Study of carbon reactivity in oxygen plasma using mass spectrometry

G. Petraconi, A.M. Essiptchouk, C. Otani, R.S. Pessoa, A.R. Coutinho,
and G. Capobianco.
Technological Institute of Aeronautics, Department of Physics – Plasma and Processes Laboratory
(LPP), 12228-900, ITA – CTA , São José dos Campos, SP, Brazil.
Abstract. Having in mind the plasma assisted carbon material activation process development,
samples of carbonized macadamia nut shells were treated in oxygen plasma. The carbonization
process was performed at 800oC in nitrogen atmosphere, and a titanium electrode Hollow
Cathode Reactor (HCR) coupled with mass spectrometry monitoring system was set up for this
experimental work. The process parameters temperature (200 to 800 oC) and time (1 to 50 min)
were varied, while the chemical composition of plasma was being monitored by mass
spectrometry technique. The analysis was focused to measure the relative concentration of
species O, O2, C, CO and CO2, due to the fact that the carbon activation process is mainly based
on surface carbon oxidation reaction. The partial pressure measurements of each specie as
functions of process parameters showed that atomic and molecular oxygen contributions to total
pressure were diminished with increase of time, temperature and flux, while the C, CO and CO2
presented an opposed trends. A correlated analysis of results obtained in this work with another
works in which the adsorption characteristics of similarly treated materials, with same precursor,
were measured, leads to the conclusion that the activation process of carbon materials by cold
plasma is based on physical sputtering and also on chemical oxidation processes of carbon
localized at surface sites.
Keywords: carbon reactivity, coal oxidation, mass spectrometry analysis
INTRODUCTION
Plasmas offer an attractive alternative to more traditional methods for the surface
modification of carbon fibers, mesocarbon micro beads, graphite, and glassy carbons
and carbon blacks [1]. Marsh et al.[2] reported that plasma assisted surface
modification process applied to activated carbons, could result in improvements of
their performance as adsorbents and/or catalyst supports. An objective of this work
was to find out the best conditions for activation of carbonized macadamia nut shells
using a Hollow Cathode Discharge (HCD) process, having in mind the applicability of
the resultant porous carbon material as molecular sieves for adsorbing molecules in
gaseous phase [3,5].
It is well established that the porous carbon properties, structure and surface
chemistry are all influenced by the nature of raw material and activation method [6].
There are several activation techniques classified as chemical or physical method. In
this work it is proposed a physical-chemistry activation process using a HCD reactor.
For monitoring the reaction products in the plasma gas phase during the treatment,
which are expected to be sensitive to the intensity of the treatment, we investigated the
applicability of mass spectrometry. The plasma is composed of a variety of species
like electrons, ions, neutrals and excited atoms, radical and molecules. Many of these
may have to be into account for a successful treatment of the coals. Thus, the relative
intensities of mass lines obtained from quadrupole mass spectrometry turned out to be
a good indicator for the intensity of the plasma activation treatment.
EXPERIMENTAL
Samples of porous carbonaceous matter, prepared by carbonization of wood
(macadamia nut shells), at 800ºC in nitrogen atmosphere, were produced to carry out
the experiments. Oxygen was the gas used as the activation medium to produce a
plasma excited by DC power in a Hollow Cathode Discharge (HCD), varying the
following process parameters: pressure, gas flow rate, DC power, temperature and the
residence time. The chamber process comprises a quartz discharge tube of 65 mm
diameter surrounded by a coaxial electrical furnace of 40 cm in length(see fig. 1).
The samples were positioned inside the hollow cathode made of titanium. A DC
electrical power supply adjusted at 300 W (600 V and 0.5 mA) was used to produce a
stable glow discharge in oxygen media. The gas pressure and mass flow rate were
maintained in order of 0.5 mbar and 60 sccm, respectively. Temperature (200 to 800
oC) and and time (1 to 50 min) were varied, while the chemical composition of plasma
was being monitored by mass spectrometry technique.
The Mass Spectrometer AccuQuad 200D (with a resolution of 1 amu) was mounted
with the sampling orifice in the vicinity of the chamber wall, out of direct sight of the
electrodes. The system was differentially pumped and the internal pressure was
regulated to be constant at 10−5 mbar. The neutral gas species are extracted from the
chamber through an orifice and undergo subsequent electron impact ionization at
constant electron energy of 70 eV. The resulting ionized gas species and fragments
enter the quadrupole RF mass filter to be detected as a function of their mass-tocharge
ratio (m/q). The collected spectra were recorded in the mass range
FIGURE 1. Schematic view of the HCD reactor.
RESULTS AND DISCUSSION
Fig. 2 shows a mass spectrum after pumping down the system and just before the
plasma ignition. The sample is inside the vacuum chamber. No additional processing
gas is introduced into the chamber. The spectrum is dominated by the water peak at 18
amu, which is caused by water evaporating out if the samples. The peaks for
molecular oxygen (32 amu) and nitrogen (28 amu) also result from degassing of the
samples but not from vacuum leaks, which would result in a different relative intensity
of these peaks. Therefore, the ignited plasma mainly consists of water vapor, hydrogen
(2 amu), oxygen, nitrogen and carbon dioxide (44 amu) at the beginning of the
process.
When the oxygen processing gas is introduced into the chamber at pressure of 0.5
torr and mass flow of 60 sccm, we observe the mass spectrum shown in fig. 3. The
spectrum is dominated by the molecular oxygen and atomic peaks at 32 and 16 amu,
respectively.
Other authors have shown in a previous study [7] that during plasma treatment of
coal, oxidation of the coal surface occurs, which should produce some of the following
gases: carbon monoxide, carbon dioxide, atomic and molecular oxygen, atomic carbon
and a variety of other organic compounds. Fig. 4 shows the development of a mass
spectrum during a dynamic plasma treatment of carbonized macadamia nut shell.
After starting the discharge, the CO, CO2 and C signals increase, whereas the water
signal slightly decreases. The reasons are the dissociation of water by the plasma,
further degassing and drying out of the coal and possibly the beginning of hydrogen
abstraction from the organic molecules at the coal surface. At the conditions showed
in fig. 3 and at temperature of 400 oC, the atomic and molecular oxygen signal reach a
maximum, and, after the plasma ignition at current discharge of 380 mA and discharge
voltage of 616 V, these signal decreases, in particular, the molecular oxygen signal
drops significantly.
At the same time, a significant increase in the CO2 signal at 44 amu and in the mass
line at 28 amu can be observed. The signal at 28 amu is mainly the sum of the nitrogen
(N2) and the carbon monoxide (CO) signal. Neither a leak nor degassing process can
explain an increase of the nitrogen signal without a corresponding increase in the
oxygen signal. Taking into account the increase of the CO2 signal, the increase in the

Intensity (a.u.)
mass (amu)
FIGURE 2. Background mass spectrum, before
oxygen injection and plasma ignition.

FIGURE 3. Mass spectrum before plasma
ignition and by inserting the oxygen gas at 0.5
torr and mass flow of 60 sccm.
signal at 28 amu is most likely to be attributed to carbon monoxide and not to the
nitrogen molecules. The development of the O2, CO2, C and CO concentrations clearly
reveals that surface of the organic carbonized material is etched and oxidized by the
oxygen in the plasma. Note that, after only 1 min of starting the discharge the signal
becomes constants.
Fig.5 shows that the reactivity of carbons towards atomic oxygen is very much
dependent upon temperature. The ratio of the partial pressure of carbon away from the
surface to the partial pressure of oxygen increases with the temperature and reaches a
maximum at temperature of order of 500 oC. At higher temperature value, the partial
pressures ratio of C/O drops significantly.
CONCLUSION
Our work demonstrates the applicability of mass spectrometry to monitor chemical
reactions during a coal oxidation process in a hollow cathode reactor operating at low
pressure oxygen discharge. Based on the results of the static treatment, mass
spectrometry was tested as a process control tool for dynamic treatments. To perform
a real online control with a good time resolution, only the mass lines at 28 amu (N2 +
CO), 32 amu (O2), 12 amu (C), 16 amu (O) and 44 amu (CO2) were monitored. From
this, we conclude that atomic and molecular oxygen adsorbs and reacts (form bonds)
with the sample without significant mass loss, additional energy being required to
break the bonds and release carbon as C, CO and CO2. The ratio of partial pressures of
C and O can be used to evaluate the plasma reactivity being probably more dependent
on the temperature than on exposure time.
REFERENCES
1. Torre, l.e.c. et all, carbon, 36(1998) 277.
2. Marsh, H; Heinz, E.A.; Reinoso, R., F. Introduction to carbon science. [s.l.]: University of Alicante, 1997.
3. Hassler, j.w. Activated carbon. London: Leonard Hill Books, 1967.
4. Bokros, j. C.; Lagrange, l. D.; Schoen, f. J. Chemistry and physics of carbon, 9, 103 – 164, 1973.
5. Noll, K. E. Adsorption technology for air and water pollution control. Chelsea: Lewis Publishers, 1992.
6. Adamson, a. W. Physical chemistry of surfaces. 5º ed.

Опубликовано в рубрике Documents