Proc. 8th Int. Conf. X-ray Microscopy
IPAP Conf. Series 7 pp.331-333
Chemical Analysis of Rust on Japanese Smoked Roof Tiles using Soft X-Ray Spectroscopy
Yasuji MURAMATSU1, Mika HIROSE2, Muneyuki MOTOYAMA3, Eric M. GULLIKSON4, and Rupert C. C. PERERA4
1
Department of Materials Science and Chemistry, University of Hyogo, 2167 Shosha, Himeji Hyogo 671-2201, Japan
Matsuoka Roofing Inc., 2039-1 Funatsu-cho, Himeji, Hyogo 679-2101, Japan
3
Center for Corporate Relations, University of Hyogo, 1-3-3 Higashikawasaki-cho, Chuo-ku, Kobe, Hyogo 650-0044, Japan
4
Center for X-Ray Optics, Lawrence Berkeley National Laboratory, 1 Cyclotron Road, Berkeley, California 94720, USA
2
A defect that rarely forms on the surface carbon films of Japanese smoked roof tile, Ibushi-Kawara, was characterized using soft x-ray absorption spectroscopy. The soft x-ray absorption spectra in the FeL region confirm that the
defect is iron (III) oxide, Fe2O3, and/or hydroxide FeOOH. Therefore, the defect is regarded as rust and not as defective carbon compounds. The mapping measurements determined by monitoring the x-ray absorption peak intensity in
the FeL, OK, and CK regions indicate that Fe is distributed at the defective spot center and spreads outward on the
surface. Hence, it is estimated that rust can be formed by the oxidation of the iron found in the basal sintered clay
when water soaks through the pinholes on the surface carbon films and then the rust spreads on the surface
KEYWORDS: soft x-ray spectroscopy , synchrotron radiation, chemical analysis, carbon films, smoked roof tiles, rust
ings with a 50-Pm exit slit. The 600 lines/mm grating was
used for the CK and OK regions, while the 1200-lines/mm
grating was used for FeL region. All the spectra were collected over a wide-range-scanning from 200 eV to 800. The
estimated energy resolution (E/¨E) of the incident x-rays
was about 2 x103, 7x102, and 7x102 in the CK, OK, and FeL
regions, respectively. The incident x-ray intensity was monitored using the photocurrent of a reference gold sample and
the x-ray beams were normal incident to the sample plane.
Scanning a knife-edge along the x and y directions indicated
that the spot size of the incident x-ray beams at the sample
position was 350 Pm in the horizontal (x) direction and 40
Pm in the vertical (y) direction, as shown in Fig. 1. This spot
size is sufficiently small to measure the element distributions of mm-order-sized samples.
1. Introduction
The Japanese smoked roof tile “Ibushi-Kawara” (hereafter referred to as Kawara) is made from a sintered clay plate
with a carbon film coating on the surface. The outstanding
features of Kawara include the metallic silver color and
weather resistance, which essentially originate from the surface carbon films deposited by chemical vapor deposition
(CVD) of hydrocarbons1). To further understand the features
from a chemical-bonding viewpoint, we investigated the
Kawara tiles by soft x-ray absorption and emission spectroscopy using synchrotron radiation. Our previous study2) revealed that the carbon films essentially consist of carbonblack-like sp2 carbon atoms that have a unique microstructure. Half of the carbon atoms form oriented clusters parallel
to the basal clay plane, which results in the metallic color.
The rest form random clusters, which rigidly connect the
layered clusters and result in the durability. Carbon films
degraded by weathering have also been analyzed. Weathered carbon films are oxidized and have a less-oriented
structure on the surface.3,4) Another important issue on
Kawara is the rare defects formed on the surface carbon
films in the water-repellent test during the manufacturing
process. The defects, which present as colorful spots on the
surface, deteriorate the quality of Kawara. Therefore, it is
important to characterize these defects and to determine the
mechanism of defect formation to further improve the
Kawara quality.
In this study, we analyzed the defects on Kawara using
soft x-ray absorption spectroscopy, including elemental
mapping measurements, and discussed the mechanism of
defect formation.
