Structure and optical properties of amorphous diamond films prepared by ArF laser ablation as a function of carbon ion kinetic energy

Article abstract of DOI:10.1063/1.122515

Amorphous carbon films with variable sp³ content were produced by ArF (193 nm) pulsed laser deposition. Electron energy loss spectroscopy and spectroscopic ellipsometry were employed to systematically study changes in the bonding and optical properties of the carbon films as a function of the kinetic energy of ablated C ions, which was measured using an ion probe. The measurements reveal that the films with the most diamond-like properties are obtained at the C ion kinetic energy of ～90 eV. In contrast to measurements made as a function of laser fluence, ion probe measurements of kinetic energy are a convenient as well as more accurate and fundamental method for monitoring deposition conditions, with the advantage of being readily transferable for interlaboratory comparisons.

Full text of DOI:10.1063/1.122515

Structure and optical properties of amorphous diamond films prepared by ArF laser
ablation as a function of carbon ion kinetic energy
Vladimir I. Merkulov, Douglas H. Lowndes, G. E. Jellison Jr., A. A. Puretzky, and D. B. Geohegan
Citation: Applied Physics Letters 73, 2591 (1998); doi: 10.1063/1.122515
View online: http://dx.doi.org/10.1063/1.122515
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/73/18?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Bonding configurations and optical band gap for nitrogenated amorphous silicon carbide films prepared by
pulsed laser ablation
J. Appl. Phys. 92, 2485 (2002); 10.1063/1.1498885
Influence of nitrogen and temperature on the deposition of tetrahedrally bonded amorphous carbon
J. Appl. Phys. 88, 1149 (2000); 10.1063/1.373790
Optical and electrical properties of nitrogen incorporated amorphous carbon films
J. Appl. Phys. 87, 2874 (2000); 10.1063/1.372271
Optical and structural properties of laser annealed Er-doped amorphous silicon thin films
J. Appl. Phys. 86, 701 (1999); 10.1063/1.370788
Pulsed-laser deposition of diamond-like carbon: Relations between laser fluence, velocity of carbon ions, and
bonding in the films
J. Appl. Phys. 85, 2915 (1999); 10.1063/1.369057
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
128.235.251.160 On: Fri, 19 Dec 2014 16:02:49

APPLIED PHYSICS LETTERS
VOLUME 73, NUMBER 18
2 NOVEMBER 1998
Structure and optical properties of amorphous diamond films prepared
by ArF laser ablation as a function of carbon ion kinetic energy
Vladimir I. Merkulov,a) Douglas H. Lowndes, G. E. Jellison, Jr., A. A. Puretzky,
and D. B. Geohegan
Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831-6056
͑
Received 6 July 1998; accepted for publication 1 September 1998͒
3
Amorphous carbon films with variable sp content were produced by ArF ͑193 nm͒ pulsed laser
deposition. Electron energy loss spectroscopy and spectroscopic ellipsometry were employed to
systematically study changes in the bonding and optical properties of the carbon films as a function
of the kinetic energy of ablated C ions, which was measured using an ion probe. The measurements
reveal that the films with the most diamond-like properties are obtained at the C ion kinetic energy
of ϳ90 eV. In contrast to measurements made as a function of laser fluence, ion probe
measurements of kinetic energy are a convenient as well as more accurate and fundamental method
for monitoring deposition conditions, with the advantage of being readily transferable for
interlaboratory comparisons. © 1998 American Institute of Physics. ͓S0003-6951͑98͒03044-7͔
The useful properties1 of hydrogen-free tetrahedral
amorphous carbon (ta-C) films have triggered great interest.
The properties include high hardness, chemical inertness, a
low coefficient of friction, moderately large optical gap of
ϳ2 eV, and perhaps low electron affinity. Consequently,
ta-C films may have various applications in mechanical and
optical coatings and electronic devices.
precise interlaboratory comparisons rather difficult. In con-
trast, in situ ion probe measurements provide accurate infor-
mation about the time of flight ͑TOF͒ of ions which allows
for reliable comparison of results obtained by different re-
search groups. An important simplification is obtained by
using the ArF ͑193 nm͒ laser for fluences of ϳ2 up to ϳ8
2
J/cm : the ablated flux consists mostly of monatomic neutral
ta-C films have been prepared by a variety of energetic-
8,9
and ionized C species. As a result, the dominant mass of
2
beam methods including filtered cathodic vacuum arc
ablated species ͑12 amu͒ is known and the KE of C ions can
be determined easily from the ion probe’s TOF measure-
ments. In addition, the KE of C species is the more funda-
mental factor controlling the properties of ta-C.
