5% is shared by 4 to 5 other elements. Therefore, the
small impurity content causes only minor side
effects as compared to the very high oxygen content
in the surface layer and its major influence on the C1s
binding energy (BE). In addition, the BEs of C—N
and C—S species are much closer to the basic C1s
284.6 eV maximum than are the BE maxima recorded
for highly oxidized carbon atoms. As a matter of fact,
the binding energies of 293 eV and higher are due to
functional groups contribute less intensively to the
whole C1s peak. Such a carbon surface, after the Na-
modification, is covered by a relatively high number
of carbon atoms oxidized so intensively that their
corresponding energy shift exceeds 290 eV. In some
papers [8—10] this intensive shift in the binding en-
ergy is ascribed to carboxylic or carbonate structures.
With regarding to the method of carbon film produc-
tion and particularly the Na-modification process
one has to discuss whether the 290.7—292.5 eV region
corresponds to: (i) residual carboxylic groups from
*
pPp electronic transitions resulting from the
‘‘aromatic nature’’ of the solid. However the heights
of the corresponding peaks do not exceed 5—10% of
the height of the basic (‘‘graphite’’) C1s sub peak at
CH COONa introduced to the precursor, (ii) newly
ꢁ
formed carbonate groups that may appear in the
*
284.6 eV [5—10]. The intensity of the pPp sub
carbon after the thermal transformation of sodium
acetate into sodium carbonate during carbonization,
and (iii) carbon dioxide adsorbed onto the carbon
surface during the storage of the samples in the open
air.
peak is usually lower than the intensity of ante-
cedent sub peaks ascribed to carbon—oxygen surface
groups. Any departure from these two normal features
*
may be treated as a superposition of a pPp sub
peak and another sub peak resulting from the pres-
ence of a carbon derivative (probably carbon—
oxygen groups). The exclusive attribution of the sub
peaks on Figs 1—4 to carbon—oxygen structures is
discussed in the next section of this paper that ana-
lyses the O1s spectra.
The C1s spectrum of a Na-modified LT carbon
shown in Fig. 2 is dramatically different from the C1s
spectrum of the pure carbon film. In this case two
forms of carbon are dominating the spectrum; namely
graphitic carbon (284.6 eV) and heavily oxidized car-
bon (290.7—292.5 eV). Moderately oxidized forms
of carbon atoms entrapped in alcohol or carbonyl
If the residual sodium acetate is present in large
quantities in the Na-modified carbons then one
should observe intensive sub peaks in the range
288—289 eV [10]. However, this region in Figs 2 and
4 contains only minor sub peaks suggesting a low
contribution from sodium acetate. Therefore one may
assume that carboxylic groups do not dominate the
chemical map of the Na-modified carbon surface. Of
particular interest to this discussion is that thermo-
gravimetric examinations confirm that sodium acetate
added to furfuryl alcohol decomposes to Na CO
ꢀ
ꢁ
during carbonization [11].
The problem of the origin of the 290.7—292.5 eV BE
region has another aspect, i.e., the form of the remain-
ing sodium carbonate in the Na-modified carbon films
provided that such residuals really exist. Particularly,
one has to consider the crystalline form of this com-
pound and the size of the Na CO crystallites. X-ray
ꢀ
ꢁ
diffraction data presented elsewhere [11] suggest that
during the carbonization of Na-modified LT and HT
carbons, sodium acetate undergoes a transformation
with the formation of Na CO microcrystallites. This
ꢀ
ꢁ
transformation only proceeds efficiently under the
specific conditions created in bulk carbon samples. In
the reported case the carbon powder had a grain size
of 0.02—0.12 mm. The possible crystallization of
Na CO in thin carbon films is still not fully under-
ꢀ
ꢁ
stood and it is under intensive examination. The
possible presence of Na CO microcrystallites in the
Figure 3 Typical C1s spectrum of pure HT carbon film.
ꢀ
ꢁ
carbon films cannot be fully excluded but their size
should be comparable to the molecular scale
(1—10 nm). It should be noted that any potential crys-
tallization of Na CO is a 3-dimensional process that
ꢀ
ꢁ
requires a certain space to exist inside the films. Thus,
any Na CO microcrystallites formed in the films
ꢀ
ꢁ
could influence the XPS detectable surface properties
of the Na-modified carbon films provided some
microcrystallites are placed just in the subsurface
region and are partly exposed to the ambient
atmosphere.
The analysis of the Na1s spectrum will be discussed
at a later point of this paper.
C1s sub peaks that have a very large BE shift (above
290 eV) are observed in the spectra of Na-modified
HT carbon films (Fig. 4) while there are only a minor
Figure 4 Typical C1s spectrum of Na-modified HT carbon film.
6065