Detection of 133 Xe from the Fukushima nuclear power plant in the upper troposphere above Germany
Hardy Simgen, Frank Arnold, Heinfried Aufmhoff, Robert Baumann, Florian Kaether, Sebastian Lindemann, Ludwig Rauch, Hans Schlager, Clemens Schlosser, Ulrich Schumann
DDetection of
Xe from the Fukushima nuclear power plant in theupper troposphere above Germany
Hardy Simgen a , Frank Arnold a , b , Heinfried Aufmhoff b , Robert Baumann b ,Florian Kaether a , Sebastian Lindemann a , Ludwig Rauch a ,Hans Schlager b , Clemens Schlosser c , Ulrich Schumann b a Max-Planck-Institut für Kernphysik, Saupfercheckweg 1, D-69117 Heidelberg, Germany b DLR Oberpfaffenhofen, Münchner Straße 20, D-82234 Weßling, Germany c Bundesamt für Strahlenschutz, Rosastraße 9, D-79098 Freiburg, GermanyEmail-addresses:
[email protected]@[email protected] (Published in Journal of Environmental Radioactivity, Volume 132 (June 2014) Pages 94-99;doi:10.1016/j.jenvrad.2014.02.002)
Abstract
After the accident in the Japanese Fukushima Dai-ichi nuclear power plant in March 2011 large amountsof radioactivity were released and distributed in the atmosphere. Among them were also radioactive noblegas isotopes which can be used as tracers to test global atmospheric circulation models. This work presentsunique measurements of the radionuclide
Xe from Fukushima in the upper troposphere above Germany. Themeasurements involve air sampling in a research jet aircraft followed by chromatographic xenon extraction andultra-low background gas counting with miniaturized proportional counters. With this technique a detectionlimit of the order of 100
Xe atoms in litre-scale air samples (corresponding to about 100 mBq/m ) isachievable. Our results provide proof that the Xe-rich ground level air layer from Fukushima was lifted upto the tropopause and distributed hemispherically. Moreover, comparisons with ground level air measurementsindicate that the arrival of the radioactive plume at high altitude over Germany occurred several days beforethe ground level plume.
Keywords:
Fukushima, Reactor accident, Low-level gas counting, Ultra-low background, Radioxenon,Xenon-133.
One of the strongest earthquakes in Japanese historyhappened close to the east coast of Honshu Islandon March 11, 2011 and was followed by a destruc-tive tsunami. This triggered a series of accidents inthe nearby Fukushima Dai-ichi nuclear power plant[1] which caused a release of large amounts of ra-dionuclides [2, 3, 4, 5], among them the radioactivexenon nuclide
Xe. The total
Xe source termwas estimated to be between 12 EBq and 19 EBq[5, 6, 7]. Such a strong release provides an oppor-tunity to test and improve global atmospheric circu-lation models using
Xe as a tracer. Since xenon isa noble gas it is inert and rarely reacts with otherelements. Thus, it is superior to other commonlyused tracers which require profound understanding of atmospheric chemistry, washout processes and rain.Moreover, its lifetime of 7.57 days [8] is comparableto the time scale for intercontinental transport in theupper troposphere. In this work we present the detec-tion of airborne
Xe in the upper troposphere aboveGermany after the Fukushima event. A second paper[9] deals with the atmospheric aspects and implica-tions of the
Xe measurements as well as with theinterpretation of simultaneously measured trace gasesand aerosols. For comparison with data obtained fromground level measurements in Europe of iodine andcesium radionuclides we refer to [10].After first information about the nuclear accidentbecame public we prepared an aircraft campaign atshort notice. The rationale of our aircraft measure-ments was to investigate pollution of the upper tro-posphere, particularly the tropopause region, by pol-1 a r X i v : . [ phy s i c s . i n s - d e t ] D ec H. Simgen et. al: Detection of
Xe from the Fukushima nuclear power plant in the upper troposphere above Germany
Figure 1 : Sketch of the miniaturized ultra-low background proportional counter used for the
