All opinions expressed herein are those of the authors and should not be reproduced, quoted in publications, or
used as a reference without the author’s consent.
An international journal of news from the stellarator community
Editor: James A. Rome Issue 167 October 2019
E-Mail: Phone: +1 (865) 482-5643
On the Web at
Enhancement of Energetic-ion
Confinement Study Using
Comprehensive Neutron
Diagnostics in LHD
In a fusion reactor, a burning plasma is sustained by alpha
particle heating [1]. Therefore, energetic ion confinement
has been intensively studied to predict the confinement of
alpha particles in a burning plasma. In the Large Helical
Device (LHD), energetic particle confinement has been
studied using neutral beam (NB) injection [2–4]. In a
hydrogen plasma regime, study of energetic ion confinement
has been mainly studied primarily by measuring
charge-exchange of energetic particles using a neutral particle
analyzer (NPA), as well as escaping energetic particles
using fast ion loss detection (FILD) [5–7]. Studies of
magnetohydrodynamic (MHD) mode-induced beam ion
transport/loss have also been reported [8–14]. In LHD,
operation with deuterium gas was initiated in March 2017
[15]. One of the goals of the deuterium plasma campaign
in LHD was the expansion of energetic particle studies. A
predictive study based on numerical simulation using the
steady-state Fokker-Planck model showed that the beamthermal
neutrons should be dominant [15]. Therefore, by
means of neutron diagnostics, we can obtain information
regarding energetic particles confined inside the plasma.
Comprehensive Neutron Diagnostics Installed
in LHD
Neutron diagnostics such as the ex-vessel neutron flux
monitor (NFM) [16–19] in order to measure total neutron
emission rate Sn, the neutron activation system (NAS) [20]
to perform shot-integrated measurement of DD and secondary
DT neutron yields, the vertical neutron camera
(VNC) [21–24] to obtain the radial profile of neutron
emission, and scintillating-fiber (Sci-Fi) detectors [25–26]
to conduct time-resolved measurements of DT neutron
yield have been developed and installed to enhance energetic
particle study in LHD, in addition to radiation safety
management. The commissioning of those diagnostics is
reported in Refs. 27–31.
Global Confinement of Beam Ions
Study of Sn on NB-heated plasmas has been performed for
various magnetic axis positions (Rax) in LHD. Beam-thermal
and beam-beam reaction fractions were evaluated in
the case of balanced NB injection. Time-resolved analysis
of thermal, beam-thermal, and beam-beam fractions was
performed by the TASK/FP code [32, 33]. It is confirmed
that the beam-thermal component is dominant, as pre-
In this issue . . .
Enhancement of Energetic-ion Confinement
Study Using Comprehensive Neutron Diagnostics
in LHD
The neutrons produced by fusion in a deuterium
plasma in the Large Helical Device (LHD) have
offered new ways to study fast ion confinement.
Global confinement, fusion gain, the beam ion confinement
in an MHD quiescent plasma, radial profile of
beam ions, and demonstration of MeV ion confinement
can be studied using comprehensive neutron
diagnostics such as the neutron flux monitor, the neutron
activation system, the vertical neutron camera,
and the scintillating-fiber detector. The result of the triton
burnup study conducted in stellarator/heliotron for
the first time was selected as a research highlight in
the July 2019 issue of Nature Physics. [50] ............. 1
Plans for EPOS: A Tabletop-sized, Superconducting,
Optimized Stellarator For Matter/antimatter
Pair Plasmas
The EPOS (Electrons and Positrons in an Optimized
Stellarator) project is a six-year program that will
launch at the end of this year. It aims to combine
results from several rapidly advancing fields –– superconductor
technology; 3D printing of advanced materials;
and stellarator optimization –– in order to produce
a high-magnetic-field (>2 T), tabletop-size (between
10 and 50 liters) device for fundamental plasma physics
and “laboratory astrophysics” investigations of
electron/positron plasmas. ....................................... 5
Stellarator News -2- October 2019
dicted by Ref. 15. The beam-beam component accounts
for approximately 1/5 of Sn. The relatively high beambeam
fraction is due to balanced NB injection. Sn dependence
on electron density was surveyed. Figure 1 shows
the dependence of the maximum Sn on ne_avg with Rax of
3.55 m to 3.90 m. As the Fokker-Planck model predicted
[15], the maximum Sn has a peak at ne_avg of around 2.5 ×
1019 m3. Maximum Sn decreases with outward shift of
Rax at the same density. To understand Sn dependence on
Rax and ne_avg, the beam-thermal neutron emission rate
was calculated by using the FIT3D-DD code [34]. The
dependence of Sn on ne_avg is successfully reproduced
though Sn was overestimated by almost two times in
FIT3D-DD, but the shape of the curve is in rough agreement
to experiment.
