Plasma and Fusion Research: Regular Articles Volume 12, 2402002 (2017)

Behavior of Non-Thermal Electrons during ECR Pre-Ionization at

Aditya Tokamak∗)
Shishir PUROHIT, Shankara JOISA, Malay Bikas CHOWDHURI, Brajkishore SHUKLA,
Jayesh RAVAL, Ranjana MANCHANDA, Nilam RAMAIYA, Umesh NAGORA, Parveen ATREY,
Rakesh TANNA, Kumarpalsinh JADEJA, Shailesh BHATT, Chetnarayan GUPTA,
Ajai KUMAR, Joydeep GHOSH and Aditya Team
Institute for Plasma Research, Bhat, Gandhinagar, 382 428, India
(Received 27 November 2015 / Accepted 30 November 2016)

The non-thermal electrons (NTE) in tokamak plasma have a strong effect over the plasma properties and
on the plasma facing components. This establishes an importance of NTE to be studied in both flux and energy
space. The NTE was studied for Aditya tokamak by monitoring X-ray spectrum using a Silicon Drift Detector
(SDD) based X-ray spectroscopic diagnostic for the ECR pre-ionization experiments. The spectra shows X-ray
line radiation band on top of the bremsstrahlung. Line radiation is primarily the combination of multiple charac-
teristics lines existed between 5 keV to 7 keV and mostly associated with multiple ionized states of iron, tokamak
wall material. This is likely due to the interaction between the NTE, generated during ECR pre-ionization ex-
periment, with the stainless steel vessel wall of the Aditya Tokamak. The behavior of NTE was studied as a
function of vertical field and pre-filled pressure. With the application of vertical field, energy carried by the NTE
is reduced due to higher vertical drift velocity in comparison to the case of without one. Significant reduction of
the X-ray line radiation strength with higher pre-filled pressure also indicates the reduction of NTE due to the
enhanced electron neutral collision.

c 2017 The Japan Society of Plasma Science and Nuclear Fusion Research

Keywords: non-thermal electron, X-ray spectra, SDD, ECR pre-ionization, Aditya tokamak
DOI: 10.1585/pfr.12.2402002

1. Introduction work has reported the X-ray spectra associated with NTE
Electron populations associated with the tail of elec- for the ECR pre-ionization experiments in Aditya toka-
tron energy distribution for tokamak plasma are referred mak [7].
as non-thermal electrons (NTE). They hold higher ener- Figure 1 shows the characteristic X-ray radiation band
gies and are potential enough to migrate into the runaway on top of the bremsstruhlang, which seems to be a combi-
regime causing the plasma to lose its energy [1]. Moreover nation of multiple characteristics lines colonized within 5
interaction of NTE with the plasma facing components to 7 keV. The band is primarily associated with the multiple
(PFC) and other inner structures of the tokamak are po- ionized states of Fe, which is main constituent of stainless
tential enough to damage them and tokamak vessel, which steel, wall material of vacuum vessel of Aditya tokamak.
is a matter of great concern [2]. These distinctive features
offer a strong candidature of NTE to be studied in depth,
both in energy and flux space. NTE are mostly observed
during the plasma startup and current collapse/disruption.
However a finite population of NTE is always present over
the plasma discharge.
The NTE are also prominently observed during the
Electron Cyclotron Resonance (ECR) pre-ionization as-
sisted tokamak plasma breakdown [3]. It serves as a useful
tool for discharge initiation which reduces the consump-
tion of Ohmic power at initial phase of discharge, thus
widely employed for tokamak plasma startup [4, 5]. NTE
has been studied by monitoring the X-ray spectrum [6, 7]
generated during the plasma experiments. Our previous
author’s e-mail: [email protected] & [email protected] Fig. 1 X-ray spectrum, integrated over few discharges for ECR
∗) This article is based on the presentation at the 25th International Toki
pre-ionization experiment for Aditya tokamak plasma.
Conference (ITC25).