Xe measurements. lutants released from fossil fuel combustion in the re-gion of North-East China, Korea and Japan. In fact,this region represents the strongest SO -source regionworldwide. Since SO is an important precursor of po-tentially climate-active atmospheric aerosol particles,above mentioned region became a matter of particu-lar interest at the Deutsches Zentrum für Luft- undRaumfahrt (DLR) [9, 11, 12, 13].Samples were taken during two flights with the Fal-con research jet aircraft [14] of DLR. The first flight onMarch 23, 2011 aimed to the intercept the Fukushima Xe plume in the upper troposphere above Ger-many, shortly after its expected arrival on March21/22, 2011. Due to strict air safety regulations onlyrather small (1 litre) and passive air samplers wereapproved in the short time before the first flight. Thesecond flight on April 14, 2011 aimed to probe thehighly diluted Fukushima plume in the upper tropo-sphere. Here larger samples of about 10 litres couldbe taken. Model simulations conducted by the Institutfür Physik der Atmosphäre of DLR prior to the flights,predicted that upper tropospheric
Xe activity con-centrations above Germany would be around 1 Bq/m during the first flight and about 10 times less duringthe second flight. Thus, extremely sensitive Xe de-tection techniques in the laboratory are required toobserve the Fukushima signal in litre-scale samples. Xe measurements with miniaturized pro-portional counters
Miniaturized ultra-low background proportional coun-ters (see Figure 1) were developed at the Max-Planck-Institut für Kernphysik to detect few atoms of Gein the
Gallex solar neutrino experiment [15]. Theywere also used for
Xe studies in
Gallex [16] andlater to detect low levels of various radioactive noblegas isotopes in the astroparticle physics experiments
Borexino [17],
Gerda [18] and
Xenon [19]. Due totheir low background and low energy threshold theyare ideally suited to detect gaseous radioactive iso-topes with high sensitivity. The key technology isa dedicated processing of gas samples before loading them into the counter. It includes removal of traceimpurities which disturb the performance of the pro-portional counter (e.g. oxygen, humidity, ...) as wellas separation of unwanted radioactive impurities. Fi-nally, the size of any gas sample has to be reducedwithout losing the target isotope and without intro-ducing contamination to fit into the active volumeof the miniaturized counter ( ∼ cm ). The samplepreparation is done by means of a dedicated gas han-dling and counter filling line made from glass. Such asystem was developed in the framework of astroparti-cle physics experiments and as indicated in [20] it maybe used for xenon processing.The main challenge is the xenon separation athigh acceptance from the air samples. This is doneby pumping the sample over a 0.6 gram activatedcarbon column held at -186 ◦ C by immersing it inliquid argon. While most of the oxygen and nitrogenis pumped away, xenon is efficiently stopped in thecolumn. Subsequently, the sample is transferred tothe top of a gas chromatography column by heatingthe activated carbon column and cooling the gas chro-matography column at liquid nitrogen temperature.Helium is used as a carrier gas to run the samplethrough the gas chromatography column which isfilled with Chromosorb 102 [21]. Nitrogen, oxygen,carbon dioxide, xenon and radon elute from such acolumn in the listed order. The temperature is variedstep-wise from liquid nitrogen temperature to -110 ◦ Cand finally to -30 ◦ C to achieve good separation of allcomponents. A big challenge is
Rn in the sampleeluting shortly after xenon. Radon is not visible inthe chromatogram which is recorded by a thermalconductivity detector with limited sensitivity. Thus,it is necessary to reject the late fraction of the xenonwhich might not be fully separated from
Rn. Thiscauses losses of
Xe between five and ten percentfor a single run. These losses are quantified via anatural xenon carrier which is added to the samplebeforehand. Xenon has no long-lived radioactiveisotopes, therefore the carrier which is several yearsold does not introduce background radioactivity. . Simgen et. al: Detection of
Xe from the Fukushima nuclear power plant in the upper troposphere above Germany Figure 2 : Sketch of the gas counting setup with passiveshield and active plastic scintillator veto system.
