Surface products of the dissociative adsorption of methane on Pt{1 0 0}-(1 × 2)

Article abstract of DOI:10.1016/S0039-6028(01)01678-8

Numerous studies have been made of the mechanism of dissociative methane adsorption on transition metals, but the surface dissociation products have seldom been conclusively identified. Using a supersonic molecular beam system we have identified the methane dissociation products on Pt{1 1 0}-(1 × 2). In the temperature range 300 to 450 K the stable dissociation fragment at all coverages is methylidyne (CH). Saturation exposure at 370 K has revealed for the first time an ordered c(2 × 4)-CH overlayer, which saturates at a coverage of 0.25 ML. Under these conditions there is no evidence for C-C coupling even at high coverages. On heating beyond 470 K this species decomposes to gaseous H₂ and adsorbed C. After adsorption to high coverages at 500 K evidence is found for C-C coupling.

Full text of DOI:10.1016/S0039-6028(01)01678-8

Surface Science 505 (2002) 49–57
www.elsevier.com/locate/susc
Surface products of the dissociative adsorption of methane
on Pt{110}-(1 ꢀ 2)
D.T.P. Watson, S. Titmuss, D.A. King
*
Department of Chemistry, University of Cambridge, Lensfield Road, Cambridge CB2 1EW, UK
Received 31 May 2001; accepted for publication 12 October 2001
Abstract
Numerous studies have been made of the mechanism of dissociative methane adsorption on transition metals, but
the surface dissociation products have seldom been conclusively identified. Using a supersonic molecular beam system
we have identified the methane dissociation products on Pt{1 1 0}-(1 ꢀ 2). In the temperature range 300 to 450 K the
stable dissociation fragment at all coverages is methylidyne (CH). Saturation exposure at 370 K has revealed for the
first time an ordered c(2 ꢀ 4)-CH overlayer, which saturates at a coverage of 0.25 ML. Under these conditions there is
no evidence for C–C coupling even at high coverages. On heating beyond 470 K this species decomposes to gaseous H
2
and adsorbed C. After adsorption to high coverages at 500 K evidence is found for C–C coupling. Ó 2001 Elsevier
Science B.V. All rights reserved.
Keywords: Alkanes; Platinum; Chemisorption
1
. Introduction
An understanding of the mechanism of methane
partial oxidation at an atomic level would be in-
valuable in improving this conversion reaction,
making it an industrially viable process. Before a
complete mechanism can be postulated, the iden-
tity of the hydrocarbon fragments formed when
methane is adsorbed at different surface tempera-
tures must be established.
Methane usually accompanies oil deposits in
large quantities, but it is one of the most under-
utilised natural chemical resources. Large quanti-
ties are continuously flared off, because it is too
dangerous and expensive to transport from remote
areas. An attractive solution to this problem is the
in situ partial oxidation to syn-gas [1], a 1:2 mix-
ture of carbon monoxide and hydrogen, which can
easily be converted to methanol [2], over a Cu/ZnO
catalyst, or higher hydrocarbons, over Fe or Ru
catalysts, by the Fischer Tropsch process [3,4].
It is widely accepted that the initial dissociation
products of methane on most transition metals are
methyl (CH ) and atomic hydrogen. Adsorbed
3
CH groups have been prepared at low tempera-
3
tures using a variety of techniques including ther-
mal dissociation of CH
methane beam [8] and a CH
3
halides [5–7], a supersonic
radical source [9,10].
3
*
The thermal chemistry of these metastable adlay-
ers has been extensively studied using RAIRS. In a
RAIRS study on Pt{1 1 1}, Stair and coworkers [9]
Corresponding author. Tel.: +44-1223-336338/336449; fax:
44-1223-762829.
E-mail address: hg243@cam.ac.uk (D.A. King).
+
0
039-6028/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved.
