Icarus 168 (2004) 53–59
www.elsevier.com/locate/icarus
Photon-stimulated desorption of Na from a lunar sample:
temperature-dependent effects
Boris V. Yakshinskiy and Theodore E. Madey ∗
Department of Physics and Astronomy and Laboratory for Surface Modification, Rutgers, The State University of New Jersey, 136 Frelinghuysen Road,
Piscataway, NJ 08854-8019, USA
Received 15 August 2003; revised 21 October 2003
Abstract
We report recent results in an investigation of source mechanisms for the origin of Na atoms in tenuous planetary atmospheres, focusing on
non-thermal processes. Experiments include photon stimulated desorption (PSD), electron stimulated desorption (ESD), and ion sputtering
of Na atoms from the surface of a lunar basalt sample. Bombardment of the sodium covered surface by 3 keV Ar+ ions removes Na from the
surface by sputtering into vacuum and by implantation into the sample bulk. Bombardment of the Na covered surface by ultraviolet photons
or by low energy electrons (E > 3 to 4 eV) causes desorption of “hot” Na atoms. These results are consistent with our previous measurements of sodium and potassium desorption from a silica surface: electron- or photon-induced charge transfer from the substrate to the ionic
adsorbate causes formation of a neutral alkali atom in a repulsive configuration, from which desorption occurs. There is a strong temperaturedependence of Na ESD and PSD signals, under conditions where the Na surface coverage is constant and thermal desorption is negligible.
The yield of Na (atoms/photon) increases by 10× from 100 to 470 K; an activation energy of ∼ 20 meV is measured. This phenomenon may
be attributed to thermally-induced changes in surface bonding sites, and will affect recent modeling of the sodium atmospheres of Mercury
and the Moon.
 2003 Elsevier Inc. All rights reserved.
Keywords: Moon; Atmosphere; Surface
1. Introduction
Earth’s Moon and the planet Mercury have extended tenuous atmospheres containing suprathermal sodium and potassium atoms (Potter and Morgan, 1985, 1988, 1997; Potter et
al., 2002), and Mercury’s atmosphere has been discovered
recently to contain traces of Ca atoms (Bida et al., 2000).
The lifetime of an alkali atom in the atmosphere is on the
order of hours due to photoionization by solar photons, so
the atmospheres must be continuously resupplied in order to
maintain the observed concentrations. Various mechanisms
and source processes have been proposed as origins of the
Na, K atmospheres (Hunten and Sprague, 1997; Killen and
Morgan, 1993; Killen and Ip, 1999), including sputtering by
the solar wind, electron, and photon stimulated desorption
(ESD and PSD), (McGrath et al., 1985; Madey and Yakshinskiy, 1998; Mendillo et al., 1999; Madey et al., 2002),
* Corresponding author.
E-mail address: [email protected] (T.E. Madey).
0019-1035/$ – see front matter  2003 Elsevier Inc. All rights reserved.
doi:10.1016/j.icarus.2003.12.007
thermal desorption (Yakshinskiy et al., 2000), micrometeorite impact (Morgan et al., 1988; Cintala, 1992; Verani et
al., 1998; Wilson et al., 1999; Smyth et al., 1999, 2001), and
thermal diffusion from the crust (Sprague, 1992; Killen and
Morgan, 1993). There are also detailed theoretical models
of the sources, sinks and transport dynamics of Na and K
in the lunar and mercurian atmospheres (Smyth and Marconi, 1995a, 1995b; Leblanc and Johnson, 2003; Wurz et al.,
2003).
Recently we presented evidence based on laboratory experiments that photon-stimulated desorption (PSD) is likely
to be the dominant source process for desorption of Na from
the Moon (Yakshinskiy and Madey, 1999, 2000). Since those
experiments were performed on a model substrate, SiO2 , it
is desirable to determine whether or not PSD is an important
process for a lunar sample, a much more complex oxide substrate. Preliminary measurements using a lunar basalt sample (Yakshinskiy and Madey, 2003) indicate that PSD of Na
is indeed an efficient source process.