Together with 10 % of the quenching gas methane thexenon is used as a counting gas for the proportionalcounters which are operated at atmospheric pressure.The radon issue is discussed in more detail in section 6.The energy calibration is done by illuminating theproportional counters with cerium X-rays according toa technique which was developed for the
Gallex ex-periment [22]. The cerium X-rays excite xenon atomsand generate a discrete low-energy line spectrum inthe proportional counter. The energy-to-channel al-location is done by fitting a second order polynomialto the positions of the three peaks that appear at 1.1keV, 5.0 keV and 9.7 keV. For some runs the full ab-sorption peak centered at 34.6 keV was included in thefit and it was confirmed that a second order polyno-mial provides a good description up to that energy.Another peak at 0.3 keV was not considered, becauseit was not always clearly resolved due to its vicinityto the threshold.
The counting system (see Figure 2) to read out theproportional counters is located at the subterraneanLow-Level-Laboratory of the Max-Planck-Institut fürKernphysik in Heidelberg, which is at a depth cor-responding to about 15 m of water equivalent. Thecounters are placed in a plastic scintillator veto sys-tem which provides almost 4- π coverage to reject cos-mic muons. The plastic scintillator itself is embeddedin a 15 cm lead shield. Xe decays with a lifetime of 7.57 days by betadisintegration with a Q-value of 427.4 keV (see Ta-ble 1). 99.12 % of all decays populate a 81 keV iso-meric state (lifetime 9.1 ns) which de-excites with 37 %
Figure 3 : Spectrum of a
Xe standard for efficiencycalibration. The bump around 5 keV is due to admixed m Xe, which is present at a few percent in the
Xestandard. probability by emission of a gamma quant and in theremaining cases by internal conversion [8]. In the pro-portional counters we measure the energy depositionof emitted electrons and of X-rays from simultaneousde-excitation of the atomic shell. Typically, only asmall fraction of the released energy is deposited inthe miniaturized counters, thus a low energy thresh-old is required. At around 150 eV the efficiency of ourdata acquisition system drops down significantly. Toavoid threshold effects a conservative 500 eV cut-offwas applied in the data analysis. The 81 keV gammaray might trigger an unwanted veto signal in the plas-tic scintillator. Thus, the counters were directly sur-rounded by a ∼ mm lead shield to which a box con-taining the frontend electronics is connected. Signalsare read out by a Flash - Analog to Digital (FADC)converter board (Struck SIS3301) at 100 MHz sam-pling rate and 14 bit resolution. The rise-time of aproportional counter pulse is defined by the time dif-ference between 10 % and 90 % of its maximum am-plitude. The plastic scintillator covers a surface ofapproximately 0.6 m perpendicular to the muon fluxresulting in a muon event rate of about 100 Hz. Eachsignal from the proportional counter that is not in co-incidence with a plastic scintillator signal is recordedwith the indication of timestamp, energy and rise-time. Xe-events have a rise-time of ∼ µ s. Aconservative rise-time cut of µ s was applied to dis-criminate against a population of slow pulses whichwas present in some of the data. H. Simgen et. al: Detection of
Xe from the Fukushima nuclear power plant in the upper troposphere above Germany
Isotope Lifetime τ Q-value [keV] Decay-mode
Xe 7.57 d 427.4 β -decay to excited levels m Xe 17.21 d 163.9 Converted gamma transition Kr 15.51 a 687.1 β -decay ( . to excited level) Rn 5.52 d 5590.3 α -decay Table 1 : Selected decay data of relevant radionuclides taken from [8].
