Polymerisation of cyanogen on graphite and graphite supported copper films

Article abstract of DOI:10.1016/S0039-6028(02)02099-X

The interaction of cyanogen with graphite and graphite supported copper film surfaces was investigated using x-ray photoelectron spectroscopy (XPS). It was found that on both the surfaces an Al Kα x-ray source induced a modification of condensed cyanogen overlayers. The results suggested that modification was x-ray induced polymerization formed via an indirect mechanism involving secondary electrons from the substrate.

Full text of DOI:10.1016/S0039-6028(02)02099-X

Surface Science 517 (2002) L563–L567
www.elsevier.com/locate/susc
Surface Science Letters
Polymerisation of cyanogen on graphite and
graphite supported copper films
a,
A.F. Carley , M. Chinn , C.R. Parkinson
b
a,1
*
a
Department of Chemistry, Cardiff University, P.O. Box 912, Cardiff CF10 3TB, UK
b
Dstl, Porton Down, Salisbury, Wiltshire SP4 0JQ, UK
Received 31 January 2002; accepted for publication 17 July 2002
Abstract
The adsorption of cyanogen, (CN)
2
, on graphite and graphite-supported copper films has been investigated using X-
ray photoelectron spectroscopy. X-ray irradiation is shown to effect a polymerisation of condensed cyanogen overlayers
on both graphite and graphite supported copper film surfaces via a substrate-mediated secondary electron process.
Ó 2002 Elsevier Science B.V. All rights reserved.
Keywords: X-ray photoelectron spectroscopy; Chemisorption; Surface chemical reaction; Copper; Carbon; Cyanogen; Polycrystalline
thin films; Metallic films
1
. Introduction
2
action of HCN and (CN) with Cu/Cr impreg-
nated carbons but this complex reaction is far
from fully understood. Furthermore, the adsorp-
tion of carbon monoxide and nitric oxide on
supported metal surfaces has been well-studied,
and since CO, NO , and CN are isoelectronic it
is interesting to compare their adsorption behav-
iour with that of cyanogen.
The adsorption of cyanogen has been studied
previously on a variety of metals. On the Ag (1 1 0)
surface only reversible molecular adsorption has
been reported [3]. This contrasts dramatically with
the adsorption of cyanogen on copper single
crystal surfaces where dissociation has been re-
ported to occur at 150 K [4]. Moreover, Rh (1 0 0)
adsorbs cyanogen in three states; a molecular
state, a dissociated state involving only Rh–C
bonding, and a dissociated state with both Rh–C
and Rh–N bonding [5].
Research into cyano compounds is in part dri-
ven by their acute toxicity and hence the require-
ment to provide a means for their destruction in
industrial and military applications. It has long
been known [1] that activated carbons impreg-
nated with copper are effective for the removal of
HCN in filtration applications. Subsequent studies
have demonstrated that one of the products from
the reaction of HCN with Cu(II) is cyanogen.
Alves and Clark[2] described to a degree the re-
þ
À
*
Corresponding author. Tel.: +44-29-2087-4139; fax: +44-
9-2087-4030.
E-mail address: carley@cardiff.ac.uk (A.F. Carley).
2
1
Present address: Department of Physics, University of
Warwick, Coventry CV4 7AL, UK.
0
039-6028/02/$ - see front matter Ó 2002 Elsevier Science B.V. All rights reserved.
PII: S0039-6028(02)02099-X

L564
A.F. Carley et al. / Surface Science 517 (2002) L563–L567
In this paper we report the first X-ray photo-
freeze-pump–thaw cycles, and no impurities were
detected with the mass spectrometer.
electron spectroscopic study of cyanogen adsorbed
on graphite and graphite supported copper films.
The workwas underta ke n as part of an ongoing
project to study the interaction of a range of
3. Results and discussion
2 2
molecules (CO, NO, O , H O) with copper clusters
and films deposited on graphite surfaces.
3.1. Low temperature adsorption of cyanogen on a
clean graphite surface
2
. Experimental
The C(1s) and N(1s) XPS spectra taken fol-
lowing the exposure of the clean graphite surface
The experimental setup used in this study has
2
to 10 L of (CN) at 80 K and annealing to various
been described in detail elsewhere [6]. Briefly,
experiments were performed on a vacuum gener-
ators ultrahigh-vacuum (UHV) photoelectron spec-
trometer equipped with a dual anode X-ray source,
concentric hemispherical analyzer, Ar ion sput-
tering gun, Knudsen molecular effusion cell (WA
Technology K-cell), and a quadrupole mass ana-
lyzer (QMA). The XPS spectra were obtained
using an Al Ka radiation and the binding energies
were calibrated relative to the C(1s) peakat 283.7
temperatures are shown in Fig. 1. At 80 K, two
carbon peaks and one nitrogen peak are observed
with binding energies of 283.7, 289.0 and 401.0 eV
respectively. The peaks at 289.0 and 401.0 eV
correspond to a C:N ratio of 1:1 and are assigned
to multilayers of adsorbed cyanogen. Following
annealing to room temperature the high binding
energy C(1s) peakand the N(1s) component at
401.0 eV disappear due to desorption/reaction of
physisorbed cyanogen, and a new N(1s) peak
emerges at a binding energy of at 399.0 eV. The
intense C(1s) peakdue to the substrate precludes
us from identifying an associated C(1s) feature.
eV or to the Cu 2p
graphite or the copper films respectively.
peakat 932.7 eV, for the
ð3=2Þ
The graphite sample (Goodfellows 99.999%)
was mounted on a UHV sample manipulator
with facilities for liquid nitrogen cooling to 80 K
and conductive heating to 700 K. Sample temper-
ature was measured using a thermocouple moun-
ted close to the sample holder. The surface was
cleaned in vacuum with cycles of Ar ion splutter-
ing followed by annealing in vacuum at 700 K.
The resulting surface was amorphous in nature as
evidenced by the loss of satellite structure at ap-
proximately 8 eV toward higher binding energy
from the main C(1s) peak[7].
Controlled quantities of copper were deposited
onto the cleaned graphite surface at room tem-
perature via the K-cell, operating at a temperature
of 1100 °C. The copper film thickness was esti-
mated from the attenuation of the C(1s) substrate
peak, whereas sub-monolayer surface atom con-
centrations were obtained using quantification pro-
cedures described elsewhere [8].
Cyanogen was prepared by the thermal de-
composition of silver cyanide (Aldridge 99.99%) in
a stainless steel cylinder mounted on a UHV gas
line. The resulting gas was purified by successive
Fig. 1. C(1s) and N(1s) spectra obtained from (A) clean
graphite surface, (B) after exposure to 10 L cyanogen at 80 K,
(C) on warming to 300 K, and (D) on heating to 700 K.

