1. Progress of Megatile Studies
Yong Liu
Johannes Gutenberg Universität Mainz
CALICE Collaboration Meeting
Mar. 23, 2017, LLR, Palaiseau
Volker Büscher, Phi Chau, Reinhold Degele, Karl-Heinz Geib, Sascha Krause,
Lucia Masetti, Ulrich Schäfer, Stefan Tapprogge, Rainer Wanke

2. In a nutshell: mass assembly of AHCAL
Scintillator Tile
SiPM
Reflective foil
PCB
Surface-mount tile design (Uni-Mainz)
Suitable for mass assembly
Optimized with Geant4 full simulation for response uniformity
Detector unit: a scintillator tile (30×30×3 mm³) with a SiPM
1st Mainz SMD-HBU with DESY
1st board built successfully in 2014
Adopted as a baseline design for the technological prototype
6 new SMD-HBUs fully assembled in 2016
SMD-HBU: Bottom View
SMD-HBU: Top View
Further simplify mass assembly?
Y. Liu et al, IEEE NSS 2014
DOI: /NSSMIC
CALICE Collaboration Meeting at LLR

3. CALICE Collaboration Meeting at LLR (yong.liu@uni-mainz.de)
Overview: megatile
Motivations: further simplify mass assembly
Scintillator plates with embedded structures
Optically isolated for individual readout
Structure optimized by Geant4 simulation
MIP response, cell-to-cell crosstalk
CALICE Collaboration Meeting at LLR

4. Megatile designs Cut out trenches (or grooves) on a scintillator plate
Fill in the trenches with white paints
To enhance mechanical stability, also mixed with glue
Reflective Foil (ESR)
Trench schematics (side view): not in scale
Design 1:
double trenches
SMD-SiPM
Optical trench
(TiO2)
SMD-SiPM
Optical trench
(TiO2)
Design 2:
single trenches
CALICE Collaboration Meeting at LLR

5. Megatile simulation (reminder)
Trench parameters are 0.3mm wide, 0.3mm offset with minimal air gap in top & bottom surfaces
Muons pass through the center cell of 12×12 cells (one HBU size)
Straight trenches
Tilted trenches (45°)
Central cell: 25.4 p.e./MIP
Neighboring cell: 0.49 p.e./MIP
Cell-to-cell crosstalk: 1.9 %
Central cell: 22.4 p.e./MIP
Neighboring cell: 0.36 p.e./MIP
Cell-to-cell crosstalk: 1.9 %
Both designs show moderately high MIP response and low optical crosstalk
CALICE Collaboration Meeting at LLR

6. CALICE Collaboration Meeting at LLR (yong.liu@uni-mainz.de)
Why tilted trenches?
Geant4 simulation
Megatile: tilted trenches
Compared to cell mean response: 22.4 p.e.
99.3% area: uniformity 60%
96.1% area: uniformity 70%
79.1% area: uniformity 80%
51.7% area: uniformity 90%
Megatile: 4 corners
Most megatile area is active
Individual wrapped tiles
Scintillator Tile
Non-sensitive area: 400µm
between each cell (simulation)
Individual tiles: 4 corners
Individually wrapped tile: 29.6 × 29.6 mm²;
Dead area per tile 23.8 mm² (~ 2.6%)
Perpendicular tracks:
megatile with tilted trenches has minimal dead area;
projective cracks with layers of individually wrapped tiles
CALICE Collaboration Meeting at LLR

7. Megatile: a first prototype (reminder)
Double trenches (straight), 3×3 cells
Scintillator: NE110 (comparable to BC408)
Depth 2.0 mm, width 0.5 mm, offset 1.0 mm
Less challenging parameters, only for the first prototyping
Expect worse performance than simulation (0.3mm wide, offset 0.3mm)
Still useful to verify simulation (new geometry adapted)
Fabricated by machine: cutting, polishing (cracks seen; difficult to polish perfectly)
3×3 cells (side view)
Megatile in a “foil cap”
3×3 cells (top view)
CALICE Collaboration Meeting at LLR

