Äîêóìåíò âçÿò èç êýøà ïîèñêîâîé ìàøèíû. Àäðåñ îðèãèíàëüíîãî äîêóìåíòà : http://rtm-cs.sinp.msu.ru/docs/borexino/pxe_2007.pdf Äàòà èçìåíåíèÿ: Thu Jan 24 13:54:11 2008 Äàòà èíäåêñèðîâàíèÿ: Mon Oct 1 20:31:31 2012 Êîäèðîâêà:

Phenylxylylethane (PXE): a high-density, high-flashp oint organic liquid scintillator for applications in low-energy particle and astrophysics exp eriments
Borexino Collaboration

H.O. Backa , M. Balatab , A. de Baric , T. Beaud , A. de Bellefond , G. Bellinie , J. Benzigerf , S. Bonettie , A. Brigattie , C. Buckg , B. Caccianigae , L. Cadonatif ,1 , F. Calapricef , G. Cecchetc , M. Chenh , A. Di Credicob , O. Dadound,9 , D. D'Angeloe,2 , H. de Haasf , A. Derbin
j,10

, M. Deutsch

k ,15

, F. Eliseil , A. Etenkom ,

F. von Feilitzschi , R. Fernholzf , R. Fordf ,3 , D. Francoe , B. Freudigerg,15 , C. Galbiatif , N. Gartneri , F. Gattin , S. Gazzanab , M.G. Giammarchie , D. Giugnie , M. G¨ oger-Neff A. Gorettib , C. Griebi , C. Hagneri,4 , W. Hampelg , E. Harding
f ,13 i,16

,

, F.X. Hartmanng , T. Hertrich i , H. Hessi ,

G. Heusserg , A. Iannib , A.M. Iannif , H. de Kerretd , J. Kikog , T. Kirsteng , G. Korga
e,12

, G. Korschineki , Y. Kozlovm ,

D. Krynd , M. Laubensteinb , C. Lendvaii , F. Loeserf , P. Lombardie , S. Malvezzie , J. Maneirae,5 , I. Mannoo ,

Preprint submitted to Elsevier Science

31 August 2007


D. Manuzion , G. Manuzion , F. Masettil , A. Martemianov

m,15

,

U. Mazzucatol , K. McCartyf , E. Meronie , L. Miramontie , M.E. Monzanie , P. Musicon , L. Niedermeieri,9 , L. Oberaueri , M. Obolenskyd , F. Ortical , M. Pallavicinin , L. Papp
e,12

, ,

S. Parmeggianoe , L. Perassoe , A. Pocarf , R.S. Raghavan G. Ranuccie , W. Raug,2 , A. Razeton , E. Resconin,6 ,

p,14

A. Sabelnikove , C. Salvon , R. Scardaonie , D. Schimizzif , S. Sch¨ onert
g ,16

, K.H Schuhbecki,7 , E. Seitzi , H. Simgeng ,

T. Shuttf , M. Skorokhvatovm , O. Smirnovj , A. Sonnenscheinf ,8 , A. Sotnikovj , S. Sukhotinm , V. Tarasenkovm , R. Tartagliab , G. Testeran , D. Vignaudd , R.B. Vogelaara , V. Vyrodovm , M. Wo jcikq , O. Zaimidorogaj , G. Zuzel
a

q

Physics Department, Virginia Polytechnic Institute and State University, Robeson Hal l, Blacksburg, VA 24061-0435, USA

b

I.N.F.N Laboratori Nazionali del Gran Sasso, SS 17 bis Km 18+910, I-67010 Assergi(AQ), Italy

c

Dipartimento di Fisica Nucleare e Teorica Universit` and I.N.F.N., Pavia, Via a A. Bassi, 6 I-27100, Pavia, Italy

d

Astroparticule et Cosmologie, 10, rue Alice Domon et Leonie Duquet, F-75025 Paris cedex 13

e

Dipartimento di Fisica Universit` and I.N.F.N., Milano, Via Celoria, 16 I-20133 a Milano, Italy
f

Dept. of Physics,Princeton University, Jadwin Hal l, Washington Rd, Princeton NJ 08544-0708, USA

