Pulsed laser ablation of Sn and SnO2 targets: Neutral composition, energetics, and wavelength dependence

Article abstract of DOI:10.1021/jp000369a

We report time-of-flight mass spectrometric studies of neutral gas-phase species generated in 532 and 355 nm laser ablation of Sn and SnO₂ targets at intensities of ≈10⁸ W cm^-2, below the plasma threshold. A wavelength-dependent yield of Sn:Sn_xO_y species is observed for the oxide target, with Sn_xO_x (x = 1-3) the primary products at 532 nm and atomic Sn dominant at 355 nm. Sn and Sn₂ are the primary products of Sn metal ablation, and the relative Sn:Sn₂ yield increases at the shorter wavelength. The speed distributions of neutrals ejected from the oxide target are well represented by unshifted time-transformed Maxwell-Boltzmann (MB) distributions, while those ejected from the metal target exhibit bimodal MB distributions. Typical most-probable speeds are (1-2) × 10⁵ cm s^-1, with peak kinetic energies (KEs) of 1-2 eV. The implications of our results for pulsed laser deposition of SnO_x films will be emphasized.

Full text of DOI:10.1021/jp000369a

5324
J. Phys. Chem. B 2000, 104, 5324-5330
Pulsed Laser Ablation of Sn and SnO2 Targets: Neutral Composition, Energetics, and
Wavelength Dependence
Scott A. Reid,*,† Wei Ho,† and F. J. Lamelas‡
Department of Chemistry and Department of Physics, Marquette UniVersity, P.O. Box 1881,
Milwaukee, Wisconsin 53201-1881
ReceiVed: January 31, 2000; In Final Form: March 27, 2000
We report time-of-flight mass spectrometric studies of neutral gas-phase species generated in 532 and 355
nm laser ablation of Sn and SnO2 targets at intensities of ≈108 W cm-2, below the plasma threshold. A
wavelength-dependent yield of Sn:SnxOy species is observed for the oxide target, with SnxOx (x ) 1-3) the
primary products at 532 nm and atomic Sn dominant at 355 nm. Sn and Sn2 are the primary products of Sn
metal ablation, and the relative Sn:Sn2 yield increases at the shorter wavelength. The speed distributions of
neutrals ejected from the oxide target are well represented by unshifted time-transformed Maxwell-Boltzmann
(MB) distributions, while those ejected from the metal target exhibit bimodal MB distributions. Typical most-
probable speeds are (1-2) × 105 cm s-1, with peak kinetic energies (KEs) of 1-2 eV. The implications of
our results for pulsed laser deposition of SnOx films will be emphasized.
I. Introduction
Over the past 30 years, a variety of thin-film materials have
been fabricated using PLD21 and many studies of laser ablation
have been reported. The ablation threshold is associated with
formation of a dense plume or plasma above the target surface
and the onset of gas-dynamic effects, which occur for material
ejection rates exceeding 0.1-1 monolayer per pulse. In PLD
nomenclature, this threshold delineates regimes of incongruent
and congruent (i.e., stoichiometric) material transfer.22 The latter
arises from nonthermal target erosion characteristic of a laser-
generated plasma. Early studies of nanosecond ablation typically
invoked a transient surface heating mechanism, based on
expected thermalization times that are short compared to the
laser pulse duration. Recent studies have shown the importance
of electronic effects,23 and for semiconductors a parameter
important in governing these effects is the ratio of photon (hν)
Thin-film materials based on optically transparent semicon-
ductors have important applications as gas sensors, in display
systems, and in solar cells.1-5 The n-type semiconductor SnO2
is a promising base material for such devices. Thin polycrys-
talline SnO2 films exhibit high sensitivity to a variety of gases,
and doping can improve selectivity.6-10 Film sensitivity is
increased for average grain sizes D comparable to the length
of the space charge region L, which is typically several
nanometers.11 When
D < 2L, entire grains are electron
depleted and an increase in gas sensitivity is expected.11 Thin
films exhibiting D < 5 nm, comparable to 2L, have recently
been grown using pulsed laser deposition (PLD).12 This is but
one of several recent reports of PLD-generated SnOx films that
have utilized a variety of substrates, targets, and deposition
conditions.12-19 The as-grown films range from polycrystal-
line13,16 to amorphous12,19 in structure and with O:Sn stoichi-
ometries ranging from near 2:116 to <1:1.13,19 Very recently,
we produced the high-pressure o-SnO2 phase in high (∼50%)
yield from postdeposition heat treatment of initially amorphous
films of near 1:1 O:Sn stoichiometry.19
to band-gap (Egap) energies. Supragap excitation (hν > Egap
)
involves single photon band-to-band transitions in the bulk solid
as well as transitions among surface states,24 while subgap
excitation primarily involves transitions at localized defect states,
except at high intensities where multiphoton band-to-band
transitions contribute. Studies of photon energy dependencies
such as reported here are important in characterizing relative
contributions from thermal vs electronic processes. In the
ablation regime, secondary electronic processes including
formation of a laser-induced plasma, ionization of emitted
neutral species, and photoabsorption by free electrons will also
contribute.22-24 Here we focus on a intensity regime near but
below the plasma threshold, but will also report initial results
concerning laser intensity dependencies.
