PERGAMON Carbon 39 (2001) 991–999

A novel laser technique for oxidation-resistant coating of

carbon–carbon composite
Lloyd Snell, Avery Nelson, Pal Molian*
Mechanical Engineering Department, Iowa State University, Ames, IA 50011, USA
Received 20 December 1999; accepted 2 July 2000

Abstract

High-temperature protection of carbon–carbon (C / C) composites against oxidation by ambient air is a serious problem
with no single, universal solution. In this paper, we present the results of a novel laser-induced chemical decomposition
(LICD) process used to apply two types of oxidation protection system (OPS) coatings on C / C composites. Silicon carbide
and iridium were selected as the candidate materials for OPS based on their chemical compatibility with carbon and
oxidation protection capabilities, and deposited on the C / C substrates by LICD of Si 3 N 4 and IrCl 3 . High-temperature
oxidation tests of the coatings in air environment at 6508C showed a mass loss improvement of about 10% for single layer
SiC and 50% for single layer iridium.  2001 Elsevier Science Ltd. All rights reserved.

Keywords: A. Carbon composites; B. Laser irradiation, Oxidation; C. Scanning electron microscopy; D. Surface properties

1. Introduction oxidation. When used as a thermal protector, the C / C

Carbon fiber-reinforced carbon (C / C) composites may
be the ultimate development in the history of carbon
technology and are extensively used in extreme perform-
ance situations that include high-temperature, high-speed,
low-friction, and highly corrosive environments [1–10].
These materials are known for their strength, light weight,
and thermal properties. Despite the numerous benefits of
C / C, their performance at high temperatures in long-term
applications is seriously affected by their susceptibility to
oxidation. At temperatures above 4008C, the carbon fibers
readily react with oxygen and rapidly burn away, resulting
in partial or complete deterioration of the composite [3,10].
The oxidative behavior of C / C composites, which begins
with gasification of carbon fibers, is followed by enlarge-
ment of pores inherently present in the composite, further
aggravating the oxidation. The oxidation rate of C / C
composites in air increases rapidly with increased tempera-
ture [10]. Water vapor in air further increases the rate of
*Corresponding author. Fax: 11-515-294-3261.
E-mail address: molian@iastate.edu (P. Molian).
composite is often made extremely thick to compensate for
its tendency to oxidize rapidly at elevated temperatures.
Research into development of an efficient oxidation
protection system (OPS) for carbon has been an ongoing
effort for the past 50 years, and many different combina-
tions of materials and processes have been attempted
[3,10]. Currently, three primary methods are used to apply
OPS to the surface of C / C composite: pack cementation,
chemical vapor deposition (CVD), and slurry method. The
problems common to these techniques include low prod-
uctivity (they are slow and tedious), low quality (thermal
cracks and porosity result), and non-flexibility (they cannot
apply coatings of varying compositions and structures). In
addition, the CVD and slurry methods occasionally require
vacuum and are difficult to apply to large-scale structures.
In this work, a novel laser-induced chemical decomposi-
tion (LICD) method of coating C / C composite was
investigated, the objectives being to develop a better
method and to enhance the usability of these composites in
extreme operating conditions.
LICD process has been previously applied [11] for
direct-writing metallization applications in microelectron-
ics using organometallic compounds. The principle of the
process is to use a high-energy laser beam to dissociate a