2402002-1 
c 2017 The Japan Society of Plasma
Science and Nuclear Fusion Research
Plasma and Fusion Research: Regular Articles Volume 12, 2402002 (2017)

It can be inferred that there exist a strong NTE and vessel

interaction. This strong interaction is mostly attributed to
the drift of NTE particle across the field lines in the ab-
sence of any compensating mechanism. This drift of NTE
is observed due to different reasons namely; drift due to the
presence of error field, grad B and curvature drift [8–10].
The drifting of NTE takes away considerable energy from
the plasma which would have been consumed in for better
ionization of the fuel particles. A study has been done to
understand the behavior of NTE of Aditya tokamak during
ECR based pre-ionization startup and to understand the ef-
fect of different parameters like pre-filled pressure and ap-
plied vertical field, over the NTE behavior.
The experimental results from Electron Cyclotron
Resonance based pre-ionization startup are presented in
this article. It was observed that vertical field and higher
pre-filled pressure reduces that NTE particle and wall in-
teraction. The experimental setup is described in Sec. II.
Result and Discussion of behavior of NTE with different
experimental conditions are discussed in Sec. III which is
followed by a summary. Fig. 2 Cross-section view of the Aditya tokamak. Coils TR1-
TR5 are part of the OT connected in series, BV1 and BV2
are the vertical field coils and FF coils are the fast feed-
2. Experimental Setup back coils.
ADITYA tokamak is a mid-sized device (R = 0.75 m,
a = 0.25 m), with two breaks in vacuum vessel toroidally
180◦ apart, carrying about 80 - 150 kA of plasma current
with a graphite poloidal limiter [11]. Equally spaced 20
coils offers 0.75 T - 1.35 T toroidal field. The cross-section
view of Aditya tokamak is shown in Fig. 2. The TF coils
are square in shape. The BV1 and BV2 are the major coils
by which the vertical field is configured for the Aditya
tokamak. The typical Aditya core plasma temperature re-
sides within 300 to 500 eV and the line averaged electron
density ranges in between 1.0 to 2.5 × 1019 m−3 . Aditya
Tokamak is having a verity of diagnostics for measuring
different plasma parameters, namely Rogowski coils for
plasma current, and magnetic diagnostic and Mirnov coils
Fig. 3 SDD spectrometer coupled with the Aditya Tokamak by
for the MHD studies, Soft X-ray detector array for chord
a 35CF connection on a radial mid plane.
average plasma temperature and MHD studies. The mi-
crowave diagnostic measures the plasma electron density
[11, 12]. Visible spectroscopy systems [13] provides infor-
toroidal field.
mation about the temporal behavior of Hα , Hβ emissions
and spectral line emissions, such as OII, CII, and CIII of
different impurities, from the Aditya tokamak plasma. The 3. Results and Discussion
Silicon Drift Detector (SDD) X-ray spectroscopic diagnos- The experiments were divided in to two sets. First set
tic records the NTE X-ray spectrum in the energy range of was to understand the NTE drift behavior under the vertical
1.5 - 26 keV [7]. The system is mounted on a radial mid field and second set for the dependence of the collisions
plane port with 35CF opening on the Aditya tokamak (see over NTE as a function of pre-filled pressure.
Fig. 3). This geometry allows the detector to view 20 cm
plasma cross-section around the center of plasma. 3.1 Effect of the vertical field
The ECR system [14] coupled with Aditya tokamak The first set of experiment was performed at fixed
is having a 42 GHz, 0.5 MW configurations. Its pulse plasma and operational parameters and by altering ap-
length for this experiment was kept ∼ 70 - 100 ms and pow- plied vertical field. The result from two representative dis-
ers 130 - 150 kW, respectively. ECR power was launched charges 26953 (blue) and 28163 (red) are shown in Fig. 4.
from the low field side of the Aditya tokamak with 0.75 T The time evolution of the plasma current (A), ECR power
2402002-2
Plasma and Fusion Research: Regular Articles Volume 12, 2402002 (2017)

