Preparation of facilities for fundamental research with ultracold neutrons at PNPI

Abstract

The WWR-M reactor of PNPI offers a unique opportunity to prepare a source for ultracold neutrons (UCN) in an environment of high neutron flux (about 3*10^12 n/cm^2/s) at still acceptable radiation heat release (about 4*10^-3 W/g). It can be realized within the reactor thermal column situated close to the reactor core. With its large diameter of 1 m, this channel allows to install a 15 cm thick bismuth shielding, a graphite premoderator (300 dm^3 at 20 K), and a superfluid helium converter (35 dm^3). At a temperature of 1.2 K it is possible to remove the heat release power of about 20 W. Using the 4pi flux of cold neutrons within the reactor column can bring more than a factor 100 of cold neutron flux incident on the superfluid helium with respect to the present cold neutron beam conditions at the ILL reactor. The storage lifetime for UCN in superfluid He at 1.2 K is about 30 s, which is sufficient when feeding experiments requiring a similar filling time. The calculated density of UCN with energy between 50 neV and 250 neV in an experimental volume of 40 liters is about 10^4 n/cm^3. Technical solutions for realization of the project are discussed.

Full text

Page 1

Preparation of facilities for fundamental research with ultracold neutrons
at PNPI
A.P. Serebrova,?, V.A. Mityuklyaeva, A.A. Zakharova, A.N. Erykalova, M.S. Onegina, A.K. Fomina, V.A.
Ilatovskiya, S.P. Orlova, K.A. Konopleva, A.G. Krivshitcha, V.M. Samsonova, V.F. Ezhova, V.V. Fedorova,
K.O. Keshyshevb, S.T. Boldarevb, V.I. Marchenkob
aPetersburg Nuclear Physics Institute, RAS, 188300 Gatchina, Leningrad District, Russia
bP.L. Kapitza Institute for Physical Problems, ul. Kosygina, 2, Moscow 119334, Russia
a r t i c l e i n f o
Available online 7 August 2009
Keywords:
Ultracold neutrons
Cold neutrons
Neutron source
a b s t r a c t
The WWR-M reactor of PNPI offers a unique opportunity to prepare a source for ultracold neutrons
(UCN) in an environment of high neutron flux (about 3?1012n/cm2/s) but still acceptable radiation heat
release (about 4?10?3W/g). It can be realized within the thermal column situated close to the reactor
core. With its large diameter of 1m, this channel allows to install a 15-cm-thick bismuth shielding, a
graphite premoderator (300dm3at 20K), and a superfluid helium converter (35dm3). At a temperature
of 1.2K it is possible to remove the heat release power of about 20W. Using 4p flux of cold neutrons
within the reactor column can bring more than a factor 100 of cold neutron flux incident on the
superfluid helium with respect to the present cold neutron beam conditions at the ILL reactor. The
storage lifetime for UCN in superfluid He at 1.2K is about 30s, which is sufficient when feeding
experiments requiring a similar filling time. The calculated density of UCN with energy between 50 and
230neV in an experimental volume of 40l is about 104n/cm3. Technical solutions for realization of the
project are discussed.
& 2009 Elsevier B.V. All rights reserved.
1. Introduction: Evolution of UCN sources and prospects for
future developments
Ultracold neutrons (UCN) are successfully used in fundamental
research: for the search of the neutron electric dipole moment,
neutron lifetime measurements, measurements of neutron decay
asymmetries, and other studies. The accuracy one may reach in
these experiments is severely limited by counting statistics.
Therefore, there is a strong activity in the development of more
intense UCN sources (Fig. 1).
After the first demonstration of UCN production in Dubna [1]
and in Munich [2] the UCN density has been increased by eight
orders of magnitude using more powerful reactors and cold
neutron sources in Gatchina (PNPI) [3] and in Grenoble (ILL) [4].
These sources were placed in extremely high neutron fluxes and
we employ liquid hydrogen (PNPI) or liquid deuterium (ILL). In the
1990s this line of UCN source development came to saturation.
Promising developments of alternative UCN sources employ
superfluid He 4 at a temperature of about 1K or below, and solid
deuterium at about 4K.
A first test experiment with solid deuterium has been
carried out in Gatchina (PNPI) in 1980 [5]. More detailed studies
of solid deuterium UCN source were realized in Gatchina at the
WWR-M reactor in 1994 [6]. The solid deuterium UCN source with
volume 6l was installed in the thermal column of the reactor. It
was demonstrated that solid deuterium gives a gain factor of
about 10 with respect to liquid deuterium. Unfortunately, it
requires an environment of lower heat release density. A pulse
mode is suitable when the average heat release is low enough,
which can be realized at neutron spallation sources. Therefore, the
project of the so-named UCN factory was proposed in Ref. [7] and
presented in the first UCN Workshop devoted to this project in
Pushkin, Russia, in 1998. The first and simplest realization of this
idea was done in LANL [8]. A more detailed project of UCN factory
was developed by PNPI and PSI [10], which is now under
construction at PSI.
UCN production in superfluid He 4 was proposed in Ref. [11].
After the first demonstration of the method at the ILL in 1980 [12],
it has been further investigated in Japan in 1992 [14], and in 2002
again at the ILL [13]. Now there are a few projects, one based on
using superfluid He with UCN accumulation in the converter
ARTICLE IN PRESS
Contents lists available at ScienceDirect
journal homepage: www.elsevier.com/locate/nima
Nuclear Instruments and Methods in
Physics Research A
0168-9002/$-see front matter & 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.nima.2009.07.078
?Corresponding author. Tel.: +78137146001; fax: +78137130072.
E-mail address: serebrov@pnpi.spb.ru (A.P. Serebrov).
Nuclear Instruments and Methods in Physics Research A 611 (2009) 276–279