0008-6223 / 01 / $ – see front matter  2001 Elsevier Science Ltd. All rights reserved.
PII: S0008-6223( 00 )00200-1
992 L. Snell et al. / Carbon 39 (2001) 991 – 999
precursor solid or liquid. For example, Si 3 N 4 and IrCl 3 can
be reduced to Si and Ir by the application of laser
radiation.
Si 3 N 4 → 3Si(l) 1 2N 2 (g)
IrCl 3 → Ir(s) 1 3Cl(g)
The purpose of this work is to demonstrate the feasibility
of using LICD technique to deposit silicon carbide (SiC)
and iridium coatings on C / C substrates and evaluate the
oxidation resistance of these coatings. SiC is the most
widely used coating material for C / C because of high
refractoriness, chemical compatibility with carbon, and
better match of coefficient of thermal expansion (CTE)
with carbon, compared with other materials. The oxidation
resistance derives from the formation of silica (SiO 2 ) scale
upon oxidation. SiO 2 has low vapor pressure and low
oxygen diffusivity. Noble metals such as iridium could be
alternative coatings because of their high melting point
(24408C) and extremely low permeability to oxygen.
Iridium has been demonstrated to offer excellent protection
against oxidation but the difficulties of producing high-
quality coatings, along with cost and availability, are
significant barriers to further development of its use [3,10].
2. Materials and methods
Two types of C / C composites were employed in this
work: Type I is a 2.5 mm thick, 2D-composite (lay-up5
2D 0 / 90M-0; fiber5Amoco T-300 3k tow; matrix5
phenol; ply58-harness satin, weave524323 tows per 25
mm); and Type II is a 6 mm thick, 3D-composite used in
friction applications. Type I was acquired from NASA–
Langley while Type II was received from BF Goodrich
Aerospace, Carbon Products Division. The samples with
dimensions of 20 mm320 mm were cut using a 1 kW CO 2
laser (1.27 m / min cutting speed, 13 psig of O 2 , 127-mm
focal length lens) and then polished in 600-grit SiC paper.
The samples were then cleaned by rinsing with isopropyl
alcohol (91%), allowed to dry, and weighed using a
balance with milligram (mg) resolution.
The LICD process consisted of two steps: first, a slurry
of precursor material was prepared and painted on the
surfaces of C / C composite; second, using a raster scanning
method, a laser was used to irradiate the slurry so as to
form the desired coating. Si 3 N 4 was chosen as the
precursor material because its thermal expansion (¯33
10 26 / K) matches closely with carbon than SiC (¯53
10 26 / K), allowing crack-free, excellent bonding of the
coating to be achieved during laser processing. When
Si 3 N 4 is subjected to laser irradiation, it decomposes into
liquid or gaseous silicon, depending on the type of laser
and the energy density of the beam [12]. The silicon
formed then reacts with carbon to form SiC, because the
energetic and process conditions in laser processing favor
SiC formation.
Si 3 N 4 (1–5 mm diameter) powder with a purity of 99%
was mixed with isopropyl alcohol (91%) and stirred with a
magnetic stirrer to obtain a homogeneous slurry. Although
no specific molar concentrations were calculated for the
mixture, enough alcohol was added to ensure that the
Si 3 N 4 could be suspended in the liquid. With an eyedrop-
per, a thin layer of Si 3 N 4 was deposited on a surface of a
C / C coupon, and then allowed to dry naturally. Drying
samples by forced airflow, as from a hair drier, caused the
coating material to buckle on the C / C coupon surface.
After the coating had dried, the coupons were weighed,
and weights were compared with initial weights to de-
termine the weight of the applied coating. The coating
thickness, in the range of 25–75 mm, was calculated by
dividing the weight gain by the density (3.3 g / cm 3 ) of
Si 3 N 4 , and then again by coating area.
For iridium metal coating, iridium chloride hexahydrate,
along with proprietary additives, was mixed with isopropyl
alcohol and blended for 1 h to form a smooth, fine slurry.
The amount of isopropyl alcohol was varied to control the
viscosity of the slurry. The purpose of the additives is to
reduce the thermal expansion mismatch of iridium (¯73
10 26 / K) with carbon that otherwise would lead to crack
formation. The samples were dipped in the slurry to form a
uniform coating of 50–100 mm in thickness.
The coated samples were mounted on an X–Y computer
controlled positioning table and then irradiated with two
types of high-power lasers, Nd:YAG and CO 2 , depending
on the type of coating. Si 3 N 4 deposited samples were
processed using a 30 W, continuous wave, Nd:YAG laser
because absorption of the YAG wavelength by Si 3 N 4 is
excellent. The beam was defocused (12 mm from the focal
point) to a larger size on the surface, using a 127 mm focal
length. The scanning speed ranged from 1 to 10 mm / s.
Argon gas was flown around the optical lens at a rate of 30
ft 3 / h in order to protect the lens from airborne discharges
and to isolate the fusion zone from oxidation during
scanning. A glass ring, 100 mm diameter and 40 mm in
height was secured to the sample platform to confine the
argon gas to the sample. Iridium coating was formed by
use of a 500 W, continuous wave CO 2 laser because of the
need for higher laser power to deposit metal coatings. All
the experiments were carried out using a lens with a focal
length of 254 mm and a defocused beam (12 mm above
the focal spot). The scanning rate was varied from 11
mm / s to 44 mm / s. Argon was again used as the assist gas.
Optimum coatings were obtained at 500 watts and a
scanning rate of 33 mm / s. In all the experiments, the
samples were laser-coated on all the sides.
For microstructural evaluation, one specimen from each
 L. Snell et al. / Carbon 39 (2001) 991 – 999 993

