Reaction of sulfur dioxide with Ni(1 0 0) and Ni(1 0 0)-p(2 × 2)-O

Article abstract of DOI:10.1016/j.susc.2005.07.003

XPS, TPD, LEED and STM have been used to probe the interaction between sulfur dioxide and the Ni(1 0 0) and Ni(1 0 0)-p(2 × 2)-O surfaces. On Ni(1 0 0) at 300 K SO₂ disproportionates according to 2SO₂ → S(a) + SO₃(a) + O(a). Sulfur and sulfite occupy sites in a p(2 × 2) arrangement, while oxygen adsorbs into c(2 × 2) domains amid Ni chains of (n√2×2√2)R45°/ (2√2×n√2)R45° (n = 7-12) periodicity that are presumed to be due to segregation of oxygen to the subsurface. On Ni(1 0 0)-p(2 × 2)-O at 300 K SO₂(a) reacts with O(a) to form SO₃(a). Sulfite adsorbs into p(2 × 2) islands encompassed by c(2 × 2)-O. TPD measurements with ¹⁸O are suggestive of a monodentate sulfite binding configuration.

Full text of DOI:10.1016/j.susc.2005.07.003

Surface Science 592 (2005) 141–149
www.elsevier.com/locate/susc
Reaction of sulfur dioxide with Ni(100)
and Ni(100)-p(2 · 2)-O
*
Ali R. Alemozafar, Robert J. Madix
Department of Chemical Engineering, Stanford University, Stanford, CA 94305-5025, USA
Received 2 May 2005; accepted for publication 6 July 2005
Available online 25 July 2005
Abstract
XPS, TPD, LEED and STM have been used to probe the interaction between sulfur dioxide and the Ni(100) and
Ni(100)-p(2 · 2)-O surfaces. On Ni(100) at 300 K SO2 disproportionates according to 2SO2 ! S(a) + SO3(a) + O(a).
Sulfur and sulfite occupy sites in a p(2 · 2) arrangement, while oxygen adsorbs into c(2 · 2) domains amid Ni chains of
p
p
p
p
(n 2 · 2 2)R45ꢀ/(2 2 · n 2)R45ꢀ (n = 7–12) periodicity that are presumed to be due to segregation of oxygen to the
subsurface. On Ni(100)-p(2 · 2)-O at 300 K SO2(a) reacts with O(a) to form SO3(a). Sulfite adsorbs into p(2 · 2) islands
encompassed by c(2 · 2)-O. TPD measurements with 18O are suggestive of a monodentate sulfite binding configuration.
ꢁ 2005 Published by Elsevier B.V.
Keywords: Scanning tunneling microscopy; Thermal desorption spectroscopy; Surface chemical reaction; Nickel; Sulfur; Oxygen;
Sulfur dioxide; Single crystal surfaces
1. Introduction
troposphere. Investigations of the interaction be-
tween SO2 and various heterogeneous (metal and
non-metal) surfaces will aid in the understanding
of atmospheric phenomena, where an assessment
of gas–surface interactions is required for accurate
climate models.
Over the past twenty years the adsorption and
reaction of SO2(g) has been investigated on a num-
ber of single crystal metal surfaces with an array of
surface analytical techniques, such as X-ray photo-
electron spectroscopy (XPS) [3,4], near-edge X-ray
absorption fine structure spectroscopy (NEXAFS)
[5], temperature-programmed desorption (TPD)
Sulfur dioxide (SO2) is an environmental pollu-
tant. It is a byproduct of the process used to gen-
erate electricity in coal-fired power plants [1]. SO2
emissions can lead to adverse health and environ-
mental effects. SO2 is an irritant and a key compo-
nent in the formation of sulfate aerosol particles
[2], which act as cloud condensation nuclei in the
*
Corresponding author. Tel.: +1 6507232402; fax: +1
6507239780.
E-mail address: bobcat@stanford.edu (R.J. Madix).
0039-6028/$ - see front matter ꢁ 2005 Published by Elsevier B.V.
doi:10.1016/j.susc.2005.07.003

142
A.R. Alemozafar, R.J. Madix / Surface Science 592 (2005) 141–149
[3,4,6], reflection absorption infrared spectroscopy
(RAIRS) [7,8], high-resolution electron energy loss
vibrational spectroscopy (HREELS) [9] and scan-
ning tunneling microscopy (STM) [4,6,10,11].