In the present paper we report new data on PSD, ESD,
and ion sputtering of Na deposited onto a lunar basalt sam-
54
B.V. Yakshinskiy, T.E. Madey / Icarus 168 (2004) 53–59
ple. The most striking new result is the observation of a
strong temperature dependence of the PSD and ESD atom
desorption yields; the yield of Na (atoms/photon) increases
by ∼ 10× from 100 to 470 K. Recently, surface temperature variations were shown to be critical in understanding the
observed variability of Mercury’s atmosphere (Leblanc and
Johnson, 2003). This new result means that not only does
surface temperature play an important role over a large area
of Mercury’s surface, but it is also important for understanding the lunar atmosphere.
2. Experimental procedures
Experiments are performed in a baked, stainless steel
ultrahigh vacuum (UHV) chamber with a base pressure
∼ 5 × 10−11 Torr; the apparatus and procedures have been
described previously (Yakshinskiy and Madey, 2000, 2003).
X-ray photoelectron spectroscopy (XPS) and low-energy ion
scattering (LEIS) are used for surface characterization. The
sample is a piece of Ti-rich lunar basalt (Sample #74275)
with area ∼ 1 cm2 and thickness of 1 mm (Fink et al., 1998).
The sample is attached to a piece of 0.5 mm thick Mo sheet
by thin Ta straps. The sample is mounted on an XYZ-rotary
manipulator, and can be cooled to ∼ 100 K using liquid nitrogen as well as heated by resistive heating of the Mo sheet
to a maximum temperature of ∼ 600 K (to maintain the integrity of the sample, we do not heat above 600 K). To reduce
the concentration of surface impurities, the sample is sputtered by 1–2 KeV Ar+ ions at 45◦ incidence. Sputtering depletes the surface of oxygen, which is restored by exposure
to 10−6 Torr of O2 in UHV. To increase the surface concentration of Na to measurable levels, Na is deposited onto the
sample surface by thermal evaporation from an SAES Getters source.
The measurement procedure for PSD and ESD of Na
from the lunar sample involves a pulsed low energy electron
source (0 to 200 eV) and a highly sensitive detector based on
surface ionization of Na on hot Ir ribbons (∼ 1800 K). The
neutral Na atoms incident on the hot Ir surface desorb as Na+
ions and are detected using a channeltron electron multiplier.
The velocity distributions of neutral atoms are measured using a time-of-flight (TOF) method. For PSD measurements,
a mechanically-chopped (3.3 kHz) photon source (500 W Hg
arc lamp) is used. Glass bandpass filters are used for wavelength discrimination.
3. Results
An X-ray photoelectron spectrum of the lightly Ar+ sputtered lunar sample is shown in Fig. 1. In the binding
energy range displayed here (0–400 eV), evidence for C and
for oxide forms of Fe, Mg, Al, Si, Ca is seen; Ti and O are
also present, as found by increasing the energy range. The
C impurity is mainly from atmospheric exposure, and is reduced by heating and Ar+ sputtering. A very small Na signal
Fig. 1. X-ray photoelectron spectrum of lunar basalt sample; photoelectron
signal vs. electron binding energy for Al Kα excitation. The major core
level features are associated with specific elements. Inset: a portion of an
XPS spectrum around the Na 1s peak, illustrating the existence of sodium
in the sample.
is observed, at a level 1 atom percent (see inset). The relative elemental concentrations in the surface region mirror
the bulk composition (Fink et al., 1998).
Electron bombardment of the lunar sample induces desorption of atomic Na from the Na low concentration present
in the surface layer, but the signal is too small for accurate measurements of desorption threshold energies. To
generate a measurable signal Na is deposited onto the surface and ESD/PSD measurements are made. The coverage
(measured in monolayers, ML) is determined using XPS,
and the ESD/PSD yields (atoms/electron, atoms/photon) are
found to scale linearly with coverages < 1 ML. At coverages < 1 ML, the XPS data indicate that Na is adsorbed
ionically.
