Measurement of Excitation Temperature of Radio-Frequency Plasma in
Water Using a High-Speed Camera
Shinobu Mukasa, Shinfuku Nomura and Hiromichi Toyota
Graduate school of Science and Engineering, Ehime University, 3 Bunkyo-cho, Matsuyama 790-8577, Japan.
Abstract: The distribution of the excitation temperature of radio-frequency
underwater plasma is measured using a high-speed camera. The image of the
plasma is split into two images by a dichroic mirror. By passing each of the
images through band-path filters, the emission images of Hα and Hβ are captured
simultaneously using the high-speed camera. The excitation temperature is
calculated from the two emission intensities. The method proposed in this
experiment improves both the time and spatial resolutions of the measurement of
the excitation temperature, and enables analysis of the relationship between the
excitation temperature and the behavior of bubbles generated by the evaporation
of the liquid around the plasma. Change of the excitation temperature coincided
periodically with repetitive bubble behavior such as appearance, growth on the
tip of the electrode or departure from the electrode. The temperature reaches a
maximum before the departure of the bubble from the electrode. At that time, the
distribution of the temperature is stretched upward from the electrode
simultaneously with the upward movement of the bubble.
Keywords: in-liquid plasma, excitation temperature, high-speed camera, bubble
behavior
1. Introduction
Efficient synthesis and decomposition of substances
by means of plasma induced by microwave or highfrequency irradiation into a liquid is expected to be
applicable across a wide field of studies. One
physical difference from the plasma in the general
gaseous phase that should be considered is the
interaction between the plasma and the behavior of
the bubbles generated by evaporation of the
surrounding liquid due to the heat of plasma. The
properties of the plasma in a liquid are changed by
the fluctuation of the bubbles.
Although the spectroscopic analysis of the plasma
[1] and observation of the behavior of the bubbles
using a high-speed camera [2] have been conducted
separately as fundamental research of the plasma in
liquid, these researches have yet to be executed
simultaneously. To aid in this, in recent years there
has been rapid development of highly powerful
computers with the ability to process huge amounts
of data and advances in solid-state technology, for
example in CCD, CMOS, etc. These enable
measurement of physical quantities with high
temporal and spatial resolution.
In this research, radio-frequency emission from
underwater plasma was split optically into two
monochromatic emissions of Hα and Hβ, and the
images of these emissions were captured
simultaneously by one high speed camera. This
method enables measurement of the distribution of
the excitation temperature in plasma with high
temporal and spatial resolution.
2. Experimental setup and procedure
The experimental apparatus is shown in Fig. 1.
Radio frequency (RF) of 130-150 W at 27.12 MHz
is supplied from the power supply to the electrode
through an adjustment circuit. The reactor vessel is
constructed as an aluminum block pierced on both
sides with a 30 mm-diameter steel pipe, and an
electrode is inserted from the bottom. Two
observation windows of transparent quartz glass are
provided in front and the rear of the vessel. The
electrode is composed of a tungsten rod with
Vacuum pump
Polycarbonate Band-path
plate
filter
Valve
Pump
Reservoir
tank
Flow meter
Bubble
Plasma
Quartz window
Stainless steel pipe
Bubble
Plasma
Matching
Box
Dichroic
mirror
spectrum
E − E 3 ⎡ ⎛⎜ I Hα
Tex = 4
a ⎢ln
⎜
kB
⎣⎢ ⎝ I Hβ
Figure 1. Experimental apparatus.
intensity (arb. unit)
spectrum
Band-path
filter
spectrum
high-speed camera
Figure 3. Schematic image of photographic system.
Electrode
1
spectrum
spectrum
Electrode
Water flow
Mirror
(1)
where E is the energy, kB is the Boltzmann constant,
transition probability and g is the statistical weight
of the energy level. The subscript of 3 or 4 expresses
the energy level of hydrogen.
OH
0.5
Hβ
O
O
0
400
−1
⎞⎤
⎟⎥ ,
⎟⎥
⎠⎦
ν is the frequency of the spectral lines, A is the
Hα
10 kPa
⎞
⎛
⎟ − ln⎜ ν 3 AHα g 3
⎟
⎜ v 4 AHβ g 4
⎠
⎝
600
wavelength (nm )
800
Figure 2. Emission spectrum of RF plasma in water.
diameter of 3 mm inserted into a transparent quartz
tube with an outer diameter of 6 mm. The tip portion
of the electrode is shaped hemispherically to
promote easy plasma generation. Pure water is used
as a liquid, and circulated through the equipment by
the pump with flow rate of 1.5 L/min for constant
plasma generation. Although the temperature of
water is not controlled, it is retained at
approximately 40°C by the thermal balance between
the heat from plasma and the overall heat dissipation
from the equipment. The pressure throughout the
system is reduced by an aspirator and controlled in
the range between 10 kPa and atmospheric pressure.
The emission spectrum of the underwater plasma
consists of mainly OH (309 nm), Hβ (486 nm), Hα
(656 nm), O (777 nm) and O (845 nm), as shown in
Fig. 2. When assuming that the excitation
distribution follows the Boltzmann distribution, the
excitation temperature of atomic hydrogen Tex is
calculated by introducing intensity ratio of Hα to Hβ
IHα/IHβ to the following equation.
A combination of optical devices allows only Hα
and Hβ to be observed from the plasma emission.
The outline of the optical system is shown in Fig.3.
