DRD Collaborations,
Prague Activities in DRD3
The timescale
HL-LHC runs into the early 2040s, FCC-ee collides in the late 2040s,
and the detector that records them has to be designed, prototyped and built well before then
A student starting a PhD today is a tenured professor before first collisions.
Whichever machine gets built, somebody has to still know how to build a detector for it.
The DRD collaborations
How the DRDs came about
European Strategy update asks for a detector R&D roadmap
aimed past the HL-LHC, where no landscape existed
ECFA Detector R&D Roadmap: nine technology domains
each with numbered R&D themes graded by which future facility needs them, plus general strategic recommendations
CERN Council and SPC decide CERN hosts the collaborations
RD42, RD50, RD51 and the other LHC-era programmes close on 31.12.2023
DRD1 to DRD8 formed and approved by the Research Board
DRD3 came out of RD42 and RD50, fully approved in June 2024
Numbering is inherited: roadmap theme DRDT n.x belongs to DRDn; task force 9, training, became the ECFA Training Panel
The eight collaborations
Three-year terms, re-approved in waves: DRD1, 2, 4 and 6 in November 2026, DRD3, 5 and 7 in March 2027, DRD8 in November 2027.
What a DRD is
Hosted by CERN. The DRDC meets twice a year and reviews each collaboration once a year.
Joining means signing the MoU. A common fund is not universal: DRD4, DRD5 and DRD6 have none at all, and where one exists part of it flows back to collaboration-wide projects.
A DRD coordinates the R&D; the money comes from national and EU projects.
2026: the strategy update lands
CERN Council, 22 May 2026: FCC-ee is the preferred next flagship collider
The Strategy names the DRDs in Recommendation E: they "must receive adequate funding", and new R&D topics should be integrated into the DRD scheme
Era 1 experiments use what is developed now; that keeps the technology and the group active until Era 2.
Czech and Slovak involvement
Who is in what
Funding
- CERN-CZ (LM2023040), 2023 to 2026, MŠMT, coordinated at FZU; a follow-up has been requested, outcome and financing not yet known
- AIDAinnova 2021 to 2025 (FZU); OTELLO, its successor: first bid lost to ATTRACT-2 (12.00 of 15), new submission June 2026, outcome not public; Charles University and CTU in WP7, WP4 co-coordinated from Prague
- TAČR TK05020011 with onsemi funded the SiC production (finished)
- FORTE (FZU and FNSPE CTU) carries the SiC and OCTOPUS work; institutional funding and grants carry the rest
- The DRDs themselves bring no money; they coordinate what national and EU money pays for.
Prague in DRD3
CTU FNSPE (CAPADS) and FZU teams
DRD3 and Prague
120 institutes, spokesperson G. Kramberger (JSI)
Prague sits in WG1 / WP1, monolithic CMOS, and WG6 / WP3, wide bandgap
Prague in DRD3: two lines
FZU: radiation effects in 4H-SiC diodes, LGADs and MOSFETs; TPA-TCT; GaN sensor characterisation before and after irradiation
FNSPE CTU: 4H-SiC LGAD design and production runs with onsemi, segmented devices, lab and beam characterisation, irradiation campaigns
FNSPE CTU: OCTOPUS ASIC design, holding the WP2 co-lead; DAQ development; lab characterisation and test beams; irradiated H2M prototypes
Intermediate goal: a new beam-telescope detector for DESY and CERN
The two lines share the beam tests and the irradiation facilities.
Wide bandgap: what it is
4H-SiC has three times the bandgap of silicon: leakage stays negligible without cooling, even after irradiation
But a MIP leaves ~57 e-h/µm against ~72 in silicon, and epitaxial layers stop near 100 µm
Thickness cannot recover the signal, so the SiC LGAD adds gain
Radiation tolerance and fast timing without cooling motivate the R&D.
Wide bandgap: why now
Monolithic CMOS: why
The requirement: a Higgs-factory vertex detector asks for ≤3 µm single-point resolution, ~0.1 % X₀ per layer, O(5 ns) and under 50 mW/cm²
Why monolithic: nothing hybrid meets that material budget, so sensor and front end share one wafer. Foundry access and design engineers are the bottleneck
Step on the way: a beam-telescope sensor for DESY and CERN, the same ≤3 µm but O(100 ns) and under 500 mW/cm²
OCTOPUS: 13 institutes, coordinated by DESY and CERN; FNSPE CTU co-leads WP2, ASIC design
The ladder: a front-end chiplet on ER3 in 2026; MPR2 in 2027 and beam tests in 2028 to 2029, first at DESY, then at CERN, are planned dates
H2M: a hybrid architecture made monolithic
H2M puts a hybrid pixel architecture into TPSCo 65 nm: 64 × 16 pixels at 35 µm, read out with Caribou
It works: 45 e⁻ noise, 17 e⁻ threshold dispersion, >99 % efficiency, unchanged after thinning to 21 µm
In-pixel response is not uniform: efficiency and timing follow the n-wells of the analogue circuitry
Prague measures what irradiation does to that, before OCTOPUS commits to a front end
OCTOPUS: Prague's part
Optimized CMOS Technology fOr Precision in Ultra-thin Silicon
Small collection electrode MAPS on high-resistivity p-type epi, in TPSCo 65 nm
1024 rows × ~1500 columns at ~20 µm, matrix 2 × 3 cm, no stitching
Prague co-leads WP2, the ASIC design work package. CAPADS designed one of the two candidate end-of-column readout schemes
The pace is set by IP-core access in the shared 65 nm runs; OCTOPUS has asked DRD7 for support
Summary
- The DRDs replaced the RD programmes with something broader and more formal, and re-approval starts in November 2026, DRD3 in March 2027
- The DRDs coordinate. National funding does the work, and CERN-CZ ends in December 2026
- Czech groups are in six of the eight, and hold one collaboration-level leadership role across all of them
- Prague in DRD3 runs two lines: 4H-SiC LGADs, and monolithic CMOS
The people who will build the FCC-ee detectors are being trained now.