The variability in results obtained for PLD-generated SnOx
films illustrates a sensitivity to growth process variables and
underscores the importance of characterizing the composition
and energetics of products ablated from Sn-containing targets
at intensities typical of deposition. We recently reported the first
study for an SnO2 target, applying time-of-flight mass spec-
trometry (TOFMS) to analyze the neutral products of ablation
at 532 nm.20 In this paper we examine neutrals produced in
532 and 355 nm ablation of Sn and SnO2 targets at intensities
of 108 W cm-2, near but below the ablation threshold for both
targets. Our goal is to understand the mechanism(s) of the initial
laser-target interaction and correlate the properties of PLD-
generated SnOx films with growth process variables.
The outline of the paper is as follows. Section II is a summary
of the experimental methodology. Sections IIIa and IIIb include
our results for ablation of Sn and SnO2 targets at wavelengths
of 532 and 355 nm and intensities of ≈108 W cm-2. These
wavelengths correspond to photon energies well below and near,
respectively, the 3.5 eV band gap of SnO2.13 A discussion of
these results, focusing on the laser-target interaction and
implications for PLD studies, is provided in section IIIC. Our
conclusions are summarized in section IV.
* To whom correspondence should be addressed. E-mail: scott.reid@
marquette.edu.
† Department of Chemistry.
‡ Department of Physics.
10.1021/jp000369a CCC: $19.00 © 2000 American Chemical Society
Published on Web 05/06/2000

Pulsed Laser Ablation of Sn and SnO2 Targets
J. Phys. Chem. B, Vol. 104, No. 22, 2000 5325
II. Experimental Section
A diagram of the experimental apparatus was given in ref
20. Pulsed Nd:YAG lasers (Continuum NY-61, Surelite II) were
used for ablation and ionization. Typical ablation pulse energies
were 1-5 mJ in a 6-8 ns pulse (fwhm). The beam struck the
target along the surface normal, with a spot size of 1 mm. The
beam area at the target was estimated using burn paper, and
calculated intensities are based on a “top-hat” spatial profile.
The ablated material passed into the interaction region of a linear
time-of-flight mass spectrometer (TOFMS) and was ionized by
118.2 nm photons generated by tripling 10 mJ of the 355 nm
output of the second Nd:YAG system in a cell filled with 0.8
mbar of Xe. The separation d between target and ionization
region was fixed at 70 ( 1 mm for the oxide target and 73 (
1 mm for the metal target. Following extraction and acceleration,
the ions traversed a distance of 118 cm before striking a dual
microchannel plate detector. The signal was amplified (×25)
using a fast preamplifier (Stanford Research 445) and recorded
by a multichannel scaler (SR 430) with 5 ns bin width. The
ablation laser Q-switch was triggered at half the rate of the
ionization laser, allowing active background subtraction. Timing
of ablation and ionization laser Q-switch triggers was set using
a digital delay generator (SR DG535). The mass spectrometer
had a typical resolution (m/∆m) of ∼200 at a mass of 120 amu.
Two types of data will be reported. The first type is mass
spectra accumulated over typically 5000 ablation laser shots at
a set ablation-ionization laser delay, or neutral arrival time.
These spectra give the relative neutral distribution at a fixed
arrival time, or speed determined by the separation d between
target and ionization region. The second type of data to be
reported is the arrival-time (speed) distributions of mass-selected
species, obtained by varying the ablation-ionization laser delay
and accumulating mass spectra over 5000 shots at each delay
setting. Intensity fluctuations over the course of an experiment
were corrected for by normalizing to spectra repetitively
acquired at a reference delay near the peak of the Sn arrival-
time distribution.
Figure 1. Mass spectrum of the neutral products of 532 nm pulsed
laser ablation of an SnO2 target.