Page 2

ARTICLE IN PRESS
(ILL/Munich [15]), or in the regime of steady-state operation
(Japan [16], Gatchina [17]).
2. Scheme of UCN source implementation in the thermal
column of the WWR-M reactor
In this article we discuss the project of an UCN source at the
WWR-M reactor of PNPI [14], for which superfluid helium shall be
used as a converter for UCN. The thermal column of diameter 1m,
situated close to the reactor core (see Fig. 2), offers a unique
opportunity to install a source for UCN in an environment of high
neutron flux (about 3?1012n/cm2/s) but still acceptable radiation
heat release (about 4?10?3W/g). In order to reduce the heat
release from g-rays from the reactor core, a bismuth shielding will
be installed. The external diameter of this cylindrical shielding
will be 990mm with a wall thickness of 95mm. The closing of the
cylinder in the direction of the reactor core will have a thickness
of 150mm. The bismuth is enclosed in a water-cooled aluminum
shell (5mm thick). Inside the bismuth shielding a graphite
moderator with thickness 150mm will be installed. It will be
cooled down to 20K by means of a helium refrigerator. This
graphite is also enclosed in an aluminum shell and this
construction is placed in a vacuum jacket. Inside the graphite
moderator a cylindrical vessel with superfluid helium at a
temperature of 1.2K is placed. Its diameter is 300mm, and
length 500mm. The thickness of the aluminum walls is 2mm. The
internal surfaces of the Al walls are coated with (3–5)?103˚A
58NiMo alloy with critical velocity 7.8m/s. UCN can be extracted
from the source by means of an UCN guide coated by58NiMo too.
UCN guide and source volume will be separated by a 100-mm-
thick Al membrane with support grid. The scheme in Fig. 2 is
shown in scale. The thicknesses of Bi shielding and graphite
premoderator have been chosen to obtain the maximum possible
neutron flux with wavelength 9˚A, at the condition that the heat
load on the superfluid source will not exceed 20W.
Calculations of neutron fluxes and heat release were done
using the MCNP code. The total heat releases at 16MW reactor
power are: 16kW in the bismuth shielding, 750W in the graphite
moderator, 13W in the aluminum shell of the helium source, and
6W in the superfluid helium. Hence the total heat load at
temperature 1.2K is 19W.
3. Premoderator
The choice of graphite as a cold moderator is a compromise
between simplicity and the goal to obtain the maximum possible
flux of cold neutrons with wavelength 9˚A. These neutrons can be
converted into UCN by means of a one-phonon process. UCN
production due to multi-phonon processes can be similarly strong
but it depends on the neutron spectrum [18]. Fig. 3 shows the
results of calculations of cold neutron spectrum for the following
materials: liquid deuterium, graphite, frozen heavy water, and
beryllium. For all cases the temperature of the moderator was
chosen to be 20K. From a practical point of view the graphite
moderator is the simplest one However, it is inferior to liquid
deuterium in the production of 9˚A neutrons by a factor of two.
4. Cryogenic scheme of source
Now we would like to discuss the practical realization of the
proposed scheme (Fig. 4). The method of heat removal is based on
thehuge thermalconductivity
superfluid component moves to the source of heat and the
normal component moves to the heat exchanger; so it is not
necessary to specially arrange the circulation of liquid helium.
of superfluid helium.The
Fig. 2. Implementation scheme of the planned UCN source in the thermal column of the WWR-M reactor (see text for description of the components).
1965
10-7
10-6
10-5
10-4
10-3
10-2
10-1
100
101
102
103
104
105
[5]
[15-16]
present PNPI project [17]
first test
experiment
with SD2
PNPI
SD2 Mainz [9]
projects
ILL [13]
ILL [12]
first test experiments
with superfluid He
test experiment
project
SD2 in pulse mode
SD2 pulse mode
LANL-PNPI [8]
SD2 reactor
PSI-PNPI [10]
PNPI [6]
SRIAR
PNPI
ILL [4]
ILL
IAE
TUM
IAE
JINR
PNPI
PNPI
PNPI
PNPI
PNPI [3]
UCN density, cm-3
years
1970 1975 1980 1985 1990 1995 2000 2005 2010 2015
Fig. 1. Development of UCN sources and future prospects. More detailed
information can be found in Refs. [3–6,8–10,12,13,15–17].
A.P. Serebrov et al. / Nuclear Instruments and Methods in Physics Research A 611 (2009) 276–279
277