Depending on the surface temperature, SO2 ad-
sorbs molecularly and dissociatively with the sur-
faces of Fe, Rh, W, Ni, Pd, Pt, Cu, and Zn to
form SOx (x = 1, 2, 3, or 4), whereas adsorption
on Ag is only molecular [12,13]. Oxygen adsorbed
on the Ag(110) surface affects the oxidation of
SO2 to SO3 (sulfite) and SO4 (sulfate) [3,6].
Ohta et al. previously probed the reaction be-
tween SO2(g) and the Ni(100) and Ni(100)-
c(2 · 2)-O surfaces using NEXAFS [5]. Their
results indicate that below 300 K SO2 dispropor-
tionates on Ni(100) according to 3SO2(a) !
S(a) + 2SO3(a). At room temperature a fraction
of the sulfite decomposes into SO2(g) and O(a),
leaving the surface partially covered with sulfur,
sulfite and atomic oxygen, i.e. 2SO2(a) ! S(a) +
SO3(a) + O(a). On Ni(100)-c(2 · 2)-O at 300 K,
SO2(a) reacts with oxygen to give SO3 and SO4
in approximately a 3:1 ratio.
evaporation onto a gold foil (ꢁ4 lA, 15 min) prior
to imaging. Unless otherwise stated, all scans were
taken in constant height mode.
The Ni(100) crystal used was aligned to within
0.5ꢀ of the (110) plane using Laue backscattering
and was mechanically polished down to 0.3 lm
alumina paste. The crystal was cleaned in vacuum
by three Ar ion sputter (2 lA, 500 eV, 15 min at
600 K) and anneal (800 K, 10 min) cycles, with
the first anneal done in an oxygen atmosphere
(1 · 10ꢀ7 Torr) to cleanse the surface of impurities
observed in STM images. Hydrogen treatment at
600 K followed by another sputter/anneal cycle re-
moved residual oxygen. A sharp p(1 · 1) LEED
pattern and AES-spectra showed a well-order
surface devoid of sulfur, carbon and oxygen
impurities.
The crystal could be cooled to 120 K with liquid
nitrogen and heated to 1100 K by electron bom-
bardment to the back of the crystal. The tempera-
ture was monitored by
a Chromel–Alumel
thermocouple spot-welded to the back of the
crystal. The STM ramp housing the crystal and
the STM scan head were allowed to thermally
equilibrate for 30 min prior to STM measure-
ments.
In this paper we report XPS, TPD, LEED and
STM studies of the reaction between SO2(g) with
clean and the oxygen-covered Ni(100)-p(2 · 2)-
O. STM has been used to elucidate the interaction
between SO2 and the Ni(100) and Ni(100)-
p(2 · 2)-O surfaces. Binding configurations and
the distribution of structures produced by the reac-
tion products are discussed.
Separate XPS measurements were made in a
second UHV system consisting of interconnected
preparation and analysis chambers. The analysis
chamber exhibited a base pressure of 2 · 10ꢀ10
Torr and was equipped with LEED optics, a Per-
kin–Elmer 04-548 dual anode X-ray source, an
EA-10-plus hemispherical energy analyzer from
SPECS and a UTI 100c QMS used for TPD mea-
surements. The ionizer of the QMS was enclosed
in a glass cap with a small hole facing the crystal
surface. A computer connected to the QMS was
used to record TPD spectra. The preparation
chamber reached a base pressure of 8 · 10ꢀ10 Torr
and was equipped with a sputter ion gun and stain-
less steel gas dosers. The two chambers were iso-
lated from each other during experiments.
2. Experimental
Experiments were performed in an ultrahigh
vacuum chamber equipped with STM, low energy
electron diffraction (LEED), Auger electron spec-
troscopy (AES) and a quadrupole mass spec-
trometer (QMS) for temperature-programmed
desorption (TPD) measurements. The chamber
was equipped with a sputter ion gun and stainless
steel gas dosers. The system exhibited a base pres-
sure of 4 · 10ꢀ10 Torr which rose to approximately
7 · 10ꢀ10 Torr during experiments.
