22nd International Symposium on Plasma Chemistry
July 5-10, 2015; Antwerp, Belgium
Plasma dissociation of water for CO 2 conversion
T.T. Belete, M.C.M. van de Sanden and M.A. Gleeson
Dutch Institute for Fundamental Energy Research (DIFFER), De Zaale 20, NL-5612 AJ, Eindhoven, the Netherlands
Abstract: Direct utilization of H 2 O and CO 2 for the production of syngas (H 2 + CO) or
hydrocarbons is appealing from a sustainability perspective. It can be carbon neutral when
driven by renewable energy and coupled to processes that capture CO 2 from the air. We
present preliminary results of plasma-induced dissociation of H 2 O in inductively-coupled
plasma (ICP) as a preliminary step on the way to an integrated conversion process.
Keywords: H 2 O, plasma-induced dissociation, CO 2 conversion, syngas, solar fuels
Introduction
Production of syngas (H 2 + CO) or hydrocarbons from
H 2 O and CO 2 using a thermo-chemical cycle offers a
path for converting renewable energy (solar, wind) into a
storable form. Utilization of H 2 O and CO 2 for fuel
production as a storage solution is the most appealing
option from a sustainability perspective. It can be a
carbon neutral when coupled to processes that capture
CO 2 from the air.
Water dissociation by plasma for hydrogen production
involves conversion of electricity into chemical energy.
Other approaches such as electrolysis, thermo-chemical
and thermo-catalytic cycles, radiolysis [1], and thermal
decomposition [2] are possible. However, there are
drawbacks associated with each. For instance, the
complexity of reaction kinetics, phase separation and
corrosion phenomena in electrolysis; and high
temperature around 3000 K required to produce radicals
and molecules - OH, H 2 , O 2 , H and O - during thermal
decomposition [2].
Fridman [1] has reported direct dissociation of water in
plasma can be more efficient than aforementioned
processes for hydrogen production, according to equation
(1) with high specific productivity.
1
𝐻2 𝑂 → 𝐻2 + 𝑂2
∆𝐻 = 2.6 𝑒𝑒/𝑚𝑚𝑚
(1)
2
It also potentially allows for in-situ production of
reactive hydrogen intermediates that are essential in any
CO 2 -based hydrocarbon production process. Since plasma
can also be used to activate CO 2 molecules, the approach
has the potential to allow for development of an
integrated reaction process.
Quite apart from its potential as a hydrogen source, the
presence of water has been demonstrated to have a
beneficial influence on the performance of several CO 2 capture materials. The extent to which this influence can
be modified by plasma-activation of the water molecules
remains an open question.
This work presents preliminary results from an
experimental setup that is designed for investigation of
plasma-activation of H 2 O and CO 2 both in the gas phase
and in conjunction with relevant solid CO 2 -capture
materials. It represents part of the commissioning and
P-II-8-1
validation of a new experimental setup. Plasma-induced
activation and dissociation of H 2 O is achieved by means
of inductively-coupled plasma (ICP). The reactor
performance is monitored in real time by means of Mass
Spectrometry, Optical Emission Spectroscopy and total
pressure response measurements. In case of plasmamaterial interaction studies, pre- and post-exposure
analysis is available ex-situ. This work is part of longer
term efforts to develop an integrated plasma process for
CO 2 conversion.
Experimental procedure
The experimental setup used is depicted in Figure1. It
consist of three different sections with different pressure
ranges. The first section had a Quadrupole Mass
Spectrometer (QMS) attached in a high vacuum region
(~10-9 mbar base pressure), separated from a second
section by 3 mm aperture. The base pressure in the second
section is typically ~10-8 mbar to ~10-7 mbar. A valve
separates these two high vacuum sections from the third
(medium-low vacuum: ~10-3 mbar-10 mbar working
pressure range) region. A 1 mm aperture is mounted
before the valve, to limit the pressure rise in the QMS
region during system operation. A heating stage is
mounted in this section for resistive heating target
material. The introduction of water vapour and plasma
generation is via a quartz tube of 30 cm length and 2 cm
in
diameter
attached
Fig. 1. Schematic of setup for dissociation of water by
plasma.
1
to this section. An Optical Emission Spectrometer (OES)
is connected via fiber optic cable to measure plasma light
(200-1100 nm) through the top of the quartz tube.
Figure 2 illustrates the two possible configurations of
the quartz tube: either horizontally for direct sampling of
the plasma species, or vertically for plasma-exposure of a
material sample that can be heated independently using a
resistive heating plate. For the current experiments the
former configuration was used.
generate an initial (no plasma) 0.55 mbar pressure for
different input powers. The dominant species produced
are H 2 and O 2 .
The generation of H 2 and O 2 from water is more clearly
illustrated in Figure 4 which shows both increasing with
input power. The ratio of O 2 to H 2 varies somewhat at
low power, but becomes effectively constant at higher
values.
0
Power (W)
50 100 150
200
250
275
0.33
0.22
H2/H2O
0.11
0.00
0.54
Fig. 2. Orientation of the quartz tube for direct plasma
sampling and for material exposure.