͑FCVA͒, mass-separated ion beam deposition ͑MSIBD͒, and
pulsed laser deposition ͑PLD͒. Their diamond-like properties
are associated with the presence of a substantial fraction
3
͑
ϳ80%͒ of sp bonded C atoms in the amorphous network.
ta-C films were prepared in a high vacuum chamber
This is in contrast to conventional amorphous carbon (a-C)
Ϫ8
with a base pressure of 3ϫ10 Torr using a Lambda
prepared by evaporation or sputtering that consists mostly of
2
3
3
Physik Compex 301i pulsed excimer laser operated with ArF
sp bonded C atoms. The amount of sp bonding in ta-C
depends upon the kinetic energy ͑KE͒ of the depositing car-
bon species. For FCVA, for example, it recently was found
͑
193 nm͒. Maximum laser fluences at the target were ϳ8
2
ϩ
J/cm and produced C KEϳ100 eV. Higher energy carbon
ions were produced using a Questek 2960 excimer laser. The
Questek has shorter pulse duration, higher pulse energy, and
3
that the highest sp fraction is obtained for C ions with KE
4
–6
of ϳ80–100 eV,
although earlier experiments using
ϩ
a sharply peaked beam profile which allowed C ions to be
FCVA and other methods have found nearly optimal
diamond-like properties for KE values of 20 eV to several
produced with kinetic energies up to 225 eV. The substrates
were kept at room temperature and placed at variable dis-
tances ͑4.8–7.3 cm͒ from the target. The ion probe consisted
of a bare coaxial-cable tip biased at Ϫ100 V and mounted on
an arm that rotated it in an arc passing through the center of
the ablation plume. More details on the experimental setup
can be found elsewhere.10
2
7
hundred eV. For PLD, the early work by Pappas et al. es-
tablished a correlation between the diamond-like properties
of ta-C and the KE of C ions, but to date the optimum KE of
3
C ions required to produce the highest sp content for laser
ablated films has not been determined. Consequently, the
interesting question of whether the PLD and FCVA methods
produce ta-C films with similar diamond-like characteristics
for a given C ion KE has remained unanswered.
In the present work we address all these issues. In the
most previous PLD experiments, the properties of ta-C films
were correlated with the laser wavelength and energy fluence
For spectroscopic ellipsometry ͑SE͒ measurements,
ϳ70–200 nm thick ta-C films were deposited on p- and
n-type ͑001͒-oriented Si wafers. The SE measurements were
made using the two-modulator generalized ellipsometer
͑
2-MGE͒11 operating from 250 to 850 nm. The SE results
͑or power density͒. However, laser fluence is a parameter
were fit using a five-medium model consisting of air/surface
roughness/ta-C film/interface/Si, where the optical proper-
ties of ta-C films were modeled using the five-parameter
Tauc–Lorentz ͑TL͒ formulation for the optical functions of
amorphous materials.12 Previous studies have shown that
this model fits SE data taken on a wide range of materials,
which is rather difficult to measure accurately, primarily due
to the large uncertainty in the laser beam spot size at the
target and a non uniform laser beam profile. This makes
13
a͒Electronic mail: merkulov@solid.ssd.ornl.gov
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
003-6951/98/73(18)/2591/3/$15.00 2591 © 1998 American Institute of Physics
28.235.251.160 On: Fri, 19 Dec 2014 16:02:49
0
1

2592
Appl. Phys. Lett., Vol. 73, No. 18, 2 November 1998
Merkulov et al.
this plasmon is proportional to the density of ta-C films:14
the higher the plasmon energy, the higher the film density.
As two reference points, one can use the plasmon peak po-
sition of graphite ͑ϳ26.5 eV͒ and of diamond ͑33.6 eV͒.