Figure 2 shows the ESD and PSD yields as a function of
excitation energy for ∼ 0.5 ML of Na adsorbed at 100 K.
The threshold energies for both ESD and PSD are in the
range 3 to 4 eV, and the desorption signals rise sharply
for higher energies. The horizontal bars in the PSD data
represent the effective filter bandwidths at 70% peak transmittance level. The electron energies are corrected for the
emitter work function and estimated electron affinity of the
lunar sample (Yakshinskiy and Madey, 2003). The data are
qualitatively similar to measurements on pure SiO2 films,
but the magnitude of the Na desorption signal is about 30%
of the corresponding signals for desorption of Na from SiO2
(Yakshinskiy and Madey, 2000); the PSD cross-section at
hν ∼ 5 eV is estimated to be ∼ 10−20 cm2 .
Figure 3 shows the velocity distribution for ESD of Na
from the lunar sample at 100 K. The peak velocity of
∼ 800 m/s is less than that observed for desorption of Na
from an SiO2 film (Yakshinskiy and Madey, 2000) but the
Photon-stimulated desorption of Na
Fig. 2. Solid line: ESD yield of neutral atomic Na from 0.5 ML of Na
deposited onto surface of lunar basalt sample at 100 K, as a function of
electron bombardment energy. Inset: PSD yield of Na atoms as a function
of photon energy. The horizontal lines represent the effective bandwidths of
optical filters.
55
(a)
(b)
Fig. 3. Velocity distribution (dN/dV vs. velocity) for neutral Na from
∼ 0.5 ML of Na on lunar basalt. dN is the number of desorbing atoms
having velocities in the range V + dV .
desorbing Na is suprathermal (∼ 900 K) with respect to the
100 K substrate. Focusing on the DIET processes, our experiments provide an average Na escape velocity ∼ 1 km/s
from an SiO2 surface, with extended tail up to ∼ 2.5 km/s;
as shown in Fig. 3 for the lunar sample, the extended tail
reaches ∼ 2 km/s. Based on our results, UV solar photons
and electrons with E < 300 eV do not contribute appreciably
to the desorbing Na flux with speed V > 2 km/s. The higher
Na ejection speeds of 2.1–2.4 km/s from the lunar surface
have been associated with micrometeorite impact vaporization (Smyth and Marconi, 1995b; Wilson et al., 1999).
Figure 4 illustrates the temperature dependence of Na
desorption yield. In Fig. 4a, the PSD Na intensity is plotted
Fig. 4. Temperature dependence of Na PSD desorption yield from lunar
basalt. Sample is irradiated with “white light” ∼ 1–5 eV. (a) Na intensity vs. time as sample is heated and cooled; light is turned off and on.
(b) Boltzmann plot of PSD intensity (hν ∼ 1 to 5 eV) and ESD intensity
(Ee = 200 eV) as function of reciprocal temperature. The activation energy
is 20 ± 2 meV.
vs. time as the sample is heated and cooled. To maximize
the signal, 1 ML of Na is irradiated with “white light,” 1–
5 eV (IR is filtered out). The incident photons do not cause
a detectable change in substrate temperature. The signal increases and decreases as light is turned on and off, and also
as the sample is heated and cooled. The Na signal varies
reversibly with temperature, and is ∼ 10× more intense at
470 K than at 100 K. In this temperature range, there is negligible thermal desorption of Na as verified using XPS; the
surface Na concentration is constant. The temperature dependence of Na PSD signal is displayed on a Boltzmann
plot, log I /I300 vs. 1/T , in Fig. 4b. Here, the data are normalized to the magnitude I300 of the signal measured at
300 K. Both PSD data and ESD data (Ee = 200 eV) dis-
56
B.V. Yakshinskiy, T.E. Madey / Icarus 168 (2004) 53–59
Fig. 5. Na signal (Na 1s XPS intensity) vs. time of bombardment by 3 keV
Ar+ ions. Grazing emission signal (60◦ from normal) corresponds to increased surface sensitivity.
play similar temperature dependences, and the best straight
line fit to the data gives an activation energy of ∼ 20 meV.
