INSTITUTE OF PHYSICS PUBLISHING
PLASMA SOURCES SCIENCE AND TECHNOLOGY
Plasma Sources Sci. Technol. 15 (2006) 501–506
doi:10.1088/0963-0252/15/3/028
A plasma needle generates nitric oxide∗
E Stoffels1 , Y Aranda Gonzalvo2 , T D Whitmore2 , D L Seymour2
and J A Rees2
1
Department of Biomedical Engineering, Eindhoven University of Technology, PO Box 513,
5600 MB Eindhoven, The Netherlands
2
Plasma&Surface Analysis Division, Hiden Analytical Ltd., 420 Europa Boulevard,
Warrington, WA5 7UN, UK
E-mail: e.stoffels.adamowicz@tue.nl
Received 10 November 2005, in final form 11 November 2005
Published 5 June 2006
Online at stacks.iop.org/PSST/15/501
Abstract
Generation of nitric oxide (NO) by a plasma needle is studied by means of
mass spectrometry. The plasma needle is an atmospheric glow generated by
a radio-frequency excitation in a mixture of helium and air. This source is
used for the treatment of living tissues, and nitric oxide may be one of the
most important active agents in plasma therapy. Efficient NO generation is
of particular importance in the treatment of cardiovascular diseases.
Mass spectrometric measurements have been performed under various
plasma conditions; gas composition in the plasma and conversion of feed
gases (nitrogen and oxygen) into other species has been studied. Up to 30%
of the N2 and O2 input is consumed in the discharge, and NO has been
identified as the main conversion product.
(Some figures in this article are in colour only in the electronic version)
1. Introduction
This work is focused on the plasma chemistry of a
cold atmospheric gas discharge for biomedical applications.
Interactions of cold plasmas with living mammalian cells,
tissues and bacteria are thoroughly investigated by the
biomedical plasma group at the Eindhoven University. A
suitable plasma source (plasma needle) has been developed [1],
characterized [2] and tested on various biological materials
[3–6]. The plasma is an atmospheric radio-frequency glow,
generated at 13.56 MHz at the tip of a sharp metal needle.
Plasma needle operates at room temperature and low voltages,
so it can be applied directly to the tissue without thermal and
electrical damage.
The intended medical application of the plasma needle
is minimum-destructive, precise removal or modification of
tissues as well as non-contact disinfection of wounds and
cleaning of dental cavities. In vitro studies on cells and artery
sections have shown that plasma can remove cells without
necrosis (accidental cell death) [3, 4], which means that no
inflammation and excessive tissue injury will occur. This also
implies that post-operative healing will be optimal and no scars
∗
This paper was submitted as part of the special section featuring papers from
the 27th International Conference on Phenomena in Ionized Gases. To view
the special section go to stacks.iop.org/PSST/15/i=2.
0963-0252/06/030501+06$30.00
© 2006 IOP Publishing Ltd
will be formed. Furthermore, pilot studies have demonstrated
the applicability of the plasma in dentistry: bacteria can be
efficiently eliminated at very low power settings, under which
dental tissue remains unharmed [5, 6]. Plasma treatment will
provide a painless and tissue-saving method for treatment of
caries and for improving oral hygiene.
Plasma acts as a well-controllable source of chemical
energy that can be deposited on the surface. There are
numerous indications that active plasma radicals play an
important role in inducing specific reactions of living cells
and bacteria. The same radicals are also produced in the
organism during certain activities, such as fighting infections
and repairing damaged tissues; the concentrations of natural
(physiological) and plasma-generated species are comparable.
Thus, the hypothesis is that the plasma can operate in a closeto-natural way, by locally reinforcing the chemical activity of
the body.
Generation of reactive oxygen species (O, OH, etc) by
the plasma needle has been demonstrated previously [2]; these
particles are mainly responsible for bacterial deactivation and
for inducing programmed cell death (apoptosis). In this work
attention is given to nitric oxide (NO). Nitric oxide plays
a very important role in the organism. It participates in
many physiological processes, such as signal transmission
between nerves (neurotransmitter), suppression of infections
Printed in the UK
501
E Stoffels et al
Figure 1. (a) A picture of the plasma needle positioned against the sampling orifice of the HIDEN QMS system; the plasma is ignited and
visible at the tip of the needle. (b) The plasma needle in a catheter (flexible plastic tube) immersed in a liquid medium.