PII: S0039-6028(01)01678-8

50
D.T.P. Watson et al. / Surface Science 505 (2002) 49–57
show that the intensity of the deformation mode of
CH is attenuated on heating to 240 K and con-
clude that this is the CH decomposition temper-
ature. The C–H stretch of CH is notoriously weak,
and they were therefore unable to determine the
nature of the resulting decomposition product at
low coverages. Some coupling to ethylidyne has
entially pumped and collimated before entering the
sample chamber. An inert flag just in front of the
crystal is used to control the dosing time and to
measure sticking probabilities [20]. The surface
3
3
temperature ðT Þ is monitored by a crystal-moun-
s
ted thermocouple referenced to an electronic ice
point and is regulated by programmed resistive
heating. The partial pressures of up to 16 indi-
vidual gases can be monitored using a fixed QMS
situated behind the crystal.
been observed using the CH radical source [10]
3
and a supersonic methane beam [8], which allow
high coverages to be obtained. These assignments
ꢁ
1
were based on the presence of dCH3 at 1339 cm
and mCꢁC at 1115 cm bands.
The Pt sample, 11 mm diameter ꢀ 1 mm
thickness, was cut to within 1° of the {1 1 0} plane.
Initial cleaning of the crystal was achieved by re-
peated cycles of ion sputtering, annealing and
oxygen treatment. Routine cleaning consisted of
annealing at 1240 K, exposure to oxygen for 5 min
while cooling from 1100 to 950 K and annealing
for 15 min at 950 K. This procedure yields a clean
Pt{1 1 0}-(1 ꢀ 2) surface which gives a sharp
LEED pattern and oxygen thermal desorption
spectra that are in good agreement with the liter-
ature [21]. It should also be noted that Auger
analysis performed during earlier studies found no
evidence for impurities such as calcium, sulphur
and silicon, and the above cleaning procedure
removed carbon which remained on the surface
ꢁ
1
The main problem with RAIRS is the lack of
sensitivity to low coverages of CH species, pre-
sumably due to screening of the C–H dipole by
metal valence electrons. To date it has not been
possible to observe CH using this technique.
However, using HREELS Ceyer and coworkers
[
11,12] did identify CH as a methane dissociation
product on Ni{1 1 1}. Density functional theory
DFT) slab calculations performed by Michaelides
(
and Hu [13], and cluster calculations performed by
Kua and Goddard [14], have predicted that CH is
energetically the most stable C1 fragment on
Pt{1 1 1}.
In this study we show how temperature pro-
grammed reaction spectroscopy (TPRS) can be
used as an alternative to vibrational spectroscopy
to identify the nature of the stable hydrocarbon
fragments on Pt{1 1 0}-(1 ꢀ 2) over a large cover-
age range, formed by a supersonic molecular
beam. This new classical technique has the ad-
vantage of being compatible with the supersonic
beam system, and the results are, in this instance,
unequivocal. This study follows a series of detailed
studies of the dynamics of methane dissociation
4 2
after experiments. The CH and O used were
>99.9995% and 99.995% pure, respectively, as
quoted by the suppliers (Messer (UK) Ltd.). The
supersonic methane beams used in these experi-
ments had a composition of 8% CH , 92% He and
4
the nozzle was heated to 790 K, giving the methane
molecular beam a translational energy of 580 meV,
estimated on the basis of an ideal supersonic ex-
pansion. All coverages are quoted in monolayers
14
ꢁ2
(ML) (1 ML ꢂ 9:22 ꢀ 10 molecules cm ). The
coverages were determined using a spinning rotor
gauge to yield absolute beam fluxes, and by com-
parison of TPD areas with earlier studies [16,22].
[15,16] and its subsequent oxidation [17,18] over
the same surface.
2
. Experimental
3. The stable surface product at 370 K
The experimental apparatus is described in de-
tail elsewhere [19]. The sample is mounted cen-
trally on a manipulator in an ultrahigh vacuum
chamber with a base pressure of <2 ꢀ 10ꢁ mbar.