Deionised water is subjected to several freeze-pumpthaw cycles before use. Water vapour is introduced via a
needle valve to achieve a target pressure in the quartz
tube. To ignite ICP inside the tube, RF power is applied to
the load coil and an alternating current oscillates inside
the coil at a rate corresponding to the frequency (13.56
MHz) of the RF generator. Dissociation of water as a
function of input power and tube pressure can be
monitored by the QMS. During water dissociation, the
input powers typically ranging from 50 W to 275 W from
the RF generator were established while minimizing the
reflected power to zero by adjusting the capacitances of
matching box.
Results
QMS signals of various species, such as H 2 O, H 2 , O 2 ,
and OH, were monitored as a function of input power. For
example, Figure 3 shows the water plasma response
0.36
O2/H2O
0.18
0.00
1.5
1.0
O2/H2
0.5
0.0
50 100 150 200 250 300 350 400 450
Time(s)
Fig. 4. Ratios of different species as a function of input
power. Initial H 2 O pressure ~0.55 mbar.
Data for dissociation of water in ICP plasma at different
pressures as a function of input power is depicted in
Figure 5. Water dissociation decreased with increasing
pressure in quartz tube. The dissociated fraction appears
to stabilize at ~69 %.
1.1
2.5x10
H2O
O2
Counts
2.0x10-9
OH
50w
100w
1.5x10-9
150w
1.0x10-9
200w 250w
275w
-10
5.0x10
H2O dissociation
0.44 mbar
0.55 mbar
0.60 mbar
0.64 mbar
0.80 mbar
0.90 mbar
1.08 mbar
1.0
Normalized count (10-9)
0w
-9
H2O Plasma at 0.55 mbar
H2
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0
0.0
50
100
150
200
250
300
Input Power (W)
0
50 100 150 200 250 300 350 400 450
Time(s)
Fig. 5. Normalized counts for dissociation of water as a
function of input power for different initial pressures.
Fig. 3. QMS signal for different species at different input
power.
2
P-II-8-1
Additional OES investigation of 100 W water plasma at
0.64 mbar and 0.91 mbar initial pressure shows less
excitation of the hydrogen and oxygen lines at the higher
pressure [Figure 7 (a-b)]. The intensity of Hα also
decreased for the higher pressure (not shown).
1000
500
Hγ
0
250
300
350
400
450
500
550
b)
50 w
100 w
150w
200w
50
200
Intensity(a.u)
OH =307.3 nm
Intensity (a.u)
100
Hb
OH
200
a)150
[email protected]
[email protected]
a)
Intensity(a.u)
In addition to QMS measurements, OES studies,
employing Ocean Optics HR 4000 Spectrometer, have
been carried out by to monitor changes of plasma species.
The emission spectra of the plasmas were monitored in
the range 200-1100 nm. The important lines for water
emission were oxygen radicals at 775.14 nm and
842.5 nm, OH radicals at 307.3 nm and Balmer series
(H γ =432.16 nm, H β =484.2 nm and H α =654.15 nm) as
plotted in Figure 6 (a-d) for 0.29 mbar pressure and input
power from 50 W to 200 W. Those values are about
~2 nm less than the values given in [2, 3, 4, 5]. The
intensities of the emissions increase with increasing input
power, consistent with the increased plasma dissociation
of water at higher input powers.
[email protected]
[email protected]
150
O
100
O
50
0
760
0
β)
280
290
300
310
320
330
340
1000
500
50 w
100 w
150w
200w
Hβ=482.2 nm
Hγ=432.16 nm
Intensity (a.u)
1500
350
440
460
480
500
wavelength (nm)
Hα=654.15 nm
c)
14000
Intensity (α.u)
12000
10000
8000
800
820
840
860
wavelenght (nm)
Fig. 7. Emission spectra of water plasma at 100 W input
power for pressure of 0.64 mbar and 0.91 mbar in quartz
tube.
Summary
Initial measurements from a new plasma setup intended
for the study of plasma-induced H 2 O and CO 2 have been
presented. The system is intended to track the evolution of
species during plasma-phase and plasma-material
interactions. The measurement capabilities have been
applied to track the power and pressure dependencies of a
water plasma.
0
420
780
50 w
100 w
150w
200w
Acknowledgement
We acknowledge gratefully Mr. Waldo Bongers and
Mr. Bobby van Hemert for their support and advice in the
provision of the ICP coil setup.
6000
4000
2000
0
645
660
600
400
300
200
100
665
50 w
100 w
150w
200w
O=842.5 nm
500
Intensity (α.u)
655
O=775.14 nm
d)
650
0
760
780
800
820
840
860
wαvelength (nm)
Fig. 6. Emission spectra of OH, H and O for different
input power for 0.29 mbar pressure in quartz tube.
P-II-8-1
References
[1] A. Fridman. Inorganic gas-plasma decomposition
processes in Plasma Chemistry. Cambridge University
Press, 259-354 (2008)
[2] N. Boudesocque, C. Vandensteendam, C. Lafon, C.
Girold, J.M. Baronnet. Hydrogen production by thermal
water splitting using a thermal plasma. 16th World
Energy Conference, 1-12, June 13-16, 2006 Lyon, France
[3] Jian Luo, Steven L. Suib, Yuji Hayashi, and Hiroshige
Matsumoto. J. Phys. Chem A. 103, 6151-6161(1999)
[4] NIST Atomic Spectra Database
[5] Qing Xiong, Anton Yu Nikiforov, Manuel A
Gonzalez, Christophe Leys and Xin Pei Lu. Plasma
Sources Sci. Technol. 22, 1-13 (2013) 015011
3