The removal of Na by Ar+ sputtering is indicated in
Fig. 5. In this case, ∼ 1 ML of Na deposited onto the lunar sample, and the Na 1s intensity is monitored using XPS
as a function of Ar+ ion beam irradiation time (3 keV Ar+ ,
0.2 µA/cm2 , rastered over sample). XPS intensities are measured for both normal emission and grazing emission (60◦
from normal); the grazing emission emphasizes the surface
contribution to the Na intensity, while the normal emission contains signal from both surface and bulk. The data
for two emission angles demonstrate that Ar+ bombardment removes Na from the surface both by sputtering into
vacuum and by implantation into the bulk; the more rapid
decrease for grazing emission (higher surface sensitivity)
indicates that the surface is depleted more effectively than
the bulk. The estimated cross-sections for removal of Na
are ∼ 2.1 × 10−16 cm2 and ∼ 2.9 × 10−16 cm2 for normal
and grazing emission angles, respectively. The experiments
of Fig. 5 represent our attempt to simulate effects of solar wind and magnetospheric ion impacts. The efficiency of
Na removal by sputtering with projectiles much lighter in
mass, i.e., He+ ions, is found to be ∼ 1.5 times less than
for Ar+ .
4. Discussion
4.1. Mechanism of ESD and PSD of Na
Sodium in lunar samples is generally associated with
Na2 O, or with Na silicates; it is in an oxidized form (Na+ ).
Fractional monolayers of Na deposited onto oxide surfaces
(SiO2 and lunar sample) also adsorb as Na+ , whereas multilayers (> 1 ML) contain metallic Na (Yakshinskiy et al.,
2000). The ESD and PSD data of Figs. 2–4 are interpreted
in terms of a model described previously to explain ESD and
PSD of Na and K from a model mineral surface, SiO2 (Yakshinskiy and Madey, 1999, 2000, 2003). Electron or photon
bombardment of the lunar surface with energies greater than
threshold values of 3 to 4 eV can induce electron transfer from an electronic state in the bulk, or from a surface
state, to the unoccupied Na+ 3s level. Once the 3s level is
occupied, the neutral Na atom has a larger radius than the
original Na+ ion; the atom is in a highly repulsive configuration. For Na adsorbed on a wide bandgap insulator, the rate
of charge transfer from the neutralized Na atom to the substrate is slow, and the atom in its repulsive state can survive
and desorb from the surface. Sodium desorbs from the lunar
sample with escape velocity peaked at ∼ 800 m/s (Fig. 3)
with corresponding temperature ∼ 900 K, which we associate with the desorption of ground state “hot” neutrals. In
independent experiments, neutral ground state K atoms have
been found in studies of the ultraviolet laser-induced desorption of potassium from Cr2 O3 (0001) (Wilde et al., 1997);
the authors use a resonance excitation method to demonstrate directly that the desorbing species are translationallyhot ground-state neutrals. Nonetheless, the possibility of a
contribution to the high velocity tail in Fig. 3 from the desorption of Na Rydberg atoms cannot be excluded (Holmlid, 2002; Yarygin et al., 2003). Note that both ESD and
PSD proceed via a similar charge transfer mechanism, especially near threshold. Of course, we have a much greater
experimental energy range available for electron-induced excitations as compared to photon-induced studies. At higher
energies, desorption can also be initiated by core hole excitations (e.g., O 2p at ∼ 25 eV) (Madey and Yakshinskiy,
1998).
