Journal of Physics D: Applied Physics

ACCEPTED MANUSCRIPT

Time-Resolved Imaging and Spectroscopy of Atmospheric Pressure

Plasma Bullet Propagation and RONS Production
To cite this article before publication: Ryan Patrick Gott et al 2020 J. Phys. D: Appl. Phys. in press https://doi.org/10.1088/1361-6463/ab876a

Manuscript version: Accepted Manuscript

Accepted Manuscript is “the version of the article accepted for publication including all changes made as a result of the peer review process,
and which may also include the addition to the article by IOP Publishing of a header, an article ID, a cover sheet and/or an ‘Accepted
Manuscript’ watermark, but excluding any other editing, typesetting or other changes made by IOP Publishing and/or its licensors”

This Accepted Manuscript is © 2020 IOP Publishing Ltd.

During the embargo period (the 12 month period from the publication of the Version of Record of this article), the Accepted Manuscript is fully
protected by copyright and cannot be reused or reposted elsewhere.
As the Version of Record of this article is going to be / has been published on a subscription basis, this Accepted Manuscript is available for reuse
under a CC BY-NC-ND 3.0 licence after the 12 month embargo period.

After the embargo period, everyone is permitted to use copy and redistribute this article for non-commercial purposes only, provided that they
adhere to all the terms of the licence https://creativecommons.org/licences/by-nc-nd/3.0

Although reasonable endeavours have been taken to obtain all necessary permissions from third parties to include their copyrighted content
within this article, their full citation and copyright line may not be present in this Accepted Manuscript version. Before using any content from this
article, please refer to the Version of Record on IOPscience once published for full citation and copyright details, as permissions will likely be
required. All third party content is fully copyright protected, unless specifically stated otherwise in the figure caption in the Version of Record.

View the article online for updates and enhancements.

This content was downloaded from IP address 137.111.162.20 on 11/04/2020 at 14:00

Page 1 of 18 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1

1
2
3
4 Time-Resolved Imaging and Spectroscopy of Atmospheric Pressure

pt
5 Plasma Bullet Propagation and RONS Production
6 Ryan P. Gott1 and Kunning G. Xu2
7 Department of Mechanical and Aerospace Engineering, University of Alabama in Huntsville, Huntsville, Alabama,
8 35899, United States
9

cri
10 An atmospheric pressure plasma jet visually appears as a continuous plasma, but is actually comprised of a
11 series of fast moving ionization waves, commonly called plasma bullets. Plasma bullets are the visible front of
12 ionization streamers. Studying their spatiotemporal formation and propagation can help understand the
13 plasma dynamics. In this work, the formation and lifetime of a high voltage pulsed dc plasma jet and the
14 subsequent bullets were observed using time-resolved synchronized measurements with an ICCD camera and
15 an ICCD spectrometer. The operating voltage, helium flow rate, and pulse width were varied to observe
16

us
changes in bullet behavior. The measured velocities of the plasma bullets largely depended on voltage but
17 showed that flow rate can have an effect. The formation of OH and excited N2 (2+) was measured at the
18
nanosecond timescale. It was seen that the formation of these species begins at the visible bullet, but the peak
19
emission and decay occur ~50-100 ns after bullet passage, indicating a finite time of reaction, excitation, and
20
emission. Observing the temporal formation of plasma bullets and reactive species provides insight into some
21
of the reactions occurring in the discharge and how to tailor the plasma operating conditions.
22
23
24
25
26
27
28
I. Introduction
an
Teschke in 2005[1] and Lu and Laroussi in 2006[2] first reported that atmospheric pressure
plasma jets (APPJ) generated by ac or pulsed dc discharges form as bullets on the nano to
microsecond timescales. When a sufficient electric field is generated at the electrodes, the gas near
dM
29 the electrodes begins to ionize and then forms ionization waves that propagate away from the
30 electrode. In an APPJ, the ionization front continues out away from the electrode and extends
31
outside of the tube. If a feed gas with a lower breakdown voltage than air (such as helium) is used,
32
33 the plasma will travel in small “bullets” along the gas channel. These bullets occur on the
34 nanosecond timescale and are observed through high speed or synchronized imaging, which can
35 give insight to the formation of the bullets. It is worth noting that “bullet” refers to the nanosecond
36 timescale ionization wave front. Other terms, such as jet or effluent, can be used to describe the
37
38 time-averaged output of an APPJ.
pte

39 The physics of APPJ bullets have been extensively studied to show that the bullets move
40 at speeds of tens of km/s[3]–[5], are affected by the gas mixture they move through[6], and are
41 controlled by the electric fields[7], [8]. Considerable work has been done to measure the time-
42
43 integrated emissions of the formation of jets[8]–[12], but time-resolved measurements of the
44 generation of important reactive species would also be of interest. Such observations provide
45 insight into the mechanisms and operating conditions that could improve the overall production of
46 reactive species for various applications.
ce

47
48
APPJs can produce reactive oxygen and nitrogen species (RONS), UV light, shock waves,
49 and electrons and ions that produce useful reactions on material surfaces. RONS include OH,
50 peroxide, ozone, NOx, N2+, and others. These species have been shown to benefit water, plants,
51 and biomaterials in unique ways. RONS can disinfect surfaces and remove dangerous bacteria by
52
damaging the cells of organisms[13]. RONS can also drive important purification processes in
Ac