3
The analysis of the sp fractions in the ta-C films was
performed in a way similar to that of Berger et al.15 The sp2
fraction, fsp
2
, is calculated by taking the ratio of integrated
intensities of the 1s-␲* and 1s-␴* peaks, I and I , and by
␲
␴
comparing it to that of arc-evaporated a-C, which is as-
2
3
sumed to be entirely sp bonded. The sp fraction is then
simply 1-fsp . A few simplifications are assumed using this
2
method. First, the matrix elements for the electronic transi-
tions occurring in C atoms with different bonding are not
known. Therefore, to employ Berger’s method, one is forced
to set the matrix elements to be the same for all C atoms.
Second, the disorder effects, such as bond rotation and varia-
tions in the spatial distribution of C atoms, are not taken into
consideration. In an amorphous C network, p-electron orbit-
als are not parallel to each other as in graphite but are rotated
at various angles. Consequently, the total overlap between
the p orbitals will be less and the number of electronic tran-
sitions to the ␲* states will decrease, which will result in
reduced intensity of the peak at 285 eV compared to that of
graphite. This problem can be solved by using an amor-
FIG. 1. EELS ͑a͒ K-edge and ͑b͒ plasmon spectra of amorphous C films
prepared by ArF laser deposition at various C ion kinetic energies (KEmc
corresponding to the maximum current of the ion probe signal.
)
2
phous, not graphitic, 100% sp calibration sample where the
amount of disorder is statistically similar to that of ta-C.
The spatial distribution of sp2 C atoms may also have a
significant effect upon the EELS results. As a substantial
yielding the thickness of each layer, the Tauc ͑optical͒ band
gap, and the optical functions of the film.
For electron energy loss spectroscopy ͑EELS͒ measure-
ments, ta-C was deposited on freshly cleaved NaCl sub-
strates which were subsequently dissolved in DI water and
the resultant free-standing ta-C films were placed on folding
Ni transmission electron microscopy ͑TEM͒ grids. The
thickness of the films was ϳ25–30 nm which is substantially
lower than the mean free path for 100 keV electrons ͑Ͼ50
nm͒. In this case the effect of multiple scattering can be
2
number of sp C atoms are not bonded to each other but are
3
surrounded by sp C atoms, which may be the case for ta-C
3
with a high sp fraction, the p orbitals do not overlap so that
the intensity of the 1s-␲* peak will be reduced and the
3
extracted sp fraction will be overestimated. Other problems
include substantial multiple electron scattering for thick
specimens and the difficulty of sample preparation. In spite
of these problems, EELS is currently the most-used tool for
1
4
neglected. The EELS measurements were carried out using
a VG HB501 scanning tunneling electron microscope
3
analyzing the sp fraction in ta-C. Partly this is due to the
absence of any more suitable alternative method, although
͑STEM͒ fitted with a high sensitivity parallel EELS and op-
UV Raman scattering1
6,17
may be a good candidate. We be-
erated at 100 keV. The energy resolution was 1.1 eV during
the measurements, as determined by the full width at half
maximum ͑FWHM͒ of the zero-loss peak. We did not ob-
serve any notable degradation or contamination of the speci-
mens throughout the experiments.
lieve that a combination of techniques, such as EELS, ellip-
sometry, Raman scattering, and possibly others, is required
to fully characterize such a complicated material as ta-C.
3
Nevertheless, EELS provides a reasonable sp estimate suit-
able for most practical applications.
The K-edge spectra of ta-C films deposited by ArF PLD
ϩ
The details of the EELS data analysis can be found
with various C ion kinetic energies (KE ) corresponding
mc
elsewhere.10 Briefly, it includes decomposition of the 1s-␲
*
to the maximum current of the ion probe signal are shown in
Fig. 1͑a͒. The spectra exhibit two main features: a peak at
ϳ285 eV and a broadband with a maximum at ϳ293 eV.
These features are due to 1s-␲* and 1s-␴* electronic tran-
and 1s-␴
*
peaks via a best fit of a Gaussian into the lower
energy edge of the latter. The value of I␴ was taken as an
integral over the energy range of 290–310 eV which is large
enough to avoid any significant systematic error. The ex-
sitions, respectively.15 The former is associated with sp
2
bonded C atoms and the latter with both sp and sp3
2
C
3
tracted sp fractions as a function of C ion KEmc are shown
3
atoms. One can see that the relative intensity of the 1s-␲*
peak changes with the C ion KEmc and becomes the lowest
for the films prepared with KEmc in the range of 70–120 eV.
in Fig. 2͑a͒. The sp fraction clearly exhibits a maximum of
ϳ73% at KEmcϳ90 eV. This is consistent with the optimum
KE obtained for FCVA deposition of ta-C.