A.F. Carley et al. / Surface Science 517 (2002) L563–L567
L565
Heating the sample further to 700 K did not affect
the shape or position of the N(1s) feature, em-
phasising the high stability of the adlayer.
3.2. Low temperature adsorption of cyanogen on a
ꢀ
graphite supported 20 A thick copper film
Such a high reactivity was not expected for the
relatively inert graphite surface, and we suspected
that a beam effect might be involved. To examine
whether overlayers of cyanogen adsorbed on a
clean graphite surface are sensitive to X-ray radi-
ation a clean graphite surface was exposed to 10 L
Fig. 3 shows C(1s) and N(1s) spectra obtained
ꢀ
following the exposure of an atomically clean 20 A
thickcopper film to 10 L of (CN)
subsequent warming to 300 K. The adsorption of
(CN) at 80 K gives rise to an intense single peak
at 80 K, and
2
2
at a binding energy of 288.2 eV in the C(1s)
spectrum. On warming to 110 K complete loss of
this peakis observed and a new peakat a binding
energy of 286.5 eV with a slight shoulder towards
a lower binding energy at 284.0 eV is revealed.
Quantification of the peakremaining at 110 K re-
2
of (CN) at 80 K, and then warmed to room
temperature, and XPS spectra were only then ac-
quired (Fig. 2). The N(1s) spectrum shows that
there are no reaction products remaining on the
surface at 298 K. Clearly, X-ray irradiation of the
cyanogen multilayers at 80 K initiates a reaction
resulting in the presence of a surface cyano-like
species. A number of studies have reported photon
induced chemical reactions [9,10], and indeed elec-
tron and photon beams have been shown to act as
polymerisation initiators of heterocycles on metal
surfaces [11]. Clearly, we must bear such processes
in mind as we move on to study the adsorption of
cyanogen on supported copper surfaces. No X-ray
effect is observed for cyanogen adsorption on
graphite at 300 K, suggesting that it is the irradi-
ation of physisorbed layers of cyanogen which
initiates this phenomenon.
15
veals a surface concentration of 3:8 ꢀ 10 atoms
À2
cm , indicative of multilayer formation. Both the
N(1s) and C(1s) peaks observed in the temperature
range 110–300 K exhibit a slight asymmetry.
The C(1s) and N(1s) spectra in Fig. 4 demon-
strate the effect of X-ray irradiation on this system.
We note that
(i) the copper film reacts with adsorbed cyanogen
even in the absence of X-ray irradiation, giving
Fig. 2. C(1s) spectra obtained, at 300 K, from (A) clean
graphite surface exposed to 10 L cyanogen at 80 K and warmed
to 300 K, (B) clean graphite surface exposed to 10 L cyanogen
at 80 K and exposed to an Al Ka X-ray beam for one hour
before warming to 300 K.
ꢀ
Fig. 3. C(1s) and N(1s) spectra obtained from (A) 20 A thick
Cu film, (B) after exposure to 10 L cyanogen at 80 K, (C) on
warming to 110 K, (D) on warming to 200 K, and (E) on
warming to 300 K.