53
54
55 1
Graduate Research Assistant, Mechanical and Aerospace Engineering, rpg0007@uah.edu
56 2
Associate Professor, Mechanical and Aerospace Engineering, gabe.xu@uah.edu
57
58 1
59
60
 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1 Page 2 of 18

1
2
3 water[14]. Plasmas generate RONS by dissociating oxygen, nitrogen, and water molecules.
4

pt
5 Different types of plasma sources have been used to produce RONS, including dielectric barrier
6 discharges (DBDs)[15], [16], plasma torches[17], and APPJ[18], [19].
7 By using intensified charge-coupled device (ICCD) imaging and optical emission
8 spectroscopy (OES), both synchronized with the voltage pulses, time-resolved images of the
9

cri
10
plasma bullet can be compared to the time-resolved emissions at fixed locations in the jet. This
11 comparison provides a unique look at the driving forces behind reactive species production and
12 provides insight into how to tailor the operation of a plasma jet for different treatments.
13
14
15
16 2. Experimental Setup

us
17 The schematic of an APPJ developed for this work is seen in Figure 1. A ¼ inch plastic
18 Tee compression fitting comprises the main manifold. A 2 mm inner diameter (ID), 3 mm outer
19
20
diameter (OD) quartz capillary tube with one closed end is inserted into the top of the Tee and held
21 with a ¼ inch Teflon tube. A stainless steel rod 0.04 inch (1 mm) in diameter is housed by the
22 quartz tube. Another quartz tube with a 4 mm ID and 6 mm OD protrudes from the bottom of the
23
24
25
26
27
28
an
Tee and out through the grounded box. Through this geometry, the box also serves as the ground
electrode. Helium gas (>99.999% purity) enters the side of the Tee via a 20 slm mass flow
controller and exits between the nested quartz tubes and through the bottom. The stainless steel
rod serves as the powered electrode and connects to a high voltage pulsed dc power system. The
power system is made of a 10 kV dc power supply (Matsusada AU-10P60), a high voltage pulse
dM
29
generator (DEI PVX-4110), and controlled by a digital delay generator (DDG). The voltage is
30
31 supplied to the stainless steel rod electrode via a high voltage coaxial connector (Kings 1765-1)
32 and RG-8 cable. Figure 2 shows a sample I-V curve that demonstrates how power is applied to the
33 system.
34
35
36
37
38
pte

39
40
41
42
43
44
45
46
ce

47
48
49
50
51
52
Ac

53
54
55
56 Figure 1. The atmospheric pressure plasma jet schematic.
57
58 2
59
60
Page 3 of 18 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1

1
2
3
4

pt
5
6
7
8
9

cri
10
11
12
13
14
15
16

us
17
Figure 2. The power is applied to the APPJ as a square voltage wave.
18
19
20
21 2.1 Diagnostics
22 The effluent can be studied using synchronized ICCD imaging to observe the formation of
23
24
25
26
27
28
an
the plasma on a nanosecond timescale. For this research, an Andor iStar 334T ICCD camera was
used to take images of the plasma jet. The camera was synchronized with the trigger signal from
the DDG which drives the pulse generator. A user set variable delay on the DDG allowed the
camera timing to be offset from the voltage pulse.
The emission intensity of the OH (A-X) and N2 second positive transition species (C3Πu–
dM
29 B3Πg) were measured using OES with a 0.5 meter Princeton Instruments SP2500 spectrometer
30
31 coupled to a PI-MAX 4 1024x256 pixel ICCD camera. Images were synchronized with the voltage
32 pulse to isolate behaviors and match with the iStar ICCD images. Both cameras have minimum 3
33 ns gate widths. A 50 mm focal length biconvex lens was placed in between the jet and a fiber optic
34 cable used to capture the signal, as shown in Figure 3. To reach steady-state conditions before each
35
36 experiment, the plasma jet was active for 10 minutes. The OES measurements were taken at four
37 evenly spaced vertical distances from 1 cm inside the tube to 2 cm below the tube exit at time steps
38 corresponding to the bullet imaging. Thus, we were able to spectroscopically observe the passage
pte

39 of the bullet and the appearance and disappearance of OH and N2 emissions at the four locations.
40
41
The 600 g/mm grating was used to average of 10 frames with 1000 CCD accumulations. The
42 spectrometer wavelength was calibrated using Princeton Instrument’s Intellical system.
43
44
45
46
ce

47
48
49
50
51
52
Ac

53
54
55
56
57
58 3
59
60
 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1 Page 4 of 18

1
2
3
4

pt
5
6
7
8
9

cri
10
11
12
13
14
15
16

us
17
18
19
20
21
22
23
24
25
26
27
28
4. Results
an
Figure 3. The optical emissions measurements of the plasma bullet were enhanced by a lens and
captured by a fiber optic cable.

