ARTICLE IN PRESS
Vacuum 80 (2005) 178–183
www.elsevier.com/locate/vacuum
XAS study of oxygen plasma-treated micronized
iron oxide pigments
I. Arčona,b,, M. Mozetičb, A. Kodreb,c
a
Nova Gorica Polytechnic, Vipavska 13, SI– 5000 Nova Gorica, Slovenia
b
Institut Jožef Stefan, Jamova 39, 1000 Ljubljana, Slovenija
c
Faculty of Mathematics and Physics, University of Ljubljana, Jadranska 19, SI– 100 Ljubljana, Slovenia
Abstract
The stability of commercially available yellow (Bayferrox 3920) and red (Bayferrox 140) micronized iron oxide
pigments against highly aggressive treatment with oxygen plasma is examined. The red pigment microparticles with
hematite crystal structure are stable, while the yellow pigment changes colour during the treatment. Structural analysis
with powder X-ray diffraction and X-ray absorption methods, EXAFS and XANES, shows that the yellow pigment
microparticles recrystallize from goethite to hematite crystal structure. Larger hematite particles are composed of small
highly anisotropic nano-domains. No nanoclusters or amorphous iron oxide phases are detected. The valence state of
iron atoms in the pigments is not affected by the strongly oxidative plasma environment.
r 2005 Elsevier Ltd. All rights reserved.
Keywords: Iron oxide pigments; Oxygen plasma; Fe K-edge EXAFS; XANES; Nano-domains
1. Introduction
Oxygen plasma treatment of materials has
become a technique widely used on both an
experimental and industrial scale. It is used in
chemistry, biology, medicine, nanotechnology,
and in the electronic, and automotive industries
[1–8]. The technologies based on the application of
Corresponding author. Nova Gorica Polytechnic, Vipavska
13, SI–5000 Nova Gorica, Slovenia. Tel.: +38653315227;
fax: +38653315240.
E-mail address: [email protected] (I. Arčon).
an oxygen plasma include non-isotropic plasma
etching, selective etching, low-temperature ashing,
degreasing, surface activation, and plasma sterilization.
A selective etching of polymer–matrix composite materials by weakly ionized plasma is often
used in chemical practice as an excellent method of
preparation of samples prior to SEM analysis in
order to detect the distribution and orientation of
particles in a polymer matrix [8–13]. High etching
selectivity can be obtained due to the different
probability for oxidation with O atoms for
different composite ingredients in the material.
0042-207X/$ - see front matter r 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.vacuum.2005.08.012
ARTICLE IN PRESS
I. Arčon et al. / Vacuum 80 (2005) 178–183
In this process, the mineral pigment particles are
often presumed immune to the treatment.
In this work, we examine the stability of two
commercially available micronized iron oxide
pigments (yellow—Bayferrox 3920 and red—Bayferrox 140) to the highly aggressive treatment with
oxygen plasma. The structural and chemical
changes in the pigments upon exposure to the
plasma were analysed by X-ray powder diffraction
(XRPD) and X-ray absorption methods—EXAFS
and XANES.
179
beam stabilization feedback control. Samples of
the iron oxide pigments and reference compounds
(Fe2O3, Fe3O4, FeSO4) were prepared as fine
powders homogeneously distributed on multiple
layers of adhesive tape. The thickness of each
sample was chosen to provide optimal absorption
of 1 above the Fe K-edge. The exact energy
calibration was established with a simultaneous
absorption measurement on the Fe metal foil.
3. Results
2. Experimental
Micronized yellow iron oxide pigment Bayferrox 3920 and red pigment Bayferrox 140 have been
exposed to oxygen plasma created in a plasma
reactor with a volume of 0.03 m3 with an
inductively coupled RF generator with frequency
of 27.12 MHz and power of 5000 W. The reactor
was pumped with a two-stage oil rotary pump with
the pumping speed of 60 m3/h. The base pressure
was about 0.3 Pa. Commercially available oxygen
(purity 99.99%) was leaked into the plasma
reactor so that the working pressure was 75 Pa.
Plasma parameters were measured with a Langmuir probe and a catalytic probe [14–16]. The
electron temperature was about 6 eV, the plasma
density about 2 1016 m3 and the density of
neutral oxygen atoms 6 1021 m3. The pigment
samples were kept in plasma for 10 min.