4.2. Temperature-dependent effects in ESD and PSD
Temperature dependences in desorption yields of ionic
and neutral species have been reported in an number of studies involving ESD of adsorbed monolayers under ultrahigh
vacuum conditions. The most common effect is an irreversible change in desorption yield as temperature increases,
due to thermally-induced desorption and/or decomposition
of an adsorbed layer (Madey and Yates, 1971). In many instances, a sharp decrease of desorption yield coincides with
thermal desorption (Faradzhev et al., 2003). However, there
are also reports of reversible changes in desorption yield as
a function of temperature; measurements are made under
conditions where the maximum temperature is lower than
that required for thermal desorption, i.e., the surface coverage of adsorbate does not change with temperature. In two
cases (Menzel, 1969; Kutsenko, 1969) the ESD positive ion
yield is reported to increase with temperature, a phenomenon that the authors associate with an increased average
Photon-stimulated desorption of Na
distance of the atom from the substrate due to population
of excited vibrational states. In contrast, a small (∼ 10%) reversible decrease of O+ ESD signal for oxygen on W (110)
is attributed to an increased neutralization rate of ions as
substrate temperature increases (Madey and Yates, 1969).
The temperature dependence of ESD yield of alkali metal
ions from a silicided iridium surface (Ageev and Yakshinskiy, 1995) is associated with the transfer of thermal energy
from the solid during the reverse motion of the desorbing
ion, and with the increasing initial escape velocity of the
ion from the surface. Somewhat larger reversible effects are
seen in ESD of neutral Eu and Sm atoms from adsorbed
monolayers on oxygen-covered W surfaces (Ageev et al.,
2003a, 2003b). In both cases, the electron-stimulated desorption mechanism involves resonant electronic excitations
of W, Sm, and Eu core levels; the temperature dependence
of the ESD atom yield is associated with the lifetimes of
core excitons. A reversible temperature dependence is also
observed for desorption of alkali atoms from oxidized W at
Ee > 25 eV (Ageev et al., 1996), and attributed to temperature dependent charge transfer.
The reversible temperature dependences shown in Fig. 4
for PSD and ESD of Na from a lunar sample are quantitatively different from the measurements described in the
previous paragraph. The excitation energy (1–5 eV for PSD)
is much lower than for that the above-mentioned ESD studies, and the desorption yield follows a Boltzmann plot over
a wide range at low temperatures; the yield increases by
∼ 10× from 100 to 470 K. The effect is too large to be
explained easily by changes in vibrational amplitudes or
electron tunneling probability (charge transfer rates). Note
also in Fig. 4b that both PSD (hν 5 eV) and ESD (Ee ∼
200 eV) at different excitation energies exhibit the same
temperature dependences. PSD proceeds via charge transfer neutralization, while the ESD signal contains contributions from core-level excitations. This suggests that the
temperature-dependent desorption yields are less influenced
by the details of electronic excitation mechanisms than by
the initial-state bonding configuration and final state survival
probability.
A possible explanation of the data of Fig. 4 is a temperature-dependent change in binding sites, where the local
atomic coordination changes and, therefore, the PSD/ESD
desorption rate is different. Recent calculations for binding of alkali atoms to amorphous SiO2 surfaces indicate
that the alkali is completely ionized, and that the two most
energetically-favorable sites involve coordination to two and
three O−− anions, respectively (Lopez et al., 1999). In this
case, the difference in adsorption energy is ∼ 0.5 eV, much
greater than the energy of ∼ 0.02 eV derived from the slope
of Fig. 4b. The lunar sample is a much more heterogeneous
material than pure SiO2 , and it is possible that there are subtle reversible changes in local bonding configuration of Na+ .
That is, Na + may diffuse locally and reversibly to O−−
sites associated with different cations (Ca++ , Si4+ , Al3+ ,
etc.). The more stable site at lower temperature could involve
57
higher O−− coordination number, and a correspondingly
low PSD yield. As population shifts from a “PSD-inactive”
site at low T to a “PSD-active” site at higher T , the measured PSD signal changes.