The image of the plasma is split into two images; the
transmitted image, and the image reflected by a
dichroic mirror (DIM-50-RED, Sigma Koki)
inclined 45° in relation to the plasma. The reflected
image is then refracted again at a right angle using a
mirror parallel to the dichroic mirror to range the
two images. By passing each of the images through
band-path filters (Hα 35nm and Hβ 8.5nm, Baadar
Planetarium), one is the emission image of Hα while
the other is that of Hβ. Finally, the two separated
images are captured simultaneously using a highspeed camera (MEMRECAM GX1-plus, Nac image
technology). To compensate for the difference of the
optical path (35.4 mm) between two images and
adjust the focus of the images, the image that
traveled longer is transmitted through a transparent
polycarbonate board (refractive index: 1.59) with a
thickness of 95 mm.
The frame rate of the high-speed camera is set to
2000 fps with the shutter open for 250 ms, and
therefore 500 sequential images are captured. The
size of the captured image is 1280×720 pixels, and
the image is 12-bit monochrome. At the same time,
spectroscopic measurement using a spectroscope
(PMA-11, Hamamatsu) is conducted.
y (spectroscope)
8000
800
Total intensity (Hα)
y = 5.04×10
-3
Total intensity (Hβ)
x
y = 6.69×10
y (spectroscope)
10000
6000
4000
-3
x
600
400
200
2000
1
0
x (high-speed camera)
2
6
[×10 ]
0.5
0
1
x (high-speed camera)
(a) Hα
5
[×10 ]
(b) Hβ
Figure 4. Relationship between two emission intensities
measured by different devices.
10 kPa
4000
2000
(a) 10 kPa
Tex (K)
70 kPa
4000
2000
101 kPa
4000
2000
0
0.1
time (s)
0.2
Figure 5. Temporal change of excitation temperature.
3. Results and discussions
Figure 4 shows the relationship of the emission
intensity between Hα and Hβ observed by two
devices; a high-speed camera and a spectroscope.
The horizontal x-axis indicates the total intensity for
all the pixels in the area containing the whole plasma
emission over 500 sequential pictures captured by
high-speed camera, while the vertical y-axis
indicates the emission intensity measured by the
spectroscope. The data obtained from the experiment,
when plotted for the changes in pressure from 10
kPa to the atmospheric pressure, and is nearly a
proportional relationship.
The excitation temperature is calculated by
conversion of the emission intensity ratio measured
using the high-speed camera into that of the
spectroscope. However, when the values calculated
(b) 101 kPa
Figure 6. Sequential images of distribution of excitation
temperature.
from the results of Fig.4 are assigned as a conversion
factor, the temperature is overestimated greatly. The
reason for this is thought to be that the actual
detectable area of the spectroscope is smaller than
that in our prediction, but the reason is still not clear.
Therefore, the conversion factor is determined to be
that for which the maximum temperature in the timeaveraged distribution becomes the same as that
measured by moving the spectroscope probe 0.5 mm
vertically and horizontally [1].
The temporal change of the excitation temperature
averaged spatially throughout an area with a width
Figure 6 shows images of excitation temperature
distribution sequenced temporally. The black area of
the background picture indicates the side-view of the
electrode 3 mm in a diameter. When the pressure is
low, the temperature distribution spreads over the
top surface of the electrode as shown in picture (a)
and fluttered irregularly. However at high pressure,
capillaceous distribution of high temperature appears
intermittently as shown in picture (b).
(a) 10 kPa
Figure 7 shows sequential images of plasma and
bubble formation for comparison of the distribution
of the excitation temperatures shown in Fig. 6 with
the behavior of the bubble. It should be noted that
the images were captured on a different day from
when the images in Fig. 6 were captured, although
the experimental conditions were exactly the same.
When the pressure is 10 kPa, the entire surface of
the electrode is covered by the bubble. When the
pressure is 101 kPa, the bubble covers only the part
of the surface touched by the plasma, while other
areas of the electrode surface are covered with water.
4. Conclusions
(b) 101 kPa
Figure 7. Sequential images of plasma and bubble.
of 2 mm and length of 1 mm at the tip of the
electrode where strong emission is observed is
shown in Fig. 5. The temperature fluctuates in the
range of approximately from 1000 to 4000 K at each
pressure, and the periodicity of the fluctuation
becomes more regular with increase of the pressure.
The cycle is approximately 16 ms at the atmospheric
pressure, and because it is almost equal to the
growth cycle of bubbles, it may be considered that
the temporal change of the excitation temperature is
synchronized with the growth of the bubbles.
Measurement of the excitation temperature in RF
plasma in water with high temporal and spatial
resolution was conducted using a high-speed camera.
The temporal change of the temperature became
more periodical with an increase of pressure. The
cycle seemed to be connected to that of the growth
of the bubble which was generated containing the
plasma.
Acknowledgements
This work was partially supported by a Grant-in-Aid
from the Ministry of Education, Culture, Sports,
Science and Technology of Japan (No. 22760153).
References
[1] S. Mukasa, S. Nomura, H. Toyota, T. Maehara, F.
Abe and A. Kawashima, J. Appl. Phys. 106, 113302
(2009).
[2] S. Mukasa, T. Maehara, S. Nomura, H. Toyota,
A. Kawashima, Y. Hattori, Y. Hashimoto and H.
Yamashita, Int. J. Heat Mass Transfer 53, 3067
(2010).