ICCD images captured the formation of plasma bullets at various flow rates, voltages, and
dM
29 pulse widths. The test matrix is shown in Table 1.
30
31
32 Table 1. The conditions of each test are shown. Test Number 1 was the baseline condition.
33
Test # Voltage (kV) Pulse Width (ns) Frequency (kHz) Helium Flow Rate (slm)
34
35 1 8 1000 6 2
36 2 6 1000 6 2
37 3 10 1000 6 2
38 4 8 500 6 2
pte

39 5 8 2000 6 2
40
6 8 1000 2 2
41
42 7 8 1000 10 2

43 8 8 1000 6 1
44 9 8 1000 6 3
45
46
ce

47
48
49
50
51
52
Ac

53
54
55
56
57
58 4
59
60
Page 5 of 18 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1

1
2
3 4.1 Bullet Behavior
4

pt
5
6
7
8
9

cri
10
11
12
13
14
15
16

us
17
18
19
20
21
22
23
24
25
26
27
28
an
dM
29 Figure 4. The plasma forms as bullets that follow the ionization front. The baseline condition (8
30
kV, 1000 ns pulse width, 6 kHz, 2 slm He) is shown. Time after voltage pulse is given in
31
32 nanoseconds above each image.
33
34 As seen in Figure 4, the plasma forms inside the tube before exiting as a bullet. The
35 formation sees nearly linear growth inside the tube until reaching the exit at 392 ns. Upon exiting
36 into open air, the ionization wave experiences a nonlinear decrease in velocity. This was observed
37 for all conditions. The bullet total travel distance (jet length) and the bullet velocity can be
38
determined from these images. First, the total distance traveled by each bullet was found. This was
pte

39
40 then used to find the average velocity of the bullets. The bullet distance was compared to the visual
41 jet length as measured by a ruler. These results are tabulated in Table 2.
42
43
44
45
46
ce

47
48
49
50
51
52
Ac

53
54
55
56
57
58 5
59
60
 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1 Page 6 of 18

1
2
3 Table 2. The average velocity of each bullet was calculated at each operating condition. Error is
4

pt
given as absolute. Each listed condition describes a variation from the baseline of 8 kV, 1000 ns,
5
6 6 kHz, and 2 slm.
7
Average Velocity Measured Distance Measured Error
8 Condition
(m/s)
Bullet Distance (cm)
(cm) (cm)
9

cri
Voltage (kV)
10
11 6 5.05E+04 1.79 1.9 +/- 0.2

12 8 7.39E+04 3.78 3.8 +/- 0.3

13 10 1.05E+05 4.19 4.2 +/- 0.2
14
15 Flow Rate (slm)

16 1 7.90E+04 1.55 1.9 +/- 0.4

us
17 2 7.43E+04 3.78 3.8 +/- 0.3
18 3 7.60E+04 5.00 4.9 +/- 0.3
19
20 Pulse Width (ns)
21 500 9.00E+04 1.68 1.7 +/- 0.5
22
23
24
25
26
27
28
1000
2000
7.39E+04
6.75E+04

an 3.78
3.59
3.8
3.6
+/- 0.3
+/- 0.3

The ruler measured distance was taken from an average of three measurements. The table
shows that each bullet travels the full distance of the observed jet, meaning that the visual effluent
dM
is indeed composed of thousands of individual bullets. The frequency determines the time between
29
30 the bullets, so does not affect each individual bullet. The pulse width also does not noticeably
31 affect the bullet formation but does appear to have a minor effect on both bullet travel distance and
32 velocity. Interestingly, the flow rate affects the overall distance traveled but not in the same manner
33 as the velocity of the bullet. The distance effect is likely tied to the channel of helium in the
34
35 effluent, so a higher flow rate leads to a longer channel. The bulk flow velocity is on the order of
36 1-3 m/s while the ionization wave front is moving at a speed of around 80,000 m/s. This means
37 that on the timescale of the bullet the flow is largely unchanged, and the gas particles can be
38 considered stationary. The flow rate serves simply to provide a channel with a high mole fraction
pte

39
40 of helium which has a lower ionization potential than the surrounding air for the ionization wave
41 (bullet) to propagate through. The voltage makes the most significant change to the bullet motion.
42 Naturally, higher voltages, and thus higher electric fields, lead to higher velocities for the bullet.
43 Higher voltages also increase the gas temperature and thus the tube temperature which can add
44
45
energy to the plasma bullets. In our previous work, we discussed the effects of operating conditions
46 on jet gas temperature[9]. This showed that both voltage and frequency cause small increases in
ce

47 temperature. Varying from 6 to 10 kV increased the temperature by about 12 K, while increasing

48 from 2 to 10 kHz increased the temperature by less than 4 K. The lower flow rate of 1 slm also
49
50
caused an increase in temperature due to a lack of gas cooling. This increase was only around 3 K,
51 but may contribute to the slightly higher velocity at lower flows.
52 Another factor that determines the bullet velocity is the voltage rise time. As shown in
Ac

53 Figure 5, the rise time increases with higher voltages. The rise times were found to be 80 ns for 6
54
kV, 104 ns for 8 kV, and 142 ns for 10 kV. When the voltage increases for a longer time before
55
56 peaking, more energy is deposited into the system. This led to higher velocities for higher voltages.
57
58 6
59
60
Page 7 of 18 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1

1
2
3
4

pt
5
6
7
8
9

cri
10
11
12
13
14
15
16

us
17
18
19
20
21
22
23
24
25
26
27
28
an
dM
29
Figure 5. Current-Voltage curves are shown for (a) 6, (b) 8, and (c) 10 kV.
30
4.2 Synchronized Spectroscopy
31
32 While the bullet behavior trends provide insight into the formation and propagation of the
33 plasma, they provide little knowledge about the RONS in the plasma. For plasma jet applications,
34 the species are typically the primary useful mechanism. The synchronized OES measurements
35
show how OH and excited nitrogen species form in time and space for each condition. The
36
37 emissions were measured from 300-340 nm. OH bands are strongest from 305-311 nm, and N2
38 excited nitrogen bands occur from 311-317 nm and 335-340 nm, as shown in Figure 6. Each OES
pte