The molecular beam is sourced from a supersonic
expansion from a 50 lm nozzle, skimmed, differ-
3.1. Temperature programmed reaction spectro-
scopy
10
After adsorbing methane using a helium seeded
beam at a crystal temperature of 400 K, the tem-

D.T.P. Watson et al. / Surface Science 505 (2002) 49–57
51
ꢁ
1
perature was ramped at 6 K s to 650 K. The
hydrogen partial pressure was monitored and a
hydrogen peak was observed at 470 K. This is
considerably higher than the desorption tempera-
ture of hydrogen from a clean platinum surface;
temperature programmed desorption investiga-
tions showed that hydrogen desorbs at 310 K from
clean Pt{1 1 0}-(1 ꢀ 2), with the peak temperature
increasing with decreasing coverage [23]. To check
hydrogen adsorbing on the out-of-beam areas of
the crystal during dosing. The magnitude of this
background signal was quantitatively determined
by repeating the above experiment with the inert
flag between the sample and the beam. This
background spectrum was then subtracted from
the first spectrum to obtain the corrected TPR
spectrum shown in Fig. 1(a). To reduce the relative
size of the background spectrum, an intense beam
with high translational energy was employed in
order to build up the adlayer in the minimum
amount of time. The relative areas of the 310 and
470 K peaks can be used to calculate the average
stoichiometry of the hydrocarbon species left on
the surface after desorption of the first peak, at
ꢃ370 K. The experiment was performed twice and
the ratio was found to be ð3:2 Æ 0:3Þ:1, implying
that the hydrocarbon decomposition product has a
2
that the H peak at 470 K originated from disso-
ciative adsorption of methane, the clean crystal
was exposed to methane which had been therma-
lised with the walls of the vacuum chamber, by
inserting an inert gold flag between the methane
beam source and the crystal. This resulted in a
ꢁ
8
methane partial pressure of 10 mbar and a hy-
ꢁ
10
drogen partial pressure of 10
ple. The dissociative sticking probability of such
thermalised (300 K) CH
on Pt{1 1 0}-(1 ꢀ 2) is
negligibly small (as distinct from rotationally
cooled CH formed in a supersonic beam source
where the sticking probability is relatively high
15]) and after exposing the crystal, which was held
mbar at the sam-
4
C
1
H
1
stoichiometry at ꢃ370 K. The experiment
was repeated using a beam diameter of 1.5 mm at
the crystal. In this case the relative size of back-
4
2
ground H desorption was greater, but the same
[
ratio, ð3:1 Æ 0:3Þ:1 was obtained. We therefore
at 400 K, under these conditions, no hydrogen was
observed at 470 K. We conclude that the 470 K
peak is associated with the decomposition of a
stable hydrocarbon fragment on the surface
formed by dissociative methane adsorption from
conclude that the stable surface product following
CH
380 K has an average stoichiometry of C
4
adsorption at surface temperatures of 340 to
1
1
H .
There have been several suggestions that C–C
coupling can occur on Pt surfaces [8–10]. In an
the supersonic beam, which we designate, C
x
H
y
.
2 2
attempt to distinguish between CH and C H and
Two independent mass balance methods were
employed to identify this hydrocarbon fragment.
In the first, a helium seeded methane supersonic
molecular beam giving a 5.6 mm diameter spot at
the crystal was fired at the crystal which was kept
at 180 K. At this temperature any H adatoms
to find out whether the identity of the stable
fragment is dependent on coverage, the supersonic
methane beam was allowed to impinge on the
crystal, held at a temperature of 370 K, for in-
creasing amounts of time. Following each dose the
hydrogen partial pressure was measured while the
crystal was ramped at 6 K s to 750 K. The re-
sulting TPR spectra are shown in Fig. 1(b). Clearly
the peak temperature and the full width at half
maximum (fwhm) do not change with exposure
(Æ2 K), while the peak intensity increases with
increasing exposure up to saturation. In each case
the carbon remaining on the surface after heating
to 750 K was titrated with an oxygen molecular
beam to form CO(g). The integrated area under
the CO curve is a direct measure of the carbon
coverage for a particular exposure [16]. The area
under the hydrogen desorption curve is plotted
against the area under the corresponding carbon
ꢁ
1
formed in the dissociation of CH
the surface, and should desorb, at low C
4
will remain on
cov-
x
H
y
erages, at ꢃ310 K. The crystal temperature was
ꢁ
1
ramped at 3 K s to 650 K and the hydrogen
partial pressure in the main chamber was moni-
tored using the fixed mass spectrometer. The hy-
drogen temperature programmed reaction (TPR)
spectrum has two peaks, the first of which, at 310
K, corresponds to the recombinative desorption
of H adatoms, while the second, at 470 K, is
x y
the reaction-limited peak arising from C H de-
composition. There is a significant contribution
to the desorption-limited peak from background

52
D.T.P. Watson et al. / Surface Science 505 (2002) 49–57
2
Fig. 2. The amount of H desorbed in the reaction-limited peak
is plotted as a function of the carbon coverage measured as the
amount of CO generated by reaction with oxygen at 750 K.