A commercial
Xe calibration standard was usedto determine the detection efficiency of the propor-tional counters for
Xe. It was first measured witha high purity germanium (HPGe) gamma spectrome-ter to identify unwanted radionuclides. Besides
Xe,the gamma rays from m Xe and Kr were visible inthe spectrum indicating contamination with these nu-clides at the percent level. Figure 3 shows an energyspectrum of the calibration standard as recorded withone of the proportional counters.Alpha-particles deposit a large amount of energy inthe counter due to the high ionization density alongtheir track. Thus, overflow channels are not consid-ered in the analysis and an upper energy threshold of27 keV is applied to remove these events.Using the selection criteria discussed in section 3good events are selected in a window between 0.5 keVand 27 keV and with a rise-time of less than µ s.To obtain the efficiency for Xe, the temporal decaycurve of these events is fitted with two exponentialdecays for
Xe and m Xe, respectively and a con-stant background component, which comprises mainlylong-lived Kr. In the fit the lifetimes of
Xe and m Xe were fixed to the literature values given in Ta-ble 1. The obtained efficiencies for
Xe vary between49 % and 56 % for the six counters in use. Such vari-ation is expected due to geometrical differences of thecounters resulting in different ratios of active to pas-sive volumes.
The recorded events of all samples were analyzed witha maximum likelihood approach which was originallydeveloped for data analysis in radiochemical solar neu-trino experiments [23]. To estimate the initial num-ber N of Xe atoms in a given sample the proba-bility density function of
Xe decays is used whichdecreases exponentially with the
Xe lifetime. Incontrast, the background distribution is assumed to beconstant in time with a rate b . The expected sourcesof background events are unvetoed muons and con-tamination of the proportional counters with traces ofradioactive nuclides (e.g. Pb) with long lifetimes. Contributions from m Xe, m Xe and
Xe whichwere also released during the accident are negligible inthe four samples of the first flight [7]. Data from theBundesamt für Strahlenschutz (see section 7) suggestthat the m Xe fraction of the signal reaches 11 %in the second flight. We checked that the inclusionof this component in the likelihood function does notchange the result for
Xe significantly. In particu-lar, since the m Xe also originates from Fukushima,it slightly improves the significance for the detectionof radioxenon from the accident. However, the effi-ciency for detection of m Xe was not calibrated, sowe decided to ignore it in the analysis.If t i are the time stamps of the n recorded eventsand if T describes the time intervals of the total mea-surement, the resulting likelihood function L can bewritten as log L ( N, b ) = − (cid:90) T (cid:18) Nτ e − t/τ + b (cid:19) dt ++ n (cid:88) i =1 log (cid:18) Nτ e − t i /τ + b (cid:19) . (1)While the first part of this equation is derivedfrom the probabilities of events occurring at the timestamps t i , the second part describes the likelihood ofno observed events in between. Finally, log L ( N, b ) isnumerically maximised to get the best estimators ˆ N and ˆ b for the free parameters. In maximum likelihoodtheory, the statistical error of a fit-parameter a is es-timated by varying the parameter around its best fitvalue ˆ a until log L (ˆ a ) − log L (ˆ a ± σ a ) = 12 (2)while log L (ˆ a ± σ a ) is maximised regarding theremaining free parameters. To consider a possibleasymmetry of the error, both sides of ˆ a are treatedindependently. However, the asymmetry turned outto be negligible in our case. . Simgen et. al: Detection of Xe from the Fukushima nuclear power plant in the upper troposphere above Germany Figure 4 : Temporal development of the measured decayrate per day for sample ’Flight A-1’. The excess due topresence of
Xe in the left part of the plot is clearly vis-ible. The black curve shows the result and the 1 σ uncer-tainties (grey band) of the maximum likelihood analysis. The
Rn concentration in ambient air fluctuates de-pending on local geological properties, on meteoro-logical influences and on the altitude. Even in theupper troposphere it might not be negligible with re-spect to the expected
Xe signal from the Fukushimaplume [24]. Thus, special care was taken to separateradon from xenon during the gas chromatography pro-cedure (see section 2). However, since the separationis difficult, it cannot be excluded that a few radonatoms entered the proportional counter. The lifetime τ = 5 .