In this system the crystal was supported by two
W wires spot-welded to the back. The temperature
was monitored by a Chromel–Alumel thermocou-
ple spot welded to the back edge of the crystal. The
surface was cleaned by three sputter-anneal cycles,
with the first anneal (800 K) done in an O2(g)
The homemade ‘‘Johnnie Walker’’ type STM
(RHK STM 100) employed in this study utilized
a Pt/Ir tip, which was conditioned via induced field

A.R. Alemozafar, R.J. Madix / Surface Science 592 (2005) 141–149
143
atmosphere (10ꢀ7 Torr). Hydrogen treatment
600 K followed by a final sputter/anneal cycle re-
moved residual oxygen. Surface cleanliness and
composition were probed with XPS using non-
monochromatic MgKa X-rays. The photoelec-
trons were collected normal to the surface by the
energy analyzer utilizing a 25 eV pass energy.
Binding energies were calibrated with respect to
the Au 4f peak (84.00 eV) and referenced to the
Ni(2p3/2) peak.
The purity of the SO2(g) (Praxair, 99.98%),
O2(g) (16O2, Praxair, 99.999%), 18O2(g) (MSD Iso-
topes, 97.7%) and D2S(g) (Cambridge Isotope
Laboratories, 98%) were monitored with the
QMS during dosing. Unless otherwise stated, all
gases were dosed from the background. Exposures
are reported in units of Langmuir (1 L =
10ꢀ6 Torr s). Typical SO2(g) dosing pressures were
on the order 1 · 10ꢀ8 Torr.
3. Results and discussion
Fig. 1. XP-spectra for SO2(g) on Ni(100)-p(2 · 2)-O and
SO2(g) on Ni(100). All S(2p) curves are doublets consisting
of S(2p3/2) and S(2p1/2) components approximately 1 eV apart,
with the former at lower binding energy (eV). All spectra were
obtained at 300 K. SO2 on Ni(100)-p(2 · 2)-O: The surface at
300 K following our cleaning procedure (a) was exposed to
oxygen to generate the p(2 · 2)-O overlayer (b) and subse-
quently exposed to SO2(g) (1 · 10ꢀ8, 10 min) to produce a
surface covered by sulfite and oxygen (c). Heating the sulfite
and oxygen-covered surface to 600 K desorbed SO2, leaving
atomic oxygen on the surface (d). SO2 on Ni(100): The surface
at 300 K following our cleaning procedure (e) was exposed to
SO2(g) (1 · 10ꢀ8 Torr, 10 min) to generate the sulfur, sulfite and
oxygen-covered surface (f), which was subsequently annealed to
600 K to desorb SO2, leaving atomic sulfur and oxygen on the
surface. Deconvolutions of O(1s) features are shown in (c)
and (f).
3.1. X-ray photoelectron spectroscopy
XPS spectra following adsorption and reaction
of SO2 on Ni(100) and Ni(100)-p(2 · 2)-O are
shown in Fig. 1 and summarized in Table 1. All
S(2p) features are doublets, consisting of 2p3/2
and 2p1/2 components (approximately 1 eV apart),
with the former at lower binding energy [14]. The
chemical state of the surfaces elucidated by XPS
is in good agreement with previous NEXAFS re-
sults (see Section 1).