SnO2 is not detected. The IPs of both O (13.64 eV)27 and O2
(12.071 eV)27 lie above our ionizing photon energy (10.49 eV),
and the observed signals probably arise from secondary (electron
impact) ionization. We note that gas-phase SnxOx (x ) 1-8)
has previously been observed in equilibrium above heated SnO2
samples28 and oxygen saturated tin melts.29
Figure 2 displays the integrated mass counts (points) as a
function of neutral arrival time for Sn, SnO, and Sn2O2. We fit
the arrival-time distributions shown in Figure 2 to unshifted
time-transformed Maxwell-Boltzmann (MB) speed distributions
of the form30
The SnO2 targets were prepared by cold pressing ap-
proximately 10 g of SnO2 powder (Aldrich, >99.9%) into a
0.75’’ long by 0.5’’ diameter pellet at a typical pressure of 6000
psi. The pellets were annealed for 60-70 h at 1473 K prior to
use. The Sn target (Aldrich, 6 mm diameter rod, 99.99%) was
polished with a fine (typically 400) grit paper and cleaned with
acetone prior to use. A motorized, computer-controlled linear-
rotary motion feed through was used to simultaneously translate
and rotate the target. During experiments the main chamber and
flight tube pressures were 3 × 10-7 and 10-7 mbar, respectively.
S(t) ) (A/d3)(tpeak/t)4 exp{-2(tpeak/t)2}
(1)
In eq 1, d is the target-ionization region separation, tpeak is
the maximum in the arrival-time distribution, and A is a scaling
constant. Nonlinear least-squares fits to eq 1 give most-probable
speeds of 1.7 × 105 cm s-1 (Sn), 1.4 × 105 cm s-1 (SnO), and
1.2 × 105 cm s-1 (Sn2O2), or peak kinetic energies (KEs) of
1.8, 1.4, and 2.0 eV. The estimated uncertainty in all most-
probable speeds and KEs reported is (2 × 104 cm s-1 and (0.1
eV, respectively. Temperature values of 1.1 × 104 K (Sn), 7.8
× 103 K (SnO), and 1.1 × 104 K (Sn2O2) are obtained.
We also fit our data to a shifted Maxwellian distribution,
which describes the speed distribution of molecules in a
supersonic jet31 and has been used to model particle speed
distributions in laser ablation.32-36 The time-transformed dis-
tribution has the form33
III. Results and Discussion
A. Laser Ablation of SnO2 at 532 and 355 nm. We first
examined the 532 nm ablation of an SnO2 target. Figure 1
displays a mass spectrum obtained at a pulse energy of 5 mJ,
or intensity of approximately 108 W cm-2, and neutral arrival
time of 65 µs. The background arises from ablated ions, is
integrated over all arrival times, species, and states, and thus is
not indicative of the true ion fraction. It is most pronounced at
the beginning of a given experiment, as is the neutral signal
from surface impurities, primarily the alkali metals Na, K, and
Ca. The presence of these impurities is attributed to their high
volatility, low ionization potentials (IPs), and presence on the
surface in ionic compounds.25
S(t) ) At-4 exp{-c1t-2 + c2t-1}
(2)
where c1 ) md/(2kT), m is the molecular mass, T is the absolute
temperature, and c2 ) 2c1u/d, where u is the stream velocity.
The fits obtained to this distribution were of a quality similar
to those obtained using eq 1, with similar derived temperatures
and near-zero stream velocities (u ≈ 0). We therefore report fit
parameters only for the unshifted distribution (eq 1).
The signal observed for Sn-containing compounds is domi-
nated by SnO and Sn2O2. Smaller signals are observed for Sn,
Sn3O3, and the hypermetallic26 compound Sn2O. Gas-phase

5326 J. Phys. Chem. B, Vol. 104, No. 22, 2000
Reid et al.
Figure 3. Pulse energy dependence of the total yield of neutral
Sn-containing species for 532 nm ablation of an SnO2 target.
Figure 2. Arrival-time distributions (points) for (a) Sn, (b) SnO, and
(c) Sn2O2 from 532 nm ablation of an SnO2 target. The lines are fits to
an unshifted time-transformed Maxwell-Boltzmann speed distribution,
from which the most probable speed, peak kinetic energy, and
temperature were derived.
We estimated relative species abundances by integrating the
fits to eq 1. SnO and Sn2O2 fragmentation to Sn and SnO,
respectively, was negligible in the ionization step (<5%).20
Relative Sn:SnO:Sn2O2 abundances of 0.22:0.65:1.0 are obtained
under the assumption of similar photoionization efficiencies,
which is questionable. The IP of Sn2O2 is unknown, while that
of SnO (10.5 ( 0.5 eV)27 lies close to the ionizing photon energy
of 10.49 eV and is far above that of Sn (7.344 eV).27 Intensity
calibration would require knowledge of vacuum ultraviolet
absorption cross sections and ionization quantum yields, yet
important relative trends may still be identified, e.g., the
dependence of neutral composition on laser intensity and
wavelength.