The XRPD pattern of the product was collected
on a Philips PW 1710 diffractometer using Cu Ka
radiation in the 2y range from 101 to 801 in steps of
0.021 with a sampling time of 1 s per step.
X-ray absorption spectra in the energy region of
the Fe K-edge were measured in the transmission
mode at the E4 beamline of the HASYLAB
synchrotron facility at DESY in Hamburg. The
station provided a focused beam from an Aucoated toroidal mirror and a Si(1 1 1) double
crystal monochromator with about 1 eV resolution
at the Fe K-edge. Higher-order reflections from
the monochromator were effectively eliminated by
a plane Au-coated mirror, and by a slight detuning
of the monochromator crystals, keeping the
intensity at 60% of the rocking curve with the
The two pigments are distinct forms of iron
oxides. The XRPD patterns (Fig. 1) confirm that
the red pigment is composed of well-crystallized
microparticles with hematite (a-Fe2O3) crystal
structure, while the microparticles of the yellow
pigment have goethite (a-FeOOH) crystal structure.
During the plasma treatment no visual changes
were observed in the red pigment. In the yellow
pigment, a continuous change of colour from
yellow through orange to dark red was observed.
Fig. 1. XRPD spectra of the yellow pigment before (top) and
after plasma treatment (bottom). Spectrum of the red pigment
(a-Fe2O3) is shown for comparison (middle).
ARTICLE IN PRESS
I. Arčon et al. / Vacuum 80 (2005) 178–183
The XRPD pattern of the dark red product
obtained after 10 min exposure to oxygen plasma
reveals that the crystal structure of the pigment has
changed completely from goethite to hematite.
However, the diffraction peaks are much broader
than in the case of an untreated red pigment,
which clearly indicates a relatively poor crystallinity of the pigment particles after the treatment
[17]. The broadening is caused by nano-size
perfect-crystal domains, either in the form of
separate nanoparticles [18,19] or mosaic domains
in relatively large microcrystallites. The width of
the lines is inversely proportional to the nanoparticle size. A detailed inspection of the peak widths
reveals that there are two subsets of diffraction
lines in the spectrum with significantly different
broadening, indicating that the microcrystallites
are highly anisotropic, i.e. exhibit significantly
different sizes in different crystallographic directions [20].
The decisive information on the morphology of
the treated pigment is provided by Fe K-edge
EXAFS analysis, elucidating the local structure
around Fe atoms in the sample. EXAFS analysis is
performed with the University of Washington
programs using the FEFF6 code for ab initio
calculation of scattering paths [21,22]. The comparison of the Fourier transforms of the EXAFS
spectra measured on the pigments before and after
the exposure to oxygen plasma (Fig. 2) shows that
plasma treatment does not change the local
structure of the red pigment. The spectra before
and after the treatment are identical, exhibiting the
characteristic local Fe neighbour shell structure of
hematite. The change of crystal structure in the
yellow pigment from goethite to hematite after the
plasma treatment, identified by XRPD, is also
clearly visible in the local neighbourhood of the Fe
atoms (Fig. 2).
A detailed comparison of the spectrum of the
plasma-treated yellow pigment with the spectrum
measured on hematite reveals identical contributions from the first Fe coordination shell and a
suppression of the signal of further shells in the
treated pigment (Fig. 2). This may indicate either
the presence of some amorphous phase which
could not be detected by XRPD analysis, or small
hematite nanocrystals with diameters of the order
24
Yellow pigment
(goethite)
20
FT magnitude (arb. units)
180
16
Red pigment
plasma treated
12
8
Red pigment
hematite
(dashed line)
4
Yellow pigment
plasma treated
0
0
1
2
3
4
5
6
R (Å )
Fig. 2. Fourier
transforms of k3-weighted EXAFS spectra
( 1 to 14 Å1) of the yellow pigment before the plasma
(k ¼ 4:5 A
treatment—goethite (top); red pigment after the plasma
treatment (middle); yellow pigment after the plasma treatment
(bottom); hematite for comparison (dashed lines).
of 1 nm, which would result in reduced average
neighbour coordination numbers around Fe atoms
[18,19]. Another possibility is the poor crystallinity
of the hematite microparticles, which would
increase the average static disorder in the Fe
neighbour shells and, consequently, as witnessed
by an increase in Debye–Waller factors, reduce the
EXAFS signal.