Fusion Gain
The equivalent fusion gain in DT plasma QDT has been
studied in N-NB heated plasmas. The ratio of fusion gain
in DD plasma QDD on equivalent QDT were evaluated
using the FBURN code [35]. Maximum QDT/QDD of 249
is obtained in the case of deuterium NB injection into a triton
plasma. The maximum equivalent QDT reached 0.11
(Fig. 2(a)), which is comparable with that in large tokamaks
with 5 MW NB injection [36–38]. It is found that the
equivalent QDT linearly increases with Te0
1.5 [Fig. 2(b)].
When beam-thermal neutrons are dominant, equivalent
QDT can be expressed as equivalent QDT ~ Sn/PNB ~ ni ×
PNB × s/PNB ~ ni × Te
1. 5/ne ~ Te
1.5. The plot shown in
Fig. 2(b) is consistent with fusion reaction origin.
NB Blip Experiments
Blip experiments with total neutron emission measurements
were performed to study beam ion confinement in
an MHD-quiescent plasma heated by positive-ion-sourcebased
NB (P-NB) injection (two beamlines with beam
energies of 60-80 keV).
A short-pulse P-NB having a pulse width of 20 ms is
injected into the electron-cyclotron-heated plasma [Fig.
3(a)]. A rapid increase in Sn is observed due to P-NB
injection. The characteristics of slowing down and transport
of beam ions appear in the Sn decay state after the PNB
is turned off. The decay time of Sn obtained in experimentally
is 215 ms. To understand beam ion transport and
loss, Global NEoclassical Transport (GNET) simulation
[39] and the FBURN simulation without radial diffusion
have been performed (Fig. 3(b)). We obtained almost the
same rise time for Sn in the experiment, FBURN calculation,
and GNET calculation. We also have obtained a relatively
short decay time for Sn with GNET (216 ms)
Fig. 1. Electron density dependence of total neutron emission rate Sn for Rax from 3.55 m to 3.90 m.
Fig. 2. Dependence of equivalent QDT on (a) NB deposition
power and (b) Te0
1.5. Equivalent QDT reached 0.11.
Fig. 3. Time evolution of Sn in P-NB blip (a) in experiment
and (b) calculated by GNET and FBURN.
Stellarator News -3- October 2019
compared with that calculated by FBURN (250 ms). It is
found that GNET can reproduce the time trace of Snn.
Hence, we can describe the transport of NB blip beam ions
in the low- MHD quiescent plasma using neoclassical
Radial Profile of Neutron Emission
Neutron emission profiles in plasmas heated with negative-
ion-source-based NBs (N-NB) (3 beamlines with
beam energy 180 keV) were measured using a vertical
neutron camera (VNC) in various Rax conditions. The time
trace of neutron counts shows that the counts are higher in
the central channel than in the edge channel. Line-integrated
neutron emission profiles at t of 3.7 s to 4.3 s in
three Rax configurations are plotted in Fig. 4. Due to the
increase in Sn with the inward shift of Rax, as shown in
Fig. 1, neutron counts of VNC increase accordingly. The
position of peak neutron counts is shifted according to Rax.
We also have calculated the line-integrated neutron profile
using numerical simulation with the MORH code [40]
based on HINT equilibrium [41] with an effective charge
of 1 assumed. Figure 4 also shows the line-integrated neutron
profiles in three different Rax configurations calculated
by MORH. Although we have obtained two times
higher neutron counts than that obtained in the experiment,
the simulation reproduces the shift in neutron count
peak according to Rax position.
