R&D activities
Gas R&D activities 2026-08-27 09:51(GMT+2:00)
Key Outcomes
Testing of ZE/HFO gas mixtures shows promising time resolution results at higher drift fields, with 20% and 40% ZE approaching CF4 performance when drift field is increased to 6 kV/cm.
The LHCb distillation system was successfully pressurized after fixing a leak and sensor miswiring, though the reflux part behavior remains under investigation.
ZIF-8 cycle tests confirm stable absorption/desorption performance by the third consecutive cycle, with isobutane pre-saturation identified as a key conditioning step.
Gas Mixture Testing – ZE/HFO Results
· Working point shifts ~40 V higher with each 5% ZE addition
· Plateau efficiency decreases slightly (~5%) as ZE concentration increases
· At 3.5 kV/cm drift field, time resolution degrades beyond 5% ZE; 10%, 20%, 40% ZE all perform worse than CF4 at this field
· Increasing drift field to 6 kV/cm recovers performance: 20% ZE reaches within ~0.5 ns of 40% CF4
· 1% isobutane addition at 20% ZE does not affect efficiency or time resolution, but improves current stability
· ZD tested but went directly from no gain to discharging — no usable operating window found
· Open question: whether the efficiency plateau decrease at high drift fields is a physical gas cross-section effect or an electronics artifact
Distillation System – LHCb Commissioning
· Leak traced to the Zimmerli pressure control valve on the reflux line; plugged by the team
· Sensor miswiring corrected; system degassed by raising temperature above zero to release trapped gases
· System pressurized from 350 mbar to 1,600 mbar; coolant reached −50°C
· No condensate observed in the reflux arm — root cause under investigation
Distillation System – ATLAS installation
· Missing hardware blocking full installation: pipe connecting chiller to system and the mass flow controller (Visada) not yet installed
· Flammable gas detection sensor not yet identified for this system
Distillation Simulation – Reflux Ratio Findings
· System designed for 95% R134A standard RPC mixture
· Reflux ratio operating window: 6–8
· Below 6: purity drops below 99%
· Above 8: column flooding risk
· Recommended operating point: 6–7 (6 most conservative, 7 optimal)
· Bottom feed ratio: ~0.57; estimated R134A recovery ~80%
· Simulation run in Aspen; a Python-based live simulation tool also developed for broader team use
· Action: repeat simulation table for the R134A + CO2 mixture to compare operating conditions and azeotrope formation behavior
ZIF-8 Adsorption Cycle Testing
· Vacuum-only desorption: releases >99% R134A, isobutane <1%, SF6 <0.01%
· With argon carrier gas: isobutane slightly elevated (~1.2–1.3%), but otherwise stable
· First absorption cycle shows elevated isobutane and SF6 uptake; stabilizes by third cycle (both <5%)
· Desorption capacity remains constant across cycles; absorption capacity lower in cycle 1, normalizes in cycles 2–3
· Cycle timing: absorption to breakthrough ~69–85 min; desorption ~90 min
· Anomaly: R134A appears in exhaust during early absorption before breakthrough in cycle 1, then is absorbed at ~90% in later phase — behavior stabilizes by cycle 3
· Pre-saturation with isobutane identified as the solution to condition the sorbent material before operational use
Action Items
· Stefania: Retest 5% and 10% ZE at 6 kV/cm drift field to complete the field-dependence picture
· Emmanuel+Ryan: Add R134A recovery efficiency and azeotrope formation analysis to the Aspen simulation; compare CO2 vs. no-CO2 mixture conditions
· Emmanuel+Ryan: Add Python tool link to shared documentation
· Emmanuel+Ryan: Resolve missing chiller pipe connection and procure/install Vaisala and mass flow controller
· Emmanuel+Ryan: Investigate reflux arm — confirm correct temperature set point (~−50°C) and verify no condensate issue
· Fabio: Leak-test Zimmerli valve on the new distillation rack (suspected similar issue to LHCb unit)
· Thomas: Locate network socket for control panel; bring Thomas to site for orientation
· Andrea: Complete desorption for third absorption cycle and finalize five-cycle dataset
Gas R&D 2026-08-20
Key Outcomes
The team reviewed ongoing gas sensor measurements for HCL and HF production at GIF, identified potential noise and offset issues requiring further investigation, and agreed on a systematic lab-first testing approach before resuming GIF measurements. A molecular simulation collaboration with Torino is planned for October. ELMB2 testing in Micromegas is progressing with a step-by-step validation strategy.