Fig. 4 The time evolution of the Plasma current (A) ECR power Fig. 5 The time evolution of the Plasma current (A) ECR power
& loop voltage (B), Hα signal (C) applied vertical field & loop voltage (B), Hα (C) applied vertical field (D), pre-
(D), pre-filled pressure (E) , line averaged density (F) and filled pressure (E), line averaged density (F) and time in-
time integrated NTE spectrum (G) for two Aditya dis- tegrated NTE spectrum (G) of pure Ohmic Aditya break-
charges. down for discharges 26754.

& loop voltage (B), Hα signal (C), applied vertical field the production of NTE during the breakdown phase data
(D), pre-filled pressure (E), and line averaged electron den- were collected from many discharges with plasma startup
sity (F) are shown. The NTE spectra for the respective dis- initiated by Ohmic power only without any ECR assisted
charges integrated over the discharge life time are shown pre-ionization. A typical Ohmic Aditya discharge 26754
in Fig. 4 (G). is shown in Fig. 5. This discharge, as like of 26953, is
The ECR pulse was launched when the vertical field having almost similar current and the vertical field at the
was almost zero for discharge 26953 (blue). The Hα signal time when Hα is peaked. The Loop voltage is also higher
for this discharge is low indicating less ionization within than the discharge 26953 (blue). The NTE spectrum for
the tokamak. However, one can see that it has also a sec- this discharge can be seen in the Fig. 5 (G). The spectrum
ond peak, corresponding to the application of loop voltage. do not exhibit any line radiation and the flux is also very
The affect of loop voltage on the generation of NTE is ex- less in comparison to the discharge 26953 (blue). Then,
plained later. The X-ray spectrum shows a considerable it can be inferred that ECR pulse becomes the dominant
line radiation sitting above the continuum (see Fig. 4 (G)), contributor to the generation of NTE as indicated by the
within the energy band of 5 - 7 keV for this discharge. This strong X-ray spectrum as shown in Fig. 4 (G).
line radiation is mostly correlated with the stronger NTE During tokamak breakdown phase, charged particle
and wall material interaction resulting from the particle intensely drift vertically due to the error field in toroidal
drift. On the other hand the ECR power for discharge field, gradient and curvature drift under zero or the non
28163 (red) launched at the ramping up phase of the ver- application of appropriate vertical field. During the pre-
tical field. The Hα signal for this discharge has increased, ionization phase, equilibrium is mainly provided by the
almost twice from the 26953 (blue), indicating higher ion- vertical field. In the presence of vertical field, the field
ization with this vertical field configuration and offers a lines are helical in nature which allows the particles flow
better plasma start-up condition. The NTE X-ray spectrum toroidally [15], which reduces the loss of particles on the
shape is also different for the discharge, 28163. The X-ray wall due to the drift [10, 16]. That is to say that vertical
line radiation contribution within the X-ray spectrum is ob- field acts as a compensatory field.
served to be less prominent, pointing towards the change in The X-ray spectrum, at higher vertical field, of the dis-
the energy of the NTE reaching to the wall under the pres- charge 28163 (red) shows a shift in the line radiation peak.
ence of vertical field. One can figure out that the centroid of the line radiation
To evaluate the possible contribute of loop voltage in has shifted towards the lower energy side on the energy
2402002-3
Plasma and Fusion Research: Regular Articles Volume 12, 2402002 (2017)