2


g

Max-Planck-Institut fur Kernphysik,Postfach 103 980, D-69029 Heidelberg, ¨ Germany

h

Dept. of Physics, Queen's University Stirling Hal l, Kingston, Ontario K7L 3N6, Canada

i

Technische Universit¨ Munchen, Physik Department E15, James Franck Straúe, at ¨ D-85747, Garching, Germany
j k

Joint Institute for Nuclear Research, 141980 Dubna, Russia

Dept. of Physics Massachusetts Institute of Technology,Cambridge, MA 02139, USA
l

Dipartimento di Chimica Universit` Perugia, Via Elce di Sotto 8, I-06123 a, Perugia, Italy
m

RRC Kurchatov Institute, Kurchatov Sq.1, 123182 Moscow, Russia

n

Dipartimento di Fisica Universit` and I.N.F.N., Genova, Via Dodecaneso,33 a I-16146 Genova, Italy
o p

KFKI-RMKI, Konkoly Thege ut 29-33 H-1121 Budapest, Hungary

Bel l Laboratories, Lucent Technologies, Murray Hil l, NJ 07974-2070, USA

q

M. Smoluchowski Institute of Physics, Jagel lonian University, PL-30059 Krakow, Poland

Abstract We report on the study of a new liquid scintillator target for neutrino interactions in the framework of the research and development program of the Borexino solar neutrino experiment. The scintillator consists of 1,2­dimethyl­4­(1­phenylethyl)­ benzene (phenyl­o­xylylethane, PXE) as solvent and 1,4-diphenylbenzene (paraTerphenyl, p-Tp) as primary and 1,4-bis(2-methylstyryl)benzene (bis-MSB) as secondary solute. The density close to that of water and the high flash point makes it an attractive option for large scintillation detectors in general. The study focused

3


on optical properties, radioactive trace impurities and novel purification techniques of the scintillator. Attenuation lengths of the scintillator mixture of 12 m at 430 nm were achieved after purification with an alumina column. A radio carbon isotopic ratio of
14

C/12 C = 9.1 â 10-
238

18

has been measured in the scintillator. Initial trace
14

impurities, e.g.

U at 3.2 â 10-

g/g could be purified to levels below 1 â 10

-17

g/g

by silica gel solid column purification. Key words: Phenyl-o-xylylethane, PXE, organic liquid scintillator, solar neutrino spectroscopy, low-background counting, Borexino PACS: 14.60.Pq, 23.40.-s, 26.65.+t, 29.40.Mc, 91.35.Lj

1

Now at Massachusetts Institute of Technology, NW17-161, 175 Albany St. Cam-

bridge, MA 02139
2

Now at Technische Universit¨ Munchen, James Franck Straúe, D-85747 Garchat ¨

ing, Germany
3

Now at Sudbury Neutrino Observatory, INCO Creighton Mine, P.O.Box 159

Lively, Ontario, Canada, P3Y 1M3
4

Now at Universit¨ Hamburg, Luruper Chaussee 149, D-22761 Hamburg, Gerat

many
5

Now at Dept. of Physics, Queen's University Stirling Hall, Kingston, Ontario K7L

3N6, Canada
6 7

Now at Max-Planck-Institut fuer Kernphysik, Heidelberg, Germany Now at Max-Planck-Institut fuer Plasmaphysik, Boltzmannstr.2 D-85748 Garch-

ing, Germany
8

Now at Center for Cosmological Physics, University of Chicago, 933 E.56th St.,

Chicago, IL 60637
9

Marie Curie fellowship at LNGS On leave of absence from St. Petersburg Nuclear Physics Inst. - Gatchina, Russia On leave of absence from Institute for Nuclear Research, MSP 03680, Kiev,

10 11

4


1

Intro duction

Organic liquid scintillators are used in large quantities for rare event detection in particle astrophysics. The main ob jective in these experiments is the real time spectroscopy of neutrinos from steady-state sources such as the Sun, nuclear reactors and from beta decays in the crust and mantle of the Earth, as well as from transient sources such as supernovae. Despite the large target mass of several hundreds of tons, the signal rates of the steady-state sources are typically in the range of a few events per day down to a few events per year at MeV or sub-MeV energies. Thus, background signal rates created by radioactivity and cosmic ray interactions need to be extremely low. Low backgrounds can be achieved by locating the detectors deep underground to suppress the cosmic ray muon flux, shielding the scintillator target against the ambient radioactivity from the surrounding rocks, and suppressing and
Ukraine
12