Measurements & Findings
· HCL and HF peaks correlate with temperature changes during CO2 flushing; HF shows a non-zero offset while HCL decays to zero.
· A small HCL peak (~0.012 ppm) was observed coinciding with the GIF source being turned back on after weekly access — at the bottom range of sensor sensitivity and possibly noise.
· Previous tests with detectors present showed HCL/HF levels of 4–6 ppm (>10× higher than current readings).
· HF lower quantification limit is ~0.3 ppm; current readings are near or below this threshold.
· CO2 flushing confirmed by vendor to not create sensor offset; however, the humidifier is a suspected source of interference.
Decisions Made
· Bypass the humidifier in the next test run to isolate its potential contribution to HF offset.
· Lab-first approach: Begin systematic testing in Lab 256 using detectors 21 or 29, flush with Argon or CO2, establish baseline/offset before any GIF measurements.
· Nitrogen to be tested as an alternative flush gas after CO2 decay is confirmed complete.
· Manuel, Ryan, Albin, and Max (with chemical background) present to learn from Maria Cristina how to use ISE.
· Calibration test with 1,000 ppm chlorine solution (Thermo Fisher recommendation) to be scheduled.
Gas Decomposition & Simulation
· R1224YD breaks down slower than ZD; YD also degrades into PFAS; ZD produces both HCL and HF, R134A produces HF.
· YD has one more fluorine atom than ZD — higher fluorine correlates with higher electronegativity, which may be beneficial; need a comparison table of YD vs. ZD decomposition products and properties.
· A molecular simulation student from Torino arrives in October for one month to study fragmentation pathways using DFT; goal is to predict new molecule behavior before physical testing.
· Bulk decomposition is driven by electric field, not radiation alone — irradiating without voltage produces negligible decomposition.
· Long-term goal: feed experimental parameters into Garfield/Markowitz simulation tools; primary ionization data from the X-ray source in QDD may support this.
Open Questions
· Whether the small HCL peak at GIF source restart is a real signal or random fluctuation — temperature correlation also needs checking.
· Lifetime of the electrochemical sensor cells under low-HF conditions is uncertain; a second sensor has been ordered (delayed from China).
· No RPC detector exists that has never been irradiated at GIF with ZD gas — complicates clean baseline measurement.
· Whether to keep detectors at working point during source-off periods or adapt voltage — deferred until lab results are available.
ELMB2 / Hardware Status
· ELMB2 works in OT but not confirmed in MDT; plan is to test first in Micromegas (2 racks) as a simpler environment.
· If Micromegas test fails by end of October/November, the Micromegas upgrade will proceed and ELMB2 will be re-tested post-PLC upgrade.
· MDT testing is more complex due to pressure sensors and a busy CAN bus — lower priority.
· Manual CAN operational intervention (heartbeat reset) may be needed for ELMB2 to become operational in LX4G.
Action Items
· Max: Bypass humidifier and rerun HF/HCL measurement; try nitrogen flush after CO2 decay confirmed
· Max: Order second HF and F2 sensor for cross-checking
· Maria Cristina, Max, …: Send one slide summarizing decomposition products (HCL/HF) for YD and ZD
· Max : Set up Lab 256 test with detectors 21 or 29; establish offset baseline before GIF campaign
· Team: Prepare case study for Torino simulation student arriving October
· Thomas / Davide: Test ELMB2 in Micromegas racks (Monday check planned)
· Miguel: Share updated documentation/guide with Gianluca and Beatrice; coordinate greenhouse gas data entry with Laura
· Jesse: Share thesis and other email for continued loop inclusion