Triton Burnup Experiment
A triton burnup experiment has been conducted for the
first time in stellarator/heliotron devices. The study of 1-
MeV triton burnup can be regarded as a proxy for DT born
alpha particle confinement. We installed Sci-Fi detectors
in order to measure the secondary DT neutron emission.
We used the triton burnup ratio, defined as total DT neutron
emission yield divided by total DD neutron emission
yield, in order to show the MeV confinement capability
[42–49]. Typical time traces from the triton burnup experiment
show that there are two decay components in Sn
after the NB is turned off. The fast component and the
slow component correspond to DD and DT neutrons,
respectively. The time trace of the relatively slow component
matches the DT neutron rate measured by a Sci-Fi
detector. The dependence of the triton burnup ratio on Rax
is plotted in Fig. 5. The result shows that we obtained a
higher triton burnup ratio in the small Rax case than in the
large Rax case. We achieved a triton burnup ratio of
0.45%. The LHD record is a similar value to that obtained
in tokamaks such as ASDEX Upgrade and KSTAR [48,
49] whose minor radius is comparable to that of LHD. Figure
5 shows the triton burnup ratio evaluated using GNET.
The GNET result also shows the larger triton burnup ratio
in the the inward-shifted configuration. The lower triton
burnup ratio in GNET may be due to re-entering effects
because in GNET, the triton orbit is calculated in Boozer
Energetic particle confinement research in stellarator/
heliotron devices is benefiting from the deuterium LHD
plasma experiments. The dependence of Sn on electron
density shows the same trend as predicted by numerical
simulations. The beam-beam components account for 20%
of Sn in N-NB-heated plasmas as shown by TASK/FP
code. We have achieved equivalent QDT of 0.11, which is
almost the same value obtained in large tokamaks with the
same NB heating power. A short NB injection experiment
into an MHD-quiescent plasma shows that the time evolution
of Sn can be described by neoclassical models. Neutron
emission profile measurement and triton burnup
experiments were performed for the first time in stellarators
and helical devices. The peak of the line-integrated
neutron profile measured by VNC shifted according to
Rax. The triton burnup ratio increases with the inward shift
of Rax, as expected on the basis of numerical simulation,
and reaches 0.45%.
Fig. 4. Line-integrated neutron emission profile obtained
for Rax of 3.60 m, 3.75 m, and 3.90 m by experiment and
MORH simulation.
Fig. 5. Triton burnup ratio dependence on Rax obtained in
experiment and GNET simulation.
Stellarator News -4- October 2019
[1] Fasoli, A. et al., 2007 Nucl. Fusion 47 S264.
[2] Osakabe M. et al., 2010 Fusion Sci. Technol. 58 131.
[3] Toi, K. et al., 2010 Fusion Sci. Technol. 58 186.
[4] Toi, K. et al., 2011 Plasma Phys. Controlled Fusion 53
[5] Isobe, M. et al., 2010 Fusion Sci. Technol. 58 426.
[6] Ogawa, K. et al., 2008 Plasma Fusion Res. 3 S1082.
[7] Ogawa, K. et al., 2009 Journal of Plasma Fusion Res.
SERIES 8 655.
[8] Osakabe, M. et al., 2006 Nucl. Fusion 46 S911.
[9] Ogawa, K. et al., 2010 Nucl. Fusion 50 084005.
[10] Isobe, M. et al., 2010 Contrib. Plasma Phys. 50 540.
[11] Ogawa, K. et al., 2012 Plasma Science and Technology
14 269.
[12] Ogawa, K. et al., 2012 Nucl. Fusion 52 094013.
[13] Ogawa, K. et al., 2013 Nucl. Fusion 53 053012.
[14] Ogawa ,K. et al., 2014 Plasma Phys. Controlled Fusion
56 094005.
[15] Osakabe, M. et al., 2017 Fusion Sci. Technol. 72 199.
[16] Isobe, M. et al., 2014 Rev. Sci. Instrum. 85 11E114.
[17] Nakano, Y. et al., 2014 Rev. Sci. Instrum. 85 11E116.
[18] Nakano, Y. et al., 2014 Plasma Fusion Res. 9 3405141.
[19] Nishitani, T. et al., 2017 Fusion Eng. Des. 123 1020.