axis. This clearly shows that the energy carried by the NTE
has reduced. This can be understood by following equa-
tion of electron velocity in Z direction and the connection
length [7, 10].
mγ(v2 + v2⊥ /2) Bv
vz  − + v , (1)
eRBϕ Bϕ
Bϕ
L=H . (2)
BV
Here v and v⊥ denote parallel and perpendicular compo-
nent of velocity to the field and γ is the relativistic fac-
tor. The first term represents the vertical drift due to the
gradient and curvature in the Bϕ . H is the particle drifted
distance, which is equal to the minor radii of the vessel.
Then to avoid the drifting of the particle along the verti-
cal direction proper compensating vertical field is needed
to be applied. However, in our experimental condition sec-
ond term is dominating than the first term, as for example,
for the electron having an energy of 3 keV and a pitch an-
gle of 80 degree, the second term becomes ∼ 1 × 105 m/s
while the first term ∼ −1.4 × 104 m/s for R = 0.75 m,
Bv = 100 Gauss and Bϕ = 0.75 T. It is basically over
compensating situation due to the application of vertical Fig. 6 The time evolution of the Plasma current (A) ECR power
field Bv of ∼ 100 G. That is to say the experimental data (B), Hα (C) applied vertical field (D), pre-filled pressure
shown here exhibit two discharges; one is without compen- (E) , line averaged density (F) and time integrated NTE
sation and other is the case of over compensation. When spectrum (G) for two Aditya discharges.
Bv was not applied (discharge no - 26953) the magnitude
of the vertical drift velocity was lower than that of during
the over compensating case having shot no of 28163. In both of the discharges. During the experiment Ohmic
the over compensating case, the electron reaches the wall power was switched off and vertical fields were kept zero
quickly without gaining much energy and they are there- as shown in Fig. 6 (D). Both discharges are having simi-
fore producing the X-ray spectrum having peak shifted to- lar plasma current as shown in Fig. 6 (A).The Hα signal for
wards relatively lower energy as reflected in the spectrum the discharges with low pressure 28392 (red) shows lower
shown in Fig. 4 (G) in red colour. Increasing vertical field signal. The high pre-filled pressure discharge 29213 (blue)
will reduce the connection length. It causes the particle exhibiting higher Hα signal, which is almost saturating for
drift quickly towards the wall. And therefore NTE is likely most of the discharge life time is indicative of improved
going to carry lower energy while interacting with the wall. ionization for this discharge.
Along with that the electron having vz = 0 will be confined, The line radiation shown in Fig. 6 (G) (blue in color),
gain energy and may convert to the runway electron and indicative of interaction of NTE and wall, has completely
will not contribute in x-ray line radiation until they reach disappeared and the total X-ray counts have also reduced
the wall, but predominantly contribute in continuum X-ray in comparison from the low pressure discharge. Peak elec-
spectrum. This can be seen in the Fig. 5 (G), here charac- tron densities are similar in both cases although Hα signal
teristics peak is not considerably apparent but continuum is lower in low pre-filled pressure discharge. This might
radiation is present. be related to the contribution of electrons coming from
higher amount of impurities entering into the plasma be-
3.2 Effect of the pre-filled pressure cause of higher NTE-wall interaction in the low pressure
The effect of the collisions within the plasma over the discharges.
NTE was investigated as a function of pre-filled pressure. The high pre-filled pressure offer higher neutral den-
The desired pre-filled pressure has been created for the hy- sity which is directly affects the electron–neutral colli-
drogen gas, by a piezoelectric valve, prior the ECR power sion frequency, νen , given by equation 3 [16, 17], where
applied into the Aditya tokamak. σen (T e ), effective electron–neutral scattering cross section
The results from two representative discharges 28392 given by equation 4 nn is the neutral density and T e is elec-
(red) and 29213 (blue) for these experiments are shown tron temperature.
in the Fig. 6. Here, the discharge 28392 (red) and 29213