Figure 2͑b͒ shows that the position of the C plasmon
peak also reaches its highest value of ϳ30.9 eV at
2
This corresponds to the materials with the lowest sp content
and hence more diamond-like properties. EELS spectra also
were measured in the low-energy range and are shown in
Fig. 2͑b͒. The broad peak around 30 eV is associated with
KE ϳ90 eV. This behavior is expected since higher plas-
mc
mon peak energy corresponds to higher film density, which
3
collective excitation, or
electrons. Using a simple Drude-like model, the energy of
a
plasmon, of the valence
is characteristic of more diamond-like films with higher sp
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
1
5
content. It should be noted that the plasmon peak position
1
28.235.251.160 On: Fri, 19 Dec 2014 16:02:49

Appl. Phys. Lett., Vol. 73, No. 18, 2 November 1998
Merkulov et al.
2593
In conclusion, the combination of ArF pulsed laser depo-
sition and in situ ion probe measurements allows one to de-
posit ta-C films with well-controlled diamond-like proper-
ties. EELS measurements show that both the sp3 fraction
and the plasmon peak energy ͑ϳ film density͒ reach their
maximum values of ϳ73% and 30.9 eV in films deposited at
KE ϳ90 eV. Scanning ellipsometry measurements reveal
mc
that the optical ͑Tauc͒ energy gap is also maximized ͑ϳ2.0
eV͒ at KE ϳ90 eV, consistent with the presence of the
mc
3
highest sp fraction. This combination of independent opti-
cal and electronic measurements provides strong evidence of
an optimum kinetic energy KE ϳ90 eV for ArF PLD of
mc
ta-C.
The authors are especially grateful to S. J. Pennycook
for providing the STEM for the EELS measurements. They
also thank P. H. Fleming for assistance with sample prepa-
ration. This research was partially sponsored by the Defense
Advanced Research Projects Agency under Contract No.
DARPA-MIPR-97-1357 with Oak Ridge National Labora-
tory ͑ORNL͒, and by the Office of Basic Energy Sciences,
Division of Materials Sciences, U.S. Department of Energy.
The research was carried out at ORNL, managed by Lock-
heed Martin Energy Research Corp. for the U.S. Department
of Energy, under Contract No. DE-AC05-96OR22464.
FIG. 2. ͑a͒ sp3 content and ͑b͒ plasmon peak position extracted from the
EELS measurements ͑Fig. 1͒ and ͑c͒ optical ͑Tauc͒ energy gap obtained by
scanning ellipsometry, as a function of C ion kinetic energy, KEmc
.
3
can serve as a useful tool for correlating the sp fractions
obtained by different research groups. Interlaboratory dis-
1
crepancies in the estimate of sp3 fractions can arise from
J. Robertson, Prog. Solid State Chem. 21, 199 ͑1991͒.
Y. Lifshitz, Diamond Relat. Mater. 5, 388 ͑1996͒.
H. Pan, M. Pruski, B. C. Gerstain, F. Li, and J. S. Lannin, Phys. Rev. B
2
2
various factors, including structural differences in sp cali-
3
bration samples, variations in the energy ranges and fitting
44, 6741 ͑1991͒.
4
5
6
7
8
procedures used to define I␲ and I , and differences in
S. Ravi, P. Silva, S. Xu, B. X. Tay, H. S. Tan, and W. I. Milne, Appl.
Phys. Lett. 69, 491 ͑1996͒.
␴
EELS resolution. The position of the plasmon peak is a more
settled parameter and therefore can be used as a reference
S. R. P. Silva, S. Xu, B. K. Tay, H. S. Tan, H.-J. Scheibe, M. Chhowalla,
and W. I. Milne, Thin Solid Films 290-291, 317 ͑1996͒.