On leave of absence from KFKI-RMKI, Konkoly Thege ut 29-33 H-1121 Bu-

dapest, Hungary
13 14

Now at Lockhead Martin Corporation, Sunnyvale CA Now at Physics Department, Virginia Polytechnic Institute and State University,

Robeson Hall, Blacksburg, VA 24061-0435, USA
15 16

Deceased Corresponding authors:

Stefan Sch¨ onert, email: stefan.schoenert@mpi-hd.mpg.de, Marianne G¨ oger-Neff, email: marianne.goeger@ph.tum.de.

5


removing radioactive impurities present in trace amounts in the detector and ancillary systems as well as in the liquid scintillator itself. This concept of background reduction has been pioneered by the Borexino collaboration [1] in the Counting Test Facility (CTF) [2­4] and is implemented in the Borexino detector, and similarly, in the KamLAND experiment [5]. This paper summarizes the study of 1,2­dimethyl­4­(1­phenylethyl)­benzene (phenyl­o­xylylethane, PXE), a new scintillator solvent the key characteristics of which are its high density (0.988 g/cm3 ) and high flash point (145o C). This scintillator solvent has been investigated as a `back-up solution' for the Borexino experiment. The design of choice is based on 1,2,4-trimethylbenzene (pseudocumene, PC), both as buffer liquid and as neutrino target. Since the density of PXE is close to that of water, even large scintillator targets can be submerged in water - serving as a shield against ambient radiation while creating only modest buoyancy forces on the scintillator containment vessel. Moreover, a detector with water as shield and PXE as target material provides a substantial higher fiducial target mass because of the improved shielding performance against external radiation compared to a detector with identical dimensions, but with both shield and target of organic liquids with standard densities ( 0.9 g/cm3 ). Finally, a PXE-water configuration reduces the overall inventory of organic liquid in the detector systems, e.g. to about one fourth in the case of Borexino. This, depending on national regulations, may have an impact on the legal classification of the detector systems and therefore on the safety and operational aspects of the experiment. A further asset of PXE is its high flash point simplifying safety systems relevant for transportation, handling and storage. According to regulations by the United Nations (UN), PXE is legally non-hazardous for transportation purposes and 6


no special United Nation number code applies [6]. In a paper by Ma jewski et al. [7] PXE has been described as a relatively safe solvent with very low toxicity compared to standard liquid scintillators. The Double Chooz reactor neutrino experiment [8], aiming at a measurement of the mixing angle 13 , will use as target a scintillator mixture based on PXE.

The study reported here of PXE as a solvent for a low-background scintillator for solar neutrino spectroscopy was carried out within the Borexino pro ject. Research on PXE scintillators started in 1995 with laboratory measurements focusing on optical properties and radio purification techniques with solid columns. After completion of the laboratory scale study, about five tons of PXE solvent for testing on prototype scale with the Counting Test Facility (CTF) of Borexino were acquired in 1996. Fluors were added on site in hall C of the Laboratori Nazionali del Gran Sasso (LNGS) and the final scintillator was purified with Module-0, a solid column purification and liquid handling system [9]. The CTF was loaded with PXE scintillator in October 1996 and first data were acquired until January 1997. The quality of data was limited because only a small fraction of the photomultipliers in the CTF were operational. The PXE scintillator was unloaded from the CTF after the shut down of the detector in July 1997, and moved back into the storage tanks of Module-0. Further batch purification operations were carried out during the period October until December 1997. Samples for neutron activation analysis were taken to monitor the achieved radio purity after each operation. After reconstruction of the CTF during 1999, the PXE scintillator was reloaded into the CTF and measured from June to September 2000. The main ob jective during this period was the analysis of radio purity and optical properties in a large volume detector. Beyond this scope, physics limits on electron instability 7


[10], nucleon instability [11], neutrino magnetic properties [12], and on violations of the Pauli exclusion principle [13] could be derived from measurements with the PXE scintillator. The paper is structured as follows: Section 2 summarizes the physical and chemical properties of the PXE solvent. Section 3 describes the optical properties of the solvent as well as the mixed scintillator. Section 4 is dedicated to the large scale test of the PXE scintillator with the CTF including the scintillator preparation, purification and analysis of trace impurities, and conclusions are given in Section 5.