[20] Pu, N. et al., 2017 Rev. Sci. Instrum. 88 113302.
[21] Ogawa, K. et al., 2014 Rev. Sci. Instrum. 85 11E110.
[22] Uchida, Y. et al., 2014 Rev. Sci. Instrum. 85 11E118.
[23] Uchida, Y. et al., 2017 Rev. Sci. Instrum. 88 083504.
[24] Ogawa, K. et al., 2018 Rev. Sci. Instrum. 89 095010.
[25] Takada, E. et al., 2016 Plasma Fusion Res. 11 2405020.
[26] Takada, E. et al., 2019 Rev. Sci. Instrum. 90 043503.
[27] Isobe, M. et al., 2018 IEEE Transactions on Plasma Science
46 2050.
[28] Isobe, M. et al., 2018 Nuclear Fusion 58 082004.
[29] Ogawa, K et al., 2017 Plasma Fusion Res. 12 1202036.
[30] Ogawa, K et al., 2018 Nucl.r Fusion 58 034002.
[31] Ogawa, K. et al., 2018 Nucl. Fusion 58 044001.
[32] Nuga, H. and Fukuyama, A. 2011 Prog. in Nucl. Sci.
Technol. 2 78
[33] Nuga, H. et al., 2019 Nucl. Fusion 59 016007.
[34] Murakami S. et al., 1995 Trans. Fusion Tech. 27 256.
[35] Ogawa, K. et al., 2018 Plasma Phys. Controlled Fusion
60 9.
[36] Jassby, D. L. et al., 1991 Phys. Fluids B: Plasma Phys.
3 2308.
[37] Keilhacker, M. and the JET Team, 1990 Phys. Fluids B:
Plasma Phys. 2 1291.
[38] Nishitani, T. et al., 1994 Nucl. Fusion 34 1069.
[39] Murakami, S. et al., 2002 Nucl. Fusion 42 L19.
[40] Seki, R. et al 2015 Plasma Fusion Res. 10 1402077.
[41] Suzuki, Y. et al., 2003 Plasma Phys. Controlled Fusion
45 971.
[42] Barnes, C. W. et al., 1998 Nucl. Fusion 38 597.
[43] Källne, J. et al., 1988 Nucl. Fusion 28 1291.
[44] Hoek, M. et al., 1996 Nucl. Instrum. Methods Phys.
Res. A 368 804.
[45] Heidbrink, W. W., Chrien, R. E., and Strachan, J. D.,
1983 Nucl. Fusion 23 917.
[46] Duong, H. and Heidbrink, W. W., 1993 Nucl. Fusion 33
[47] Batistoni, P. et al., 1987 Nucl. Fusion 27 1040.
[48] Hoek, M., Bosch, H. S. and Ullrich, W., 1999 “Triton
burnup measurement at ASDEX Upgrade by neutron
foil activation,” IPP-Report IPP-1/320.
[49] Jo, J. et al., 2016 Rev. Sci. Instrum. 87 11D828.
[50] Reichert, S., Nature Phys. 15, 622 (2019);
Ogawa, K., et al., 2019 Nucl. Fusion 59 076017.
Mitsutaka Isobe and Kunihiro Ogawa
National Intitute for Fusion Science
Toki, Japan
Stellarator News -5- October 2019
Plans for EPOS: A Tabletopsized,
Superconducting, Optimized
Stellarator For Matter/
antimatter Pair Plasmas
Why make a plasma out of electrons and positrons?
The large mass asymmetry between positive and negative
species is a cornerstone of the physics of traditional electron/
ion plasmas. It is not surprising, then, that quasineutral
plasmas comprising positively and negatively charged
particles of identical mass (“pair plasmas”) are expected to
exhibit a number of properties qualitatively different from
those of electron/ion plasmas. Without the inherent separation
of time scales and length scales that arises from the
large mass ratio, many standard plasma physics phenomena
vanish — e.g., sheaths, Faraday rotation, whistler
waves, and lower hybrid waves. To consider pair plasmas
necessitates rederiving plasma physics from the ground
up, a project that theory and simulation first began to
tackle more than 40 years ago [1] and that to date has
resulted in hundreds of predictions. Among these, for
example, are changes to the physics of reconnection, soliton
solutions, and turbulence; in certain regimes and magnetic
configurations, pair plasmas are expected to exhibit
“remarkable stability” to anomalous transport modes that
tend to dominate electron/ion plasmas [2].