of the coating systems was sectioned in the laser-coated
condition, mounted, polished, and examined by optical and
scanning electron microscopy (SEM) to characterize the
coating microstructure and degradation of coatings due to
oxidation. Energy dispersive spectroscopy (EDS) was also
performed for determining semi-quantitative chemical
composition of coatings. X-ray photoelectron spectroscopy
(XPS) was used to identify various phases formed in the
coating.
After laser processing, samples were exposed to high
temperatures in air environment to test the effectiveness of
OPS. Effects of thermal exposure were evaluated using
mass-versus-time plots at 6508C. Isothermal oxidation
testing utilized samples mounted on a work-holding de-
vice, made of machinable ceramic, in the naturally occur-
ring environment of the furnace. Five slots were milled in
the ceramic material to allow the test samples to stand
upright. No attempt was made to control any other
variables during the temperature exposure period. Samples
were removed at 30-min intervals, observed by use of an
optical microscope, and weighed.
3. Results and discussion
Fig. 1 shows a scanning electron micrograph of SiC-
coated Type I C / C substrate. Rough, continuous, non-
uniform coatings with an average thickness of 200 mm

Fig. 1. Scanning electron micrograph showing the transverse section of SiC-coated Type I C / C substrate (1003).
994 L. Snell et al. / Carbon 39 (2001) 991 – 999
Fig. 2. X-ray photoelecron spectrum of SiC-coated showing the peaks for SiC, SiO, Si and others.
were observed. X-ray photoelectron spectroscopy analysis
of laser-processed layers revealed the formation of SiC on
the surface of C / C composite (Fig. 2). In addition to SiC,
metal silicates and silica were also formed. Fig. 3 shows
the oxidation test data of uncoated and laser-coated
samples at 6508C. Regression statistical analysis was used
to plot the data for a reliability of 95% based on 11 data
points. The laser-coated samples were produced using a
50-mm thick precursor Si 3 N 4 at a power of 30 W and scan
rate of 2 mm / s. The oxidation loss was reduced by
approximately 10% through the laser coating process. The
small improvement in oxidation resistance is not highly
unexpected because the oxidation data of unidirectional
Fig. 3. Oxidation test data of uncoated and SiC-coated Type I
C / C substrate at 6508C.
C / C composites, reported by Fitzer [13], indicates that the
weight loss improvement was only marginal when SiC was
deposited as a single layer. However, substantial improve-
ments were obtained if the SiC coated composite was
impregnated with boron oxide and zinc phosphates. The
beneficial effects of impregnators are attributed to the
filling of cracks in the SiC formed during oxidation [10].
Examination of the uncoated samples during oxidation
testing revealed a sequence of events beginning with
separation of the fibers at the edges, followed by surface
deterioration and appearance of visible strands of indi-
vidual carbon fibers, and ending with complete delamina-
tion. Fig. 4 shows the oxidized surfaces of laser-coated
C / C where the individual fibers, large cracks, and delami-
nated layers are readily seen. Fig. 4(a)–(c) show the
successive stages of delamination during oxidation. First,
the fibers are exposed through the coating (Fig. 4a) and
then cracks develop and further expose the fibers to the
oxygen (Fig. 4b). Finally, the fibers tend to lift the coated
layer (Fig. 4c). Although the laser-scanned shell remained
intact, the substrate was degraded by delamination, which
could be due to the porosity of the coating, which allows
oxygen to penetrate. During the oxidation tests, the layers
started to delaminate, although the edges continued to hold
them together. The delamination caused the center of each
sample to bow outwards.
Fig. 5 shows the scanning electron micrograph of a
transverse section of CO 2 laser-coated Ir on Type II C / C.
The coating is relatively continuous, with few cracks, and
 L. Snell et al. / Carbon 39 (2001) 991 – 999 995