39 measurement at a given location and time step produced a separate intensity vs. wavelength plot,
40
which resulted in a lot of spectrums. To make it easier to analyze and display, we calculated the
41
42 total intensity of the species at each spatiotemporal step by taking the integral of the species
43 emissions over their respective wavelength ranges.
44
45
46
ce

47
48
49
50
51
52
Ac

53
54
55
56
57
58 7
59
60
 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1 Page 8 of 18

1
2
3
4

pt
5
6
7
8
9

cri
10
11
12
13
14
15
16

us
17
18 Figure 6. Example OH and N2 bands captured in one spatiotemporal point at 1200 g/mm grating.
19
20
21 4.2.1 Spatial Measurements
22
23
24
25
26
27
28
an
Figure 7 and Figure 8 show the integrated OH and N2 emissions measured at distances
inside and outside the tube from the start of the discharge until the emissions fell below the noise
floor. Negative distances correspond with measurements inside the tube. Spatial measurements
demonstrate how the OH and excited nitrogen species counteract one another. It is clear that OH
is dominant inside the tube, before it can interact with the nitrogen outside the tube. The OH also
dM
29 last much longer inside the tube than outside of it, as seen in the -1 and 0 cm lines in Figure 7. It
30 is notable that OH rapidly dissipates at distances of 1 cm and 2 cm below the tube exit. Nitrogen
31 is a key loss mechanism in OH reactions, as described in our previous paper[9]. As the bullet
32 travels farther through the exterior helium channel, the air begins to mix in and add nitrogen to the
33
34 helium channel. This causes a reduction in OH and an increase in N2*. In Figure 8, the peak
35 production of N2* occurs 1 cm below the tube. At this location, the bullet has had time to interact
36 with the surrounding air and produce excited N2. At 2 cm downstream the bullet has begun to
37 dissipate, and the ionization wave is weaker, thus less able to excite the surrounding nitrogen.
38
A comparison of the bullet propagation from the ICCD images with the species emission
pte

39
40 from OES shows the relative timing. As shown in Figure 8, the bullet (represented by markers)
41 reaches each location (-1, 0, 1, and 2 cm, shown on right axis) 0-10 ns before the emissions start
42 to appear. The peak total emission occurs ~100 ns after the bullet passage, which indicates some
43
44
finite excitation and emission time for the N2 particles. Similar behaviors are observed for OH,
45 though the lower emission intensity outside the tube makes the behavior less clear.
46
ce

47
48
49
50
51
52
Ac

53
54
55
56
57
58 8
59
60
Page 9 of 18 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1

1
2
3
4

pt
5
6
7
8
9

cri
10
11
12
13
14
15
16

us
17
Figure 7. The OH emissions change over space, with most emissions occurring inside the tube.
18
19 Measurements were taken at conditions of 8 kV, 6 kHz, 1000 ns, and 2 slm of helium.
20
21
22
23
24
25
26
27
28
an
dM
29
30
31
32
33
34
35 Figure 8. The timing of excited N2 emission intensity (lines, left y-axis) is shown along with the
36
37
timing of the bullet motion (markers, right y-axis) at the baseline condition (8 kV, 6 kHz, 1000 ns,
38 2 slm of helium). The markers indicate when the bullet reaches the given location, and are color
coded to match the N2* emission curves taken at the same locations.
pte

39
40
41
42 4.2.2 Operating Conditions
43 Each operating condition also produced variations in the species production. A summary
44
45
of the trends at each operating condition can be seen in Figure 9 and Figure 10.
46
ce

47
48
49
50
51
52
Ac

53
54
55
56
57
58 9
59
60
 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1 Page 10 of 18

1
2
3
4

pt
5
6
7
8
9

cri
10
11
12
13
14
15
16

us
17
18
19
20
21
22
23
24
25
26
27
28
an
Figure 9. The N2* emissions vary with operating conditions. In each of the figures, when one
parameter was changed, the other three were held at their baseline values (8 kV, 6 kHz, 1000 ns,
2 slm of helium). The measurements were taken 1 cm below the tube exit.
dM
29
30
31
32
33
34
35
36
37
38
pte

39
40
41
42
43
44
45
46
ce

47
48
49
Figure 10. The OH emissions vary with operating conditions. In each of the figures, when one
50
51 parameter was changed, the other three were held at their baseline values (8 kV, 6 kHz, 1000 ns,
52 2 slm of helium). The measurements were taken 1 cm below the tube exit.
Ac

53
54
55
56
57
58 10
59
60
Page 11 of 18 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1

1
2
3
4

pt
5
6
7
8
9

cri
10
11
12
13
14
15
16 Figure 11. At 2 cm below the tube, an increase in flow rate resulted in an increase in emissions.