52 d [8] of
Rn is in the range of
Xe, there-fore radon can not be treated as a part of the constantbackground b in the likelihood analysis.To investigate possible radon contamination ofthe samples, a delayed coincidence analysis wasperformed: The Rn decay chain includes
Biwith its short-lived progeny
Po (lifetime µ s [8]). Thus, a Bi-
Po (BiPo) coincidence event ischaracterized by a β -decay of Bi followed by an α -decay of Po on a short time scale. With themeasurement of a
Rn standard sample the detec-tion of one BiPo event was calibrated to correspondto an average of . ± . radon events in the energyregion of interest between 0.5 keV and 27 keV.From the six measured samples five showed no BiPoevent, while there were two BiPo events in the sample’Flight A-1’ (see section 7). Since the gas handlingprocedure was similar for all six samples we assumethe contamination risk to be the same for all sam- Sample ˆ N [atoms] ˆ b [cpd]Flight A-1 72 ±
13 3.5 ± ±
18 8.8 ± < ± ±
12 4.1 ± ±
16 5.1 ± ±
32 30.9 ± Table 2 : Results of the maximum likelihood analysis forthe six samples. The errors are 1- σ statistical uncertaintiesand the upper limit is given at 90 % confidence level. ples. Thus, the number of radon atoms entering thecounter is expected to be Poisson distributed. Us-ing the quoted numbers above, a likelihood analysisof this scenario leads to a correction of the number N of Xe atoms in each sample by . +3 . − . . In total six samples were obtained during the twoflights. The first flight took place on March 23, 2011and four samples were collected in evacuated stainlesssteel containers of one litre volume. On the secondflight on April 14, 2011 we combined five samples fromvarious places in Northern Germany and five samplesfrom various places in Southern Germany to obtaintwo larger samples, each of about 10 litres air. Allsamples were taken inside the aircraft’s cabin wherea constant pressure of 800 mbar is maintained. Thecabin air is permanently exchanged with outside airand a complete exchange takes only a few minutes.The samples were counted in our gas countingsetup differently long, but at least several months,such that the constant background rate could be pre-cisely determined after the decay of
Xe. As an ex-ample Figure 4 shows the data from sample ’FlightA-1’. The result of the maximum likelihood analysisis given in Table 2. For each sample the number ofinitial atoms at the start of the measurement ˆ N andthe constant background rate ˆ b as described in section5 are given. Sample ’Flight B-2’ suffered from an un-expected high background which cannot be explainedeasily. Five of the six samples show a signal above thedecision threshold. For sample ’Flight A-3’ an upperlimit at 90% confidence level is given.The fit results are corrected for the Rn contami-nation and then converted into a
Xe activity at thetime of sampling by using the measured
Xe detec-tion efficiency of the six proportional counters. Alsothe losses as determined with the xenon carrier (seesection 2) are taken into account. Finally, the cor-responding
Xe activity concentration is calculatednormalized to air at standard temperature and pres-
H. Simgen et. al: Detection of
Xe from the Fukushima nuclear power plant in the upper troposphere above Germany
Sample Date Coordinates Altitude
Xe activity[km] [mBq/m (STP)]Flight A-1 23.3.11 48 ◦ N, 11 ◦ E 8.1 540 ± ◦ N, 11 ◦ E 9.2 950 ± ◦ N, 11 ◦ E 9.2 < Flight A-4 23.3.11 51 ◦ N, 11 ◦ E 11.8 210 ± ± ± Table 3 : Measured
Xe activity concentrations and coordinates / altitude of the sampling position.