Studying plasmas with reduced mass ratios is a way to
improve our understanding of “normal” plasmas, e.g., in
such aspects such as reconnection and tokamak heat fluxes
[3]. Testing predictions of how plasma physics changes in
the limit of a mass ratio of unity will thus confirm our
understanding of fundamental aspects of plasma phenomena,
validating our ability to simulate plasmas in various
regimes. This study will also improve our knowledge of
our universe, in which pair plasmas dominated for 1–10
seconds after the Big Bang (a period known as the Lepton
Era), and in which electron-positron plasmas are still
being generated (e.g., in the neighborhood of pulsars and
active galactic nuclei). Hence, there is a compelling need
for experiments to complement the ever-growing body of
theoretical/computational literature about pair plasmas,
which can be thought of as the “hydrogen atom of plasma
physics” (i.e., a comparatively simple system that nevertheless
informs understanding of its far more complex
Electron/positron plasmas, which can be magnetized easily,
are of particular interest, and there has recently been
significant progress in experimental methods to study
them in the laboratory—a significant experimental challenge
to which there are several approaches that span a
range of target densities, temperatures, and confinement
times. Among these is the approach (Fig. 1) of the APEX
(A Positron Electron eXperiment) collaboration [4], which
is bringing together people, expertise, and experimental
hardware from the Max Planck Institute for Plasma Physics
(IPP, where the project is based) and universities in
Germany, the United States, and Japan. APEX has been
has making strides toward its goal of magnetically confining
a cold electron/positron pair plasma in a dipole magnetic
field, recently demonstrating proof-of-principle
results in the areas of lossless injection across flux surfaces
and subsequent long confinement of injected positrons
[5]. In parallel with the next steps of development of
existing programs (positron beam optimization, positron
pulse accumulation, and the construction of the levitated
dipole trap), a new branch of the collaboration’s experimental
program will be launched at the end of this year,
involving the design, construction, and operation of a second
magnetic confinement device for the pair plasma: a
As shown in Fig. 1, the final configuration for the APEX
collaboration will bring together a world-class positron
beam (from the NEPOMUC NEutron-induced POsitron
source MUniCh); a series of Penning-Malmberg-type
traps to collect the steady-state NEPOMUC beam into
pulses, then combine these pulses as non-neutral positron
plasmas in ever greater numbers; and two complementary
toroidal traps (APEX and EPOS) in which the positrons
will be confined with electrons to make the pair plasmas.
How to make a plasma out of electrons and
In order to create an electron/positron pair plasma in the
laboratory, one first needs access to sufficient amounts of
antimatter (which, needless to say, is notoriously hard to
come by). Specifically, for a plasma with a temperature of
0.1–5 eV in a volume of 5–50 liters (a tabletop device),
between 109 and 1011 positrons are required to achieve 10
Debye lengths—i.e., the classic threshold above which a
collection of charged particles constitutes a “plasma,” as
defined by the relevance of collective forces in comparison
to individual particles’ kinetic energies [4, 9].
These conditions can be reached using a world-class e+
beam that can produce up to 109 e+/s [6], plus a series of
non-neutral plasma traps to accumulate the steady-state
beam into cool, dense, tailorable pulses [7]. Finally, the
positrons need to be combined with a comparable number
of electrons; this is best done in a magnetic trap that provides
excellent confinement for both positive and negative
species, either for fully quasineutral plasmas or for nonneutral
plasmas, without requiring plasma current (which
these extremely low-density plasmas would not to be able
to generate). Two obvious choices are a levitated dipole
trap and an optimized stellarator. Although they have in
Stellarator News -6- October 2019
common that they both meet the requirements described
above, the plasma physics in these two configurations
tends to be very different, due to the disparate magnetic
field topologies (Table 1); they have complementary technical
strengths and weaknesses, as well. Comparing the
behavior of pair plasmas in the APEX dipole (with
axisymmetry and strong flux expansion) to those with
approximately the same parameters in the EPOS stellarator
(with magnetic shear and long connection lengths) can
be expected to deeply enrich the findings of both.