8. Megatile prototype: measurement vs simulation
Geant4 simulation
Comparisons: data vs MC
Similar MIP response in central cell in data and MC
MC also predicts similar crosstalk
Worse uniformity of crosstalk observed in prototype measurements
Perfect quality of trench cutting assumed in MC
Central Cell
Muons hitting boundary regions
The first prototype (not an optimal design) is only to verify simulation
Promising performance if optimal designs can be realized
Bright: top trenches
Dark: bottom trenches
Response Map (Data)
Cosmic-ray measurements
Geant4 simulation
Response Map (MC)
Data with cosmic rays
Central cell: 15.4 p.e./MIP
2-cell crosstalk: 24~34 %
CALICE Collaboration Meeting at LLR

9. Simulation: megatile single trenches
Minimal air gap in top & bottom surfaces assumed
Reflective Foil (ESR)
Design 2:
single trenches
3.0 mm
SMD-SiPM
Optical trench
(TiO2)
Trench schematics (side view): not in scale
ZOOM out: response uniformity
map of the central cell
Geant4 simulation
Central cell: 28.8 p.e./MIP
Cell-to-cell crosstalk: 0.1 %
Excellent response uniformity; similar good performance to individually wrapped tiles
High MIP response and fairly low optical crosstalk predicted in simulation
CALICE Collaboration Meeting at LLR

10. Megatile: production process
Design 2:
single trenches
4~6 mm
Optical trench
(TiO2+glue)
Scintillator Plate
Fill in glue+TiO2,
Mill down to 3mm
Optical trench
(TiO2)
3.0 mm
Reflective Foil (ESR)
Assembly:
SiPM and foil
3.0 mm
Optical trench
(TiO2)
SMD-SiPM
Trench schematics (side view): not in scale
CALICE Collaboration Meeting at LLR

11. Tests with glue and white powder
TiO2 powder
Mixing powder and glue
Scraping mixture into trenches
Scintillator strips:
~ 6 mm thick
Connected to a vacuum pump
Mill down to 3mm
CALICE Collaboration Meeting at LLR

12. Megatile adapter board to HBU
(2nd) Adapter Board with SiPMs before baking
Pin soldering
7 new MPPCs from DESY
Reuse old HBU originally designed for SiPM with pins
Megatile prototype
Megatile on the Adapter Board
CALICE Collaboration Meeting at LLR

13. Preparations for cosmic-ray measurements
2 generations of HBU boards in stack
HBU5
LED data
HBU2
Adapter board concept working
1st adapter board with 2 MPPCs (wo trenches): LED data taken successfully
2nd adapter board with 7 new MPPCs (with trenches): prepared for data taking
Megatile adapter board
plugged beneath
CALICE Collaboration Meeting at LLR

14. Full-size prototypes: 12×12 cells
3 large prototypes have been produced: more tests coming
CALICE Collaboration Meeting at LLR

15. CALICE Collaboration Meeting at LLR (yong.liu@uni-mainz.de)
Summary and plans
Megatile R&D: to further simplify mass assembly
Simulation studies for megatile
Various designs tried; promising performance foreseen
Simulation verified by a small megatile prototype
Megatile: glue tests
Glue mixed with white paints
Small prototypes produced: mechanically stable
To be measured with HBU adapter board
Full-size prototypes (12×12 cells) produced
More tests to be done
Thank you!
CALICE Collaboration Meeting at LLR

16. CALICE Collaboration Meeting at LLR (yong.liu@uni-mainz.de)
Backup
CALICE Collaboration Meeting at LLR

17. Megatile: applications in the past and at present
CDF End Plug Upgrade HCAL (1994)
D0 Run II Inter Cryostat Detector (1999)
CMS HCAL (1996)
STAR Barrel EMC (2002)
NIU Integrated Readout Layer (2009)
Note: this list is not meant to be exhaustive; the year corresponds to the earliest one appearing in the documents at hand
CALICE Collaboration Meeting at LLR

18. Efforts of MegaTile development at Mainz (1)
MegaTile with steel grids
BC408 scintillator
Steel grids coated with chrome
1x1mm² HPK MPPC
17.8 p.e./MIP
Prototype with metal grids and individual tiles
Cosmic-ray measurement
Idea: quickly produce metal grids
A first prototype worked well with steel strips and individually machined tiles
Many manufacturers tried, but could not produce the steel grids with sub-mm thickness at the size ~ 36x36 cm²
CALICE Collaboration Meeting at LLR