Figure 4. Thickness distribution for SnOx (x ≈ 1) films grown using
532 nm PLD from an SnO2 target. Distributions were determined by
analyzing interference fringes, as described in the text, and are given
(points) for cuts along (a) semimajor and (b) semiminor axes of the
elliptically patterned film. The lines in each figure represent fits to a
cosn+3 θ distribution, from which sharpness parameters (n) of (a) 6
and (b) 15 were derived.
profile, and Figure 4 displays thickness distributions for cuts
along the semimajor and semiminor axes. These distributions
were determined from the interference fringes using the equa-
tion38
Figure 3 displays the neutral signal dependence on laser pulse
energy in the range 2.1-20.1 mJ, corresponding to intensities
of approximately 5 × 107-5 × 108 W cm-2. The signal
increases exponentially at low pulse energies (<8 mJ pulse-1
)
and then flattens. We observed similar behavior for the Sn target,
and in that case the break in the curve correlates with the onset
of visible emission from atomic neutrals and ions, and occurs
at a fluence similar to the reported ablation threshold at 248
nm.37 We therefore correlate this point with the onset of ablation,
an assumption supported by estimates of the neutral angular
dependence determined via the thickness distribution of SnO
films grown on Si(001) substrates using 532 nm PLD at a pulse
energy of 10 mJ. The as-grown films exhibit an elliptical
interference pattern, the shape arising from the laser spatial
2t ) (m + 1/2)λn
(3)
where m ) 0, 1, 2, ... and λn ) λ0/n. In our analysis n was
taken as 1.5, and the violet fringes were used and λ0 was
therefore taken as 400 nm. Given an angular flux distribution
of form cosn θ, where n is the sharpness parameter, the
distributions can be represented by the following equation:39
D(θ) ) A cosn+3θ
(4)

Pulsed Laser Ablation of Sn and SnO2 Targets
J. Phys. Chem. B, Vol. 104, No. 22, 2000 5327
Figure 5. Mass spectrum of the neutral products of 355 nm ablation
of an SnO2 target. Note the relative increase in the ratio of signals for
Sn:SnxOy species compared to ablation at 532 nm (Figure 1).
The fits to eq 4 are shown as solid lines in Figure 4, and
from these fits n values of 6 ( 1 and 15 ( 2 are obtained for
the semimajor and semiminor axis distributions, respectively.
These values are significantly larger than expected in a thermal
regime (n ) 1) and are consistent with the presence of gas-
dynamic effects.
Figure 6. Arrival-time distributions (points) for (a) Sn and (b) SnO
following 355 nm ablation of an SnO2 target. The lines in each plot
are fits to an unshifted time-transformed Maxwell-Boltzmann speed
distribution.
The photon-energy dependence of the ablation products was
investigated by collecting a second set of data using a
wavelength of 355 nm. Figure 5 displays the mass spectrum
obtained at a pulse energy of 2.5 mJ, or intensity of ap-
proximately 5 × 107 W cm-2, and neutral arrival time of 40
µs. Marked changes in peak intensities and an increased
background are observed compared to ablation at 532 nm
(Figure 1). A lower pulse energy was used here to reduce the
background signal, yet our preliminary fluence dependence
studies show that the neutral composition and energetics vary
little between 2.5 and 5 mJ. Hence the differences between
Figures 1 and 5 cannot be attributed to this factor. The arrival-
time distributions at 355 nm are still well described by a single-
component MB distribution. The Sn and SnO distributions
(points) and fits to eq 1 (lines) are shown in parts a and b,
respectively, Figure 6. Most-probable speeds of 2.2 × 105 cm
s-1 (Sn) and 1.7 × 105 cm s-1 (SnO), peak KEs of 3.0 and 2.0
eV, and temperatures of 1.8 × 104 and 1.2 × 104 K, respectively,
are obtained. The relative Sn:SnO abundance is 5.2:1.0, an
increase of approximately 15 times compared to ablation at 532
nm. Comparing the relative yield of Sn to all tin oxide species,
i.e., Sn:SnxOy, the increase is approximately 40 times.
B. Laser Ablation of Sn at 532 and 355 nm. We conducted
a similar set of experiments for the Sn target at intensities of
approximately half the ablation threshold. Figure 7 displays the
mass spectrum obtained following 532 nm ablation of the Sn
target at 2.1 mJ pulse-1, corresponding to a intensity of
approximately 5 × 107 W cm-2, and neutral arrival time of 55
µs. The primary Sn-containing compounds observed are of form
Snx (x ) 1-3), with intensities that decrease with increasing x.
Figure 8 displays nonlinear least-squares fits of the observed
mass peaks for Sn and Sn2 based on natural abundance isotopic
distributions,27 with the width and overall intensity used as fit
parameters.