Fig. 4. Scanning electron micrographs showing the oxidized surfaces of SiC-coated C / C substrate (a) 153 (b) 103 (c) 303.

exhibited good adherence to the substrate. The cracks were
confined to the iridium layer without propagating into the
C / C substrate. The coating-substrate interface exhibited
fine solidification grain structure of the coating with no
evidence of delamination. Fig. 6 shows the weight loss as a
function of exposure time at 6508C. The laser-coated
samples were produced using a 50-mm thick precursor
IrCl 3 at a power of 500 W and scan rate of 33 mm / s. Both
coated and uncoated samples did not undergo delamination
due to the 3D woven composite. The weight gain in the
coated sample was caused by oxide scale formation and
closing of the cracks. After 15 h, the specimen started
losing weight because of exfoliation of oxide scale. In
contrast, the uncoated sample showed a continuous weight
loss caused by oxidation; the sample disintegrated after 3
h. These results demonstrate that the laser-deposited
coatings substantially increase the oxidation resistance of
C / C composites. Fig. 7a and b are the top and transverse
996 L. Snell et al. / Carbon 39 (2001) 991 – 999

Fig. 5. Scanning electron micrograph showing the transverse section of Ir-coated Type II C / C substrate.

section of oxidized surfaces of coated samples showing
several large cracks and oxygen penetration.
Iridium coatings have been of great interest to the US
space programs for protecting the graphite nose cones on
reentry [14]. The problems confronted with iridium coat-
ings were volatile oxide formation, poor adhesion to
carbon substrates and mismatch in CTE. However, dense,
crack-free coatings were successfully produced by methods
such as slurry dipping, plasma deposition followed by hot
isostatic pressing, and electroplating [14,15]. Wright et al.
[16] reported that iridium coatings could be used to protect
graphite for an hour or more in air at 20508C based on
plasma torch tests. Harding and Fry showed [17] that
iridium coatings deposited by CVD prevented oxidation
for more than 24 h at 20008C.
It is well known that cracks are induced during cooling

Fig. 6. Oxidation test data of uncoated and Ir-coated Type II C / C substrate at 658C.
 L. Snell et al. / Carbon 39 (2001) 991 – 999 997