erage range studied. The result strongly supports
the conclusion that the species formed on the
surface is independent of coverage. The constant
hydrogen desorption peak temperature and fwhm
(Fig. 1(a)), together with this constant carbon to
hydrogen ratio, demonstrates that the identity of
the adsorbed fragment is independent of coverage:
only a single C –H species is formed. Since CH
x
y
coupling is extremely unlikely at the lowest cov-
erages (ꢄ0.01 ML) and there is no experimental
evidence to support coupling on Pt{1 1 1} at all
but the highest coverages [8,10], we conclude that
coupling does not occur on Pt{1 1 0}-(1 ꢀ 2) at 370
K, over the coverage range studied.
Afurther experiment was performed to pro-
vide an independent check of the ratio of carbon
to hydrogen in the stable hydrocarbon fragment
formed at 370 K over an extended coverage range.
The scheme of this experiment is demonstrated by
the following sequence:
Fig. 1. (a) Raw data, background and corrected hydrogen TPR
after methane was dosed at ꢄ200 K and the crystal heated at 3
ꢁ1
K s to 650 K. (b) Hydrogen TPR signal after increasing ex-
posures to a supersonic methane beam with Etrans ¼ 580 meV
and T
s
¼ 370 K.
3
70 K
4 ꢁ y
CH þ Ptf110g-ð1 ꢀ 2Þ ! C H þ
H ðgÞ
4
x
y
2
5
80 meV
2
ð1Þ
ð2Þ
monoxide curve, and results in a good straight
line, shown in Fig. 2. This shows that the carbon
to hydrogen ratio is constant over the whole cov-
TPR
!
y
2
C þ H
2
ðgÞ

D.T.P. Watson et al. / Surface Science 505 (2002) 49–57
53
background subtraction is required; and (iii) at 370
K C does not decompose to C. The ratio of the
areas under the two peaks can be used to calculate
the average stoichiometry of the C species for
x
H
y
x
H
y
the coverage achieved. The isothermal and the
TPR peaks are clearly distinguished and the rela-
tive areas were found to be ð3:17 Æ 0:2Þ:1, again
showing that at 370 K the average stoichiometry
over the coverage range studied is C
this method does not rule out C
surface product.
1
H
1
. However,
as a possible
2
H
2
3.2. Reaction of the surface product with intense
deuterium beams
Fig. 3. The hydrogen signal when methane is dosed isother-
mally at a surface temperature of 370 K (a) beam into the main
chamber, striking inert flag; (b) inert flag removed, beam
striking crystal at 370 K; (c) beam terminated; (d) decomposi-
tion of the stable hydrocarbon fragment formed by isothermal
The following observation discriminates be-
2 2
tween CH and a coupled product such as C H as
the stable methane decomposition fragment at 370
K. The hydrocarbon fragment formed at 370 K
can be completely rehydrogenated to gaseous CD
ꢁ1
dosing, when the crystal is heated at 6 K s to 750 K.
4
Asupersonic methane beam is fired at the crystal
held at 370 K and the increase in the hydrogen
partial pressure due to the decomposition of
and CHD at this temperature using high beam
3
fluxes of deuterium. This is shown in Fig. 4. The
CD product is formed by exchange of adsorbed
4
methane to C
x
H
y
is measured. The beam is then
ꢁ
D, formed by D dissociation, with CH prior to
2
1
shut off and the crystal heated at 6 K s to 750 K
and the hydrogen formed by the decomposition of
C H is measured. Typical results are shown in
rehydrogenation; a detailed analysis of this process
is given in a later paper [24]. Analysis of the re-
sidual carbon, by reaction with O after this re-
x
y
2
Fig. 3. The advantages of this experiment over that
described above are that: (i) no assumptions are
made about the identity of the initial dissocia-
tion product; (ii) at 370 K background hydrogen is
not significantly adsorbed on the crystal and no
action with D to CD is complete, shows that
2
4
more than 95% reacted to form the gaseous C1
species. If C H were the stable fragment, hydro-
2
2
genation of this species should yield either gaseous
C D , which was not observed, or surface-bound
2
6
Fig. 4. Rehydrogenation of CH with an intense D
2
beam to produce CD
4
3
(g) which is the larger peak, and CHD (g), the smaller peak.