Figure 7. Mass spectrum of the neutral products of 532 nm pulsed
laser ablation of a Sn target.
by two-component distributions with a dominant fast component
and weak slow component peaking near 110 µs. The data and
fits for Sn and Sn2 are shown in Figure 9. For Sn, the fit gave
tpeak values of 44.7 and 115.9 µs for the two components and a
ratio of fast to slow component of ≈5:1. The primary (fast)
component corresponds to a most-probable speed of 1.6 × 105
cm s-1 or peak KE of 1.7 eV and a temperature of 1.1 × 104
K. The Sn2 fit gave tpeak values of 55.5 and 107 µs and ratio of
fast to slow component of ≈3:1. The primary component
corresponds to a most-probable speed of 1.3 × 105 cm s-1 or
peak KE of 1.8 eV and a temperature of 1.1 × 104 K. We
estimated the relative abundance using the integrated fits and
find a Sn:Sn2 ratio of 1.4:1.0 under the assumption of similar
photoionization efficiencies.
Figure 10 displays the mass spectrum obtained at a neutral
arrival time of 50 µs following 355 nm ablation of the Sn target
at a pulse energy of 1.2 mJ or intensity of approximately 3 ×
107 W cm-2. The relative cluster signal is reduced and a higher
background is observed. The Sn distribution was fit to a two-
The arrival-time distributions and relative abundances were
investigated. The Sn and Sn2 distributions were not well fit by
single-component MB distributions, but were better represented

5328 J. Phys. Chem. B, Vol. 104, No. 22, 2000
Reid et al.
Figure 8. Time-of-flight mass spectra (points) and fits (lines) for (a)
Sn and (b) Sn2 mass peaks, taken from the spectrum in Figure 7. The
fits demonstrate near-natural-abundance isotopic distributions.
Figure 9. Arrival-time distributions (points) for (a) Sn and (b) Sn2
from 532 nm ablation of Sn target. The distributions are bimodal and
were fit to the sum of two unshifted time-transformed MB speed
distributions. The solid line in each plot is the two-component fit, and
the individual components are shown as dashed lines.
component distribution, shown in Figure 11a, which gave peak
arrival times of 39.2 and 75.7 µs, and a ratio of fast to slow
component of ∼2:1. The Sn2 arrival-time distribution, shown
in Figure 11b, was fit to a single-component distribution, since
the small signal level made it difficult to discriminate between
a single- and two-component distribution. The fit gave a peak
arrival time of 66.3 µs. These values for Sn and Sn2 correspond
to most-probable speeds of 1.7 × 105 cm s-1 and 1.1 × 105 cm
s-1, peak KEs of 1.8 and 1.5 eV, and an equivalent temperature
of 1.1 × 104 K, considering only the primary Sn component.
The relative Sn:Sn2 abundance determined as above is 3.5:1.0,
an increase of 2.5 times over that at 532 nm.
C. Discussion. In examining our results and their implications
for PLD-generated SnOx films, we begin with the neutral
composition and its wavelength dependence for the oxide target.
At a subgap wavelength (532 nm), SnxOx (x ) 1-3) are the
dominant ablation products. A marked change is observed at
the near-gap (355 nm) wavelength, where the relative Sn:SnxOy
and SnO:Sn2O2 abundances increase and Sn is the dominant
product. This is a key result of our work, which may signal a
fundamental change in mechanism for the laser-target interac-
tion, e.g., the onset of electronic processes. Alternatively, laser-
induced processes in the gas phase may become important due
to the higher photon energy and increased absorption. For
example, 355 nm laser irradiation overlaps the strong D1Π r
X1Σ+ system of SnO.40 For both targets 355 nm ablation
produces a higher yield of ionic products, observed as an
increase in background signal. Mass-resolved detection of
ablated ions should give additional insight into the origin of
the wavelength dependence.
Figure 10. Mass spectrum of the neutral products of 355 nm pulsed
laser ablation of a Sn target.
following flooding of the ICR cell with resonant electrons
showed a significant population of neutral Snx (x ) 3-6). In
contrast, ablation at 248 nm was found to produce Sn, Sn+,
and some Sn2+ with no evidence of clusters.37 These results in
combination with ours reveal a consistent decrease in cluster
yield at shorter wavelengths, which may result from laser-
induced processes in the gas phase.