X-ray intensity (arb. units)
horizontal, x
0.35 mm
-3
-2
-1
0
1
2
3
0
0.1
0.2
0.3
vertical, y
0.04 mm
2. Experimental
We used total-electron-yield (TEY) x-ray absorption
spectroscopy to analyze the defects on Kawara since the
TEY method is generally surface-sensitive and can elementselectively identify the chemical bonding states. Spectral
measurements of the TEY x-ray absorption were performed
at beamline BL-6.3.25) at the Advanced Light Source (ALS).
The x-ray absorption spectra were measured by monitoring
the sample photocurrent using 600- or 1200-lines/mm grat-
-0.3
-0.2
-0.1
Position / mm
Fig.1 Beam size profile at the sample position, which was
measured by scanning a knife edge in the horizontal (x) and
vertical (y) directions.
331
Proc. 8th Int. Conf. X-ray Microscopy
IPAP Conf. Series 7
Figure 2 shows a measured Kawara piece (15 mmx x 11
mmy) where a defective spot was formed on the surface in
the water-repellent test during the ordinary manufacturing
process of smoked Kawara. The center of the spot is brightly
colored and gradually darkens the further from the center.
Areas outside of the spot show the typical oxidized-silver
color of the original Kawara. In the x-ray absorption measurements, three positions (at the spot center (denoted by A
in the figure), the darker-colored border (B) of the spot, and
the area outside of the spot (C)) were selected to investigate
the difference in the chemical states among the spot positions. Triangle-shaped carbon tapes were attached around
the spot border as markers. The distribution of the elements
(C, O, and Fe) was measured by scanning the sample position in the x and y directions while monitoring the x-ray
absorption peak intensity in the CK, OK, and FeL regions
(mapping measurements). The photon energies of the monitored x-ray absorption peaks were tuned to 293.4 eV (for
CK region), 539.6 eV (OK), and 709.7 eV (FeL), where the
most intense absorption peaks were observed. Commercially
available iron compounds of Fe metal, FeO, Fe3O4, Fe2O3,
and D-FeOOH were prepared as reference samples.
marker
shoulders near 709 eV and 707 eV. The FeL3 spectral shape
of the defective Kawara almost corresponds to the spectra of
the reference Fe2O3 and a-FeOOH, but does not correspond
to that of Fe metal, FeO, and Fe3O4. It is well known that
iron (III) oxide Fe2O3 can be formed by heating iron (III)
hydroxide FeOOH and that Fe2O3 displays various color.
Hence, it is elucidated that the defect on Kawara is the oxide
and/or hydroxide of Fe(III), which can be regarded as rust,
but not degraded carbon compounds.
FeL
Total electron yield (arb. units)
332
FeL
FeL
2nd
3rd
OK
2nd
OK
A
CK
B
C
A
ref
B
marker
C
mark of the contacted
carbon tape
200
FeL3
FeL2
Fe
Total electron yield (arb. units)
Figure 3 shows the TEY x-ray absorption spectra
scanned in a wide-energy range (200 eV – 800 eV) at three
positions (A, B, and C) on the defective Kawara and reference Kawara. The x-ray absorption peaks of CK, OK (1st
and 2nd orders), and FeL (1st, 2nd, and 3rd orders) are observed in the defective Kawara. On the other hand, the FeL
peaks are not observed in the reference Kawara. In addition,
the relative intensity of the OK in the defective Kawara is
stronger than that of the reference Kawara. Thus, it is elucidated that the defect contains both iron and oxygen. The
relative peak intensity of FeL and OK in the defective
Kawara is the strongest at the spot center (A) and decreases
outward (B and C). This suggests that the defect forms at the
center and spreads outward on the surface.
To analyze the chemical states of the defect, the x-ray
absorption spectral features in the FeL region of the defective Kawara are compared to the reference Fe compounds,
as shown in Fig. 4. The FeL3 absorption peak of the defective Kawara shows a peak at 710 eV with the lower-energy-
800
Fig. 3 TEY x-ray absorption spectra in a wide-energy-range
of 200 eV – 800 eV for the defective Kawara and reference
Kawara.
Fig. 2 Photographs of sample Kawara pieces with a defective spot.
3. Results and Discussion
400
600
Energy / eV
FeO
Fe3O4
Fe2O3
D-FeOOH
A
B
C
700
710
720
Energy / eV
730
Fig. 4 TEY x-ray absorption spectra in the FeL region for
the defective Kawara and reference Fe compounds.