Fig. 7. Scanning electron micrographs showing the oxidized surfaces of Ir-coated Type II C / C substrate. (a) Longitudinal. (b) Transverse.
998 L. Snell et al. / Carbon 39 (2001) 991 – 999
down from the processing temperatures due to thermal
expansion mismatch. In this work, the use of Si 3 N 4
minimized the CTE mismatch and reduced the crack
formation in SiC coatings. However, the use of iridium as
a coating did produce cracks that could be attributed to the
large CTE mismatch. To provide effective oxidation
resistance, coatings should not only minimize cracks or
pores, but should be of uniform thickness. Coatings
obtained by the laser process showed few cracks and are
not uniform in thickness. In addition, coatings were not
formed in some isolated locations on the sample. These
defect regions provided pathways for oxygen to penetrate
the coating and attack the C / C composite substrate.
SiC was much less effective compared to Ir in protecting
the C / C composites from oxidation. This could be ex-
plained by the oxidation reactions of these two materials as
well as by the presence of chlorine compounds on the
surface of Ir-coated samples. Oxidation of C / C occurs at
temperatures higher than 5008C primarily because of its
affinity to oxygen and the porous structure. Silicon carbide
protects the composite up to 15008C through the following
oxidation reaction:
2SiC 1 3O 2 → 2SiO 2 (s) 1 2CO(g)
The presence of SiO 2 (s) reduces the oxidation rate and acts
as a glass sealant to fill the cracks developed due to
thermal expansion mismatch. However, pinhole formation
occurs in SiC because of vaporization and some corrosion.
Hence, a single layer of SiC is generally not protective at
high temperatures unless another coating such as boron
oxide is added. This is confirmed in the present study as
well as from the data reported by Fitzer [13] on single
layer SiC coating. Iridium could be effectively used at
these temperatures because it is a very good carbon and
oxygen barrier. In addition, the chlorine compounds
formed during LICD process are effective retardants for
the carbon–oxygen reaction, leading to reduced gasifica-
tion rates. In other words, the chlorine ‘‘poisons’’ the
active sites of carbon where oxygen attack occurs. While
improved effects were observed in this work with iridium
compared to SiC, the difficulty in obtaining excellent
oxidation-resistant performance is due to the arbitrary
process parameters used in the work.
4. Conclusions
The feasibility of forming oxidation-resistant, single
layer coatings of SiC and Ir by LICD process has been
demonstrated. Oxidation tests revealed that the coated C / C
substrates were protected more effectively by iridium than
SiC coating. A mass loss improvement of about 10% for
single layer SiC and 50% for single layer iridium were
obtained, implying that laser processing has the potential
to develop into a successful technique. Despite excellent
adhesion of the coating, oxygen still penetrates at sufficient
rates to be detrimental and causes difficulty in achieving
substantial improvements at high temperatures. Methods to
reduce the cracking and obtain uniform coatings will
greatly assist in producing high-performance laser coat-
ings.
Acknowledgements
The authors extend thanks to the following supporters of
this project: Iowa Space Grant Consortium, NASA Lan-
gley Research Center (Dr. Howard G. Maahs, Head of
Environmental Interactions Branch, and Mr. Craig Oh-
lhorst, Materials Division), Hampton, VA 23681; Carbon–
Carbon Advanced Technologies, Inc., (Mr. Doug Epper-
son), 5144 Southwest Loop 820, Fort Worth, TX 76140;
and Karta Technologies Inc., (Dr. Rao Govindaraju) San
Antonio, TX.
References
[1] Zhu YC, Ohtani S, Sato Y, Iwamoto N. Carbon
1999;37:1417–25.
[2] Zhu YC, Ohtani S, Sato Y, Iwamoto N. Carbon
1998;36:929–34.
[3] Westwood ME, Webster JD, Day RJ, Hayes FH, Taylor R. J
Mater Sci 1996;31(6):1389–97.
[4] Tsou HT, Kowbel W. J Adv Mater 1996;27:9–15.
[5] Fergus JW, Worrell WL. Carbon 1995;33:537–44.
[6] Buchanan FJ, Little JA. Carbon 1995;33:491–7.
[7] Kowbel W, Withers JC, Ransone PO. Carbon 1995;33:415–
20.
[8] Sheehan JE, Buesking KW, Sullivan BJ. Ann Rev Mater Sci
1994;24:19–29.
[9] Savage G. Carbon–carbon composites, London: Chapman &
Hall, 1993.
[10] McKee DW. Oxidation protection of carbon materials. In:
Thrower PA, editor, Chemistry and physics of carbon, vol.
23, New York: Dekker, 1991, pp. 173–222.
[11] Gupta A, Jagannathan R. Appl Phys Lett 1987;51:2254–7.
[12] Maruo H, Miyamoto I, Ooie T. Processing mechanism of
ceramics with high power density lasers. In: Matsunawa A,
Katayama S, editors, Proceedings of laser advanced materials
processing, High-Temperature Society of Japan, 1992, pp.
293–8.
[13] Fitzer E. The future of carbon–carbon composites. Third
distinguished lecture on materials technology. In: Genisio M,
editor, Southern Illinois University, Carbondale: Materials
Technology Center, 1986, p. 20.
[14] Criscione JM, Mercuri RA, Schram AW, Smith AW, Volk HF.
 L. Snell et al. / Carbon 39 (2001) 991 – 999 999

High temperature protective coatings for graphite. AFML-
TDR-64-173, Parts 1 to 4, 1967.
[15] Macklin BA, Owens FR, LaMar PA. Development of
improved methods of depositing iridium coatings on
graphite. AFML-TR-67-195, 1967.
[16] Wright TR, Weyand JD, Kizer DE, Simmons WC. Proceed-
ings of 4th Space Congress, Cocoa Beach, Florida, 1967,
15-45–15-57.
[17] Harding JT, Fry VR. In: Proceedings of 10th International
Precious Metals Conference, 1986, pp. 431–5.