For insights into how changes in bonding can affect desorption of alkalis, consider a phenomenological description
of ESD and PSD that is based on models of Menzel and
Gomer (1964) and Bass and Sanche (2003). For excitation
energies near threshold, the cross-section for desorption of
an alkali atom can be expressed as
σ = σc (E)Ps ,
(1)
where σc is the cross-section for excitation and capture of
an electron by a surface Na+ cation, i.e., the cross-section
for charge transfer from the substrate to the cation resulting
in formation of Nao in a repulsive configuration. Ps is the
survival probability of the Nao atom against autodetachment
of the electron via hopping back to the substrate. Ps is given
by
Zc
Ps = exp − R(z)/v dz ,
(2)
Z0
where R(z) is the rate of electron hopping (s−1 ) from Nao
to the substrate, v is the velocity of the atom away from
the surface along the excited state potential curve, zo is the
equilibrium substrate-cation bond length, and zc is a critical
distance for electron hopping (if the atom survives beyond
zc , it will desorb as Nao ).
We can simplify the integration by defining an average
lifetime τa = R(z)−1 and letting τ ∼
= (zc − zo )/v, to give
Ps ≈ exp[−τ/τa ]
and σ ≈ σc exp[−τ/τa ].
(3)
Thus, the desorption cross-section has an exponential dependence on the lifetime τa of the transient atom and on its
velocity v away from the surface. For a given value of τ ,
if τa is very short (fast electron hopping time from atom to
substrate) the cross-section σ is smaller than for the case of
long τa (slow hopping time) where the cross-section is large.
If there are multiple adsorption sites on the surface for which
τ and τa are very different, then this lifetime argument can
be used to provide insights into the temperature dependent
data for ESD/PSD of Na from a lunar sample.
Assume there are two adsorption sites (1 and 2) separated in energy by E ∼ 20 meV and located in close
proximity to one another, so that adsorbed Na can diffuse or “hop” between them. The lower energy site (1)
is preferentially populated by Na+ at 100 K (concentration N1 , atoms/cm2 ), and as temperature increases, the
Na+ population (N2 , atoms/cm2) in the higher energy site
(2) increases according to the Boltzmann factor N1 /N2 ≈
exp{−E/kT }. If during ESD or PSD, the Na desorption
cross-section at site 2 is higher than at site 1, a temperature
dependent desorption yield is measured. That is, if the lifetime τa of the transient Na atom formed in site (1) is shorter
58
B.V. Yakshinskiy, T.E. Madey / Icarus 168 (2004) 53–59
than the lifetime τa of the transient Na atom formed in site 2,
the cross-section for desorption from site 1 (Eq. (3)) is correspondingly smaller than from site 2. As indicated above,
subtle changes in bonding configuration and coordination
number may be the cause of such lifetime effects, giving rise
to “PSD-inactive” and “PSD-active” sites.
Future measurements will focus on PSD studies of Na on
a binary system, SiO2 , for which the adsorption chemistry
will be easier to characterize than that of the complex lunar
sample.
These data have important implications for attempts to
model the source processes for Na in tenuous planetary atmospheres. Whereas thermal desorption of Na is expected
to contribute to Mercury’s atmosphere due to the high surface temperature (∼ 700 K at the subsolar point), the present
measurements indicate that thermally-assisted PSD can also
occur. That is, under conditions such that thermal desorption
of Na is negligible (e.g., at the lunar surface), variations in
surface temperature can have a large effect on the PSD rate.
Acknowledgment
The authors acknowledge valuable discussions and correspondence with Prof. Robert Johnson. This work has been
supported in part by NASA Planetary Atmospheres Division.
References
Ageev, V.N., Yakshinskiy, B.V., 1995. Electron-stimulated desorption of
Li+ and Na+ ions from lithium and sodium layers adsorbed on an
iridium surface coated with a silicon film. Phys. Solid State 37, 261–
265.
Ageev, V.N., Kuznetsov, Yu.A., Potekhina, N.D., 1996. Electron-stimulated
desorption of alkali metal and barium atoms from an oxidized tungsten.
Surf. Sci. 367, 113–127.