Adsorption and reaction of SO2 on Ni(100)-
p(2 · 2)-O produced a sulfite-covered surface
(Fig. 1). The clean surface (Fig. 1a) was initially
exposed to O2(g) to generate the p(2 · 2)-O over-
layer (Fig. 1b, see STM section) and subsequently
exposed to SO2(g). The XPS spectrum of this sur-
face (Fig. 1c) shows a S(2p3/2) peak at 165.8 eV,
which is assigned to SOx. Deconvolution of the
O(1s) features reveals peaks at 529.0 and
530.2 eV, which are attributable to atomic oxygen
and SOx, respectively [4]. The SOx O(1s) and S(2p)
peak area ratio, following correction for the S(2p)
and O(1s) photoionization cross-sections (1.64)
[15], is 2.98:1, suggesting SOx is SO3. This result
is consistent with the binding energy of SO3 ob-
served on other surfaces (Table 1). Thus, the reac-
tion between SO2(a) and O(a) gives
SO2ðaÞ þ OðaÞ ! SO3ðaÞ
ð6Þ
From the O(1s) peak areas and the initial coverage
of oxygen in the p(2 · 2)-O covered surface
(0.25 ML), the O(a) and SO3(a) coverages after
reaction are 0.15 ML and 0.10 ML, respectively;
approximately 40% of the O(a) reacted with SO2

144
A.R. Alemozafar, R.J. Madix / Surface Science 592 (2005) 141–149
Table 1
From the 1:1 sulfur and sulfite peak area ratio,
b = c. Since no S(2p) features were observed with-
in the range 168–170 eV, where SO4(a) has been
observed on other surfaces [3], ÔdÕ is set to zero,
reducing Eq. (1) to
Summary of XPS S(2p) binding energies observed in this work
(last row) and other relevant investigations
S
SO2
SO3
SO4
–
Cu(110) [4]
Cu(110) [23]
Cu2S [24]
CuSO4 [25]
Ag(110) [3]
Ag(111) [26]
Ag2S [27]
161.4
161.3
161.3
–
165.3
166.1
2bSO2ðaÞ ! bSðaÞ þ bSO3ðaÞ þ bOðaÞ
ð4Þ
–
–
–
–
–
169.5
167.9
168.4
–
–
–
Eq. (4) has infinitely many solutions, with the sim-
plest non-trivial solution being the one for which
b = 1, namely
–
165.4
–
–
165.8
–
166.1
–
–
–
161.8
161.6
–
Ni [24]
Ni(100) (this work)
2SO2ðaÞ ! SðaÞ þ SO3ðaÞ þ OðaÞ
ð5Þ
161.5
165.8
All S(2p) XPS peaks are doublets consisting of 2p3/2 and 2p1/2
components. The table entries correspond to the 2p3/2
component.
The stoichiometry of Eq. (5) is in agreement with
the NEXAFS investigation. Heating the surface
to 600 K desorbed SO2(g) and gave a surface cov-
ered by S(a) and O(a). Using the p(2 · 2)-S struc-
ture at 0.23 ML for reference and correcting for
the sulfur and oxygen photoionization cross-sec-
tions (1.64) [16], the S(a), SO3(a) and O(a) cover-
ages are determined to be 0.20 ML, 0.20 ML and
0.18 ML, respectively.
to form SO3. Heating the sulfite-covered surface to
600 K desorbed SO2(g), leaving approximately
0.25 ML O(a) on the surface (Fig. 1d).
Following adsorption and reaction of SO2 on
Ni(100), XPS features attributable to sulfite,
atomic sulfur and atomic oxygen were observed.
The surface following cleaning was devoid of sul-
fur and oxygen-containing impurities (Fig. 1e).
XP-spectra following exposure of this surface to
SO2(g) showed S(2p) peaks at 161.3 eV and
165.9 eV (Fig. 1f), which are ascribed to S(a) and
SOx(a), respectively. The O(1s) spectrum (after
deconvolution) consisted of peaks at 528.9 and
530.6 eV, which are attributed to photoemission
from O(a) and SOx(a), respectively. Following cor-
rection for the sulfur and oxygen photoionization
cross-sections, the SOx O(1s) and S(2p) peak area
ratio is approximately 2.99:1, which suggests SOx
is SO3. The S(a) and SO3(a) peak area ratio is
0.99:1; The S(a) and O(a) peak area ratio, follow-
ing correction for S(2p) and O(1s) photoionization
cross-sections, is 1.16:1, or nearly one-to-one. The
reaction between SO2 and the clean Ni(100) sur-
face can be written generally as
3.2. Temperature-programmed desorption
TPD measurements of the sulfite-covered sur-
face shows SO2(g) evolving in separate states at
380 K and 450 K (Fig. 2a). SO2(g) was introduced
to the clean surface at 300 K through a needle
doser directed at the surface with the opening
approximately one inch from the crystal. The sur-
face temperature was ramped at approximately
1 K/s and various gases were detected with the
QMS directly in front of the crystal surface. With
heating, only SO2(g) evolved from the surface.
TPD measurements following SO2 adsorption
onto Ni(100)-p(2 · 2)-O at 300 K showed an
SO2(g) desorption peak at 350 K and a shoulder
at 440 K (Fig. 2b). The p(2 · 2)-O surface was ini-
tially prepared via O2(g) exposure (1 · 10ꢀ8 Torr,
10 min) at 300 K followed by annealing to
500 K. The surface was then cooled to 300 K and
SO2(g) was dosed (1 · 10ꢀ8 Torr, 10 min) to gener-
ate the sulfite-covered surface. TPD measurements
showed only SO2(g) evolving from the surface.
