The TORCH PMT: A close packing, multi-anode, long life MCP-PMT for Cherenkov applications James Milnes Tom Conneely 1 page 1
Photek MCP-PMTs Photek currently manufacture the fastest PMTs in the world in analogue mode : Whole pulse is captured by an oscilloscope or digitiser Applications often single-shot, high intensity, e.g. Fusion Research We have several detectors on the diagnostics at the National Ignition Facility 2
Photek MCP-PMTs MCP-PMTs are also the leading detector for time-resolved photon counting Jitter in photon arrival measurements ~ 30 ps FWHM Significantly better for multi-photon events Excellent fit for Cherenkov-based particle detection Drawbacks: Detector Lifetime Most models are round, single anode and not close-packing 3
The TORCH PMT In November 2012 Photek started the 3-year development of the TORCH (Timing Of internally Reflected CHerenkov photons) PMT A collaboration with CERN and the Universities of Oxford and Bristol for the LHCb upgrade See talk by Roger Forty, Thursday 14.30 Technical aims: A lifetime of 5 C/cm 2 of accumulated anode charge or better A multi-anode readout of 8 x 128 pixels Close packing on two apposing sides with a fill factor of 88% or better 53 mm working width within a 60 mm envelope 4
The TORCH PMT Three main aims: 1. Lifetime 2. High granularity multi-anode 3. Square 5
1. MCP-PMT Lifetime Standard MCP detectors suffer from sensitivity loss after prolonged exposure: An MCP has a very large surface area Prolonged electron bombardment of this surface releases material that is ionised These ions are drawn back to the photocathode and reduce sensitivity Previous solutions have involved barrier films to prevent the ions reaching the photocathode Limited success Lowers MCP efficiency and overall sensitivity 6
1. MCP-PMT Lifetime Recent technology of ALD (Atomic Layer Deposition) coating on MCP has significantly reduced out-gassing Two PMT samples produced in 2011: Double-MCP 10 mm diameter working area One with coated MCPs, One control with standard MCPs Independently verified by Photek and others: Britting et al (PhotoDet 2012), Conneely et al (VCI 2013) 7
Photocurrent (A/W) Photocurrent (A/W) Gain 1. MCP-PMT Lifetime ALD coating results in no detectable sensitivity loss after > 5 C/cm 2 Gain is reduced by ~ 30% Gain reduction could be reduced by better MCP scrubbing (pre-conditioning) 3x10 6 2x10 6 10 6 9x10 5 8x10 5 7x10 5 6x10 5 5x10 5 4x10 5 3x10 5 2x10 5 10 5 ALD Coated MCP-PMT Gain at MCP Voltage: 1500 V 1450 V 1400 V 0 1 2 3 4 5 Accumulated Anode Charge (C/cm 2 ) 60.0m Uncoated MCP-PMT Accumulated Anode Charge: 0 C/cm 2 50.0m 40.0m 30.0m 20.0m 0.13 C/cm 2 0.17 C/cm 2 0.20 C/cm 2 0.25 C/cm 2 0.28 C/cm 2 0.30 C/cm 2 0.32 C/cm 2 0.36 C/cm 2 60.0m ALD Coated MCP-PMT 50.0m 40.0m 30.0m 20.0m Accumulated Anode Charge: 0 C/cm 2 0.38 C/cm 2 0.71 C/cm 2 1.13 C/cm 2 1.95 C/cm 2 2.49 C/cm 2 3.18 C/cm 2 3.71 C/cm 2 5.11 C/cm 2 10.0m 10.0m 0.0 300.0n 400.0n 500.0n 600.0n 700.0n 800.0n 900.0n Wavelength (m) 8 0.0 300.0n 400.0n 500.0n 600.0n 700.0n 800.0n 900.0n Wavelength (m)
Gain 1. MCP-PMT Lifetime 1 st year objective: produce 5 long-life MCP-PMTs 1 st build cycle produced 6 double-mcp devices plus 1 control ALD also gives major gain enhancement A 2 nd build cycle will attempt to produce a photon counting device with 1 MCP 10 7 10 6 10 5 10 4 10 3 10 2 800 1000 1200 1400 1600 1800 2000 Voltage across MCP pair (V) G1130614 (modified scrub) G2130614 (modified scrub) G1130510 G2130510 B1130419 B4130419 (control) B5130419 9
Counts (normalised) Voltage (V) 1. MCP-PMT Lifetime In other respects, coated MCPs behave as normal Jitter measurement made with 40 ps laser source Life-testing of sample devices to begin shortly at CERN and Photek PMT225 G1130510 PHD Voltages: 200 600 600 1000 200 650 650 1000 200 700 700 1000 200 750 750 1000 0.0-0.1-0.2-0.3-0.4-0.5-0.6 PMT225 G1130510 200 750 750 1000 LPG-1 @ 10 KHz 6 db Attenuator Averaged 16 times -0.7 10 5 10 6 10 7 10 8 Gain 10 0.0 2.0n 4.0n 6.0n 8.0n 10.0n Time (s)
2. High granularity multi-anode Traditional multi-anode manufacturing uses multiple pins brazed through a solid ceramic Prone to leaking, also unrealistic for a 128 x 8 array! Our aim is to use multilayer ceramic with filled vias Much smaller pad size allows for finer pitch The pads on this design are 0.75 mm wide on a 0.88 mm pitch Vacuum side Air side 11
2. High granularity multi-anode Contact made to anodes by Anisotropic Conductive Film (ACF) ACF are filled with conductive particles which provides electrical Detector interconnection between pads through the film thickness (z- ACF direction) The conductive particles are PCB distributed far apart thus not electrically conductive in the plane direction (x & y) of the film PCB could contain front-end electronics and/or connectors z y ACF is insulating in x and y but conducting in z 12 x
2. High granularity multi-anode - readout NINO ASIC 32 channel differential amplifier /discriminator developed at CERN 10 ps RMS jitter on the leading edge >>10 MHz maximum rate The time-over-threshold technique uses the discriminator output pulse width to determine the event charge High Performance Time-to-Digital Convertor (HPTDC) A programmable TDC developed for ALICE time-of-flight RPCs at the LHC Two modes of 100 ps LSB resolution with 32 channels, or 24.4 ps LSB resolution with 8 channels Default maximum rate is 2.5 MHz per channel, can be increased beyond 10 MHz using higher logic clock 13
3. Square Tube Development Square tube manufacturing is new to Photek Currently developing methods of Square body brazing Square MCP locating Square photocathode sealing Square anode sealing Current status: Producing leak-tight square test cells 14
3. Square Tube Development Anode Seal Traditional method of sealing anode welding is unusable due to close packing requirements We are experimenting with Indium seal Brazing Indium seal Fritting 2 inch square body with solid ceramic anode indium sealed 15
Summary TORCH PMT in development at Photek 3 year development aims to finish in November 2015 1 st year task to produce long-life demonstrators 2 nd year task to produce high-granularity multi-anode demonstrator Final year task: Fully functioning detector 16
Thank you for listening CERN, 3rd July 2013 17