54
D.T.P. Watson et al. / Surface Science 505 (2002) 49–57
ethylidyne, which does not rehydrogenate fur-
ther on platinum surfaces under UHV conditions.
Therefore, we conclude that the adsorbed frag-
ment is CH.
have increased relative to those of the overlayer
spots as the electron beam energy (penetration
depth) is increased. We can also use a plausible
argument to rule out the lifting of the recon-
struction. If the 0.25 ML CH structure were due to
an ordered c(2 ꢀ 4) overlayer on the (1 ꢀ 1) sub-
strate, every other top layer Pt row would not be
occupied by CH. It is difficult to envisage why
methane dissociation should not occur when the
4
. Conclusion:CH the stable product at 370 K
Thus, based on the combined results of the
three independent experiments described above we
conclude that adsorbed CH is the stable dissocia-
tion fragment on Pt{1 1 0}-(1 ꢀ 2) over the surface
temperature range from 350 to 470 K and for all
4
CH impinges on a vacant ridge atom; and also
difficult to understand why the CH–CH interac-
tions should lead to such a structure.
DFT calculations performed in our group by
Petersen et al. [27], demonstrate that the threefold
ridge site is energetically favoured for CH on the
(1 ꢀ 2) surface which suggests the structure for the
ordered overlayer shown in Fig. 5. However, STM
or LEED IðV Þ studies are necessary to provide
more direct evidence for this structure. The pres-
ence of an ordered overlayer shows that there are
strong intermolecular repulsive interactions be-
tween the CH species and helps to explain why
coupling does not take place under the conditions
studied. It should be noted that an analogous or-
dered ethylidyne p(2 ꢀ 2) overlayer was observed
on Pt{1 1 1} [28].
coverages up to 0.25 ML, at which coverage CH
dissociation is effectively terminated [25].
4
These results on Pt{1 1 0}-(1 ꢀ 2) are fully
consistent with DFT slab calculations of Michae-
lides and Hu on Pt{1 1 1} [13] and Nørskov and
coworkers on Ni{1 1 1} [26], which show that CH
is the most stable surface species. The former re-
3
ports that CH is a metastable species, with a
significant barrier to its dissociation, explaining
why it is formed at temperatures below about 200
K [9]. When the surface is heated, CH þ 2H forms.
At higher temperatures carbon is formed, even
though it is energetically less favourable; here the
process is driven by the entropy of the gaseous
hydrogen produced, which is analogous to the
endothermic desorption-limited hydrogen forma-
tion process from adsorbed H on Pt{1 1 0}-(1 ꢀ 2)
at 310 K.
6. The surface products at 500 K:C–C coupling?
As shown in Figs. 1–3, adsorbed CH, the pro-
duct of methane dissociation at temperatures be-
tween 240 and 370 K, is itself decomposed to
surface carbon on further heating, producing the
hydrogen TPR peak at 470 K. After heating to 500
K or higher, adsorbed carbon is the remaining
surface species and no further hydrogen desorp-
tion is observed. However, when the surface is held
at temperatures between 450 and 500 K and is
exposed to a methane beam the behaviour is
completely different.
In Fig. 6, we compare TPR spectra for near-
saturation methane beam exposures at surface
temperatures between 370 and 500 K; the beam
exposure in each case is high enough to result in
saturation of the surface. We see that at higher
crystal temperatures a new hydrogen TPR peak
develops, with peak temperatures between 650 and
5
saturation exposure to CH at 370 K
. An ordered c(2 ꢀ 4)-CH structure formed at
4
Methane exposure at a surface temperature of
3
70 K reveals a sharp c(2 ꢀ 4) LEED pattern as
the sticking probability falls off; this saturation
coverage of CH was found to be 0.25 ML as de-
termined using the titration method described by
Walker and King [16]. To determine whether the
CH lifts the reconstruction, we looked at the en-
ergy dependence of the LEED spot intensities. As
the beam energy is increased the (1 ꢀ 2) spots re-
main intense while the overlayer spots decrease
in intensity. We note that if CH lifted the recon-
struction the intensity of the (1 ꢀ 1) spots would

D.T.P. Watson et al. / Surface Science 505 (2002) 49–57
55
Fig. 5. Proposed structure for the c(2 ꢀ 4)-CH overlayer. The CH coverage is 0.25 ML.