IPAP Conf. Series 7
Proc. 8th Int. Conf. X-ray Microscopy
To investigate the mechanism of rust formation on
Kawara, elemental distributions of Fe, O, and C were measured by scanning the sample position and by monitoring the
x-ray absorption peaks the FeL, OK, and CK regions. Figure
5 shows the mapping spectra for Fe, O, and C in a x-y plane
and in a three-dimensional x-y plane with absorption intensity. The Fe spectra clearly show that the Fe concentrates at
the center of the defective spot and radiates outward. The O
spectra show that the O also concentrates around the spot
and gradually spreads outward. However, O is more widely
distributed on the surface than Fe due to rust and the oxidized surface carbon films4). The carbon film on a typical
Kawara is slightly oxidized during the manufacturing process as shown in Fig. 3 (the OK absorption peak is observed
in the reference Kawara). The C spectra show that the surface C is distributed outside of the defective rust spot, which
is reasonable if the surface carbon films are covered by rust.
333
From the x-ray absorption measurements and the mapping measurements, it is plausible that the defective rust is
formed by the oxidation of Fe with water during the waterrepellent test. The sintered clay substrates contain Fe from
minerals, which may be sufficiently reduced during the
high-temperature sintering (1000oC). If pinholes form on
the surface carbon films of Kawara, then the reduced Fe can
react with water through the pinhole during the waterrepellent test to form oxidized Fe compounds (Fe2O3 and/or
FeOOH). These oxidized Fe compounds can ooze through
the pinhole and spread onto the Kawara surface.
4. Conclusion
We analyzed a defect that is formed on the surface carbon
films of Japanese smoked roof tile, Ibushi-Kawara, using
soft x-ray absorption spectroscopy. By comparing the surface-sensitive TEY x-ray absorption spectra of the defective
Kawara piece to various iron oxides, the defect is confirmed
to be iron (III) oxides Fe2O3 and/or hydroxide FeOOH.
Therefore, the defect is regarded as rust, but not as degraded
carbon compounds. The mapping measurements from monitoring the x-ray absorption peaks in the FeL, OK, and CK
regions indicate that Fe is distributed at the defective spot
center and spread outward. Therefore, it is estimated that the
defective rust is formed by the oxidation of the Fe from the
minerals contained in the basal clay with the water that
penetrates the pinholes of the carbon films. Then the Fe oxide/hydroxide oozes to the surface through the pinholes and
spreads on the surface carbon films. Hence, it is concluded
that a thick carbon film without pinholes is essential for improving Kawara quality control.
Acknowledgments
We express thanks to Dr. Susumu Kawai of Hyogo Prefectural Institute of Technology for the useful discussion on
the manufacturing process of Kawara. This study was supported by a Grant-in-Aid from the Japanese Ministry of
Education, Culture, Sports, Science, and Technology under
contract No. 15550081 and 17550090, and by the US Department of Energy under contract No. DE-AC0376SF00098.
1) M. Motoyama, M. Tanaka, K. Ishima and G. Hashizume: Yogyo-KyokaiShi 86 (1978) 51-57.
2) Y. Muramatsu, M. Motoyama, J. D. Denlinger, E. M. Gullikson, and R.
C. C. Perera: Jpn. J. Appl. Phys. 42 (2003) 6551-6555.
3) Y. Muramatsu, M. Yamashita, M. Motoyama, J. D. Denlinger, E. M.
Gullikson, and R. C. C. Perera: Spectrochimica Acta B59 (2004) 13171322.
4) Y. Muramatsu, M. Yamashita, M. Motoyama, M. Hirose, J. D. Denlinger,
E. M. Gullikson, and R. C. C. Perera: X-Ray Spectrometry (in press).
5) J. H. Underwood, E. M. Gullikson, M. Koike, P. J. Batson, P. E. Denham, K. D. Franck, R. E. Tackaberry, and W. F. Steele: Rev. Sci. Instrum. 67 (1996) 3372.
Fig. 5 Mapping spectra of Fe, O, and C measured by monitoring the x-ray absorption peaks in the FeL, OK, and CK
regions and by scanning the sample position in the horizontal (x) and vertical (y) directions. The upper photograph
shows the measured Kawara sample.