L566
A.F. Carley et al. / Surface Science 517 (2002) L563–L567
species generated by these methods has yet to be
fully established. Cyanogen is known to form two
distinct polymers: polycyanogen, a polymer with a
backbone of –C@N– units with a pendant nitrile
on the carbon; and paracyanogen, whose structure
was generally thought to be:
However, there is little direct evidence for this
structure, and recently Maya [14] has re-examined
the structure and chemical reactivity of paracya-
nogen and has concluded that the paracyanogen
2
is mainly made of sp carbons in a disordered
structure [15] incorporating nitrogen, but without
an extended delocalised p electron system:
ꢀ
Fig. 4. C(1s) and N(1s) spectra obtained, at 300 K, from 20 A
thickCu film (A) exposed to 10 L cyanogen at 80 K and
warmed to 300 K, (B) exposed to 10 L cyanogen at 80 K and
exposed to an Al Ka X-ray beam for one hour before being
warmed to 300 K.
rise to a species at 300 K which we identify as
CN(a);
(
ii) compared with the un-irradiated overlayer, ir-
radiation of the condensed cyanogen layer
with X-rays at 80 K leads to a much higher
coverage of cyanogen remaining on the surface
at 300 K and significantly higher C(1s) and
N(1s) binding energies. The N(1s) binding en-
ergy is identical to that observed for analogous
experiments on the clean graphite surface.
The observed thermal behaviour of the X-ray in-
duced overlayer generated in this work(Fig. 1) is
not inconsistent with the thermal stability of para-
cyanogen, which only decomposes at temperatures
above 950 K, to yield cyanogen [14].
The XPS data in Figs. 1 and 3 also give some
insight into the mechanism by which the cyanogen
overlayer is modified by the X-ray irradiation.
Two photochemical processes are known to occur
as a result of irradiation of an adsorbate covered
surface, classified as being either direct or indirect.
Direct photoexcitation involves the direct excita-
tion of the adsorbate to a higher level. In contrast,
indirect photoexcitation either arises from the ini-
tial absorption of a photon into the underlying
substrate, from which a photoejected electron or
electronic energy transfer results in the excitation
Furthermore, we find that the final surface cover-
age obtained at 300 K via irradiation of the
physisorbed cyanogen is independent of the cop-
per coverage, indicating a non-specific X-ray in-
duced transformation, which occurs even on the
unmodified graphite surface.
3
.3. Nature of the X-ray modified overlayer
The most likely explanation is an X-ray induced
polymerisation. It has long been known that cya-
nogen can be polymerised by pyrolysis [12] or by
UV photolysis [13]; however the form of polymeric

A.F. Carley et al. / Surface Science 517 (2002) L563–L567
L567
of the adsorbate. Since the appearance of the
Acknowledgements
modified cyanogen species is only observed after
desorption of the upper overlayer (70% of the
overlayer is observed to desorb on warming), it
can be concluded that the indirect photoexcita-
tion mechanism most likely results in the over-
layer modification observed in this study. If direct
photoexcitation were the mechanism we would
expect complete polymerisation of the multilayer
cyanogen film at 80 K, whereas in the case of in-
direct photoexcitation the extent of reaction is
limited by the mean free path of photoelectrons in
the condensed film.
We are grateful to the EPSRC and DERA for
support of this work.
References
[1] R.E. Wilson, J.C. Whetzel, US patent 1,519,470 (1921).
[
[
2] B.R. Alves, A.J. Clark, Carbon 24 (1986) 287.
3] M.E. Bridge, R.A. Marborow, R.M. Lambert, Surf. Sci. 57
(1976) 425.
4] D.A. Outka, S.W. Jorgensen, C.M. Friend, R.J. Madix,
J. Mol. Cat. 21 (1983) 375.
[
[
[
[
5] N.J. Gudde, R.M. Lambert, Surf. Sci. 124 (1983) 372.
6] R.W. Joyner, M.W. Roberts, Surf. Sci. 87 (1979) 501.
7] A.F. Carley, L.A. Dollard, P.R. Norman, C. Pottage,
M.W. Roberts, J. Electron. Spectrosc. Related Phenom. 98
4
. Conclusion
(
1999) 223.
8] A.F. Carley, M.W. Roberts, Proc. Roy. Soc. London A
63 (1978) 403.
9] T.E. Madey, Science 234 (1986) 316.
[
[
The interaction of cyanogen with graphite and
3
graphite supported copper film surfaces has been
investigated using XPS. On both surfaces we find
that an Al Ka X-ray source induces a modification
of condensed cyanogen overlayers, as evidenced by
C(1s) and N(1s) core level shifts. The modification
is suggested to be X-ray induced polymerisation
formed via an indirect mechanism, involving sec-
ondary electrons from the substrate.
[10] M.L. Knotek, P.J. Feibelman, Surf. Sci. 90 (1979) 78.
[11] T.A. Land, J.C. Hemminger, Surf. Sci. 268 (1992) 179.
[
[
12] L.J. Gay Lussac, Ann. Chim. (Paris) 95 (1815) 175.
13] T.R. Hogmess, Liu-Sheng Tsai, J. Am. Chem. Soc. 54
(
1932) 123.
14] L. Maya, J. Polym. Sci. Pt. A, Polym. Chem. 31 (1993)
595.
[15] T.K. Brotherton, J.W. Lynn, Chem. Rev. 59 (1959) 841.
[
2