us
17
18
N2* emissions are shown for conditions of 8 kV, 6 kHz, and 1000 ns.
19
The OH and N2* emissions followed the same trends, but OH had a much lower magnitude.
20
21 Because of this, more noise is apparent in the OH intensity values. Changes in voltage had the
22 most significant effect on the reactive species. At 1 cm below the tube exit, the 10 kV case
23
24
25
26
27
28
an
produced double the intensity of the 8 kV case, which saw an order of magnitude increase
compared to the 6 kV emissions. As the electric field increases, the velocity does as well. Flow
rate changes produced minor variations in the timing of emissions but did reduce the intensity at
3 slm for distances of -1 cm, 0 cm, and 1 cm below the tube. At 2 cm below the tube, the 3 slm
case showed the highest intensity, as shown in Figure 11. These variations are due to the flow
dM
29
channel expansion caused by higher flows. The plasma bullet begins to dissipate at shorter
30
31 distances for lower flow rates. When the flow rate increases, the helium channel expands and
32 pushes out the air that produces the nitrogen species. This also allows the bullet to travel farther
33 and produce more reactions at a longer distance.
34 The pulse width shows minimal variations between 1000 and 2000 ns, but a significant
35
36 drop is seen at 500 ns. It has been observed that pulse widths below 1000 ns shorten the length of
37 the jet. This is due to the timing of the plasma bullet formation. At these voltages, the bullet will
38 last around 1000 ns after the initial pulse. At shorter pulse widths, the trailing edge of the pulse
pte

39 causes a charge balancing that shortens the lifetime, and thus traveled distance, of the bullet. This
40
41 causes the reduction in species production. The frequency measurements show minimal variation
42 from bullet to bullet, which indicates the produced reactive species dissipate before the next bullet
43 forms. Since the measured intensities are averages of an equal number of individual bullets, the
44 frequency should not affect the measurements. The frequency simply controls the number of
45
46
bullets that occur per second. In application, this would mean that higher frequencies provide more
reactive species per second.
ce

47
48
49 4.2.3 Secondary Ignition
50
51
Another interesting result was observed in the bullet images and the OES measurements
52 inside the tube. On the leading edge of the pulse (shown in Figure 2), the initial plasma formation
Ac

53 begins and leads to the generation of the bullet. However, inside the tube, the trailing edge of the
54 pulse causes a secondary discharge formation and associate secondary OH formation and emission.
55
56
57
58 11
59
60
 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1 Page 12 of 18

1
2
3 This is seen below in Figure 12 as a second peak. Longer pulse widths produced stronger secondary
4

pt
5 emissions, and larger total integrated emissions.
6
7
8
9

cri
10
11
12
13
14
15
16

us
17
18
19
20
Figure 12. A secondary emission of OH occurs at the trailing edge of each pulse and produces a
21 strong emission at the electrode (-3 cm inside the tube). The baseline condition of 8 kV, 6 kHz, and
22 2 slm of helium was held for each pulse width shown.
23
24
25
26
27
28
an
This indicates that a time averaged emission at the electrode is actually a combination of
these two discharges. However, the emission is only produced near the electrode. Figure 7 showed
that even at -1 cm inside the tube, the secondary emission is much weaker and blends into the main
discharge emission. This emphasizes the need to minimize the distance away from the electrode
dM
29 interface when OH is desirable.
30
31 5. Discussion
32
33
34 5.1 Flow Behaviors
35 While the average velocity of the bullets shows the time-averaged effect of each operating
36 condition, it is important to note that the bullet velocity changes over the course of the bullet
37
38
lifetime. As shown in Figure 13, the acceleration changes multiple times as the bullet goes through
each phase of formation.
pte

39
40
41
42
43
44
45
46
ce

47
48
49
50
51
52
Ac

53
54
55
56
57
58 12
59
60
Page 13 of 18 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1

1
2
3
4

pt
5
6
7
8
9

cri
10
11
12
13
14
15
16

us
17
18
19
20 Figure 13. The local bullet velocity changes over the lifetime of the bullet at different flow rates.
21 Each curve begins at the first ICCD frame where the bullet separates from the electrode and ends
22 when the bullet dissipates. The baseline condition of 8 kV, 6 kHz, and 1000 ns was held for each
23
24
25
26
27
28
flow rate.

an
At each flow rate, the plasma discharge rapidly separates from the electrode then drops in
velocity as the bullet begins to form. Then, as the electric field from the pulse continues to drive
the ionization front, the bullet accelerates until it exits the tube. For each flow rate, the bullet exits
dM
29 the tube approximately 100 ns before the peak velocity. The subsequent drop in velocity is likely
30 a combination of the mole fraction of air increasing, therefore decreasing the mole fraction of
31 helium, and the electric field decreasing as the distance from the electrode increases. It has been
32
33
shown that the ionization of helium requires a minimum mole fraction of 0.45-0.5, which is also
34 affected by the turbulence of the jet[6]. At flow rates above 3 slm, the gas flow becomes turbulent
35 downstream. This is also shown in Figure 13. At flow rates of 1 and 2 slm, the bullet travels to the
36 edge of the gas channel and dissipates (around 500 and 800 ns, respectively) without reaching a
37
38
steady velocity. For a 3 slm flow rate, the bullet briefly achieves steady state behavior at 600 ns
before rapidly slowing down at 700 ns. When the flow channel is turbulent, a different behavior is
pte

39
40 observed. Above 4 slm, the bullet reaches a steady velocity around 700 ns after the pulse. It then
41 lasts about 300 ns before losing velocity. This indicates that the air mixture and turbulence
42
influence the bullet.
43
44 Additionally, temperature of the tube may also influence the bullet velocities. Our previous
45 work found that lower flow rates led to slightly higher temperatures (3 K) [9]. This means that as
46 the initial bullet forms inside the tube, lower flow rates would allow more energy to be present.
ce