Quantitative EXAFS analysis was used to
determine the type of structural difference. The
FEFF model based on crystal structure of
hematite (a-Fe2O3) [23] was constructed. The Fe
atom is octahedrally coordinated to six oxygen
atoms (three at 1.94 Å and three at 2.11 Å) in the
first coordination shell, followed by alternate shells
of iron and oxygen neighbours. The FEFF model
comprised all single and multiple scattering paths
ARTICLE IN PRESS
I. Arčon et al. / Vacuum 80 (2005) 178–183
up to 3.9 Å. We calibrated the model by the
EXAFS spectrum of the red pigment (a-Fe2O3),
obtaining a good fit (Fig. 3) for the R region from
1.1 to 4.0 Å with the following set of variable
parameters: the amplitude reduction factor (S20)
and the distances and Debye–Waller factors of
individual neighbour shells. The shell coordination
numbers were kept fixed at their crystal structure
values.
The same model was used to describe the
EXAFS spectrum of the plasma-treated yellow
pigment. Again, a very good fit is obtained with
the same neighbour distances and neighbour
coordination numbers as in hematite. The Debye–Waller factors of neighbour shells beyond the
nearest one are found to be significantly larger
than in hematite. Thus, we may conclude that the
181
crystal structure contains imperfections which
produce a large static disorder in the crystal
structure of hematite microcrystals. The disorder
does not affect the oxygen octahedra around iron
atoms, but rather the correlations between the
octahedra. There is no evidence for the presence of
small hematite nanoparticles or amorphous iron
oxide phases.
In addition, we checked for the change of the
iron valence state possibly induced in the highly
oxidizing atmosphere of the oxygen plasma. The
change can be deduced from the energy shift of the
Fe absorption edge [24]. The Fe XANES spectra
of the pigment samples and reference compounds
(FeSO4 and a-Fe2O3) with known iron oxidation
states are plotted in Fig. 4. The reference spectra
6
16
Fe foil
Normalised absorption
FT magntiude (arb. units)
12
Red pigment
8
(hematite)
4
FeSO4
Fe3 O4
Fe2O3
2
red pigment
plasma tr.
red pigment
(hematite)
4
Yellow pigment
plasma tr.
Yellow pigment
plasma treated
Yellow pigment
(goethite)
0
-30
0
1
2
3
4
5
6
R (Å)
Fig. 3. Fourier transforms of k3-weighted EXAFS spectra of
hematite and of yellow pigment after the plasma treatment
(experiment—solid line, EXAFS model—dashed line).
-20
-10
0
10 20
E (eV)
30
40
50
60
∇
0
Fig. 4. Normalized Fe K-edge XANES spectra of yellow and
red pigment before and after the plasma treatment and of
reference samples: FeSO4, Fe2O3, Fe3O4 and Fe metal. Zero
energy is taken at the 1 s ionization threshold in the Fe metal
(7112 eV).
ARTICLE IN PRESS
I. Arčon et al. / Vacuum 80 (2005) 178–183
182
show a K-edge shift of about 3.0 eV per valence
state. The edge positions of the treated samples,
which are best observed in the derivative spectra
(Fig. 5) are evidently identical with that in
the standard hematite sample. The shape of the
Fe K-edge, which is very sensitive to the changes
in local symmetry of the investigated atom site,
also confirms that after the plasma treatment
of the pigments, iron atoms are located at sites
with the same local octahedral symmetry as in
hematite.
In conclusion, the change in yellow pigment
colour during the oxygen plasma treatment can
be ascribed to recrystallization of microparticles
from goethite to hematite crystal structure. Larger
hematite particles are composed of small highly
anisotropic nano-domains. No nanoclusters or
amorphous iron oxide phase were detected.