Naturally, one of the first questions that comes up when
discussing a matter/antimatter pair plasma is, “Won't it
just annihilate?” There are indeed several different channels
through which this will occur, and each has a wellknown
cross section that can be used to determine how
long the plasma would persist if that were the limiting factor
[10]. For pair plasmas in the target density and tem-
Fig. 1. The final configuration for the APEX collaboration will bring together a world-class positron beam (from the NEutroninduced
POsitron source MUniCh); a series of Penning-Malmberg-type traps to collect the steady-state NEPOMUC beam
into pulses, then combine these pulses as non-neutral positron plasmas in ever-greater numbers; and two complementary
toroidal traps (APEX and EPOS) in which the positrons will be confined with electrons to make the pair plasmas.
(*Note: Since EPOS hasn't been designed yet, this is a placeholder image, borrowed from Ref. [8].)
Table 1. A stellarator and a levitated dipole have very different and complementary physics, as well as different technical
advantages and disadvantages.
Stellarator Levitated dipole
Steady state Steady state on typical plasma time scales
Fusion relevance Astrophysical relevance
Negligible flux expansion Strong flux expansion
Irrational as well as rational flux surfaces Each B-field line closes after one pass
Long B-field connection lengths Short B-field connection lengths
parallel force balance counteracts instabilities Parallel force balance does not counteract instabilities
drift orbit confinement requires optimization Drift orbits always confined (due to axisymmetry)
Passing particles most easily confined Passing particles most sensitive
to departures from axisymmetry
Indefinite operation with permanently attached electrical
and thermal contacts
Inductive charging, cooling/warming cycles,
repeated making/breaking of thermal contacts
Many 3D coils 2-3 planar coils (floating, lifting, charging)
Requires positron pulses for fueling Possibility for steady-state fueling, using inward transport
processes unique to dipoles
Stellarator News -7- October 2019
perature ranges, these all come out to lifetimes longer than
a day. Although annihilation is not therefore expected to
limit plasma lifetime, it will still provide a tool for diagnosing
the plasma (especially when enhanced by gas or
pellet injection) [5].
Preliminary design/optimization considerations
The design and optimization for EPOS will start in earnest
in a couple of months, but it is abundantly clear from the
outset that it will be very different from that for a fusion
 Small device size (between 5 and 50 liters) but strong
magnetic fields (>2 T, so as to be able to take advantage
of cyclotron cooling, a valuable tool in non-neutral
plasma physics experiments).
 Low temperatures (0.1–5 eV) and extremely low densities
(1011–1013 m3).
 Debye length > Larmor radius.
 Very far from the regime of MHD.
 No ions (except when deliberately introduced).
 No pressure, divertor, heat loads, or neutrons.
 Significant flexibility with respect to the design of the
coils/current sheets.
Nevertheless, the EPOS experiment will be provide a
meaningful test bed for next-generation stellarator design
tools that are also used for fusion, in particular with an eye
toward maximizing construction tolerances [11]—something
that is especially important for a small device—and
minimizing neoclassical transport. Because low-density
pair plasmas are predicted to be turbulence-free in wellchosen
geometries [2,12], and because quasi-neutral pair
plasmas will have zero ambipolar field due to mass symmetry,
thereby eliminating the effect of “orbit healing” due
to radial electric fields, EPOS can explore the upper
bounds of what neoclassical optimization can provide—-
albeit in a system with different types of collisions than a
fusion plasma. Positrons also provide an extremely sensitive
probe for where and when charged particles are lost.
From the engineering side, high-temperature superconductors
(HTSCs) are an attractive option for their ability to
provide higher fields while relaxing the cooling requirements,
but it must first be assessed whether they would
demand overly restrictive constraints on the bending radii
of the coils/wires. Like UST-2 [13], EPOS aims to use 3Dprinted
winding frames/surfaces. In addition, it will seek
to take advantage of new technologies for 3D printing
metal structures suitable for UHV conditions and cryogenic
An avenue will also be needed for getting the positrons
into the device from the NEPOMUC beam line. EB drift
injection has worked extremely well for getting
NEPOMUC’s positrons into proto-APEX and then trapping
them there [5]. In order to use a similar scheme for
injecting them into EPOS, a region of “close approach” to
the plasma from the e+ beam line will be needed. In some
exploratory designs for stellarator fusion reactors, an
extended helical coil has been proposed, in order to create
large access ports; in the “stray field” of such a coil could
be the right place to hook up the positron beam line.