Page 3

ARTICLE IN PRESS
Superfluid helium around the UCN guide is used as a guide of heat
release. Its cross-section is 630cm2, and length is about 3m. The
difference of temperature along superfluid helium will be
1.2?10?2K at a heat load of 20W. Taking into account a jump
of temperature on the surface of the heat exchanger, we estimate
that the temperature difference between the helium in the source
and in the pumping bath will be 2?10?2K. The temperature in
the pumping bath is kept at 1.2K, corresponding to a pressure of
0.55mbar. A heat load of 20W corresponds to a consumption of
liquid helium of about 38l/h. For the pumping of gaseous helium,
two vacuum pumps with pumping rate 8m3/s will be required. It
corresponds to pumping 7.5l/s of helium gas at atmospheric
pressure. These estimations show that the proposed project can
be realized.
5. Estimation of UCN density
MCNP calculations gave the following results for the full reactor
power of 16MW. The neutron flux incident on superfluid helium is
3.2?1012n/cm2/s, and the differential neutron flux around 9˚A is
3.2?1010n/cm2/s/˚A. The rate R of UCN production due to the one-
phonon process can be calculated by the theoretical formula [13]
R ¼ 4.55?10?8dF/dl(9˚A)/cm3/s1. Taking into account the multi-
phonon processes and neutron spectrum in the source [18], we expect
the rate of UCN production will be 2.9?103n/cm3/s. The total
number of UCN produced in the source is about 1.0?108n/s. The UCN
lifetime in superfluid helium at a temperature of 1.2K is about 30s,
but after taking into account UCN losses on the trap walls, it is about
20s. This estimation was done using a rather pessimistic assumption
of 10?3losses per collision. Thus the UCN density in a closed trap with
superfluid helium could reach 5.8?104n/cm3. However, we are
interested in the UCN density in an experimental trap placed in the
experimental hall. For this purpose Monte-Carlo calculations have
been done for an experimental scheme, that includes a trap with
superfluid helium with a volume of 35l, an UCN guide with diameter
140mm and length 3m, and an experimental trap with a volume of
35l. In addition, a scheme with an experimental trap of 350l has been
calculated too. The volume with superfluid helium is separated from
the UCN guide by an aluminum membrane with thickness 100mm on
a supporting grid. The vacuum of the guide for cold neutrons is also
isolated from the warm UCN trap by means of a similar membrane to
avoid freezing of residual gas on the cold guide and the source
window. The calculations take into account the losses in the walls of
the UCN guide and the experimental traps, which were chosen as
3?10?4per collision. A mirror reflectivity of 99.3% for the guides was
used, which was obtained from measurements of the transmission
factor for replica guides [19]. The critical velocity of the helium trap
and the UCN guide was 7.8m/s (58NiMo), but for the experimental
trap it was chosen to be 6.8m/s (Be, BeO). Losses in the aluminum
membranes were taken as defined by the capture cross-section. The
coefficient of UCN reflection from the aluminum membranes was
calculated taking into account the critical velocity 3.2m/s and albedo
reflection due to the inhomogeneity of density inside the foils [20]. It
is well known that reflection plays an important role and can strongly
reduce the transmission coefficient. The calculations show that the
UCN density in the 35l experimental trap is less by 4.5 times than the
hypothetic UCN density in the closed helium trap, i.e. rtrap
35l¼ 1.3?104n/cm3. The time of filling of the trap in this scheme
is 22s. For the experimental trap with a volume of 350l the UCN
densityrtrap 350l¼ 7.7?103n/cm3. The filling time constantof the big
trap is 55s. In principle, this UCN density can be increased by a factor
of 3 by changing the configuration of the reactor active core but the
heat load will be increased correspondingly. Certainly, obtaining an
UCN density of order 104n/cm3is very important for fundamental
physics experiments with UCN.
6. Plan of realization and conclusion
Presently the first steps for the realization of this project have
been done. A refrigerator with a cooling power of 3kW at 20K to
cool the graphite moderator was put into operation in the
cryogenic laboratory. In 2009 we have to put into operation a
helium liquefier, which has already arrived at PNPI. The time table
of the project strongly depends on the program of the thermal
0
107
108
109
1010
1011
Φ (λ), arb. units
C
Be
LD2
D2O
λ, Å
246810121416
Fig. 3. Neutron spectrum incident on the helium converter, for different
premoderators.
LHe supply
pumping of He (P = 0.6 mbar)
superfluid filter
membrane
membrane
T0
T1
T2
heat exchanger
Q
D
d
reflector of UCN
Fig. 4. Cryogenic scheme of the UCN source.
A.P. Serebrov et al. / Nuclear Instruments and Methods in Physics Research A 611 (2009) 276–279
278