Ageev, V.N., Kuznetsov, Y.A., Madey, T.E., 2003a. Electron-stimulated
desorption of rare-Earth metal atoms from oxidized tungsten. Surf. Sci.
In press.
Ageev, V.N., Kuznetsov, Y.A., Madey, T.E., 2003b. Temperature dependences in electron-stimulated desorption of Eu atoms. J. Electron Spectrosc. Rel. Phenom. In press.
Bass, A.D., Sanche, L., 2003. Reactions induced by low energy electrons in
cryogenic films (review). Low Temp. Phys. 29, 202–214.
Bida, T.A., Killen, R.M., Morgan, T.H., 2000. Discovery of calcium in Mercury’s atmosphere. Nature 404, 159–161.
Cintala, M.J., 1992. Impact-induced thermal effects in the lunar and mercurian regolith. J. Geophys. Res. 97, 947–973.
Faradzhev, N.S., Kusmierek, D.O., Yakshinskiy, B.V., Solovev, S.M.,
Madey, T.E., 2003. Electron-stimulated desorption of F+ and F− from
SF6 on Ru(0001): structure and bonding. Surf. Sci. 528, 20–26.
Fink, D., Klein, J., Middleton, R., Vogt, S., Herzog, G.F., Reedy, R.C.,
1998. 41 Ca, 26 Al, and 10 Be in lunar basalt 74275 and 10 Be in the
double drive tube 74002/74001. Geochim. Cosmochim. Acta 62, 2389–
2402.
Holmlid, L., 2002. Conditions for forming Rydberg matter: condensation
of Rydberg states in the gas phase versus at surface. J. Phys. Condens.
Matter 14, 13469–13479.
Hunten, D.M., Sprague, A.L., 1997. Origin and character of the lunar and
mercurian atmospheres. Adv. Space Res. 19, 1551–1560.
Killen, R.M., Morgan, T.H., 1993. Diffusion of Na and K in the uppermost
regolith of Mercury. J. Geophys. Res. 98, 23589–23601.
Killen, R.M., Ip, W.-H., 1999. The surface-bounded atmospheres of Mercury and the Moon. Rev. Geophys. 37, 361–406.
Kutsenko, E.N., 1969. The influence of adsorbed temperature on the probability of desorption of O+ ions by slow electrons. Soviet Phys. Tech.
Phys. 14, 706–707.
Leblanc, F., Johnson, R.E., 2003. Mercury’s sodium exosphere. Icarus 164,
261–281.
Lopez, N., Illas, F., Pacchioni, G., 1999. Adsorption of Cu, Pd and Cs atoms
on regular and defect sites of the SiO2 surface. J. Am. Chem. Soc. 121,
813–821.
Madey, T.E., Yates Jr., J.T., 1969. Temperature effects in desorption by electron impact: molecular vibrations in the chemisorbed layer. J. Chem.
Phys. 51, 1264–1265.
Madey, T.E., Yates Jr., J.T., 1971. Electron-stimulated desorption as a tool
for studies of chemisorption: a review. J. Vac. Sci. Technol. 8, 525–
555.
Madey, T.E., Yakshinskiy, B.V., 1998. Desorption of alkali atoms and ions
from oxide surfaces: relevance to origins of Na and K in atmospheres of
Mercury and the Moon. J. Geophys. Res. 103, 5873–5887.
Madey, T.E., Johnson, R.E., Orlando, T.M., 2002. Far-out surface science:
radiation-induced surface processes in the Solar System. Surf. Sci. 500,
838–858.
McGrath, M.A., Johnson, R.E., Lanzerotti, L.J., 1985. Sputtering of sodium
on the planet Mercury. Nature 323, 694–696.
Mendillo, M., Baumgardner, J., Wilson, J., 1999. Observational test for
the solar wind sputtering origin of the Moon’s extended sodium atmosphere. Icarus 137, 13–23.