and inflammations, tissue renewal and many others. It
also regulates the functioning of blood vessels, causing
vasodilation (relaxing the vessel and improving circulation)
and protecting the artery wall against cholesterol intake. In a
healthy cardiovascular system, NO is produced by endothelial
cells; impaired endothelial function and NO deficiency lead
to formation of atherosclerotic lesions [7]. Treatment of
cardiovascular diseases is often based on the administration of
systemic NO-releasing drugs (e.g. nitroglycerin). However,
the common problem with systemic drugs is that they do
not address only the target area, but the whole organism,
and they are usually not free from side effects. If plasma
needle proves to be an efficient NO releaser, it can be applied
topically to cure an atherosclerotic lesion. There are of course
technical difficulties in administering plasma internally—the
device must be adapted for catheterization. However, the proof
of principle has been recently demonstrated—a functioning
plasma catheter has been constructed and gas–fluid separation
has been made feasible. Besides cardiovascular operations,
NO produced by the plasma needle can be used for other
therapies, where local improvement of blood circulation is
necessary.
Gas-phase NO formation has been extensively studied in
relation to atmospheric pollution [8–10]. Direct formation
from N2 and O2 has activation energy of about 300 kJ mol−1
(3 eV) and requires gas heating to 2000–3000 K [10].
However, such an energetic barrier can be easily overcome at
a much lower gas temperature, owing to special properties of a
non-equilibrium gas discharge. Non-equilibrium atmospheric
plasmas contain electrons with temperatures on the order of
1 eV [11], which provide the necessary energy to initiate
gas conversion at close-to-room temperature. This gives
confidence that the cold plasma needle will generate substantial
amounts of nitric oxide. In this work we present mass
spectrometric studies on gas conversion in the plasma
generated by the needle. Neutral chemistry, nitric oxide
production and gas temperature effects are discussed.
2. Experiment
2.1. Plasma needle
Measurements have been performed in a configuration similar
to the prototype plasma needle [1]. Atmospheric plasma is
generated by applying radio-frequency voltage to a tungsten
502
wire (50 mm long, 1 mm diameter) inserted in a quartz tube.
Frequency generator (fixed 13.56 MHz frequency), matching
network and power meter are home-built. The plasma is
ignited in a gas mixture, directed through the tube at a flow
rate of about 1 litre min−1 . Gas composition is determined
from mass spectra; the feed gas contains 15% He, 12% O2
and 73% N2 . The thickness of the wire electrode is important
for the plasma properties: thin needles produce sub-millimetre
glows with 0.1–0.5 W power input, while thicker electrodes
result in larger discharges with a power consumption of several
watts. However, in all the cases the voltage, power density
and optical properties of the plasma are about the same. The
plasma source used in the current experiment can be described
as a ‘robust’ version of the plasma needle: it generates an
elongated jet of 4 mm length, and the power consumption is
from 1 to 10 W. A picture of the plasma is shown in figure 1(a).
Naturally, this type of plasma needle is only suitable for
mass spectrometric diagnostics and not for medical treatment.
For the latter application, a specially adapted catheter has
been constructed. This device is supplied with a 0.3 mm
thick flexible electrode inserted in a flexible (plastic) catheter.
Gas is supplied through the catheter and removed through
an additional tube placed parallel to the wire electrode. In
this configuration the gas flow can be maintained so low that
no bubbles are produced when the catheter head is placed
under liquid. It makes gas-liquid separation possible, which
consequently enables internal treatment. A picture of such an
adapted device is shown in figure 1(b).
2.2. Mass spectrometry
The tip of the wire electrode is positioned against the orifice
of a HIDEN EQP mass/energy analyser system (figure 2), at
the (variable) distance of several millimetres. Species created
in the discharge are sampled using a triple stage differentially
pumped molecular beam inlet system. With this arrangement
it is possible to sample particles from an atmospheric discharge
and analyse them with a sensitivity of 0.1 mPa partial pressure.
The mass spectrometer can be operated in the SIMS mode
to extract positive ions and record their energy spectra or in
the RGA mode to detect neutrals. The mass spectrometer
is equipped with an internal electron source with variable
electron energy, which allows ionization of species (positive
RGA) as well as electron attachment (negative RGA or EAMS,
electron attachment mass spectrometry). In this work mainly
Plasma needle generates nitric oxide
Electrical and Water
Services
Atm. RF Plasma
Source or
Ozone Generator
Mass Spectrometer
Ion Source
P1,P2 and P3 = 1st, 2nd and
3rd stage pressure reduction
Mass and Energy
Analysis
Scale
25mm
1 inch
P3
P1
P2
Figure 2. A scheme of the molecular beam mass spectrometer analyser.