19. Efforts of MegaTile development at Mainz (2)
MegaTile with carbon-fiber
Built a prototype of grids
Carbon-fiber: many thin layers glued together
Mechanically fragile
A small part fractured
CALICE Collaboration Meeting at LLR

20. Megatile full simulation: overview
A scintillator plate (BC408) segmented for 12×12 cells
Cells separated by trenches, filled in with white paints
Each cell individually read out by an SMD-SiPM
Trenches filled in with TiO2, presumed to be ideally diffuse
Top/bottom surfaces covered with ESR foil
Muons pass through the central cell perpendicularly
Muons: hit positions
12×12 cells
SMD-SiPM
Response of each SiPM is read out and averaged by the number of events
CALICE Collaboration Meeting at LLR

21. Simulation: straight trenches
Rendered by G4RayTracer
Top and bottom trenches
Different trench depths, widths, offset between top and bottom
Including technical constraints
2.0 mm deep, 200 µm and 300µm wide (trapezoid), 300µm offset
2-cell crosstalk: 1.9 %
Geant4 simulation results
Central cell: 25.4 p.e./MIP
Neighboring cell: 0.49 p.e./MIP
2-cell crosstalk: 1.9 %
Same in 4 direct neighboring cells:
No cell boundary effects (cut off hit positions within 2 mm from cell boundary)
Also interesting to see what are boundary effects (next page)
CALICE Collaboration Meeting at LLR

22. Straight trenches: boundary effects
Special MC runs: muons only hit corners of 4 cells
x: -0.6~0.3 mm; y: -0.6~0.3mm; step size: 30 µm
Muons: hit positions
Cell Position in X
Cell Position in Y
Cell 67
Solid and dashed lines indicate top and bottom trenches (borders)
Boundary area: 3.6 % of a cell
Lower response: ~ 8 p.e./MIP
~ 30% of each cell response
Due to trench geometry:
only 1mm thick scintillator in these regions, not nominal 3mm
Sum of 4 cells
Small dead area: 0.01% of a cell
(overlapping of top and bottom trenches)
Individually wrapped tiles
Tile size: 29.6 × 29.6 mm²
Dead area per tile: 2.6%
(23.84mm²)
CALICE Collaboration Meeting at LLR

23. A megatile design: tilted trenches
Straight double trenches
Boundary area: 3.6 % per cell
Active, but with lower response (only ~30% of center area)
Geometry effect: 1 mm scintillator material left in the area
Dead areas (small): 0.01% per cell
Depend on trench width
Trenches tilted by some angle
To increase response of boundary areas
Tilted trenches: only one design shown
Tilted 45°, 2mm depth (vertical projection)
2.0 mm
Rendered by G4RayTracer
Constructed in SU
CALICE Collaboration Meeting at LLR

24. Simulation: tilted trenches
2.0 mm
Rendered by G4RayTracer
Geant4 simulation results
Central cell: 22.4 p.e./MIP
Neighboring cell: 0.36 p.e./MIP
2-cell crosstalk: 1.9 % in all 4 neighboring cells
No boundary effects (cut off hit positions within 2 mm from cell boundary)
MC suggests promising low crosstalk level and moderate MIP response
2-cell crosstalk: 1.9 %
CALICE Collaboration Meeting at LLR

25. Simulation of tilted trenches: uniformity map
x: -15~15mm; y: -3~2mm; step size: 100 µm
Compared to cell mean response: 22.4 p.e.
99.3% area: uniformity 60%
96.1% area: uniformity 70%
79.1% area: uniformity 80%
51.7% area: uniformity 90%
Megatile: 4 corners
All boundary area is active and most (>96%) has >70% response
Comparison with current tile design
Nominal size: 30.0 ×30.0 mm²
Current tile size: 29.6 × 29.6 mm²
Dead area per tile: mm² (~ 2.6%)
Megatile has such a potential of almost zero dead area
Improved size also exists: 29.7 × 29.7 mm²;
Dead area per tile mm² (~ 2.0%)
Non-sensitive area: 400µm
between each cell (simulation)
x: -15~15mm; y: -3~2mm; step size: 100 µm
Individual tiles: 4 corners
CALICE Collaboration Meeting at LLR