Bias-enhanced nucleation of diamond on silicon dioxide

Article abstract of DOI:10.1063/1.119839

The results of surface characterization of bias-pretreated SiO₂ using x-ray photoemission spectroscopy and Raman spectroscopy are presented. These results suggest that the formation of SiC on the substrate surface during bias-enhanced nucleatio

Full text of DOI:10.1063/1.119839

Bias-enhanced nucleation of diamond on silicon dioxide
M. D. Irwin, C. G. Pantano, P. Gluche, and E. Kohn
Citation: Applied Physics Letters 71, 716 (1997); doi: 10.1063/1.119839
View online: http://dx.doi.org/10.1063/1.119839
View Table of Contents: http://scitation.aip.org/content/aip/journal/apl/71/5?ver=pdfcov
Published by the AIP Publishing
Articles you may be interested in
Microwave plasma enhanced chemical vapor deposition of nanocrystalline diamond films by bias-enhanced
nucleation and bias-enhanced growth
J. Appl. Phys. 115, 024308 (2014); 10.1063/1.4861417
Ultrananocrystalline diamond nano-pillars synthesized by microwave plasma bias-enhanced nucleation and bias-
enhanced growth in hydrogen-diluted methane
J. Appl. Phys. 112, 124307 (2012); 10.1063/1.4769861
Systematic studies of the nucleation and growth of ultrananocrystalline diamond films on silicon substrates
coated with a tungsten layer
J. Appl. Phys. 111, 124328 (2012); 10.1063/1.4729798
Growth, microstructure, and field-emission properties of synthesized diamond film on adamantane-coated silicon
substrate by microwave plasma chemical vapor deposition
J. Appl. Phys. 107, 103305 (2010); 10.1063/1.3427436
Role of cyclic process in the initial stage of diamond deposition during bias enhanced nucleation
J. Appl. Phys. 91, 9752 (2002); 10.1063/1.1480116
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: 130.236.82.7
On: Fri, 19 Dec 2014 11:52:56