Experiments with isotopically labeled oxygen
(18O2) suggest that the sulfite species has monoden-
tate coordination to the surface (Fig. 2c). The
sulfite-covered surface was prepared by exposing
aSO2ðaÞ ! bSðaÞ þ cSO3ðaÞ þ dSO4ðaÞ þ eOðaÞ
ð1Þ
Sulfur and oxygen mass balances give
a ¼ b þ c þ d
2a ¼ 3c þ 4d þ e
ð2Þ
ð3Þ

A.R. Alemozafar, R.J. Madix / Surface Science 592 (2005) 141–149
145
O–metal bonds to be equivalent and to be intra
convertible among the other S–O bonds, the latter
two species would yield SOO to SO18O peak area
ratios of 1:2, respectively. With the same assump-
tions, the monodentate sulfite formed by the
addition of SO2(g) to the surface containing pre-
adsorbed 18O would yield an infinite ratio of
SOO to SO18O, as it is based solely on the proba-
bility of breaking and making in convertible O–Ni
bonds. The observed ratio of the peak areas of
approximately ten to one is indicative of the dom-
inance of a monodentate sulfite that decomposes
primarily by cleavage of the S–O bond formed be-
tween SO2 and preadsorbed oxygen.
3.3. Scanning tunneling microscopy
3.3.1. SO2(g) on Ni(100)-p(2 · 2)-O
The p(2 · 2)-O overlayer was prepared by
exposing the clean surface at 300 K to O2(g)
(1 · 10ꢀ8 Torr, 10 min) followed by annealing to
500 K. The surface prepared in this manner
Fig. 2. TPD spectra for (a) SO2(g) on Ni(100), (b) SO2 (g) on
Ni(100)-p(2 · 2)-O and (c) SO2(g) on Ni(100)-p(2 · 2)-18O.
Only the mass-to-charge (m/z) = 64 signal is shown in (a) and
(b). The average heating rate was 1 K/s. The vertical axis
corresponds to the QMS signal for the species with the
indicated mass-to-charge (m/z) ratios.
Ni(100)-p(2 · 2)-18O to SO2(g) at 300 K. TPD
measurements were subsequently conducted to
assess the distribution of evolving gases. Of the
possible isotopes for sulfur dioxide which could
evolve upon heating, only mass-to-charge ratios
of 64 (SO2) and 66 (SO18O) were detected; 68
(S18O18O) was not detected. Previous investigators
have suggested that sulfite coordinates to Ag(110)
through either one (monodentate) or two oxygen
atoms (bidentate) [9] and to Cu(100) or Cu(111)
by three oxygen atoms [17,18]. Assuming the
Fig. 3. STM micrographs of a p(2 · 2)-O covered surface. (a)
Large scale scan showing a large terrace and bunched steps;
tunneling conditions: ꢀ97.5 mV, 0.97 nA. (b) A zoom-in on a
portion of the surface in (a), showing the p(2 · 2)-O structure
(box); tunneling conditions: ꢀ94.3 mV, 1.33 nA. (c) Two-
dimensional model of the p(2 · 2)-O structure. Arrows 1–3
represent features described in the text (see Section 3.3.1).

146
A.R. Alemozafar, R.J. Madix / Surface Science 592 (2005) 141–149
exhibited a sharp p(2 · 2) LEED pattern. STM
measurements showed the surface to consist large
terraces and bunched steps (Fig. 3a). A zoom-in
on a terrace location shows the p(2 · 2)-O struc-
ture (Fig. 3b). In this structure, depressions (arrow
1) are interpreted as oxygen local density of states
(LDOS), following the work of Kopatzki and
Behm [19], less intense depressions (arrow 2) are
interpreted as hollow sites centered in-between
the p(2 · 2)-O structure, and areas between oxygen
atoms (arrow 3) are ascribed to hollow sites be-
tween oxygen atoms. The two-dimensional model
(Fig. 3c) accompanying the STM scan depicts
these features.