Results for the Sn target can be compared to studies at other
wavelengths. Jackson et al.41 observed Snx (x ) 1-7) species
using Fourier transform ion cyclotron resonance (ICR) mass
spectrometry following 1064 nm ablation of tin foil at a intensity
of 100 J cm-2. Changes in the intensities of negative ion peaks
The KE distributions of ejected species from the oxide target
are consistently MB-like. Typical most-probable speeds are (1-
2) × 105 cm s-1, with peak KEs of 1-2 eV and temperatures
of 104 K. The peak KEs exhibit little variation with wavelength,
with the exception of atomic Sn ejected at 355 nm, which

Pulsed Laser Ablation of Sn and SnO2 Targets
J. Phys. Chem. B, Vol. 104, No. 22, 2000 5329
component may thus reflect thermal vaporization, or result from
fragmentation of higher clusters (e.g., Sn3) in the ionization step.
We next consider the specific implications of our results for
PLD-generated SnOx films. Vispute et al.13 observed an oxygen-
deficient SnOx (x < 2) film produced in 694 nm PLD from an
SnO2 target under high vacuum conditions. Similarly, our
results19 show a reduced O:Sn stoichiometry of films grown at
532 nm which correlates with that of the neutral gas-phase Sn-
containing products and indicates that the undetected oxygen
is not ultimately incorporated into the deposited film. It was
assumed in ref 13 that oxygen loss occurred at the target;
however, we can draw no conclusions about this from our
results.
Additional insight is gained from the work of Hu et al.,12
who reported polycrystalline SnO2 films of similar orientation
and average grain size (4-5 nm) grown using 248 nm PLD
from SnO2 and Sn targets, followed in each case by postdepo-
sition annealing in an oxygen atmosphere. Our results suggest
that UV ablation of both types of targets produces primarily
atomic Sn, at least for the neutral products, with similar
energetics. Therefore, it is not surprising that similar as-grown
films are produced by Hu et al. with metal and oxide targets.
The broader implication of our work is that UV PLD from oxide
targets may produce a more severe oxygen deficiency in the
general case.
To date only one PLD study has reported the growth of
polycrystalline near-stoichiometric (i.e., SnO2) films on room-
temperature substrates under high vacuum conditions,16 where
a wavelength of 1064 nm and energy densities of 20-150 J
cm-2 were used. It may be that the trend we observe, an increase
in oxygen content with increasing wavelength, continues to
longer wavelength. Alternatively, the high intensities used may
have led to stoichiometric deposition. Our preliminary studies
at 532 nm do show a modest increase in yield of oxide species
at higher intensity, but further studies are needed to sort out
the influence of wavelength and intensity effects in producing
the interesting result of ref 16. In general, the growth of
polycrystalline SnO2 films from either 532 or 355 nm PLD will
require either postdeposition oxidation or reactive deposition
in an oxygen atmosphere.
Figure 11. Neutral arrival-time distributions (points) for (a) Sn and
(b) Sn2 from 355 nm ablation of a Sn target. The Sn distribution was
fit to a two-component MB distribution, while that for Sn2 was fit to
a single component distribution. The lines represent the fits following
the notation used in Figure 9.
exhibits the highest most-probable speed of 2.2 × 105 cm s-1
and peak KE of 3 eV or temperature of ≈1.8 × 104 K. This
may indicate that Sn is not a primary ablation product but
originates from laser-induced processes in the gas phase (e.g.,
photodissociation of SnO monomers or clusters), a hypothesis
supported by the dramatic increase in Sn:SnxOy at 355 nm. We
note that agreement with a MB distribution is not evidence for
a thermal mechanism. Studies of 248 nm laser desorption of
Au from a metal target shows MB-like distributions, but with
mean speeds that are inconsistent with a thermal model.42,43
Moreover, Au produced at intensities near and above the ablation
threshold also show MB-like distributions, suggesting that 10-
20 collisions per particle are insufficient to impart a significant
stream velocity.42 Our inferred neutral angular dependence
(Figure 4) also suggests the presence of gas-dynamic effects in
a intensity regime where the arrival-time distributions are still
MB-like.
Neutral species ejected from the metal target consistently
exhibit bimodal MB distributions. Our results are qualitatively
similar to those of Theim et al.,32 who examined the speed
distributions of neutral atoms from ablation of metal and
compound targets at intensities below plasma threshold. Atomic
Cu produced in 532 nm ablation from a Cu target at intensities
of 108 W cm-2 exhibited an MB-like speed distribution
corresponding to a temperature of 1.12 × 104 K, similar to that
obtained here for atomic Sn from the metal target, while at
higher intensities bimodal distributions were observed. This may
reflect a hybrid electronic/thermal mechanism,44 and indeed the
influence of electronic processes in desorption from metals at
low photon energies has recently been demonstrated.45 The slow
IV. Conclusions
We have probed neutral species generated in laser ablation
of SnO2 and Sn targets at both 532 and 355 nm and at intensities
of 108 W cm-2. For the oxide target, the major (>85%) Sn-
containing species observed at 532 nm are of composition SnxOx
(x)1-3). The ratio of Sn:SnxOy species is increased by
approximately 40 times for ablation at 355 nm, which may arise
from the onset of electronic processes in the target or from laser-
induced processes in the gas phase. The primary products of
ablation from Sn targets at intensities near threshold are Sn and
Sn2, with an increase in the Sn:Sn2 ratio of 2.5 times at the
shorter wavelength.
