, 2026

Rn Tagging

B13

45 of 46

CWODS 2026, UC Riverside

July 1, 2026

Rn Tagging

B14

Tagging becomes much easier in CXe with little to no atom transit!
Rn diffusion qualitatively constrained in LBNL setup (see right)

<1mm shift in 3 mins, consistent with 218Po half life
Implies <1cm shift in 214Bi over 20 mins

Quantitative measurements will be determined with the UT test bed

1 of 46

2 of 46

3 of 46

4 of 46

5 of 46

6 of 46

7 of 46

8 of 46

9 of 46

10 of 46

11 of 46

12 of 46

13 of 46

14 of 46

15 of 46

16 of 46

17 of 46

18 of 46

19 of 46

20 of 46

21 of 46

22 of 46

23 of 46

24 of 46

25 of 46

26 of 46

27 of 46

28 of 46

29 of 46

30 of 46

31 of 46

32 of 46

33 of 46

34 of 46

35 of 46

36 of 46

37 of 46

38 of 46

39 of 46

40 of 46

41 of 46

42 of 46

43 of 46

44 of 46

45 of 46

46 of 46