Bias-enhanced nucleation of diamond on silicon dioxide
M. D. Irwina) and C. G. Pantano
Department of Materials Science and Engineering, The Pennsylvania State University, University Park,
Pennsylvania 16802
P. Gluche and E. Kohn
Department of Electron Devices and Circuits, University of Ulm, Ulm/Donau D-89069, Germany
͑
Received 6 March 1997; accepted for publication 20 May 1997͒
Characterization of amorphous SiO surfaces after biasing pretreatments, which induce nucleation
2
of diamond, has been carried out using x-ray photoelectron spectroscopy and Raman spectroscopy.
A mixture of silicon carbide, silicon oxycarbide, and diamond are formed upon exposure of biased
SiO surfaces to a CH ϩH plasma used for diamond deposition. It is concluded that nucleation of
2
4
2
diamond on amorphous SiO surfaces is promoted by formation of a SiC surface layer. Textured
2
diamond films have been fabricated on bulk SiO substrates using biasing pretreatments to induce
2
diamond nucleation. © 1997 American Institute of Physics. ͓S0003-6951͑97͒03829-1͔
In order to grow diamond thin films via microwave
plasma chemical vapor deposition, the substrate surface must
be pretreated to promote diamond nucleation. The nucleation
pretreatment is critical since high nucleation densities are
required in order to obtain thin, smooth, textured diamond
films. Bias-enhanced nucleation ͑BEN͒ generates nucleation
densities which are orders of magnitude higher than conven-
tional means such as diamond abrasion. Nucleation densities
treatment, the process parameters were adjusted during the
growth stage to obtain a textured diamond surface. Typical
7
system parameters used during the biasing pretreatments and
subsequent growth steps are summarized in Table I.
The biasing pretreatments produced nucleation densities
8
2
approaching 1ϫ10 nuclei/cm . A scanning electron micro-
scope ͑SEM͒ micrograph of crystallites produced during the
biasing pretreatments is shown in Fig. 2͑a͒. The ‘‘ball-like’’
morphology was typical for crystallites produced during bi-
1
0
2
1
as high as 10 /cm have been achieved on Si and SiC.
asing pretreatment of SiO . After the pretreatment, the pro-
Although BEN has been shown to be effective on substrate
materials consisting of carbide-forming metals or metal
2
cessing conditions were changed ͑see Table I͒ such that the
continuous film of the ball-like crystallites produced by bi-
asing was grown out and a textured diamond film ͓Fig. 2͑b͔͒
was formed. Raman spectra taken from a textured diamond
film and a continuous film of ball-like crystallites on a bias-
pretreated surface are shown in Fig. 3. The characteristic
1
–4
carbides such as Ti, W, Ta, Nb, and Hf as well as Si and
SiC, nucleation and growth of diamond on oxides such as
silicate glasses remains relatively unexplored. It was shown
5
in a recent study that biasing pretreatments promote nucle-
ation of diamond on SiO substrates. In this letter the results
2
of surface characterization of bias-pretreated SiO2 using
x-ray photoemission spectroscopy ͑XPS͒ and Raman spec-
troscopy are reported. It is suggested that formation of SiC
on the substrate surface during BEN facilitates diamond
nucleation. Also, it is confirmed that relatively high diamond
nucleation densities can be achieved on fused SiO by BEN.
2
Deposition experiments were performed in two different
ASTeX™ microwave plasma chemical vapor deposition sys-
tems ͑MPCVDs͒ both of which were modified in order to
enable in situ biasing of the substrates. The initial biasing
experiments were carried out in a quartz-tube chemical vapor
deposition ͑CVD͒ reactor as shown in Fig. 1. The experi-
ments were then reproduced in a larger MPCVD resonant
6
chamber system which has been described elsewhere. Sub-
strates of high purity fused SiO2 plates 1 mmϫ1.5 cm
ϫ1.5 cm were used. Each plate was cleaned prior to deposi-
tion by sonification in acetone, methanol, and 2-propanol for
5
min. intervals, respectively, and then rinsed in de-ionized
water. The deposition process consisted of an in situ biasing
pretreatment to promote diamond nucleation followed by a
growth step in which no bias was applied to the sample. The
nucleation and growth steps were carried out using a source
gas mixture of hydrogen and methane. After the biasing pre-
FIG. 1. Schematic of the quartz tube MPCVD reactor used for biasing
experiments.
a͒Electronic mail: mdi2@psu.edu
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: 130.236.82.7
16 Appl. Phys. Lett. 71 (5), 4 August 1997 0003-6951/97/71(5)/716/3/$10.00 © 1997 American Institute of Physics
On: Fri, 19 Dec 2014 11:52:56
7