The neutral arrival-time distributions for species ejected from
the oxide target were well fit by unshifted time-transformed
MB speed distributions, while those for species ejected from
the metal target were consistently bimodal. The fits yield typical
most-probable speeds of 105 cm s-1, peak KEs of 1-2 eV, and
temperatures of ∼104 K. The observed KEs exhibit little
variation with wavelength, with the exception of Sn ejected from
the oxide target. The angular distribution of ejected neutrals
inferred from analysis of the thickness distribution of amorphous
SnO films deposited using 532 nm ablation from the oxide target
gives sharpness parameters n of 6-15, indicating the presence

5330 J. Phys. Chem. B, Vol. 104, No. 22, 2000
Reid et al.
(14) Godbole, V. P.; Vispute, R. D.; Chaudhari, S. M.; Kanetkar, S.
M.; Ogale, S. B. J. Mater. Res. 1990, 5, 372.
(15) Lal, R.; Grover, R.; Vispute, R. D.; Viswanathan, R.; Godbole, V.
P.; Ogale, S. B. Thin Solid Films 1991, 206, 88.
(16) Dai, C. M.; Su, C. S.; Chuu, D. S. Appl. Phys. Lett. 1990, 57, 1879.
(17) Morosova, N. V.; Gas’kov, A. M.; Kuznetsova, T. A.; Putilin, F.
N.; Rumyantseva, M. N.; Shatnov, V. I. Inorg. Mater. 1996, 32, 292.
(18) Hu, W. S.; Liu, Z. G.; Zheng, J. G.; Hu, X. B.; Guo, X. L.; Go¨pel,
W. J. Mater. Sci. Mater. Elec. 1997, 8, 155.
(19) Lamelas, F. J.; Reid, S. A. Phys. ReV. B 1999, 60, 9347.
(20) Reid, S. A. Chem. Phys. Lett. 1999, 301, 517.
(21) See, e.g.: Pulsed Laser Deposition of Thin Films; Chrisey, D. B.,
Hubler, G. K., Eds.; Wiley: New York, 1994.
(22) Lowndes, D. H. In Laser Ablation and Desorption; Miller, J. C.,
Haglund, R. F., Jr., Eds.; Academic: San Diego, 1998.
(23) See, e.g.: (a) Laser Ablation: Principles and Applications; Miller
J. C., Ed.; Springer-Verlag: Berlin, 1994. (b) Laser Ablation: Mechanisms
and ApplicationssII; Miller, J. C., Geohegan, D. B., Eds.; AIP: New York,
1994.
(24) Haglund, R. F., Jr.; Itoh, N. In Laser Ablation: Principles and
Applications; Miller J. C., Ed.; Springer-Verlag: Berlin, 1994.
(25) Gill, C. G.; Allen, T. M.; Anderson, J. E.; Taylor, T. N.; Kelly, P.
B.; Nogar, N. S. Appl. Opt. 1996, 35, 2069.
(26) See, e.g.: Wang, S. L., Ledingham, K. W. D.; Singhal, R. P. J.
Phys. Chem. 1996, 100, 11282, and references therein.
(27) Radzig, A. A.; Smirnov, B. M. Reference Data on Atoms,
Molecules, and Ions; Springer-Verlag: Berlin, 1985.
(28) Colin, R.; Drowart, J.; Verhaegen, A. G. Trans. Faraday Soc. 1965,
61, 1364.
of gas-dynamic effects in a intensity regime where the arrival-
time distributions are still MB-like.
Our results are helpful in understanding previous studies of
PLD-generated SnOx films, and illustrate the importance of
fundamental characterization of pulsed laser ablation in under-
standing the properties of PLD films. However, a complete
understanding of laser-induced particle emission from these
targets should include the intensity dependence of neutral
composition and energetics at different wavelengths, which will
be important in deconvoluting wavelength and intensity effects.
A detailed study of this will be the subject of a future report.
We also plan to explore other parameters including ion
composition and energetics, spectroscopic characterization of
gas-phase species in the plume, and direct measurement of the
angular distribution of neutral and ionic species.