6
70 K. The result is particularly clear for a crystal
temperature of 500 K: the TPR H product from
the decomposition of CH, previously observed at
70 K, is now completely desorbed during the
dosing of methane, and the new peak at 670 K is
clearly observed. This is a reaction-limited H de-
2
4
2
sorption peak which appears only at high cover-
ages and indicates the presence of a new surface
product from the dissociative adsorption of meth-
ane at a surface temperature of 500 K and high
coverages.
Masel and coworkers [29] studied acetylene
decomposition on Pt{1 1 0}-(1 ꢀ 2) and report a
2
H TPR peak at 650 K which they attribute to the
decomposition of acetylide, CCH. Asimilar peak
was observed by Smirnov et al. [30] after the hy-
drogenation of small carbon clusters on Pt{1 1 1}
and again this peak was attributed to CCH(a). We
also note that Fairbrother et al. [10] report IR
evidence for C–C coupling on Pt{1 1 1} using a
3
CH radical source, at coverages near saturation
and a surface temperature of 300 K. We therefore
postulate that a C þ CH ! CCH surface coupling
reaction occurs during the dissociative adsorption
of methane on Pt{1 1 0}-(1 ꢀ 2), at high coverages,
at a surface temperature 500 K, and that the
acetylide species decomposes at ꢃ650 K. By
titrating the carbon produced with oxygen, we
found that the carbon coverage at 500 K was 1.9
times larger than the saturation coverage of 0.25
Fig. 6. Hydrogen TPR signal after exposures to a methane
beam for (a) 90 min at 370 K, (b) 60 min at 400 K, (c) 60 min at
450 K, (d) 70 min at 500 K, (e) 35 min at 500 K.

56
D.T.P. Watson et al. / Surface Science 505 (2002) 49–57
ML CH at 370 K. Acomparison of the integrated
peak areas of the hydrogen TPR peaks at 470 and
Acknowledgements
670 K (Fig. 6) shows that the new species, prob-
ably CCH, constitutes 20% of the carbon at 500 K.
We acknowledge an equipment grant from the
UK EPSRC, a TMR Studentship from the Euro-
pean Community (D.W.) and a Research Fellow-
ship from Selwyn College (S.T.).
7
. Summary
References
The stoichiometry of the stable hydrocarbon
fragment formed by dosing methane at 370 K or
by dosing at a lower temperature and heating to
[
[
1] D.A. Hickman, L.D. Schmidt, Science 259 (1993) 343.
2] J.M. Thomas, W.J. Thomas, Principles and Practice of
Heterogeneous Catalysis, VCH, Weinheim, 1997, Chapter
8.
3
70 K is C H . It is concluded that the species is
1 1
adsorbed methylidyne, CH. The three main pieces
of evidence for this are: (i) the peak temperature
and peak width of the reaction-limited hydrogen
peak at 470 K do not change with initial C-species
coverage; (ii) the carbon to hydrogen ratio, 1:1, is
independent of coverage;and (iii) the C-species can
be quantitatively deuterated to CD and CHD ,
[
[
[
3] F. Fischer, H. Tropsch, Brennst. Chem. 7 (1926) 97.
4] F. Fischer, H. Tropsch, Chem. Ber. 59 (1926) 830.
5] M.A. Henderson, G.E. Mitchell, J.M. White, Surf. Sci. 184
(1987) L325–L331.
6] F. Zaera, Langmuir 7 (1991) 1998.
[
[7] Z.M. Liu, S. Akhter, B. Roop, J.M. White, J. Am. Chem.
Soc. 110 (1988) 8708.
4
3
[8] D.J. Oakes, M.R.S. McCoustra, M.A. Chesters, Faraday
Discuss. 96 (1993) 325.
with no detectable production of gas phase C or
2
C
3
molecules.