(blue) are operated with low and high pre-filled pressures, eT e
respectively, The ECR power, ∼ 150 kW, was launched for ven = σen (T e )nn , (3)
me
2402002-4
Plasma and Fusion Research: Regular Articles Volume 12, 2402002 (2017)
⎡   ⎤⎥
⎢⎢⎢ T e ⎥
− 0.1 ⎥⎥⎥⎥
⎢⎢⎢ mak X-ray spectrum were recorded by SDD based spec-
⎢ ⎥
σen = 6.6 × 10−1 ⎢⎢⎢⎢⎢  4  1.6 ⎥⎥⎥⎥⎥ .
troscopic system. It was seen that with the increase in the
(4)
⎢⎢⎢ T ⎥⎥⎥ vertical field and the pre-filled pressure the energy of NTE
⎣ 1+ e ⎦
4 reaching to the wall is reduced as indicated by the X-ray
spectrum and the improvement of ionization is occurred.
The loss time are defined as follows
With the application of vertical field energy carried by the
d NTE is reduced due to higher vertical drift velocity en-
τ= . (5)
vz abling the electron to reach the wall quickly without gain-
Where d is the half height of the vacuum vessel, vz is ing much energy in comparison to the case in which the
the drift velocity as defined in the first term of the equa- vertical field is not applied In high pre-filled pressure, NTE
tion no. 1 since Bv was not applied. For 10 eV elec- is suppressed due to the higher electron-neutral collision
trons, i.e., one can say thermal electron in ECR produced frequency.
pre-ionization plasma; typical life time becomes ∼ 5 ms,
which is one order higher than the electron-neutral colli- Acknowledgement
sion times. Estimated collision times are 34 and 103 μs for Two of the authors (S. Purohit, and M. B. Chowdhuri)
the discharge with high and low pressures, respectively. would like to thank anonymous reviewer for the valuable
However in the case of ECR produced non thermal elec- suggestions to improve the discussion in the paper.
tron having T e of 3 keV, the life time reduces to ∼ 50 μs,
which is comparable to the electron-neutral collision time
[1] H. Knoepfel et al., Nucl. Fusion 19, 785 (1979).
(∼ 60 μs) at higher fill pressure, on the other hand, electron [2] R. Yoshino et al., Nucl. Fusion 39, 151 (1999).
neutral collision times becomes ∼ 170 μs in the case lower [3] S. Tashima et al., J. Plasma Fusion Res. SERIES 9, 316
fill pressure case. That is why NTE losses its energy in (2010).
higher pressure case before reaching the wall and did not [4] V. Erckmann et al., Plasma Phys. Control. Fusion 36, 1869
produced any significant X-ray photon which is apparent (1994).
from the Fig. 6 (G) (in blue color & solid line). Whereas [5] B. Lloyd et al., Nucl. Fusion 31, 2031 (1991).
[6] R. Baskaran et al., Rev. Sci. Instrum. 70, 2637 (1999).
NTE did not suffer any energy losses due to collision in
[7] S. Purohit et al., Rev. Sci. Instrum. 85, 11E419 (2014).
lower pressure case and produces prominent X-ray spec- [8] Wesson, Tokamak (Oxford, UK: Clarendon 2004).
trum as shown in the Fig. 6 (G) with red color and dotted [9] P.K. Chattopadhyay et al., Nucl. Fusion 36, 1205 (1996).
dash line. Thus the weak X-ray spectrum and disappear- [10] I.H. Hutchinson et al., Nucl. Fusion 14, 649 (1974).
ance of the line radiation from the X-ray spectrum for dis- [11] S. Bhatt et al., Indian J. Pure Appl. Phys. 27, 710 (1989).
charge 29213 (blue) is most likely attributed to the increase [12] P.K. Atery et al., Proc. 20th IAEA FEC 2004, EX/p4-1,
pg-41.
in the collision frequency at high pre-filled pressure sug-
[13] R. Manchanda et al., Phys. Plasmas 17, 072515 (2010).
gesting the reduction of the NTE. [14] B.K. Shukla et al., SOFE (2015), Austin USA, pp.1-4.
[15] S. Nishi et al., Plasma Phys. Control. Fusion 52, 125004
(2010).
4. Summary [16] Mikelides et al., AlAA Paper 01-3355 (2001).
To understand the behavior of the NTE during the [17] D. Mueller, PPPL-4836 (2012).
ECR pre-ionization assisted startup in the Aditya toka-

2402002-5