Acknowledgment. S.A.R. acknowledges the National Sci-
ence Foundation CAREER program (CHE-9702803), the Arnold
and Mabel Beckman Foundation, the Marquette Center for
Sensing Technology, and the Petroleum Research Fund admin-
istered by the American Chemical Society for partial support
of this research, and thanks Binyong Liang and Jason P. Schmidt
for assistance in construction and calibration of the experimental
apparatus. F.J.L. acknowledges C. Pugh-Smith for assistance
in the production of tin oxide targets and N. Schook for technical
support. The authors acknowledge useful discussions with J.
Hossenlopp and thank her for stimulating our interest in tin oxide
films.
(29) Zimmermann, E.; Ko¨nigs, S.; Neuschu¨tz, D. Z. Phys. Chem. 1996,
193, 195.
(30) van Ingen, R. P. J. Appl. Phys. 1996, 79, 467.
(31) Anderson, J. B.; Andres, R. P.; Fenn, J. B. AdV. Chem. Phys. 1966,
10, 175.
(32) Thiem, T. L.; Watson, L. R.; Dressler, R. A.; Salter, R. H.; Murad,
E. J. Phys. Chem. 1994, 98, 11931.
(33) Utterback, N. G.; Tang, S. P.; Friichtenicht, J. F. Phys. Fluids 1976,
19, 900.
(34) Namiki, A.; Kawai, T.; Ichige, K. Surf. Sci. 1986, 166, 129.
(35) Zheng, J. P.; Huang, Z. Q.; Shaw, D. T.; Kwok, H. S. Appl. Phys.
Lett. 1989, 54, 280.
(36) Kools, J. C. S.; Baler, T. S.; De Zwart, S. T., Dieleman, J. J. Appl.
Phys. 1992, 71, 4547.
(37) Timm, R.; Willmott, P. R.; Huber, J. R. J. Appl. Phys. 1996, 80,
1794.
References and Notes
(1) Iida, H.; Shibata, N.; Mishuka, T.; Ito, A.; Karasawa, H.; Hayashi,
Y. IEEE Electron DeVice Lett. 1983, EDL-4, 157.
(2) Schroder, D. K. IEEE Trans. Electron DeVices 1978, 25, 90.
(3) Go¨pel, W. Sens. Actuators, B 1994, 18/19, 1.
(4) Takahata, K. In Chemical Sensor Technology; Seiyama, T., Ed.;
Elsevier: Amsterdam, 1988; p 39.
(38) Jenkins, F. A.; White, H. E. Fundamentals of Optics; McGraw-
Hill: New York, 1950; pp 254-278.
(5) Heiland, G. Sens. Actuators 1982, 2, 343.
(6) Semancik, S.; Fryberger, T. B. Sens. Actuators, B 1990, 2, 97.
(7) Sberveglieri, G.; Groppelli, S.; Nelli, P.; Camanzi, A. G. Sens.
Actuators, B 1991, 5, 343.
(8) Kohl, D. Sens. Actuators, B 1990, 1, 158.
(9) Matsyshima, S.; Teraoka, Y.; Miura, N.; Yamazoe, N. Jpn. J. Appl.
Phys. 1988, 27, 1798.
(10) Simianu, V. C.; Hossenlopp, J. M. Appl. Organomet. Chem. 1997,
11, 147.
(11) Serrini, P.; Briois, V.; Horrillo, M. C.; Traverse, A.; Manes, L.
Thin Solid Films 1997, 304, 113.
(12) Hu, W. S.; Liu, Z. G.; Wu, Z. C.; Feng, D. Mater. Lett. 1996, 28,
369.
(13) Vispute, R. D.; Godbole, V. P.; Chaudhari, S. M.; Kanetkar, S.
M.; Ogale, S. B. J. Mater. Res. 1988, 3, 1180.
(39) Saenger, K. L. In Pulsed Laser Deposition of Thin Films; Chrisey,
D. B., Hubler, G. K., Eds.; Wiley: New York, 1994.
(40) Dube, P. S.; Rai, D. K. J. Phys. B: At. Mol. Phys. 1971, 4, 579.
(41) Jackson, P.; Dance, I. G.; Fisher, K. J.; Willett, G. D.; Gadd, G. E.
Int. J. Mass Spectrom. Ion Processes 1996, 157/158, 329, and references
therein.
(42) Bennett, T. D.; Grigoropoulos, C. P.; Krajnovich, D. J. J. Appl.
Phys. 1995, 102, 726.
(43) Bennett, T. D.; Krajnovich, D. J.; Grigoropoulos, C. P. Phys. ReV.
Lett. 1996, 76, 1659.
(44) See, e.g.: Lee, I.; Callcott, T. A.; Arakawa, E. T. Phys. ReV. B
1993, 47, 6661.
(45) Hoheisel, W.; Vollmer, M.; Tra¨ger, F. Phys. ReV. B 1993, 48, 17463.