47
While the change was minor, it may contribute to the increased initial velocity of the bullet shown
48
49 in Figure 13.
50
51 5.2 Emission Trends
52 As discussed, lower flow rates cause a slight increase in velocity until turbulence is
Ac

53
54 reached. The lifetime of each bullet is determined by a combination of three factors: the initial
55 input energy, the time of exposure to this energy, and the local mole fraction of the gas in which
56 it is traveling. As discussed by Karakas, at a given voltage the pulse width controls the overall
57
58 13
59
60
 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1 Page 14 of 18

1
2
3 length of the plasma jet until the helium mole fraction drops below a threshold [6]. The mole
4

pt
5 fraction of the helium plume is controlled by the flow rate but is limited by turbulence. At flow
6 rates above 3 slm, the flow becomes turbulent. This mixes in high amounts of air and causes the
7 mole fraction of helium to drop. With a higher mole fraction of air, the necessary breakdown
8 voltage increases causing the bullet to dissipate. At higher input voltages, this can be overcome.
9

cri
10
Table 3 shows this effect. For voltages of 8 and 10 kV, the jet continues to grow in length despite
11 the high turbulence at 5 and 6 slm. The higher voltage provides more energy to ionize the reduced
12 mole fraction of helium, thus extending the length of the jet.
13
14
15
Table 3. High flow rates cause turbulent behaviors that shorten the bullet distance. Error shown
16 is absolute. Each listed condition describes a variation from the baseline of 8 kV, 1000 ns, 6 kHz,

us
17 and 2 slm.
18
19 Bullet Exit Distance Measured Error
Condition Average Velocity (m/s) Measured Distance (cm)
20 (cm) (cm)
21
22
23
24
25
26
27
28
6 kV, Flow Rate (slm)
4

6
8 kV, Flow Rate (slm)
5.04E+04

4.39E+04

2.61E+04
an 2.26

1.54

0.91
1.9

1.4

0.9
+/- 0.2

+/- 0.3

+/- 0.2
dM
4 7.55E+04 4.85 5.1 +/- 0.3
29
30 5 7.61E+04 5.30 5.7 +/- 0.4
31 6 7.75E+04 5.46 5.5 +/- 0.4
32 10 kV, Flow Rate (slm)
33
4 1.00E+05 6.61 6.9 +/- 0.4
34
35 5 1.02E+05 8.60 8.9 +/- 0.4
36 6 1.02E+05 9.92 10.2 +/- 0.5
37
38
pte

39 While the flow largely minimizes these changes with pulse width, the pulse width does
40 determine the timing of a secondary discharge. The presence of secondary emissions observed in
41
42
Figure 12 is an important behavior to note. While it has been noted previously that a secondary
43 discharge occurs at the trailing edge of each voltage pulse due to charge balancing, the
44 measurement of emissions is novel and can be utilized. Specifically, when trying to increase OH
45 production, the secondary discharge can nearly double the emissions at the electrode for higher
46
pulse widths even if the jet itself is not larger.
ce

47
48
49 5.3 First Bullet Behavior
50 The measurements shown and discussed so far represent imaging of a steady state plasma
51
jet wherein tens of thousands of discharges have occurred. On the timescale of the bullet, the
52
neutral gas can be considered a fixed background. From the ionization wave’s point of view,
Ac

53
54 whether a neutral particle is laminar or turbulent does not matter since all particles appear
55 stationary. That would then suggest the flow rate and turbulence should have minimal effect on
56
57
58 14
59
60
Page 15 of 18 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1

1
2
3 the jet and ionization wave except for helium mole fraction. However, the results and literature
4

pt
5 clearly indicate flow turbulence does affect the bullet both inside and outside the tube. Inside the
6 tube is pure helium, thus there is no mole fraction effect. There thus must be an intermediating
7 effect that acts between the ionization wave and the bulk fluid dynamics.
8 The propagation of the ionization wave into neutral gas, assuming constant applied voltage,
9

cri
10
is due to either photoionization of particles ahead of the wave, or pre-seeded charged particles in
11 the gas formed in the previous wave. Photoionization is unlikely to be affected by fluid dynamics,
12 so the fluid dynamics must affect the distribution of pre-seeded charges. We can check if pre-
13 seeded charges interact with the fluid structure by allowing the gas to fully clear between bullets,
14
15
effectively observing the first bullet.
16 Since the bulk flow travels at a speed of about 2 m/s at a helium flow rate of 2 slm, it takes

us
17 about 30 ms for the flow to fully flush the 6 cm long quartz tube. Thus, the jet was operated at 1
18 Hz frequency, which allows a full second between each bullet and each discharge effectively
19
20
propagates through a new unseeded flow channel. The results are shown below in Figure 14.
21
22
23
24
25
26
27
28
an
dM
29
30
31
32
33
34
35
36
37
38 Figure 14. First bullet images. The bullet behaves differently when the gas channel is cleared
pte

39 between each pulse. The white box represents the quartz tube outside of the grounded box.
40
41 The 392 ns images show that an individual bullet propagating into an un-seeded gas
42 channel will form inside the tube at the same rate, regardless of flow rate. This is different than the
43
high frequency velocity results in Figure 13, where the different flow rates cause noticeably
44
45 different bullet velocities inside the tube, and the lower flow rates caused the bullet to exit the tube
46 sooner. This confirms that each bullet pre-seeds the gas channel building a path for the next
ce