2

Physical and chemical prop erties of PXE

PXE is a clear, colorless liquid with an aromatic odor. It is an industrial product with different applications, as for example: insulating oil in high voltage transformers and capacitors and as oil for pressure sensitive paper. PXE is produced by reacting styrene and xylene using an acidic catalyst. It is then washed with water and distilled to improve the purity. Its final industrial purification step uses a solid column. PXE has the molecular formula C16 H18 with a weight of 210.2 g/mol. Its molecular structure is shown in Figure 1. Chemical and physical details which are of relevance for detector design, safety and operational aspects are listed in Table 1. The key features are its high density of 0.988 g/cm3 , its low vapor pressure and high flash point of 145 o C. It is therefore classified as a nonhazardous liquid. For our test, the standard production scheme at Koch Chemical Company, 8


Corpus Christi, Texas, USA

2

was modified omitting the final column purifi-

cation at the company, since it was expected that the clay column material might leach off radio active impurities, such as uranium and thorium. Instead a column purification system was built and operated using silica gel in the underground laboratories of the LNGS. Details are discussed below.

3

Optical prop erties

The optical and scintillation properties of the pure PXE solvent, of selected fluors, and of PXE-fluor mixtures have been investigated by UV/Vis spectrometry, by fluorimetry and by excitation with ionizing radiation. The ob jective was to optimize the scintillator performance by maximizing its light yield and attenuation length and minimizing its scintillation decay time. Methods to remove optical impurities were studied, since impurities can potentially reduce light yield and attenuation length. All optical properties presented in this section are derived from laboratory size samples up to a few hundred ml's. Attenuation length measurements typically are done in `one-dimension' only. Scattered light, elastic or inelastic, is undetectable in these measurements whereas in large volume applications with 4 geometry, the scattered photons are not necessarily lost. The performance of the PXE scintillator in a large volume detector, taking scattering into account, has been studied in CTF and is discussed in Sec. 4.4.
2

Koch Special Chemical Company stopped production of PXE in 2002. Nippon

Petrochemicals Co., Ltd., Japan, produces PXE as an equal mixture of ortho, para and meta isomers

9


3.1 Solvent properties

PXE diluted in cyclohexane shows an absorption maximum at 267 nm. The emission spectrum after excitation at this wavelength peaks at about 290 nm as displayed in Figure 2. The fluorescence lifetime after excitation at
exc

=

267 nm measured in diluted solutions of cyclohexane shows an exponential decay with a lifetime of 22 ns. Preliminary samples obtained from Koch had various optical impurities with absorption bands around 300, 325, 360 and 380 nm. They could be observed both by UV/Vis and fluorimetric measurements. Passing the solvent through a column with acidic alumina reduced the absorption peaks. The band at 380 nm was reduced most efficiently. The PXE used for the 5 t test, as described Section 4, still had various optical impurities, but at much reduced levels. The attenuation length at 430 nm (430 ), defined as I (x) = I0 · exp(-x/430 ), with the initial intensity I0 and attenuated intensity I (x) after the optical path length x, was
430

= 10.2 m

in the pure PXE solvent. Attenuation lengths were measured with a Varian Cary 400 UV/Vis photospectrometer in 10 cm cuvettes.

3.2 Scintil lator properties

A tertiary scintillator system was chosen in order to shift the emission wavelength to about 430 nm, well above the absorption bands of the residual optical impurities. This is achieved by using 1,4-diphenylbenzene (para-Terphenyl, p-Tp) as primary solute and 1,4-bis(2-methylstyryl)benzene (bis-MSB) as secondary solute. Absorption and emission spectra of PXE in cyclohexane and those of the fluors in PXE are shown in Figures 2 and 3. The latter shows 10


that the emission spectrum of p-Tp is satisfactorily matched by the absorption spectrum of the wavelength shifter bis-MSB, and therefore an efficient energy transfer is expected in this solvent. The scintillator properties were tested with p-Tp concentrations at 2 and 3 g/l (close to the solubility limit) and with bis-MSB at 20 mg/l. A scintillation yield of 88 (93) % for PXE with an addition of p-Tp (2.0 (3.0) g/l)/bisMSB(20 mg/l) with respect to a scintillator based on 1,2,4-trimethylbenzene (PC) and 1.5 g/l of 2,5-diphenyloxazole (PPO) has been found (uncorrected for the PMT sensitivity). Attenuation lengths of
430