DENSITY FRACTION
0.2
1.5 mm
2.5 mm
3.5 mm
0.15
He
0.1
0.05
0
Figure 3. An example of the mass spectrum of the plasma, at 5 W,
1.5 mm from the needle.
time series are performed: several mass channels (4 amu—He,
28 amu—N2 , 32 amu—O2 and 30 amu—NO) are monitored
at 70 eV electron energy, and signal responses to varying
plasma parameters are determined. Ambient gas composition
is determined from mass spectra at 70 eV. An example of the
mass spectrum taken in the plasma is shown in figure 3.
3. Results and discussion
3.1. Gas temperature
The RGA mode allows sampling neutral gas from the plasma
and its surroundings. The initial composition of the gas flow
from the needle tube (without plasma) has been determined
to be 15% He, 12% O2 and 73% N2 . The first observation is
2
4
6
POWER (W)
8
10
Figure 4. Helium density in the plasma as a function of input power
at various distances between the needle and the QMS. The unit is the
fraction of the ambient atmospheric density (1 = 2.69 × 1025 m−3 ).
that the signals of these species decrease when the plasma is
ignited. Typical behaviour of He is shown in figure 4. The
effect is larger at high plasma powers, and becomes smaller
when the distance between the needle and the QMS increases
(i.e. in the downstream region of the plasma). Such behaviour
can be easily explained by gas heating. Assuming constant
pressure p, the relation p = nkB T , where n is the density, T the
temperature and kB the Boltzmann constant, directly provides
the gas temperature in the plasma. The relative decrease in
the helium signal has been used to scale down the density; this
is because helium does not undergo any significant chemical
loss in the plasma and thus it reflects the total gas density.
Temperature has been determined in dependence of plasma
503
E Stoffels et al
0.2
DENSITY FRACTION
o
TEMPERATURE ( C)
150
120
90
60
1.5 mm
2.5 mm
3.5 mm
30
0
0
2
4
6
8
10
POWER (W)
Figure 5. Gas temperature as a function of input power at various
positions.
1.5 mm
2.5 mm
3.5 mm
0.15
O2
0.1
0.05
0
2
4
6
POWER (W)
0
Figure 7. Oxygen density in the plasma as a function of input power
at various distances between the needle and the QMS. The unit is the
fraction of the ambient atmospheric density.
MODE CHANGE
DENSITY FRACTION
0.8
N2
0.6
1.5 mm
2.5 mm
3.5 mm
0.4
0.2
0
2
4
6
POWER (W)
8
10
Figure 6. Nitrogen density in the plasma as a function of input
power at various distances between the needle and the QMS. The
unit is the fraction of the ambient atmospheric density. Arrows
indicate the threshold powers for mode transition.
power and needle to QMS position, as shown in figure 5.
The temperature in the plasma can rise up to 150 ◦ C. These
results are in very good agreement with previous measurements
involving optical emission spectroscopy [1]. For medical
applications, such a temperature is generally too high, unless
tissue de-vitalization is desired. However, this problem can be
solved by operating downstream, increasing gas flow rate (only
in external operations), reducing plasma power (using thinner
electrodes) and shortening treatment times (e.g. pulsing the
plasma).
3.2. Gas conversion
The signals of nitrogen and oxygen, at the same settings as
helium signals (figure 4), are displayed in figures 6 and 7.
The depletion of N2 and O2 is much more drastic than the
total density decrease due to thermal effects. This implies
that molecular gases undergo significant chemical conversion
in the plasma. Typical ‘threshold behaviour’ can be seen in
nitrogen and oxygen signals: depletion begins above a certain
power level, which is dependent on the distance from needle
to orifice. This correlates with a mode change of the plasma.
504
Below the threshold the plasma is sustained at the tip of the
needle (unipolar mode) and does not make contact with the
surface. Above the threshold, the surface starts to act as a
grounded electrode. This transition can be easily observed
visually: the glow extends towards the grounded object and a
thin sheath at the object’s surface is formed.