47 discharge until some steady state is reached. Assuming each discharge has the same energy, due
48
to having the same voltage and pulse width, then each bullet should produce approximately the
49
50 same number of pre-seeded charges at all flow rates. If the pre-seeding is confined to the helium
51 gas, then over time the lower flow rates with shorter helium gas channels will have a higher density
52 of charges, and thus stronger electric field for the same applied voltage. This could explain the
Ac

53 higher initial velocities of the lower flow rate bullets in the steady state data of Figure 13. At higher
54
55 flow rates, the pre-seed charges are distributed farther along the helium channel. This causes a
56
57
58 15
59
60
 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1 Page 16 of 18

1
2
3 lower electric field and causes the bullets to move slower through the helium channel but propagate
4

pt
5 farther.
6 Upon exiting the tube in the 642 ns images, the 1 slm bullet begins to shift to the left.
7 Because helium is less dense than air, the 1 slm helium gas channel begins to curl upward. The
8 bullet follows the gas channel, but likely produces some pre-seeded air below the helium channel
9

cri
10
that would remain between bullets. This would increase the local electric field and lead to the
11 higher velocity for the 1 slm bullet observed at the tube exit. The 1 slm bullet then dissipates before
12 the next image at 892 ns due to the small size of the helium channel. The 2 and 3 slm cases result
13 in nearly identical behaviors until each bullet reaches the end of the helium channel at around 900
14
15
and 1000 ns, respectively. This also shows that even the first bullet travels the entire steady state
16 length of the jet. Interestingly, upon exiting the tube, the 6 slm bullet is slower at 642 and 982 ns

us
17 compared to the 2 and 3 slm cases. This suggests the external fluid structure affects even the first
18 bullet when there are no pre-seeded charges. At 6 slm the jet flow is turbulent, which increases the
19
20
mixing with ambient air. This would have the effect of reducing the helium mole fraction even at
21 the jet exit. Since there are no pre-seeded charges, the flow turbulence and mixing are likely
22 affecting the photoionization via a higher average ionization potential due to the increased
23
24
25
26
27
28
presence of air.

V. Conclusion
an
Novel measurements of the time-resolved emissions, thus formation of OH and N2*
reactive species, from an atmospheric pressure plasma jet have been conducted. Voltage has the
dM
29
largest effect on species production at this timescale. Higher voltages lead to more species being
30
31 produced and these species occur faster than at lower voltages. Higher pulse widths extend the jet
32 until the limit determined by flow rate, but the amount of produced species remains similar for all
33 pulse widths. Additionally, the repetition frequency does not affect steady state bullet formations
34 or species productions. This indicates that time-averaged emissions should linearly increase with
35
36 frequency.
37 Helium flow rate had an unexpected effect on the steady state bullet formation inside and
38 outside the tube. Lower flow rate bullets had higher initial velocities and exited the tube sooner.
pte

39 This was theorized to be due to a buildup of pre-seeded charges from previous bullets mediated by
40
41 the fluid dynamic structure. Measurements of the first bullet formation confirmed that with a fresh
42 gas channel, thus no pre-seeded charges, the ionization wave behaves the same inside the tube at
43 different voltages. Outside the tube, the turbulence of the jet and the mixture with ambient air also
44 affect the bullet propagation, indicating that photoionization is also affected. These results of the
45
46 first bullet are a unique physical phenomenon that merit further study.
ce

47 This work provides insight into how to tailor treatments for biomaterials. For example, if
48 the goal is water purification, OH is the key species of interest. The emission and thus presence of
49 OH is strongest inside the tube, and there is also a secondary emission due to the secondary
50
51
discharge on the pulse down stroke that is exclusively inside the tube. This means that the distance
52 between the electrode and the water surface needs to be minimized, and preferably submerged to
Ac

53 prevent OH destruction from nitrogen interaction.

54 Moving forward, more needs to be understood about the interaction between plasma and
55
56
the surfaces of treated materials. Synchronized spectroscopy could also be applied to the plasma-
57
58 16
59
60
Page 17 of 18 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1

1
2
3 material interface to determine if secondary species are created. Modeling these interactions would
4

pt
5 also be important. Expanding the theory of what species are produced, how they are formed, and
6 what the driving forces are behind these reactions will greatly aid the application of APPJ
7 technology. Additionally, the first bullet results on seeding of the channel demonstrate unique
8 physics that could be further explored. Future studies could include analysis of the number and
9

cri
10
frequency of bullets required to sufficiently seed the gas channel to reach a steady discharge.
11 Answering these questions could improve models and increase efficiency of APPJs.
12
13 Acknowledgements
14
15
This material is based upon work supported by the NSF EPSCoR RII-Track-1 Cooperative
16 Agreement OIA-1655280. It is also supported by AL EPSCoR Graduate Research Scholars