= 2.6 m to 3.2 m of the

scintillator mixture have been measured depending on the sample treatment (cf. Section 4.2). The fluorescence decay time of the p-Tp(2.0 g/l)/bis-MSB(20 mg/l) scintillator mixture measured by fluorimetry after excitation at 267 nm shows a fast component of 3.7 ns. At 3 g/l p-Tp, the decay constant is shortened to 3.2 ns. The photon emission probability density function (pdf ) was further studied with ionizing radiation in order to investigate the contribution of long lived triplet states and the possibility to use the emission time for discrimination of alpha versus beta particles (pulse shape discrimination). The response to gamma radiation has been measured for a concentration of 2.0 g/l and 3.0 g/l p-Tp with a 3.0 g/l using a
210 137

Cs source, and to alpha radiation at a concentration of

Po alpha source [14,15]. The pdf can be parametrized by the
i

weighted sum of four exponentials

(qi /i ) exp(-t/i ) with the parameters

given in Table 2. The emission time distribution is shown in fig. 5. In addition to an increased slow component which can be used for pulse shape discrimination, alpha particles emit less light because of the high ionization density compared to electron or gamma radiation. The quenching factor is mainly sol11


vent dependent and has been measured for different alpha decays in the chain, using
222

238

U

Rn loaded scintillator. The results are given in Table 3 [16].

Light attenuation of the standard mixture (PXE/(2.0 g/l)p-Tp/(20 mg/l)bisMSB) which has been used in the CTF (cf. Sec. 4) has been measured at various steps during the preparation of the scintillator in 1996 and after filling the CTF. After several purification steps of the mixed scintillator through a silica gel column in Module-0 in order to remove radio-impurities (cf. sections below), attenuation lengths in the range between 2.6 and 3.0 m at 430 nm were measured. After completion of the PXE measurements with CTF and subsequent batch purification operations, the scintillator has been stored in barrels under nitrogen atmosphere. In 2003, this scintillator was used in the frame of the LENS pro ject [17] and optical properties were remeasured to check for degradation with time. The attenuation length as well as the light yield were unchanged with respect to the measurements in 1996. Passing this scintillator mixture through a weak acidic alumina column increased the attenuation lengths to 12 m as displayed in Figure 4 while retaining the scintillation yield [18].

4

Large scale test of PXE with the Borexino Counting Test Facility

The main ob jectives of the large scale test in the Borexino prototype detector (Counting Test Facility, CTF [2]) were the study of 1) optical properties on a large scale, 2) achievable levels of radioactive trace contaminations of the PXE based scintillator, 3) the performance of scintillator purification with a silica gel column, and 4) of the liquid handling system, Module-0. The purity levels required for detection of low-energy solar neutrinos, in particular neu12


trinos from the 7 Be electron capture, are at the µBq/m3 level corresponding to concentration of
238

U and

232

Th of 10-

16

g/g. For detailed specifications

of solar neutrino rates and background requirements, the reader is referred to Refs. [1,19]. As the
238

U­progenies

222

Rn,

210

Pb,

210

Bi and

210

Po are not

necessarily in equilibrium with the progenitor activity, we studied the contamination levels of
238

U (as well as

232

Th and several other isotopes) directly
-16

with neutron activation analysis (NAA) at levels of 10

g/g and below. The

short lived progenies are not accessible with NAA. Their contamination levels were studied in the CTF together with other backgrounds, as for example the radiogenic produced
14

C. New techniques to remove radioactive isotopes by

solid column purification were tested for the first time in a ton scale experiment. Furthermore, the preparation, handling and purification of several tons of liquid scintillator with Module-0 was the first test of a subsystem of the liquid handling system to be used in the scintillator operations in Borexino.