While oxygen and nitrogen signals decrease, other
molecular species emerge. Note that N2 and O2 depletion
is not stoichiometric for any specific conversion product. The
fractions of removed nitrogen and oxygen are about the same,
which means that absolute density reduction for N2 is higher
than for O2 . However, in an open system the total amount of
molecules is not conserved—diffusion from the surrounding
ambient air would smooth the density gradients if the depletion
of one species was too drastic.
NO signal is the most prominent feature of the mass
spectrum. Other possible conversion product is N2 O (mass 44),
but its signal coincides with the persistent CO2 peak, which
makes quantitative analysis difficult. Furthermore, in the
downstream zone, at a distance of 3.5 mm from the needle,
some NO2 is present. However, this is visible only at high
power levels (9–10 W) and the signal is 20 times smaller
than the NO signal. Surprisingly, no ozone has been found.
From the medical point of view, this is good news—ozone
and (especially) NO2 are toxic and their concentrations should
be minimized. In contrast, nitrous oxide (N2 O) is harmless
and nitric oxide is the desired species that determines the
effectiveness of plasma therapy. These are the most probable
stable products of gas conversion. Besides, quite significant
quantities of shorter living species are present, resulting from
the dissociation of N2 and O2 . Threshold ionization data
indicate that densities of atomic nitrogen and oxygen are in the
range of sub-per cent of the total gas density (about 1022 m−3 ).
The parametric dependence of the NO signal is shown in
figure 8. It can be seen that NO reaches its maximum density
not in the closest vicinity of the needle, but at a distance of
2.5 mm. Only at very high power levels the peak NO density
shifts back towards the needle. This behaviour is determined
by the reaction kinetics. The exact reaction mechanism in
the plasma is unknown; however, it is expected that it will
take a few milliseconds to synthesize a NO molecule. For
example, if NO is formed by dissociation of N2 and O2 and
Plasma needle generates nitric oxide
0.25
NO
1.2
4
COUNTS (10 )
DENSITY FRACTION
1.5
1.5 mm
2.5 mm
3.5 mm
0.9
0.6
0.3
0
0
2
4
6
POWER (W)
0
Figure 8. Nitric oxide signals (NO+ counts in positive RGA at
70 eV) as a function of input power at various distances between the
needle and the QMS.
DENSITY FRACTION
0.3
product
0.2
1.5 mm
2.5 mm
3.5 mm
0.1
0
0
2
4
6
POWER (W)
8
10
Figure 9. Density of conversion products of the feed gases (product
gas density) as a function of input power at various distances
between the needle and the QMS. The unit is the fraction of the
ambient atmospheric density.
subsequent three-body recombination N + O + X → NO + X,
the latter step is quite time-consuming. The recombination
rate (k) at close to ambient temperature is about 10−45 m6 s−1
(http://kinetics.nist.gov), so the production term becomes
P = kn0 [N] [O] ≈ 2.7 × 10−20 m3 s−1 [N] [O]
for the ambient density n0 of 1 atm (2.69×1025 m−3 ). Another
possibility is the reaction N + O2 → NO + O, which has a
rate of 1.4 × 10−21 m3 s−1 at 400 K [10]. Taking an estimated
atomic density of 1022 m−3 , the time constant of the threebody recombination is about 4 ms. Since the gas speed at
1 litre min−1 in the needle configuration is about 0.5 m s−1
[2], the peak NO density should be reached few millimetres
away from the place in which electron-impact dissociation is
most efficient (rf-sheath or the direct vicinity of the needle
tip). Alternatively, one can argue that efficient NO production
requires extra influx of oxygen from the surroundings. This
would also result in peak NO density shifted away from the
gas-supplying needle tube. This is in fact a likely explanation
because the initial concentration of oxygen in the feed gas is
too low to account for such high NO densities.
Subtraction of remaining N2 and O2 densities (figures 6
and 7) from the total gas density yields a good estimate of
product
[NO] scaled
0.2
0.15
0.1
0.05
0
0
2
4
6
POWER (W)
8
10
Figure 10. Scaled NO density and product gas density in fractions
of ambient atmospheric density, at 1.5 mm from the needle tip, as a
function of plasma power.
the total amount of reaction products. This ‘product gas
density’ is shown in figure 9. One can immediately note that
its behaviour is very similar to this of NO density (figure 8).