0.8
Acknowledgements
Support by the Ministry of Education, Science
and Sport of the Republic of Slovenia within the
National Research Program, by the European
Community – Research Infrastructure Action
under the FP6 ‘‘Structuring the European Research Area’’ Programme (through the Integrated
Infrastructure Initiative ‘‘Integrating Activity on
Synchrotron and Free Electron Laser Science’’)
and the bilateral project BI–DE/03–04–004 by
Internationales Buero des BMBF. Access to
synchrotron radiation facilities HASYLAB (Project II–01–44) is acknowledged. K. Klementiev of
HASYLAB provided expert advice on E4 beamline operation. Laboratory for Inorganic Chemistry and Technology from National Institute of
Chemistry is acknowledged for providing XRPD
spectra of the pigments.
References
Absorption derivative dA/dE (arb. units)
Fe foil
FeSO4
0.6
Fe3 O4
0.4
Fe2O3
red pigment
plasma treated
0.2
red pigment
(hematite)
yellow pigment
plasma treated
0.0
yellow pigment
(goethite)
-30
- 20
-10
0
10
20
30
40
50
∇
E (eV)
Fig. 5. Absorption derivative of the Fe K-edge profile of
pigment and reference samples in Fig. 4.
[1] Boenig HV. Plasma science and technology. London:
Cornell University Press; 1982.
[2] Richard A. Reactive plasmas. Paris: Societe Franc- aise du
Vide; 1996.
[3] Wertheimer MR, Marinu L, Liston EM. Plasma sources
for polymer surface treatment. In: Shah SI, editor.
Handbook of thin film process technology. Bristol:
Institute of Physics Publishing; 1998.
[4] Šorli I, Petasch W, Kegel B, Schmidt H, Leibel G. Inf
Midem 1996;26:35.
[5] Moreau S, Tabrizian M, Barbeau J, Moisan M, Leduc A,
Pelletier J, et al. In: Richard A, editor. Proceedings of the
twelfth international colloid on plasma processes. Paris:
Socience Franc- aise du Vide; 1999.
[6] Holy CE, Cheng C, Davies JE, Shoichet MS. Biomaterials
2001;22:25.
[7] Montes–Moran MA, Martinez–Alonso A, Tascon JMD,
Young RJ. Composites A 2001;32:361.
[8] Mozetič M, Zalar A, Panjan P, Bele M, Pejovnik S, Grmek
R. Thin Solid Film 2000;376:5.
[9] Kunaver M, Mozetič M, Gunde MK. Thin Solid Film
2004;459:115.
[10] Kunaver M, Gunde MK, Mozetič M, Horvat A. Surf Coat
Internat B 2003;3(68):175.
[11] Kunaver M, Gunde MK, Mozetič M, Horvat A. Dye
Pigment 2003;57:235.
[12] Gunde MK, Kunaver M. Appl Spectrosc 2003;57:
1266.
[13] Gunde MK, Kunaver M, Mozetič M, Pelicon P, Simčič J,
Budnar M. Surf Coat Internat B 2002;85:115.
ARTICLE IN PRESS
I. Arčon et al. / Vacuum 80 (2005) 178–183
[14] Šorli I, Ročak R. J Vac Sci Technol A 2000;18:338.
[15] Poberaj I, Babič D, Mozetič M. J Vac Sci Technol A
2001;20:189.
[16] Vesel A, Mozetič M. Vacuum 2001;61:373.
[17] Langford JI, Delhez R, de Keijser ThH, Mittemeijer EJ.
Aust J Phys 1988;41:173–87.
[18] Barbier A, Tuel A, Arčon I, Kodre A, Antonin Martin G.
J Catal 2001;200:106–16.
[19] Arčon I, Tuel A, Kodre A, Martin G, Barbier A. J Synch
Radiat 2001;8:575–7.
183
[20] Louer D. Advances in X-ray analysis. In: Gilfrich JV, et al.,
editors. vol. 37. New York:Plenum Press; 1994. p. 27–35.
[21] Stern EA, Newville M, Ravel B, Yacoby Y, Haskel D.
Physica B 1995;208–209:117–20.
[22] Rehr JJ, Albers RC, Zabinsky SI. Phys Rev Lett
1992;69:3397–400.
[23] Maslen EN, Streltsov VA, Streltsova NR, Ishizawa N.
Acta Crystallogr B 1994;50:435–41.
[24] Ristić A, Novak Tušar N, Arčon I, Zabukovec Logar N,
Thibault-Starzyk F, Czyzniewska J, et al. Chem Mater
2003;15:3643–9.