4.1 Scintil lator preparation

About five tons of PXE solvent was procured from Koch Special Chemical Company, Texas, USA. To ensure a controlled procedure, the solvent was loaded by us at the company site in three specially modified and cleaned stainless steel transport containers and shipped to the Gran Sasso underground laboratories. The PXE solvent was transferred from the transport containers into Module-0, a liquid handling and purification system, specially built for the PXE test. It can be used for volumetric loading and unloading of liquid scintillator to/from the CTF Inner Vessel (IV), for fluor mixing, purification of liquid scintillator 13


with a silica gel column, Rn degassing of liquid scintillator by nitrogen sparging, and spray degassing as well as for water extraction. Module-0 consists of a high and low pressure manifold system which are connected by pumps to build up the pressure difference of typically 2 atm. The manifolds are connected to tanks, columns, filters etc. to allow a variety of flow paths and operations. The system includes two 7 m3 electropolished pressure tanks (EP1/2), two 1 m3 process tanks (BT1/2) equipped with nitrogen spargers at the bottom and spray nozzles at the top (for a turbulent injection of liquid together with nitrogen gas), pumps, flow meters, Millipore filters (0.5, 0.1 and 0.05 micron) and one 70 l high pressure column purification unit. All tanks are connected to a high purity nitrogen gas system to provide a nitrogen blanket at typically 30 mbar overpressure. The complete system has been designed according to ultra-high-vacuum standards to avoid contaminations from the environmental air, in particular of
222

Rn and

85

Kr. Only stainless

steel and Teflon are in contact with the liquids. Metal surfaces are electro polished and welds carried out with thorium free welding rods. Module-0 has been constructed in a class 1000 clean room. After completion, all surfaces exposed to scintillator were cleaned following a detailed procedure to remove surface contaminations as radon progenies up-graded system are given in [9]. The fluors (p-Terphenyl and bis-MSB, both from Sigma-Aldrich Co., scintillation grade) were sieved and then added without further purification to EP1 through a glove box connected to the top inlet flange which was flushed with nitrogen while the PXE solvent was agitated with a nitrogen flux from the bottom inlet to facilitate dissolving p-Tp. To accelerate the solvation of p-Tp, the EP1 tank was heated to 38 o C for four days under continuous nitrogen 14
210

Pb,

210

Bi and

210

Po. Details about the


agitation. The final scintillator mixture had a concentration of 2.0 g/l p-Tp and 20 mg/l bis-MSB.

4.2 Purification with silica gel

One ob jective of the CTF test was to study the performance of scintillator purification with a solid silica gel column. Preceding laboratory tests of the column purification with PXE/p-Tp with and without the addition of radio tracer showed a clear reduction of metal impurities [20]. The silica gel used during the CTF test has been radio assayed with HP germanium spectrometry and by direct measurements of the emanated radon with proportional counters. The results are listed in Table 4. Despite the high bulk impurities of the Merck silica gel, we did not observe any measurable carry over into the scintillator apart from
222

Rn. For further use in Borexino we have found silica

gel material with improved radio purity, in particular a radon emanation rate of 0.13 ± 0.07 mBq/kg. The scintillator components had been mixed as received from the supplier; no special pretreatment had been given to the fluor and the wavelength shifter. The ready mixed PXE scintillator was then passed through a column filled with silica gel (Merck, Silica Gel 60, 25-70 mesh ASTM) at a flow rate of typically 100 l/hour. Radon was removed by purging the scintillator after passing the column as displayed schematically in Figure 6. Samples for neutron activation analysis were taken prior and after all ma jor steps of the scintillator preparation and purification: Sample (1) was collected from Module-0 after addition of p-Tp and bis-MSB to the PXE solvent (without purification). Sample (2) was collected during the filling of the IV. The complete scintillator 15


had passed once a column of about 80 cm height (15 kg) which had been exchanged against fresh silica gel after the first two tons had run through the column, and once with a height of about 50 cm, i.e. in total two times. Sample (3) has been taken after the scintillator was circulated from the IV through the column (re-filled to a height of 80 cm) and back into the IV. The flow rate was adjusted such that one cycle (5000 l) took two days. The scintillator was circulated for four days, i.e. two cycles. Subsequently the scintillator was unloaded from the IV into the EP1-tank. Sample (4) was taken after completion of a water extraction in the EP1 tank. Sample (5) was collected in Module-0 after circulating the scintillator from EP1 through the column back to EP1. The column packing had not been exchanged (same as in (E)). This final operation lasted for eight days with a flow rate of 3600 l/day, i.e. about 6 cycles.