To make a comparison easier, two scaled signals are plotted
in figure 10. The similarity between these curves gives a
very strong indication that NO is the major reaction product
of plasma reactions at distances of up to 2.5 mm from the
needle tip. The density of NO may thus reach up to 20%
of the ambient density. This result is supported by the
comparison of NO+ count rates with the count rates of N+
from N2 . The respective cross-sections for the ionization of
NO and for the dissociative ionization of N2 are known (see
http://physics.nist.gov/PhysRefData/Ionization and the data
provided by Cook and Peterson [12]). By comparing the count
rates one can conclude that the density ratio of NO to N2 is
about 0.3. This implies that the NO contribution to the total
density is in the range 10–20%.
In the downstream zone (at 3.5 mm position) the NO
density is clearly lower (see figure 8). However, the product gas
density is higher than at 1.5 and 2.5 mm. Besides nitric oxide,
other reaction products must be present in this region. The
product gases may contain N2 O and also some NO2 due to NO
oxidation. It is not completely clear what causes such efficient
gas conversion in the downstream region. However, one can
expect that long-living energetic plasma species (e.g. helium
metastable atoms) are present there, so they can dissociate the
ambient gases.
These ‘downstream’ conditions (long distances from the
needle) and high powers are not very interesting for medical
plasma treatment. As determined in tests on cultured cells,
the optimal performance is achieved at distances shorter than
2 mm [3–6]. Besides, the power should be kept as low as
possible to avoid gas heating. Since NO density increases with
increasing power, in a practical situation, compromises will
have to be made. On the positive note, large amounts of nitric
oxide are not necessary—the physiological concentration is
less than 1 µM [13], and in a therapy, this value may be
increased at most 10 times. Such a concentration of active
species in a 0.4 ml liquid sample can be induced by 0.5–1 min
of plasma treatment, when the gas-phase radical density is only
1019 m−3 [14]. Production of NO by the ‘robust’ plasma needle
is in fact too efficient—in practice, there will be always enough
505
E Stoffels et al
NO to achieve the desired treatment results. This leaves much
freedom to adjust other plasma parameters, to end up with a
safe, fast and efficient medical procedure.
4. Conclusion
Atmospheric plasma can be an efficient source of nitric
oxide. The studied case is a small-size radio-frequency driven
glow (plasma needle), which may be used for local medical
treatment. Mass spectrometric measurements show significant
conversion of feed gases (nitrogen and oxygen), and NO is
found to be the dominant reaction product. Its density can
reach up to 20% of the total density. Furthermore, thermal
effects of the plasma have been studied and, dependent on
conditions, gas heating up to 150 ◦ C has been observed.
This is in agreement with previous results obtained using a
spectroscopic method.
References
[1] Stoffels E, Flikweert A J, Stoffels W W and Kroesen G M W
2002 Plasma Sources Sci. Technol. 11 383
506
[2] Kieft I E, van der Laan E P and Stoffels E 2004 New J. Phys.
6 149
[3] Kieft I E, Broers J L V, Caubet-Hilloutou V, Slaaf D W,
Ramaekers F C S and Stoffels E 2004 Bioelectromagnetics
25 362
[4] Kieft I E, Darios D, Roks A J M and Stoffels E 2005 IEEE
Trans. Plasma Sci. 33 771
[5] Sladek R E J, Stoffels E Walraven R, Tielbeek P J A and
Koolhoven R A 2004 IEEE Trans. Plasma Sci. 32 1540
[6] Sladek R E J and Stoffels E 2005 J. Phys. D: Appl. Phys. 38
1716–21
[7] Rekka E A and Chrysselis M C 2002 Mini Rev. Med. Chem. 2
585–93
[8] Raine R R, Stone C R and Gould J 1995 Combust. Flame 102
241
[9] Miller R, Davis G, Lavoie G, Newman C and Gardner T 1998
SAE Paper No 980781
[10] Hewson J C and Williams F A 1999 Combust. Flame 117
441
[11] Brok W J M, Bowden M D, Van Dijk J, Van der Mullen J J A M
and Kroesen G M W 2005 J. Appl. Phys. 98 013302
[12] Cook C J and Peterson J R 1962 Phys. Rev. Lett. 9 164
[13] Hirota Y, Ishida H, Genka C, Obama R, Matsuyama S and
Nakazawa H 2001 Japan. J. Phys. 51 455
[14] Kieft I E, van Berkel J J B N, Kieft E R and Stoffels E 2005
Plasma Processes and Polymers ed P Favia and R
d’Agostino (Weinheim: Wiley VCH) p 295