Exposure of the oxygen-covered surface at
300 K to 6 L SO2(g) produced sulfite islands amid
c(2 · 2)-O (Fig. 4a). The islands are interpreted as
domains of SO3(a) because (1) they are the domi-
nant structure, (2) XPS showed sulfite on the sur-
face and (3) they are more corrugated than
Fig. 4. STM scans following exposure of the Ni(100)-p(2 · 2)-O covered surface at 300 K to SO2(g). (a) Wide-area scan, showing
sulfite islands on top of the oxygen overlayer; tunneling conditions: ꢀ99.0 mV, 0.77 nA (constant current). (b) A scan of the surface in
(a) at higher magnification; tunneling conditions: ꢀ9.58 mV, 1.68 nA. (c) A magnification of an SO3 island, showing p(2 · 2)-SO3
structures, scattered c(2 · 2)-SO3 arrangements, domain boundaries (arrow d) and vacancies (v); tunneling conditions: ꢀ9.81 mV,
ꢀ
1.63 nA. Sulfite on opposite sides of the domain boundary (arrow d) are offset by one lattice unit in the direction of the ½011ꢂ vector
(broken lines 1 and 2). (d) A magnification of the c(2 · 2)-O adjoining the sulfite island in (c); tunneling conditions: ꢀ9.77 mV, 1.40 nA.
Scattered oxide nuclei (arrow p) were observed on this surface.

A.R. Alemozafar, R.J. Madix / Surface Science 592 (2005) 141–149
147
domains of oxygen. Similar results were observed
3.3.2. SO2(g) on Ni(100)
for sulfite on Cu(110)-p(2 · 1)-O [4]. A zoom-in
on a portion of the surface in Fig. 4a (box) resolves
the SO3(a) island and the adjoining oxide structure
(Fig. 4b). Magnification of the SO3 (a) island
reveals a p(2 · 2) structure, scattered c(2 · 2)
arrangements, vacancies (v) and domain bound-
aries (d) (Fig. 4c). The sulfite islands are encom-
passed by areas of the uncovered c(2 · 2)-O and
scattered protrusions (Fig. 4d arrow p), which
are due to oxide nucleation [20,21]. Evidently, sul-
fite adsorption is accompanied by compression of
the p(2 · 2)-O into the c(2 · 2). The c(2 · 2)-O
did not react significantly with SO2(g) at the expo-
sures employed in this study, which is similar to
what was observed for SO2(g) exposure to a satu-
rated Cu(110)s-p(2 · 1)-O surface [4]. In that
study, the saturated p(2 · 1)-O overlayer at
0.5 ML coverage was unreactive toward SO2(g)
exposure.
The clean Ni(100) surface at 300 K was ex-
posed to SO2(g) (1 · 10ꢀ8 Torr, 10 min). As noted
above, the surface consisted of a mixture of sulfur,
sulfite and oxygen. STM measurements were made
on terraces several thousand angstroms in width
(Fig. 6a) and at step edges (Fig. 6b). An enlarge-
ment of a scan taken at a terrace location
(Fig. 6a, inset) reveals two electronically distinct
species. Taken as protrusions, the two species oc-
cupy equivalent sites. The coverage-weighted ratio
of the two protrusions was approximately 1:1 (5%
error). These observations suggest that the protru-
sions are due to sulfur and sulfite. We exclude oxy-
gen from the area in Fig. 6a on the basis that O(a)
appears as depressions on the surface amid SO3
protrusions (Fig. 4) and the structures imaged in
Fig. 6b are attributable to oxygen (see below).
The more corrugated protrusions are interpreted
as sulfite, an assignment consistent with a previous
study on Cu(110) in which the electronic corruga-
tion of SO3 was greater than that of adjoining sul-
fur atoms [4]. In the present investigation, sulfite
adsorbed into scattered p(2 · 2) and c(2 · 2) struc-
tures. The surface structure consisted of defects in
which SO3(a) was one lattice unit away (along
[001] and [010]) from adjoining sulfur atoms
(arrow d). Extrapolation from the binding site of
the sulfur adatoms—which occupy 4-fold hollow
sites—the sulfite occupy 4-fold hollow sites, an
assignment consistent with binding site of SO3(a)
on the oxygen-covered surface (Fig. 4).
A two-dimensional model depicting the oxygen
and sulfite structures imaged by STM is shown in
Fig. 5. The monodentate sulfite binding configura-
tion is consistent with TPD measurements (see
above).