M. Chhowalla, J. Robertson, C. W. Chen, S. R. P. Silva, C. A. Davis, G.
A. J. Amaratunga, and W. I. Milne, J. Appl. Phys. 81, 139 ͑1997͒.
D. L. Pappas, K. L. Saenger, J. J. Cuomo, and R. W. Dreyfus, J. Appl.
Phys. 72, 3966 ͑1992͒.
A. A. Puretzky, D. B. Geohegan, G. E. Jellison, Jr., and M. M. McGibbon,
in Film Synthesis and Growth Using Energetic Beams, edited by H. At-
water, J. T. Dickinson, D. H. Lowndes, and A. Polman ͓Mater. Res. Soc.
Symp. Proc. 388, 145 ͑1995͔͒.
A. A. Puretzky, D. B. Geohegan, G. E. Jellison, Jr., and M. M. McGibbon,
Appl. Surf. Sci. 96-98, 859 ͑1996͒.
D. H. Lowndes, V. I. Merkulov, A. A. Puretzky, D. B. Geohegan, G. E.
Jellison, Jr., C. M. Rouleau, and T. Thundat, in Advances in Laser Abla-
tion of Materials, Mater. Res. Soc. Symp. Proc. ͑MRS, Pittsburgh, 1998͒
͑in press͒.
3
point for extracting the sp fraction.
The EELS results were correlated with scanning ellip-
sometry ͑SE͒ measurements of the optical properties of ta-C
films. The optical energy gaps extracted from the SE mea-
surements are shown in Fig. 2͑c͒ as a function of the C ion
3
KE . The trend is very similar to that observed for the sp
mc
fraction and for the position of the plasmon peak. The energy
9
gap is lower for the films prepared at low KE , reaches its
mc
maximum at KE ϳ90 eV, and decreases for higher values
mc
10
3
of KE . This is consistent with the variation of the sp
fraction in these materials: sp rich films are expected to
mc
3
have a larger optical gap ͑5.5 eV for pure crystalline dia-
11
2
G. E. Jellison, Jr. and F. A. Modine, Appl. Opt. 36, 8184 ͑1997͒; 36, 8190
mond͒ and the gap disappears for predominantly sp bonded
͑
1997͒.
carbon.
12
G. E. Jellison, Jr. and F. A. Modine, Appl. Phys. Lett. 69, 371 ͑1996͒; 69,
137 ͑1996͒.
It also has to be mentioned that some additional mea-
surements show that for C ion energies well above ϳ200 eV,
the diamond-like properties of C films increase again. This
result implies that the presence of neutrals and nonmona-
tomic C species, as well as the width of C ion energy distri-
bution, cannot be neglected completely. More detailed TOF
mass spectroscopic studies are required to confirm and ex-
plain their effects. Nevertheless, ion probe measurements
provide direct correlation between the C ion energy and the
properties of ta-C films and therefore remain the most con-
2
13
G. E. Jellison, Jr., D. B. Geohegan, D. H. Lowndes, A. A. Puretzky, and
V. I. Merkulov, Advances in Laser Ablation of Materials, Mater. Res. Soc.
Symp. Proc. ͑MRS, Pittsburgh, 1998͒ ͑in press͒; G. E. Jellison, Jr., F. A.
Modine, P. Doshi, and A. Rohatgi, Thin Solid Films 313-314, 193 ͑1998͒.
R. F. Egerton, Electron Energy Loss Spectroscopy in the Electron Micro-
scope ͑Plenum, New York, 1986͒.
14
1
1
5
6
S. D. Berger and D. R. McKenzie, and P. J. Martin, Philos. Mag. Lett. 57,
285 ͑1988͒.
V. I. Merkulov, J. S. Lannin, C. H. Munro, S. A. Asher, V. S. Veerasamy,
and W. I. Milne, Phys. Rev. Lett. 78, 4869 ͑1997͒.
K. W. R. Gilkes, H. S. Sands, D. N. Batchelder, J. Robertson, and W. I.
17
This article is copyrighted as indicated in the article. Reuse of AIP content is subject to the terms at: http://scitation.aip.org/termsconditions. Downloaded to IP:
venient technique for in situ control of the latter.
Milne, Appl. Phys. Lett. 70, 1980 ͑1997͒.
1
28.235.251.160 On: Fri, 19 Dec 2014 16:02:49