Looking ahead to pair plasmas
The estimated timeline for EPOS involves spending
approximately the first year of the program developing the
design of the device, in close collaboration with stellarator
optimization experts at IPP and in the Simons “Hidden
Symmetries” Collaboration, as well as in consultation with
superconducting magnet researchers in Germany and the
United States. Construction will start in the second year.
In the third year, commissioning will be performed with
electrons. This will be followed by installation at
NEPOMUC (which is operated at the FRM II neutron
source, next door to IPP Garching) and the addition of
positrons, leaving approximately three years for electron/
positron experiments and cross-comparisons between
As a primary method of diagnosing the positron and pair
plasmas, both devices will be taking measurements of the
511 keV gamma rays produced by annihilation events.
Several types of gamma ray detection are planned: coincident
detection such as that used in positron emission
tomography (PET) scans, high-energy-resolution spectroscopy
with high-purity Ge detectors, and lower-resolution
detection of line- or volume-integrated annihilation rates
with scintillator/PMT (photomultiplier tube) assemblies.
(Variations on this last technique have been used in proto-
APEX to date and worked extremely well, especially
when supported with simulations [5].) The information
gathered will be then used to calculate important qualities
of the pair plasma, such as its density and temperature [4]
— thereby allowing us to determine when we have
achieved multiple Debye lengths, a key milestone.
The first things to study with the pair plasmas will be the
confinement time and stability. If electron/positron plasmas
in these devices are indeed uniquely stable, that will
be encouraging, since historically the magnetic confinement
of plasmas has repeatedly turned out to be more
complex than anticipated. But either way, it will be an
opportunity to explore a fascinating new frontier in plasma
The EPOS project is jointly funded by the Helmholtz
Association and the Max Planck Institute for Plasma Physics,
within the framework of the Helmholtz Young Investigator
Groups program.
Stellarator News -8- October 2019
The APEX collaboration receives/has received support
from the European Research Council (ERC) under the
European Union's Horizon 2020 research and innovation
programme; the Deutsche Forschungsgemeinschaft
(DFG); the Helmholtz Postdoc Programme; the UC San
Diego Foundation; the Japan Society for the Promotion of
Science (JSPS); and the National Institute for Fusion Science
[1] V. Tsytovich and C. B. Wharton. Comments Plasma
Phys. Controlled Fusion, 4 91–100 (1978).
[2] P. Helander. Phys. Rev. Lett. 113 135003 (2014).
[3] A. Le et al., Phys. Rev. Lett. 110 135004 (2013); N. T.
Howard et al., Nucl. Fusion 56 014004 (2016).
[4] T. Sunn Pedersen et al., New J. Phys. 14 035010 (2012).
[5] E. V. Stenson et al. Phys. Rev. Lett. 121 235005 (2018);
J. Horn-Stanja et al., Phys. Rev. Lett. 121 235003
[6] C. Hugenschmidt et al., New J. Phys. 14 055027 (2012).
[7] J. R. Danielson et al., Rev. Mod. Phys. 87 247 (2015).
[8] P. R. Garabedian and G. B. McFadden, J. Res. Natl.
Inst. Stand. Technol. 114 229 (2009).
[9] E. V. Stenson et al., J. Plasma Phys. 83 595830106
[10] C. J. Crannel et al., Astrophys. J. 210 582 (1976); R. G.
Greaves and C. M. Surko, AIP Conf. Proc. 606 10
(2002); P. Wallyn et al., Astrophys. J. Supp. Ser. 92 551
[11] J.-F. Lobsien et al., Nucl. Fusion 58 106013 (2018).
[12] T. Sunn Pedersen et al., Fusion Sci. Technol. 46 200
[13] V. M. Queral,
Eve V. Stenson
Max Planck Institut für Plasmaphysik
Garching, Germany

Current View