Figure 5 : Comparison of the measured
Xe concentration during the flights and in ground level air recorded by theBundesamt für Strahlenschutz on the Schauinsland mountain in Southern Germany. sure (STP: 273.15 K and Pa). Again the asymme-try of the 1 σ uncertainties is ignored since it turnedout to be negligible. The final results are shown inTable 3.The Xe concentration in ambient air is usuallywell below 50 mBq/m with rare short-term excep-tions [25]. Consequently, we can conclude that in thefirst flight on March, 23 2011 clear evidence of Xefrom the radioactive plume of Fukushima Dai-ichi inthe upper troposphere above Germany was observed.The evidence from the second flight about three weekslater is weaker, since
Xe has decayed and is fur-ther diluted. Nonetheless, our results for the secondflight are still above the usual ambient activity con-centration indicating that we were still detecting theemission from Fukushima.The Bundesamt für Strahlenschutz (BfS) oper-ates a station for monitoring the environmental ra-dioactivity at the Schauinsland mountain in South- ern Germany. As Radionuclide Station 33 it is partof the International Monitoring Network (IMS) ofthe Comprehensive Nuclear-Test-Ban Treaty Organi-zation (CTBTO) [26, 27] and a SPALAX noble gassystem [28] is installed to monitor continuously theactivity concentration of radioactive xenon in groundlevel air. It is based on fully automated samplingof large air samples, purification, concentration ofxenon on activated carbon columns and detection bygamma ray spectroscopy with a high purity germa-nium detector. With this system 24 h measurementsof four xenon radioisotopes (
Xe, m Xe, m Xeand
Xe) in ambient air are performed. For
Xethe minimum detectable activity concentration liesaround 450 µ Bq/m (STP). Note that the air samplesize is typically 60 m (STP) which is about timeslarger than the air samples taken during the flights inour study. In Figure 5 results from the Schauinslandstation in the period after the Fukushima accident are . Simgen et. al: Detection of Xe from the Fukushima nuclear power plant in the upper troposphere above Germany
Xefrom Fukushima is visible on March 24, 2011. How-ever,
Xe activity concentrations at the 1 Bq/m level are detected for the first time on March 29, 2011.The relatively high Xe activity concentrationsmeasured during the first flight on March 23, 2011in the upper troposphere show that the arrival of theFukushima plume at high altitudes was earlier thanat ground level. Also remarkable is the evidence fortraces of
Xe in the tropopause (sample Flight A-4, green diamond in Figure 5). Both observationsindicate that air masses from the Japanese regionswere quickly lifted to high altitudes. For more infor-mation of atmospheric implications and comparisonswith ground-level measurements the reader is referredto [9]. The
Xe activity concentrations in Flight2 were significantly lower than simultaneous groundlevel air measurements at Schauinsland. This may beexplained by settling of the heavy xenon in the at-mosphere and is also confirmed by model calculationspresented in [9].
We have reported a new technique successfullycombining
Xe detection using miniaturized pro-portional counters with airplane sampling campaigns.In particular, we have showed that small air samplesof litre scale (STP) are sufficient to achieve a com-petitive detection limit in the range of 100 mBq/m (STP). This is due to both the ultra-low backgroundrate of the applied proportional counters which is inthe range of few events per day at 0.5 keV thresholdand the highly efficient and contamination-free gassample preparation procedure.Our measurements reveal that a part of theFukushima Xe plume was lifted to the upper tro-posphere and to the tropopause. There it was carriedby the fast jet stream to Europe. The earlier arrivaltime in Germany at high altitude compared to groundlevel was unambiguously demonstrated by a compari-son with data from the CTBT Radionuclide Station 33of the Bundesamt für Strahlenschutz on the Schauins-land mountain. We also got evidence for
Xe tracesfrom the Fukushima accident in the upper troposphereduring our second flight on April 14, 2011, althoughradioactive decay, dilution and settling of xenon re-duced the
Xe activity concentration significantly.
References
Journalof Environmental Radioactivity 111 (2012) 13-17.[3] P. Bailly du Bois et al.:
Estimation of marinesource-term following Fukushima Dai-ichi acci-dent.
Journal of Environmental Radioactivity 114(2012) 2-9.[4] M. Schöppner et al.:
Estimation of thetime-dependent radioactive source-term from theFukushima nuclear power plant accident using at-mospheric transport modelling.