Page 4

ARTICLE IN PRESS
column modernization, which will require 3–5 years. Thus an
optimistic time estimate of source launching is 2012.
This neutron source will produce cold and very cold neutrons
as well. The general scheme of the neutron guide halls is
presented in Fig. 5. The realization of this project will create an
UCN source with record UCN density for fundamental studies. In
particular the accuracy of the neutron EDM measurement can be
increased by more than an order of magnitude. Besides, cold and
very cold neutrons can be used for studies of nanostructures and
other research in the field of condensed matter.
Acknowledgement
This work has been carried out with the support of RFBR Grant
07-02-00859.
References
[1] F.L. Shapiro, et al., JETP Lett. 9 (1969) 23 (Preprint JINR P3-4127).
[2] A. Steyerl, Phys. Lett. B 29 (1969) 33.
[3] I.S. Altarev, et al., JETP Lett. 44 (1986) 269.
[4] A. Steyerl, et al., Phys. Lett. A 116 (1986) 347.
[5] I.S. Altarev, et al., Phys. Lett. A 80 (1980) 413.
[6] A.P. Serebrov, et al., JETP Lett. 59 (1994) 757;
A.P. Serebrov, et al., JETP Lett. 62 (1995) 785;
A.P. Serebrov, et al., Nucl. Instr. and Meth. A 440 (2000) 658.
[7] A.P. Serebrov, et al., J. Phys. Soc. Jpn. 65 (Suppl. A) (1996) 259;
A.P. Serebrov, et al., JETP Lett. 66 (1997) 802;
A.P. Serebrov, Nucl. Instr. and Meth. A 440 (2000) 653.
[8] R.E. Hill, et al., Nucl. Instr. and Meth. A 440 (2000) 674;
A. Saunders, et al., Phys. Lett. B 593 (2004) 55.
[9] W. Heil, in: Proceedings of the sixth UCN Workshop on Ultra Cold and Cold
Neutrons. Physics and Sources, St. Petersburg, Moscow, Russia, 1–7 July 2007,
/http://cns.pnpi.spb.ru/ucn/articles/Heil1.pdfS.
[10] A. Fomin, et al., PSI Report TM-14-01-01 2000.
[11] R. Golub, J.M. Pendlebury, Phys. Lett. A 62 (1977) 337.
[12] R. Golub, et al., Z. Phys. B—Conden. Matter 51 (1983) 187.
[13] C.A. Baker, et al., Phys. Lett. A 308 (2003) 67.
[14] H. Yoshiki, et al., Phys. Rev. Lett. 68 (1992) 1323.
[15] O. Zimmer, et al., Phys. Rev. Lett. 99 (2007) 104801.
[16] Y. Masuda, et al., Phys. Rev. Lett. 89 (2002) 284801.
[17] A. Serebrov, in: Proceedings of the Sixth UCN Workshop on Ultra Cold and
Cold Neutrons. Physics and Sources, St. Petersburg, Moscow, Russia, 1–7 July
2007,
/http://cns.pnpi.spb.ru/ucn/articles/Serebrov1.pdfS;
M.S. Onegin, A.P. Serebrov, Preprint PNPI 2776, 2008.
[18] E. Korobkina, et al., Phys. Lett. A 301 (2002) 462.
[19] A. Serebrov, et al., Ultracold-neutron infrastructure for the PNPI/ILL neutron
EDM experiment, Nucl. Instr. and Meth., these proceedings, doi:10.1016/j.
nima.2009.07.084.
[20] A. Serebrov, et al., Phys. Lett. A 309 (2003) 218.
A.N.Erykalov,
Fig. 5. Neutron guide halls and beam positions around the new source of cold and ultracold neutrons at the WWR-M reactor.
A.P. Serebrov et al. / Nuclear Instruments and Methods in Physics Research A 611 (2009) 276–279
279