4.3 Radio assay with neutron activation analysis

Combining neutron activation and low level counting methods, we developed a novel analytical method with sensitivities of 10
232 -16

g/g and below for

238

U and

Th in liquid scintillators. For this purpose about 250 g samples of liquid
12

scintillator were irradiated at a neutron flux of (10

- 1013 ) s-1 cm

-2

up to

100 hours at the research reactor in Garching. The long lived primordial radio nuclides are transformed into short-lived radio nuclides (e.g.
232 238

U 239 Np,

Th

233

Pa), thus providing a higher specific activity with respect to their

progenitors. After irradiation, liquid-liquid and ion exchange techniques were applied to separate the radio nuclides of interest from interfering activities. Coincidence counting methods [21], e.g. - -conversion electron for 16
239

Np


and - for

233

Pa, further increase the sensitivity. Details can be found in

Refs. [22,23,19]. The main results of the neutron activation analysis of the scintillator samples (1) to (5) are summarized in Table 5 [24]. We report the concentrations of
238

U,

232

Th and other long-lived isotopes which contribute to the back-

ground in Borexino. In order to illustrate the performance of the silica gel column, also the concentrations for several non-radioactive metal isotopes are given. An overall reduction between 1 and 3 orders of magnitude has been reached for all elements where a positive value (above the detection limit) could be measured before the purification. Except for potassium, where the purity needed for Borexino is below the detection limit of the NAA, the requirements for Borexino are met for all of the long lived radio nuclides. New purity records in organic liquid scintillators have been achieved for uranium at c(
238

U) < 1 · 10-17 g/g, and for thorium at c(

232

Th) < 1.8 · 10

-16

g/g.

4.4 Measurements of PXE scintil lator in the CTF

The Borexino prototype detector CTF is a highly sensitive instrument for the study of backgrounds in liquid scintillators at energies between a few tens of keV and a few MeV. It consists of a transparent nylon balloon with 2 m diameter containing 4.2 m3 of liquid scintillator (Inner Vesssel, IV). 100 PMTs with light concentrators mounted on a 7 m diameter support structure detect the scintillation signals (optical coverage 20 %). The whole system is placed inside a cylindrical steel tank (11 m in diameter, 10 m height) that contains 1000 tons of ultra-pure water in order to shield against external rays from the PMTs and other construction material as well as neutrons from 17


the surrounding rock. In the upgraded version of the CTF detector (after 1999), 16 additional PMTs mounted on the floor of the water tank detect the Cherenkov light created by muons transversing the water buffer and are used as a muon veto. Details of the CTF detector and results of first measurements with pseudocumene scintillator are given in Refs. [2,4].

Sequence of measurements Direct counting after first loading of the PXE scintillator into the CTF in 1997 did not provide conclusive results because of photo-tube and electronics problems of the detector system that hindered data evaluation. After an upgrade of the CTF detector and the liquid handling system during 1998 and 1999, the same scintillator was filled again into the CTF in summer 2000, after passing once through a 40 l silica gel column. The loading of the scintillator into the Inner Vessel was done in four batches of 1 ton each separated by short periods of data taking. Undisturbed data taking with 4.2 tons of PXE was going on from July 16 until September 5, 2000, in total 52 days. Afterwards, a series of calibration measurements with a
222

Rn point source were carried out where

the source was moved inside the Inner Vessel to map the detector response and tune the position reconstruction software.

Detector performances The pulse height - energy relation can be derived from the
214

Po alpha peak (cf.

Figure 7) together with the measured quenching factors as given in Table 3. Due to its short half life of 164 µs, the The
214 214

Po decay can easily be tagged.

Po stems mainly from

222

Rn introduced during scintillator loading and 18


therefore is homogeneously distributed in the scintillator volume. The alpha particle has an energy of 7.69 MeV which in the scintillator is quenched to an equivalent beta energy of (950 ± 12) k