Journal of Envi-ronmental Radioactivity 114 (2012) 10-14.[5] A. Stohl et al.:
Xenon-133 and caesium-137 re-leases into the atmosphere from the FukushimaDai-ichi nuclear power plant: determination ofthe source term, atmospheric dispersion, and de-position.
Atmospheric Chemistry and Physics 12(2012) 2313-2343.[6] A. Stohl, P. Seibert, G. Wotawa:
The total re-lease of xenon-133 from the Fukushima Dai-ichinuclear plant accident.
Journal of EnvironmentalRadioactivity 112 (2012) 155-159.[7] T. W. Bowyer et al.:
Elevated radioxenon de-tected remotely following the Fukushima nuclearaccident.
Fukushima radionuclide
Xeaircraft measurements reveal cyclone inducedrapid lift of polluted planetary boundary layer airto the tropopause: Implications for aerosol forma-tion.
To be submitted to Atmospheric Chemistryand Physics.[10] O. Masson et al.:
Tracking of Airborne Radionu-clides from the Damaged Fukushima Dai-Ichi Nu-clear Reactors by European Networks.
Environ-mental Science and Technology 45 (2011) 7670-7677.[11] V. Fiedler et al.:
East Asian SO pollutionplume over Europe - Part 1: Airborne trace gasmeasurements and source identification by par-ticle dispersion model simulations. AtmosphericChemistry and Physics 9 (2009) 4717-4728.[12] V. Fiedler et al.:
East Asian SO pollution plumeover Europe - Part 2: Evolution and potentialimpact. Atmospheric Chemistry and Physics 9(2009) 4729-4745.
H. Simgen et. al: Detection of
Xe from the Fukushima nuclear power plant in the upper troposphere above Germany [13] H. Schlager et al.:
Cyclone induced transport ofEast-Asian reactive sulfur and nitrogen gases tothe lowermost stratosphere. Part A: First detec-tion by HALO-aircraft.
The miniaturized proportionalcounter HD-2(Fe)/(Si) for the G
ALLEX solarneutrino experiment.
Nuclear Instruments andMethods in Physics Research A 329 (1993) 541-550.[16] S. Pezzoni:
Non solar-production of Ge in
Gallex . PhD thesis, Universität Heidelberg(1995).[17] W. Rau, G. Heusser:
Rn emanation measure-ments at extremely low activities.
Applied Radi-ation and Isotopes 53 (2000) 371-375.[18] G. Zuzel. H. Simgen:
High sensitivity radon ema-nation measurements.
Applied Radiation and Iso-topes 67 (2009) 889-893.[19] S. Lindemann, H. Simgen, G. Zuzel:
Behaviour of
Rn at cryogenic temperatures.
AIP ConferenceProceedings, Volume 1338 (2010) 156.[20] H. Simgen, G. Heusser, G. Zuzel:
Highly sensitivemeasurements of radioactive noble gas nuclidesin the
Borexino solar neutrino experiment.
Ap-plied Radiation and Isotopes 61 (2004) 213-217.[21] Johns-Manville:
A unique porous packing support.Chromosorb 102.
Analytical Chemistry 39, 12(1967) 89A.[22] A. Urban:
Analyse von Proportionalzählrohrim-pulsen zum Nachweis von solaren Neutrinos.
Dissertation, Technische Universität München(1989).[23] B. T. Cleveland:
The analysis of radioactive de-cay with a small number of counts by the methodof maximum likelihood.
Nuclear Instruments andMethods 214 (1983) 451-458.[24] M. Ramonet et al.:
Rn measurements dur-ing the Tropoz II campaign and comparison witha global atmospheric transport model.
Journal ofAtmospheric Chemistry 23 (1996) 107-136.[25] P.R.J. Saey et al.:
Environmental RadioxenonLevels in Europe - a Comprehensive Overview.
Monitoringground-level air for trace analysis: methods andresults.
Analytical and Bioanalytical Chemistry379,2 (2004) 234-241.[28] J.-P. Fontaine et al.: