1726
J. Chem. Phys., Vol. 116, No. 4, 22 January 2002
Z. W. Deng and R. Souda
from a hot filament surface are accelerated to, e.g., 70 eV
and then collide with the residual molecules, thereby leading
to the ionization of the latter, during which the resultant posi-
tive ions are analyzed. By scanning the electron energy and
recording the ion yield as a funciton of the electron impact
energy, the electron energy threshold for the formation of a
positive ion from a specific residual species ͑gas͒ could be
determined by extrapolating the electron energy spectrum to
a zero signal.18 The obtained electron energy thresholds help
to clarify the origin of the specific positive ions, as discussed
below.
For negative ion RGA measurement, another HAL
EPIC/IDP analyzer was used. This analyzer is a quadrupole
mass spectrometer designed for secondary positive and nega-
tive ion mass spectrometry analysis as well as positive and
negative ion RGA measurement. It was used in this work for
experiments on electron attachment to thermally desorbed
C2N2 molecules in the gas phase, during which the negative
ion yield was recorded as a function of the electron energy
when scanning the electron energy. The obtained electron
energy spectrum provided information on negative ionization
via low-energy electron attachment and/or electron impact.
The sample holder used in this work was custom de-
signed, which allowed resistive heating of the sample in
UHV. A pyrolytic graphite sample was prepared by cleavage
in air to a thickness of 0.2–0.3 mm and then held between
two electrodes connected to two high-current feedthrough.
The sample holder was supported by a turnable manipulator,
allowing the sample surface to face either the ion beam or
the analyzer. After mounting the sample holder into the
chamber, the whole system was baked until a base pressure
of Ͻ1ϫ10Ϫ10 Torr was reached.
The sample was heated to 1200 °C in UHV for several
cycles to clean the surface. After cooling down to room tem-
perature ͑RT͒, it was irradiated by 800 eV Nϩ2 ions with a
dose of 1–1.5ϫ1018 ions/cm2. Then, the sample was relo-
cated in front of the EQP/EQS analyzer and in situ heated up
to 1200 °C, during which the possible emission of positive
and negative ions was monitored in the SIMS mode, and the
desorption of neutrals was monitored in the ͑positive ion͒
RGA mode.
FIG. 1. Positive ion RGA mass spectrum of thermally desorbed neutrals
during heating of nitrogen ion-irradiated graphite.
uncertain why the authors failed to detect the m/e 52 species.
However, we found that N2 desorption would contribute
much more than CO to the RGA signal at m/e 28 after care-
ful treatment of the sample and monitoring of the desorbed
species both before and after nitrogen ion irradiation.20 The
N and CN species of m/e 14 and 26, respectively, are mostly
a consequence of the dissociative ionization of N2 and C2N2
molecules, as demonstrated below.
The CN species may result from either direct desorption
or dissociative ionization of C2N2 or both. This can be clari-
fied by comparing the electron energy thresholds (Eth) for
the formation of CNϩ and C2Nϩ2 ions.18 Positive ion forma-
tion during RGA measurement involves ejection of a valence
electron from its precursor molecule by electron impact. This
process results in the formation of a parent ion, and some-
times also daughter ions as a result of bond breaking induced
by the electron impact. For the formation of the parent ion,
the electron energy threshold depends on the first ionization
potential of the relevant precursor molecule. Accordingly, if
CNϩ originates from the direct ionization of the CN radical,
its Eth should be close to that of C2Nϩ2 due to their very
similar first ionization potentials ͑13.60 eV for CN and 13.37
21
eV for C2N2 ͒. The daughter ions should appear at higher
electron energies than that of a parent ion because they re-
quire additional energy for bond breaking. As such, if CNϩ is
a result of the dissociative ionization of C2N2, Eth of CNϩ
should be several eV’s higher than that of C2Nϩ2 .22 If CNϩ
originates from both channels, then we should observe two
Eth values for CNϩ. Such an interpretation is also applicable
to Nϩ and Nϩ2 .
III. RESULTS AND DISCUSSION
For the sake of clarity, we first present results on the
desorption of neutrals upon heating of the nitrogen ion-
irradiated graphite, which are important for the interpretation
of the origin of CNϪ ions. Figure 1 shows the positive ion
RGA mass spectrum of desorbed neutral molecules obtained
by the EQP/EQS analyzer in the positive RGA mode. The
spectrum was accumulated during heating of the sample be-
tween 800 °C and 1200 °C. Two dominant neutral species,
N2͑m/e 28͒ and C2N2͑m/e 52͒, were identified, as well as
small amounts of N͑m/e 14͒, CN͑m/e 26͒, and HCN͑m/e 27͒.
The neutral species that desorbed upon heating of the nitro-
gen ion irradiated graphite were monitored by other authors
and assigned to N and CN radicals.19 The m/e 28 species
were attributed to COϩ as a result of the reaction of hot
graphite with residual oxygen in the UHV chamber. We are
Figure 2 shows the RGA CNϩ, C2N2ϩ, Nϩ, and N2ϩ ion
yields as a function of electron energy, yielding a higher Eth
for the formation of CNϩ and Nϩ than, respectively, that of
C2Nϩ2 and Nϩ2 . It is thus believed that most of the CN and N
species may be produced by cracking of the C2N2 and N2
molecules, respectively. In addition, the RGA yields of CNϩ
and C2Nϩ2 showed an identical temperature dependence, in
agreement with the above conclusion. Hence, the dominant
thermally desorbed nitrogen-containing species from a nitro-
gen ion-irradiated graphite surface are ascribed to N2 and
C2N2 molecules, while the direct desorption of CN and N
155.247.166.234 On: Sat, 22 Nov 2014 23:20:04