This result provides the first experimental
[9] D.H. Fairbrother, X.D. Peng, M. Trenary, P.C. Stair,
J. Chem. Soc. Faraday Trans. 91 (1995) 3619.
identification of methylidyne on Pt{1 1 0}-(1 ꢀ 2),
which supports density functional theory slab
calculations demonstrating that CH is the most
stable species from methane dissociation on a
Pt{1 1 1} surface. Although C is less energetically
stable, it is formed at temperatures above 450 K,
as the entropy for gaseous hydrogen formation
becomes the driving force for methylidyne de-
composition. On saturation of the surface by ex-
[10] D.H. Fairbrother, X.D. Peng, R. Viswanathan, P.C. Stair,
M. Trenary, J. Fan, Surf. Sci. Lett. 285 (1993) L455–L460.
[11] Q.Y. Yang, K.J. Maynard, A.D. Johnson, S.T. Ceyer,
J. Chem. Phys. 102 (1995) 7734.
[12] M.B. Lee, Q.Y. Yang, S.T. Ceyer, J. Chem. Phys. 87 (1987)
2
[13] A. Michaelides, P. Hu, J. Am. Chem. Soc. 122 (2000) 9866.
724.
[
14] J. Kua, W.A. Goddard III, J. Phys. Chem. B 102 (1998)
9
492.
15] A.V. Walker, D.A. King, Phys. Rev. Lett. 82 (1999) 5156.
16] A.V. Walker, D.A. King, J. Chem. Phys. 112 (2000)
[
[
4
posure to a CH beam at 370 K, the surface
4
739.
[17] A.V. Walker, D.A. King, J. Chem. Phys. 112 (2000)
937.
18] A.V. Walker, D.A. King, J. Phys. Chem. B 104 (2000)
462.
saturates with the formation of a sharp c(2 ꢀ 4)
LEED pattern at 0.25 ML CH. This structure has
been evaluated by DFT slab calculations, which
show that the stable CH site is a threefold site
involving two top layer Pt atoms.
This provides an easy route to a high purity,
high coverage adlayer of CH, which we have ex-
ploited in subsequent studies of its chemical reac-
tivity. It is an important and reactive intermediate,
which may play a key role in catalytic processes
such as Fischer Tropsch and methane oxidation.
The appearance of a high temperature hydro-
gen desorption peak, when methane is dosed at
1
[
[
6
19] A. Hopkinson, X.-C. Guo, J.M. Bradley, D.A. King,
J. Chem. Phys. 99 (1993) 1.
[20] D.A. King, M.G. Wells, Surf. Sci. 29 (1972) 454.
[
[
[
[
[
[
21] A.V. Walker, B. Klotzer, D.A. King, J. Chem. Phys. 109
1998) 6879.
(
22] T.E. Jackman, J.A. Davies, D.P. Jackson, W.N. Unertl,
P.R. Norton, Surf. Sci. 120 (1982) 389.
23] J.R. Engstrom, W. Tsai, W.H. Weinberg, J. Chem. Phys.
87 (1987) 3104.
24] D.T.P. Watson, Q. Ge, D.A. King, J. Chem. Phys. (in
press).
25] D.T.P. Watson, J.J.W. Harris, D.A. King, Surf. Sci. 505
5
00 K, is attributed to C–C coupling with the
(
2002) 58.
formation of a C
ages.
2
H species at high carbon cover-
26] R.M. Watwe, H.S. Bengaard, J.R. Rostrup-Nielsen, J.A.
Dumesic, J.K.N. ørskov, J. Catal. 189 (2000) 16.

D.T.P. Watson et al. / Surface Science 505 (2002) 49–57
57
[
[
27] M. Petersen, S. Jenkins, D.T.P. Watson, D.A. King (in
preparation).
[29] E. Yagasaki, A.L. Backman, R.I. Masel, J. Phys. Chem. 94
(1990) 1066.
28] U. Starke, A. Barbieri, N. Materer, M.A. Van Hove, G.A.
Somorjai, Surf. Sci. 286 (1993) 1.
[30] M.Y. Smirnov, V.V. Gorodetskii, A.R. Cholach, D.Y.
Zemlyanov, Surf. Sci. 311 (1994) 308.