Observation of Ions Emitted from Hot Graphite, Including
Excited Molecular Ions
P. Gee, T. Ehrenrich, Q.C. Kessel, E. Pollack and W.W. Smith
Department of Physics and the Institute of Materials Science, University of Connecticut, Storrs, CT 06269
Abstract. Thermal desorption of K ions from positively biased graphite (grafoil) containing K impurities has been
investigated by measuring the energies of the emitted ions with a hemispherical electrostatic analyzer and the masses
with a residual gas analyzer under ultra-high vacuum conditions. Potassium ions are seen to be emitted at temperatures
above 800°C. The present data provide evidence for the emission of K+ ions for sample bias voltages from 1V to 6V
and the emission of K2+ ions for biases above 10V. The emission of excited K2+ ions are inferred from the measurement
of the K+ ion energies which for sample biases above 10 volts with energies which correspond to approximately half of
the bias (accelerating) voltage. From this it is deduced that the K2+ ions are produced and disassociate after leaving the
surface. The K ion emission can be greatly enhanced by prior bombardment with an ion beam (not K ions) and
quenched by ion beam doses that amorphize the graphite surface. This would appear to indicate the emission
phenomenon depends upon the degree and crystallinity nature of the graphite surface.
INTRODUCTION
In perhaps the first experiment to measure the
energies of ions thermally desorbed from positivelybiased graphite (1-3), energy values are observed
which do not always correspond to their expected for
the bias (acceleration voltage) applied to the graphite
sample. For biases below 6V most ions are found to
have the expected energies, for example, a 3V bias
results in an ion with 3eV of energy. For bias energies
above 10V, however, the ion energies are found to be
approximately half of that expected. In the present
experiment, graphite containing K as an impurity is
heated above 800oC and ions are released from the
surface. (The sample material used is "grafoil" (4), a
form of graphite known to contain both K and other
impurities.) For bias energies below 6V the measured
ion energies correspond directly with the bias voltage.
For biases greater than 10V the majority of K ions
detected have energies, which are expected for
fragments from a K2+ ion, leaving the surface and
immediately breaking apart. This implies that for the
higher biases, the electric field present at the surface
results in the emission of excited K2+ ions, which
promptly dissociate.
3
1
4
2
6
5
FIGURE 1 Experimental setup.
1) Electrostatic energy analyzer. 2) Sample and heater
assembly. 3) Manipulator arm. 4) Ion gun. 5) Residual
gas analyzer (quadrupole mass spectrometer). 6)
Transmission grid.
CP680, Application of Accelerators in Research and Industry: 17th Int'l. Conference, edited by J. L. Duggan and I. L. Morgan
© 2003 American Institute of Physics 0-7354-0149-7/03/$20.00
207
Although Richardson wrote a text on the emission
of "positive and negative electricity" from hot bodies
in 1916 (5), most subsequent investigations have
concentrated on the emission of alkali halide ions from
hot tungsten (6) and thermal desorption spectroscopy
(7). The latter experiments generally consist of using
a residual gas analyzer (RGA) and the observation of
the species desorbed as the sample is heated. Figure 1
outlines the apparatus used in the current experiment.
It includes a 100mm radius hemispherical electrostatic
analyzer, an RGA and an ion gun. A 90 percent
transmission grid to which a bias may be applied is
positioned in front of the RGA. This permits an
approximate determination of the energies of the ions
entering the RGA. The grafoil is heated with a 0.320
"button heater" (8) and the heated sample may be
moved to face either the electrostatic analyzer or the
RGA.
1000000
6V
4V
Yield (a.u.)
100000
10000
1000
8V
2V
100
a.)
10
0 1 2 3 4 5 5 6 7 8 9 10 11 12
Energy (eV)
250
20V
Yield (a.u.)
200
THE EXPERIMENT
30V
35V
150
100
12V
50
Figure 2 shows two energy spectra of the emitted
positive ions obtained at a temperature of about 840oC.
Fig 2a shows single peaks whose energies correspond
to the applied bias as would be expected. For higher
bias voltages, presented in Fig. 2b, instead of a single
peak appearing at the energy corresponding to the bias
voltage, two peaks are observed. For these data it is
the sum of the energies of the two peaks that add to
approximately the energy corresponding to the bias
voltage. Unfortunately the electrostatic analyzer
measures only the ionized particles' energies and gives
no information about the corresponding masses. The
RGA by itself provides a determination of the masses,
ordinarily of neutral residual gases, but gives no
information about the energies of the ions from the
graphite. However, by placing a biased screen
between the source of the ions and the RGA, it is
possible to determine the stopping potential for the
ions from the graphite. Fig. 3a shows such a stoppingpotential curve for K ions from graphite biased at 10V.
A point on this curve represents the number of ions
(which pass through the screen) having energies above
those corresponding to the screen’s bias.
By
differentiating this stopping-potential curve and
plotting its negative, Fig. 3b, you have a plot of the
number of ions detected by the RGA as a function of
their energies, i.e., you develop a plot such as that in
Fig. 2b, but for the specific mass being detected by the
RGA. We assume the energy spectrum in Fig. 3b is a
measure of the energies of K ions from the graphite
surface biased at 10 V. The observed peak at 10 V
b.)