us
17 Program Track 14 funding.
18
19 References
20 [1] M. Teschke, J. Kedzierski, E. G. Finantu-Dinu, D. Korzec, and J. Engemann, “High-speed
21 photographs of a dielectric barrier atmospheric pressure plasma jet,” IEEE Trans. Plasma
22
23
24
25
26
27
28
an
Sci., vol. 33, no. 2 I, pp. 310–311, Apr. 2005.
[2] X. Lu and M. Laroussi, “Dynamics of an atmospheric pressure plasma plume generated by
submicrosecond voltage pulses,” J. Appl. Phys., vol. 100, no. 6, 2006.
[3] J. P. Boeuf, L. L. Yang, and L. C. Pitchford, “Dynamics of a guided streamer ('plasma
bullet’) in a helium jet in air at atmospheric pressure,” J. Phys. D. Appl. Phys., vol. 46, no.
1, p. 015201, Jan. 2013.
dM
29 [4] Y. Xian et al., “On Plasma Bullet Behavior,” IEEE Trans. Plasma Sci., vol. 37, no. 10, pp.
30
31
2068–2073, 2009.
32 [5] X. Lu, M. Laroussi, and V. Puech, “On atmospheric-pressure non-equilibrium plasma jets
33 and plasma bullets,” Plasma Sources Sci. Technol., vol. 21, no. 3, p. 034005, Jun. 2012.
34 [6] E. Karakas, M. Koklu, and M. Laroussi, “Correlation between helium mole fraction and
35 plasma bullet propagation in low temperature plasma jets,” J. Phys. D. Appl. Phys., vol. 43,
36 no. 15, p. 155202, 2010.
37 [7] X. Lu, G. V. Naidis, M. Laroussi, and K. Ostrikov, “Guided ionization waves: Theory and
38
experiments,” Phys. Rep., vol. 540, no. 3, pp. 123–166, Jul. 2014.
pte

39
40 [8] N. Mericam-Bourdet, M. Laroussi, A. Begum, and E. Karakas, “Experimental
41 investigations of plasma bullets,” J. Phys. D. Appl. Phys., vol. 42, no. 5, 2009.
42 [9] R. P. Gott and K. G. Xu, “OH Production and Jet Length of an Atmospheric-Pressure
43 Plasma Jet for Soft and Biomaterial Treatment,” IEEE Trans. Plasma Sci., vol. 47, no. 11,
44 pp. 4988–4999, 2019.
45
[10] S. Bornholdt, M. Wolter, and H. Kersten, “Characterization of an atmospheric pressure
46
plasma jet for surface modification and thin film deposition,” Eur. Phys. J. D, vol. 60, no.
ce

47
48 3, pp. 653–660, Dec. 2010.
49 [11] G. D. Stancu, F. Kaddouri, D. A. Lacoste, and C. O. Laux, “Atmospheric pressure plasma
50 diagnostics by OES, CRDS and TALIF,” J. Phys. D. Appl. Phys., vol. 43, no. 12, p. 10,
51 2010.
52 [12] E. Karakas and M. Laroussi, “Experimental studies on the plasma bullet propagation and
Ac

53
54
its inhibition,” J. Appl. Phys., vol. 108, no. 6, Sep. 2010.
55 [13] Ferric C. Fang, “Antimicrobial reactive oxygen and nitrogen species: concepts and
56 controversies,” Nat. Rev. Microbiol., vol. 2, no. 10, pp. 820–832, 2004.
57
58 17
59
60
 AUTHOR SUBMITTED MANUSCRIPT - JPhysD-123606.R1 Page 18 of 18

1
2
3 [14] J. Foster, B. S. Sommers, S. N. Gucker, I. M. Blankson, and G. Adamovsky, “Prespectives
4
on the Interaction of Plasmas With Liquid Water for Water Purification.pdf,” IEEE Trans.

pt
5
6 Plasma Sci., vol. 40, no. 5, pp. 1311–1323, 2012.
7 [15] J. E. Foster, G. Adamovsky, S. N. Gucker, and I. M. Blankson, “A comparative study of the
8 time-resolved decomposition of methylene blue dye under the action of a nanosecond
9 repetitively pulsed dbd plasma jet using liquid chromatography and spectrophotometry,”

cri
10 IEEE Trans. Plasma Sci., vol. 41, no. 3, pp. 503–512, 2013.
11 [16] D. Ziuzina, S. Patil, P. J. Cullen, K. M. Keener, and P. Bourke, “Atmospheric cold plasma
12
13
inactivation of Escherichia coli in liquid media inside a sealed package,” J. Appl.
14 Microbiol., vol. 114, no. 3, pp. 778–787, 2013.
15 [17] J. Heinlin et al., “Plasma-Medizin: Anwendungsmöglichkeiten in der Dermatologie,”
16 JDDG - J. Ger. Soc. Dermatology, vol. 8, no. 12, pp. 968–977, Dec. 2010.

us
17 [18] K. D. Weltmann et al., “Antimicrobial treatment of heat sensitive products by miniaturized
18 atmospheric pressure plasma jets (APPJs),” J. Phys. D. Appl. Phys., vol. 41, no. 19, p.
19
194008, Oct. 2008.
20
21 [19] A. Van Deynse, P. Cools, C. Leys, R. Morent, and N. De Geyter, “Surface modification of
22 polyethylene in an argon atmospheric pressure plasma jet,” Surf. Coatings Technol., vol.
23
24
25
26
27
28
276, pp. 384–390, Aug. 2015.

an
dM
29
30
31
32
33
34
35
36
37
38
pte

39
40
41
42
43
44
45
46
ce

47
48
49
50
51
52
Ac

53
54
55
56
57
58 18
59
60