Backup
What the roadmap says
- Detector R&D Themes crossed against the facilities that need them, and graded
- must happen · important for several goals · desirable · already covered
- The collaborations were built around the grading rather than around the list of themes
The Prague 4H-SiC programme
- Produced with onsemi on their Czech 6" SiC line; epi 30–100 µm, doping to ~5·10¹³ cm⁻³
- Strips come as both PN and LGAD; the 55 and 110 µm pixels are LGAD-only. Four lots since 2022
- Lot 4 carries the first pixel and strip segmented 4H-SiC LGADs, on 50 µm epi
- IV/CV and TCT in Prague, TPA-TCT at ELI ERIC
- Protons at CERN IRRAD, neutrons at JSI TRIGA, gammas at ÚJP Praha
- Beam test at CERN SPS H8 with the Nikhef Timepix4 telescope
- Funded through TAČR, CERN-CZ (LM2023040), MSCA COFUND Physics for Future and FORTE — not through DRD3
Segmentation
- Lot 4 carries 45 designs on 49 dies: strips at 80 µm pitch, pixels at 55 and 110 µm, with oxide-filled trenches or plain geometric separation
- TPA-TCT indicates gain around 6-10 at 300 V, uncalibrated; TCAD gives 5-8 without a trench and ~6 with one
- Trenches only isolate partially — the gain implant runs continuous underneath, so devices with zero separation still share charge
- Gap matters more: zero gap breaks down below 100 V, a gap of 1 µm or more holds to 700 V, in both strategies
- Leakage to 700 V: ~0.1 nA for PN, ~30 nA for LGAD
What a Higgs-factory vertex detector needs
- Single-hit resolution ≤ 3 µm
- Material budget ~0.1 % X₀ per layer, sensor plus support plus services
- Time resolution of order 5 ns, hit rates of order 100 MHz/cm²
- Power below 50 mW/cm², because cooling adds material
- Nothing hybrid meets that material budget, so sensor and front end share the wafer — which turns sensor development into mixed-signal chip design, and moves the bottleneck to
- foundry access, and engineers who can design in them
DRD6: the Czech calorimetry work
- WP1 AHCAL, FZU — DAQ, data interface and concentration, LED calibration; the CALICE line, still running in 2026
- J. Kvasnička, I. Polák
- WP2 noble-liquid ECAL for ALLEGRO, Charles U — simulation, geometry, combined reconstruction, particle flow
- J. Faltová, M. Mlynáriková, F. Sopková, O. Mezhenska
- WP3 TileCal, Charles U and FZU — scintillator tile response, 64-channel SiPM readout, SiPM expansion boards
- M. Haviernik, I. Polák, J. Kvasnička
DRD3 member countries
DRD3 organisation (June 2025)
ECFA roadmap needs: DRD1, gaseous detectors
ECFA Detector R&D Roadmap 2021
ECFA roadmap needs: DRD2, liquid detectors
ECFA Detector R&D Roadmap 2021
ECFA roadmap needs: DRD3, solid state detectors
ECFA Detector R&D Roadmap 2021
ECFA roadmap needs: DRD4, photon detectors and PID
ECFA Detector R&D Roadmap 2021
ECFA roadmap needs: DRD5, quantum sensors
ECFA Detector R&D Roadmap 2021
ECFA roadmap needs: DRD6, calorimetry
ECFA Detector R&D Roadmap 2021
ECFA roadmap needs: DRD7, electronics
ECFA Detector R&D Roadmap 2021
ECFA roadmap needs: DRD8, mechanics and cooling
ECFA Detector R&D Roadmap 2021
Who is in what: the long form
| DRD | Czech and Slovak institutes, with the contact person | what they do there | source type |
|---|---|---|---|
| DRD1 | IEAP CTU — H. Natal da Luz | Timepix readout of MPGDs and TPCs. Natal da Luz co-leads WP4 — the one Czech name on any collaboration-level management page. UWB Pilsen applied in June 2026; no admission published. | expression-of-interest list |
| DRD2 | FZU — J. Zálešák | Liquid detectors for neutrinos and rare events. | membership list |
| DRD3 | FNSPE CTU — P. Švihra · FZU — M. Mikeštíková · IEAP CTU — B. Bergmann · Charles U — P. Kodyš · UWB Pilsen — V. Georgiev | Wide-bandgap sensors and pixel readout. Five institutes, the widest Czech presence of any DRD. | membership list (Dec 2024) |
| DRD4 | FZU — J. Zálešák · IEAP CTU — B. Bergmann | GaAs sensors, SciFi and transition radiation. | proposal author list (2023) |
| DRD5 | FNSPE CTU — M. Matas · Charles U · UWB Pilsen | Quantum sensing, and scintillating nanocomposites. | country map |
| DRD6 | CTU — V. Čuba · Charles U — J. Faltová · FZU — I. Polák · IEP SAS Košice — P. Stríženec | AHCAL readout and DAQ, noble-liquid ECAL and TileCal for ALLEGRO. Čuba is at FNSPE, though the list says only "Czech Technical University". | membership list (Jan 2026) |
| DRD7 · DRD8 | none | Nothing on DRD7's list of 68. DRD8 publishes no list. | list of 68 / no list |
OCTOPUS: the official chip view
