Validation of detection efficiency of alpha particles in commercial liquid scintillation counters Liquid Scintillation Users' Forum 14th October 2008 Andy Pearce Radionuclide and Neutron Metrology Group NPL
Validation of detection efficiency of alpha emitters in commercial liquid scintillation counters Purpose of work Overview of work st Experimental details (1 stage) Results (1st stage) Experimental details (2nd stage) Results (2nd stage) Interpretation of data Conclusions
Purpose of Work Develop a more costeffective method of standardisation of alpha emitters coincidence counting expensive, time consuming + limited resources limited radionuclides covered many nuclides have no gammas limited range of activities necessitates dilution stages more cost, more uncertainty
Purpose of Work Provide data to back up an assumption commonly used in environmental and radioanalytical measurements LS sample preparation simple: reduce costs improve traceability determine limitations
Overview of Work Determine efficiency in controlled conditions efficiency directly measured with coincidence counting ideal conditions: clean source with low quench and understood signal processing chain Test validity to real counters and sources effect of chemical quench performance of commercial counters scintillant effects volume effects
Stage 1: Principle of Measurement Simple principle: use established coincidence counting to determine alpha efficiency Alpha efficiency given by ratio of recorded coincidences to recorded gammas Requires nuclide with strong gamma emission and uncomplicated decay scheme 249Cf Requires use of system which resembles commercial LS counter, adapted to include gamma detector
Stage 1: Measurement Equipment PMTs Hamamatsu from Rackbeta light guide from Rackbeta lead shield from LSC2
Stage 1: Measurement Technique Coincidence gate Amplifier 1 Amplifier 2 Width/Delay Preamp 1 Preamp 2 PMT 1 PMT 2 BADGER BOX (PC) NaI(Tl) PMT Amplifier 3 Width/Delay
Results from Part I Mean measured value 1.00017 (23) Detection efficiencies measured between 0.998 < ε < 1.002 Largest uncertainty due to dead-time corrections (ICRU 52) All values statistically compatible with 100 % detection
Stage 2: Principle of Measurement How do these results transfer to a commercial counter? measure same sources on commercial counters What factors may take detection efficiency away from 100 %? measure range of sources on commercial counters and look for efficiency variations What is a reasonable detection efficiency to assume on commercial counters?
Commercial Counters Packard Tricarb 2700TR Beckman 6500LS
Stage 2: Measurement of 249 Cf (5.8 MeV) Alpha counts per second determined using custom-built coincidence counting rig Value of 127.51 (3) ks-1 g-1 (type A uncertainties) Alphas per second determined using commercial counter (2700TR) Value of 127.44 (10) ks-1 g-1 (type A uncertainties) Values are statistically compatible; difference is 0.05 %
Stage 2: Measurement of Scintillant PerkinElmer Ultima Gold AB PerkinElmer Ultima Gold AB National Diagnostics Ecoscint A National Diagnostics Ecoscint A Volume /ml 10 15 10 15 208 Po (5.1 MeV) Specific Count Rate s -1 g -1 Tricarb 2700TR Beckman 6500LS 2.7326 (7) 2.7336 (7) 2.7335 (7) 2.7348 (7) 2.7345 (7) 2.7319 (7) 2.7335 (7) 2.7353 (7) NPL 2.7346 (27) - Same sources measured on each counter Each number is a mean of measurements of a set of 6 sources quenched to varying degrees with 10 % Nitromethane/Ethanol Range of quench from 100 to 460 (tsie)
Conclusions Detection efficiency for alpha emitters has been determined to be close to 100 % Value of 0.998 (2) recommended when using commercial counters Efficiency is constant under normal conditions Sample volume changes do not affect counting efficiency between 10 ml and 15 ml Insensitive to chemical quench Colour quench not tested Physical quench not tested Effect of volumes >15 ml not tested
Simplified scheme for standardisations Simplified scheme for standardisation of alpha emitters is now available Can standardise directly at level required Cost is ~1/20th of that by coincidence counting with efficiency tracing
Finally... Thanks to colleagues Arzu Arinc, Chris Gilligan and Andy Stroak who were also involved with this project!