STM measurements of the surface at 300 K at
an area in the vicinity of that imaged in Fig. 6a re-
veal chains oriented in the direction of the [001]
vector (Fig. 6b). The chains, which are of step
height (Fig. 6b, inset), were arranged into struc-
p
p
p
p
tures with (n 2 · 2 2)R45ꢀ or (2 2 · n 2)R45ꢀ
(n = 7–12) periodicity. A scan at higher magnifica-
tion (Fig. 6c) reveals c(2 · 2)-O moieties in-
between the chains. The chains consisted of
depressions (arrow d) and protrusions (arrow p),
which are two lattice units apart in the direction
of the [001] vector. The protrusions are of step
height (Fig. 6b, inset), which suggests they are
comprised of Ni atoms. Similar structures were ob-
served following oxygen adsorption on Cu(100),
which produced a Cu–O missing row structure
Fig. 5. Two-dimensional model depicting the structures
observed following exposure of the Ni(100)-p(2 · 2)-O surface
at 300 K to SO2(g).

148
A.R. Alemozafar, R.J. Madix / Surface Science 592 (2005) 141–149
Fig. 6. STM micrographs of a Ni(100) surface following exposure to SO2(g) at 300 K. (a) Scan at a terrace location, showing (inset)
p(2 · 2)-S, scattered p(2 · 2) and c(2 · 2) SO3(a) domains, and SO3(a) situated in-between two sulfur atoms (arrow d); tunneling
conditions: ꢀ198 mV, 0.97 nA; inset, ꢀ198 mV, 1.22 nA. (b) Scan of the surface next to a step edge, showing chains oriented in the
direction of the [001] vector with protrusions of step height (line scan); tunneling conditions: 0.69 V, 2.21 nA. (c) An enlargement of
p
p
the area indicated in (b), showing chains with (7 2 · 2 2)R45ꢀ periodicity consisting of depressions (arrow d) and protrusions (arrow
p); tunneling conditions: 0.69 V, 2.33 nA.
p
p
with (2 2 · 2)R45ꢀ Cu–O periodicity [22]. The
formation of this structure was concomitant with
the precipitation of Cu into islands of step height.
between SO2(g) and the p(2 · 2)-O surface (see
above). Evidently, domains of c(2 · 2)-O are ener-
getically stable toward SO2(g) oxidation.
p
p
On that surface, protrusions with (2 2 · 2)R45ꢀ
periodicity were interpreted as Cu atoms. Effective
medium theory calculations suggested that the
oxygen atoms were directly situated on top of Cu
atoms below. The Cu(100) work suggests that
4. Summary
Reaction of SO2 with Ni(100) and Ni(100)-
p(2 · 2)-O has been investigated with STM, XPS
and TPD and LEED. On Ni(100)-p(2 · 2)-O at
300 K SO2 reacts according to SO2(a) + O(a) !
SO3(a). Sulfite adsorbs into islands of a p(2 · 2)
structure while the remainder of the unreacted
p(2 · 2)-O is compressed into a c(2 · 2) structure.
The islands persist to 400 K, above which temper-
p
p
p
p
the (n 2 · 2 2)R45ꢀ/(n 2 · 7 2)R45ꢀ (n = 7–
12) chains are due to oxygen, which may be in-
duced by sulfur adsorption. A two-dimensional
model depicting the structures observed in the
STM micrographs is shown in Fig. 7.
The lack of reactivity of the c(2 · 2)-O toward
sulfite formation is consistent with the reaction

A.R. Alemozafar, R.J. Madix / Surface Science 592 (2005) 141–149
149
References
[1] H.C. Allen, J.M. Laux, R. Vogt, B.J. Finlayson-Pitts, J.C.
Hemminger, J. Phys. Chem. 100 (1996) 6371.
[2] P.R. Buseck, M. Posfai, Proc. Natl. Acad. Sci. 96 (1999)
3372.
[3] D.A. Outka, R.J. Madix, Surf. Sci. 137 (1984) 242.
[4] A.R. Alemozafar, X.-C. Guo, R.J. Madix, J. Chem. Phys.
116 (2002) 4698.
[5] T. Ohta, T. Yokoyama, S. Terada, A. Imanishi, Y.
Kitajima, Res. Chem. Intermediat. 26 (2000) 29.