Menzel, D., 1969. Inelastic interactions of slow electrons with adsorbed
particles: dependence of the interaction cross sections on temperature.
Surf. Sci. 14, 340–350.
Menzel, D., Gomer, R., 1964. Desorption from surfaces by slow electron
impact. J. Chem. Phys. 40, 1164–1165.
Morgan, T.H., Zook, H.A., Potter, A.E., 1988. Impact-driven supply of
sodium and potassium to the atmosphere of Mercry. Icarus 75, 156–
170.
Potter, A.E., Morgan, T.H., 1985. Discovery of sodium in the atmosphere
of Mercury. Science 229, 651–653.
Potter, A.E., Morgan, T.H., 1988. Discovery of sodium and potassium vapor
in the atmosphere of Moon. Science 241, 657–680.
Potter, A.E., Morgan, T.H., 1997. Evidence for suprathermal sodium on
Mercury. Adv. Space Res. 19, 1571–1576.
Potter, A.E., Killen, R.M., Morgan, T.H., 2002. The sodium tail of Mercury.
Meteorit. Planet. Sci. 37, 1165–1172.
Smyth, W.H., Marconi, M.L., 1995a. Theoretical overview and modeling of
the sodium and potassium atmospheres of Mercury. Astrophys. J. 441,
839–864.
Smyth, W.H., Marconi, M.L., 1995b. Theoretical overview and modeling of
the sodium and potassium atmospheres of the Moon. Astrophys. J. 443,
371–392.
Smyth, S.M., Wilson, J.K., Baumgardner, J., Mendillo, M., 1999. Discovery of the distant lunar sodium tail and its enhancement following
the Leonid meteor shower of 1998. Geophys. Res. Lett. 26, 1649–
1652.
Smyth, S.M., Mendillo, M., Wilson, J.K., Baumgardner, J., 2001. Monitoring the Moon’s transient atmosphere with an all-sky imager. Adv. Space
Res. 27, 1181–1187.
Sprague, A.L., 1992. Mercury’s atomospheric bright spots and potassium
variations—a possible cause. J. Geophys. Res. 97, 257–264.
Verani, S., Barbieri, C., Benn, C., Cremonese, G., 1998. Possible detection
of meteor stream effects on the lunar sodium atmosphere. Planet. Space
Sci. 46, 1003–1006.
Photon-stimulated desorption of Na
Wilde, M., Beauport, I., Al-Shamery, K., Freund, H.-J., 1997. Photoinduced
processes on alkali covered surfaces: NO desorption from K/Cr2 O3
(0001). Surf. Sci. 390, 186–193.
Wilson, J.K., Smyth, S.M., Baumgardner, J., Mendillo, M., 1999. Modeling an enhancement of the lunar sodium tail during the Leonid meteor
shower of 1998. Geophys. Res. Lett. 26, 1645–1648.
Wurz, P., Lammer, H., 2003. Monte-Carlo simulation of Mercury’s exosphere. Icarus 164, 1–13.
Yakshinskiy, B.V., Madey, T.E., 1999. Photon-stimulated desorption as a
substantial source of sodium in the lunar atmosphere. Nature 400, 642–
644.
59
Yakshinskiy, B.V., Madey, T.E., 2000. Desorption induced by electronic
transitions of Na from SiO2 : relevance to tenuous planetary atmospheres. Surf. Sci. 451, 160–165.
Yakshinskiy, B.V., Madey, T.E., 2003. DIET of alkali atoms from mineral
surfaces. Surf. Sci. 528, 54–59.
Yakshinskiy, B.V., Madey, T.E., Ageev, V.N., 2000. Thermal desorption of
sodium atoms from thin SiO2 films. Surf. Rev. Lett. 7, 75–87.
Yarygin, V.I., Sidel’nikov, V.N., Kasikov, I.I., Mironov, V.S., Tulin, S.M.,
2003. Experimental study on the possibility of formation of a condensate of excited states in a substance (Rydberg matter). JETP Lett. 77 (6),
280–284.