0
0 2 3 5 6 8 10 11 13 14 16 18 19 21 22 24
Energy (eV)
Figure 2 The ion yield is plotted versus the ion a.) low
energy spectra indicating the emission of ions with an energy
equal to the bias voltage applied. b.) higher energy spectra
for which the ions arrive in the detector with energies less
than what would correspond to the bias voltage.
therefore corresponds to the emission of K+ from the
surface and the peaks at energy, with the bias voltage
indicated for each curve. The points near 5 V are likely
to correspond to K+ resulting from the dissociation of
the K2+ (or something close to it in mass, such as
K2H+) Operating in its more usual mode (without any
bias applied to the screen) the RGA fails to detect a
significant number of ions near mass 78 which would
correspond to the K2+. For this reason we attribute the
K+ peaks at about 5V to dissociation products from an
excited molecular ion which fragments quickly upon
leaving the surface. More accurate measurements of
the energies are required in order to carry this analysis
much further. We are now preparing a Wien filter to
use in tandem with the electrostatic analyzer. It has
been observed that the ion emission may be
significantly affected by prior ion bombardment, but
this has yet to be quantified with regard to
implantation species and dose.
Implantation of
approximately 1016 2 keV Ar ions per cm2 at room
temperature can greatly enhance emission, while it has
been observed that implantation of 1021 ions per cm2
208
destroys the emission. For this latter dosage, the
graphite surface loses its crystalline sheen and appears
to have become amorphized. This aspect of the
experiment suggests that the emission of these ions
depend on the crystalline nature, perhaps the crystal
size, of the hot graphite. Grafoil is not a wellcharacterized material in terms of its crystal structure
and may be considered to be a collection of oriented
graphite crystallites. The premium grade of grafoil has
a carbon content of 98% and may contain up to 450
PPM of sulfur as well as other contaminants. For this
reason experiments in the energy range of Fig. 2b have
also been carried out with highly oriented pyrolytic
graphite (HOPG) which should not have a significant
number of impurities. However, the resulting data
were very similar, suggesting that the vacuum
chamber has been contaminated with potassium,
perhaps from the potassium that is sometimes used in
the manufacture of tungsten filament and heater wire.
Alkali ions are known to become ionized on
certain
hot
surfaces,
e.g.
beta-eucryptite
(Li2O.Al2O3.2SiO2) (6) and Holmlid and coworkers
have observed not only the emission of K ions, but
also the emission of Kn ionic molecules for n values of
1 to 61 (magic numbers) (9-11). They attribute much
of their data to the creation of Rydberg ions on the
surface. The current experiment does not detect Kn
ionic molecules and the current data appear to be
unique in their dependence on the crystalline nature of
the hot graphite and the emission of molecular ions in
excited states.
ACKNOWLEDGMENTS
We have greatly benefited from conversations
with and the prior efforts of Dr. Juan Lozano who used
this apparatus in his Ph.D. research. Michael Newman
and Ryan Sears assisted with the assembly of the
apparatus and also participated in the preliminary
measurements. The current research was supported by
the Connecticut Space Grant Consortium under NASA
EPSCOR grant No. NCC5-601.
70000
60000
Yield (a.u.)
50000
40000
30000
REFERENCES
20000
1. Kessel, Q., R. Sears, E. Pollack and W.W. Smith. "The
emission of charged particles with eV energies from hot
graphite." Application of Accelerators in Research and
Industry, edited by J.L. Duggan and I.L. Morgan, AIP Conf.
Proc. CP475, Denton, Texas 1999. 178-180.
2. Lozano, J., Kessel, Q., Pollack, E., and Smith, W.W., "Ion
Beam Enhanced Emission of Charged Particles from Hot
Graphite", Application of Accelerators in Research and
Industry, edited by J.L. Duggan and I.L. Morgan, AIP Conf.
Proc. CP576, Denton, Texas 2000, pp. 1044-1046.
3. Juan Lozano, Ph.D. thesis, The University of Connecticut,
2001.
4. Grafoil is the tradename of graphite based paper
manufactured by UCAR Carbon Company, Inc.
5. Richardson, O.W., The Emission of Electricity from Hot
Bodies, London, Longmans, Green and Co. 1916.
6. Blewett, J. P. and Jones, E. J. Phys. Rev. 50, 464 (1936).
7. Siegle, R., Davies, J.A., Forester, J.S. and Andrews, H.R.,
Nucl. Instr. and Meth. D 90, 606 (1994).
8. Manufactured by HeatWave, Watsonville, CA.
9. Wang, J., Engvall, K., Holmlid, L., J. Chem. Phys. 110
(1999) 1212-1220.
10. Wang, J., Andersson, R., Holmlid, L., Surface Sci, 399
(1998) L337-341.
11. Wang, J., Holmlid, L., Chem. Phys. Letters, 295 (1998)
500-508.
10000
0
a.)
0
1
2
3
4
5
6
7
6
7
8
9
10
Bias (V)
18000
16000
14000
Yield (a.u.)
12000
10000
8000
6000
4000
2000
b.)
0
0
1
2
3
4
5
8
9
10
Bias (V)
Figure 3 a.)Ion yield is plotted versus the stopping potential.
b.) Energy spectra of 10 eV ions.
209