[6] A.R. Alemozafar, X.-C. Guo, R.J. Madix, N. Hartmann,
J. Wang, Surf. Sci. 504 (2002) 223.
[7] C.M. Pradier, P. Dubot, J. Phys. Chem. B 102 (1998)
5135.
[8] C.M. Pradier, E. Jacqueson, P. Dubot, J. Phys. Chem. B
103 (1999) 5028.
[9] D.A. Outka, R.J. Madix, G.B. Fisher, C. DiMaggio, J.
Phys. Chem. B 90 (1986) 4051.
[10] A.R. Alemozafar, X.-C. Guo, R.J. Madix, Surf. Sci. 524
(1–3) (2003) L84.
[11] A.R. Alemozafar, X.-C. Guo, R.J. Madix, Surf. Sci. 526
(1–2) (2003) L127.
[12] J. Haase, J. Phys.: Condens. Mater 9 (1997) 3647.
[13] H.W. Wassmuth, J. Ahner, M. Hofer, H. Stolz, Prog. Surf.
Sci. 42 (1993) 257.
[14] T. Jirsak, J.A. Rodriguez, S. Chaturvedi, J. Hrbek, Surf.
Sci. 418 (1998) 8.
[15] J.H. Scofield, J. Electron Spectrosc. Relat. Phenom. 20
(1976) 333.
[16] J.H. Scofield, J. Electron Spectrosc. Relat. Phenom. 8
(1976) 129.
Fig. 7. Two-dimensional model depicting the structures
observed following exposure of the Ni(100) surface at 300 K
to SO2(g) (Fig. 6). The depressions and protrusions in Fig. 6c
(arrows d and p) have been indicated by gray circles.
[17] G.J. Jackson, S.M. Driver, D.P. Woodruff, N. Abrams,
R.G. Jones, M.T. Butterfield, M.D. Crapper, B.C.C.
Cowie, V. Formoso, Surf. Sci. 459 (2000) 231.
[18] T. Nakahashi, S. Terada, T. Yokoyama, H. Hamamatsu,
Y. Kitajima, M. Sakano, F. Matsui, T. Ohta, Surf. Sci. 373
(1997) 1.
[19] E. Kopatzki, R.J. Behm, Surf. Sci. 245 (1990) 255.
[20] E. Kopatzki, R.J. Behm, Phys. Rev. Lett. 74 (1995)
1399.
[21] M. Baeumer, D. Cappus, H. Kuhlenbeck, H.J. Freund, G.
Wilhelmi, A. Brodde, H. Neddermeyer, Surf. Sci. 253
(1991) 116.
[22] F. Besenbacher, I. Stensgaard, in: D.A. King, D.P.
Woodruff (Eds.), The Chemical Physics of Solid Surfaces,
Elsevier Science Publishers, 1993 (Chapter 15).
[23] A.F. Carley, P.R. Davies, R.V. Jones, K.R. Harikumar,
G.U. Kulkarni, M.W. Roberts, Surf. Sci. 447 (2000) 39.
[24] C.R. Brundle, A.F. Carley, Faraday Disc. Chem. Soc. 60
(1975) 51.
ature the evolution of SO2(g) reverts the surface to
its initially oxidized state. TPD measurements with
18O are suggestive of a monodentate sulfite bind-
ing configuration. On Ni(100) at 300 K SO2
disproportionates according to 2SO2 ! S(a) +
SO3(a) + O(a). Sulfur and sulfite adsorb into scat-
tered p(2 · 2) structures and oxygen adsorbs into a
c(2 · 2) structure. Oxygen adsorption is concomi-
tant with the formation of structure with
p
p
p
p
(n 2 · 2 2)R45ꢀ/(2 2 · n 2) R45ꢀ (n = 7–12)
periodicity, which are interpreted as Ni chains
induced by oxygen.
Acknowledgment
[25] B.R. Strohmeier, D.E. Leyden, R.S. Field, D.M. Hercules,
J. Catal. 94 (1985) 514.
[26] G. Tibbetts, J. Vac. Sci. Technol. 15 (1978) 497.
[27] D. Lichtman, J.H. Craig, V. Slater, M. Drinkwine, Appl.
Surf. Sci. 7 (1981) 325.
We are grateful to the National Science Foun-
dation (NSF CHE 9820703) for support of this
work.