TABLE I. Typical process parameters used for biasing pretreatments and
growth of diamond films on amorphous SiO substrates.
2
Parameter
Bias pretreatment
190–210 V
Growth
Bias voltage
Bias current
CH /H
none
none
1.5%
30 Torr
700 W
680 °C
variabled
a
8–10 mA
2.5%–5%
17 Torr
140 W, 700 W
4
2
Pressure
b
c
Microwave power
Substrate temperature
Time
ϳ900–1000 °C
30–120 min
a
Biasing current changes as a function of pretreatment time.
Quartz-tube MPCVD system.
Resonant-chamber MPCVD system.
The time required to grow out textured films depends on the deposition
rate.
b
c
d
peak at 1332 cmϪ1 indicates that the crystallites formed dur-
ing biasing pretreatments consisted of diamond.
A cross-sectional view of a textured diamond film grown
on SiO via BEN is shown in Fig. 2͑b͒. Pyramidal features
indicative of a ͑111͒ surface texture are observed. Ball-like
diamond crystallites are seen in cross-section at the diamond
2
FIG. 3. Raman spectra taken from ͑111͒ textured diamond film and a con-
tinuous film of ball-like diamond crystallites formed during a biasing pre-
treatment.
SiO interface as well as columnar, diamond grains that have
2
grown over this layer of crystallites. The interface between
the diamond film and the underlying SiO substrate is intact
2
indicating that the diamond adheres well to the surface; how-
ever, in some cases the SiO fractured below the interface
2
causing spallation of the film. This could be due to thermal
stresses which develop due to mismatch in the coefficients of
thermal expansion of diamond and SiO at high temperatures
2
͑surface temperatures of ϳ900–1000 °C during biasing pre-
treatments͒. The substrates were also relatively thick, 1 mm,
allowing for a significant thermal gradient to develop. It may
be possible to avoid spallation of the film by reducing pro-
cessing temperatures and using thinner substrates.
Analysis of bias-pretreated surfaces using XPS indicated
that a mixture of SiC, silicon oxycarbide, and a carbon phase
͑diamond and/or amorphous carbon͒ are formed during the
biasing pretreatment. XPS silicon 2p and carbon 1s spectra
taken from a sample which was subjected to a 30 min bias
pretreatment are shown in Fig. 4. The Si 2p and C 1s peaks
lying at 100.1 and 282.6 eV, respectively, in Figs. 4͑a͒ and
4
͑b͒ clearly indicate that SiC has been formed on the sub-
strate surface. The peak appearing at 103.7 eV in Fig. 4͑a͒ is
due to the Si 2p signal from the underlying SiO substrate.
2
The central peak seen in Fig. 4͑a͒ at 102 eV could be attrib-
uted to the existence of a solution of oxycarbide compounds
formed at the SiC/SiO interface, all with binding energies
2
lying between those of SiC and SiO ͑rather than to a single
2
compound with a discrete stoichiometry͒. The intermediate
peak observed in the C 1s spectrum at 283.8 eV is also
consistent with the presence of silicon oxycarbide com-
pounds. The C 1s peak at 284.6 eV is primarily attributed to
formation of diamond ͑as is confirmed by Raman, Fig. 3͒;
however, other carbonaceous species such as amorphous car-
bon and/or graphite may also contribute to this signal. The
data indicate that a silicon oxycarbide phase is formed upon
exposure of the biased SiO surface to the CH ϩH plasma
FIG. 2. ͑a͒ SEM micrograph showing ball-like diamond crystallites formed
during biasing in the resonant chamber MPCVD reactor. ͑b͒ SEM micro-
2
4
2
possibly via an O for C exchange reaction. As this reaction
progresses a SiC layer is formed leaving an oxycarbide phase
graph showing a cross-sectional view of a diamond film grown on SiO . The
2
diamond/SiO interface as well as a ͑111͒ textured diamond surface can be
2
seen.
at the interface between the SiC layer and the SiO substrate.
2
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: 130.236.82.7
Appl. Phys. Lett., Vol. 71, No. 5, 4 August 1997 Irwin et al. 717
On: Fri, 19 Dec 2014 11:52:56

energetically favorable sites and form diamond nuclei. Al-
though heterogeneous nucleation of diamond on a solid sur-
face from the gas phase is not well understood it has been
demonstrated that diamond readily nucleates on biased
-SiC.1
1,12
A surface layer which consists of a mixture of
␤
amorphous SiC and microcrystalline SiC is seen to form on
Si during biasing.1 These SiC microcrystallites could serve
as nucleation sites for diamond. Similarly, crystallites in the
,13
SiC surface layer formed on SiO surfaces during biasing
2
might act as diamond nucleation sites. A difference in crys-
tallinity or morphology of the SiC overlayers could explain
the lower diamond nucleation densities induced by BEN ob-
8
2
served in this study on SiO ͑ϳ10 /cm ͒ compared to those
2
11
2
1
commonly reported for nucleation on Si (10 /cm ). The
crystallinity of the SiC surface layer formed on an amor-
phous SiO surface would be expected to be lower than that
2
of the SiC layer formed on single-crystal Si surfaces. Thus it
may be possible to increase the nucleation density of dia-
mond produced on SiO via BEN by optimizing process con-
2
ditions to increase the crystallinity of the carbide surface
layer formed.
Bias-enhanced nucleation has been demonstrated to be a
viable in situ process for deposition of diamond on SiO2;
however, more work is required to develop this technique for
production of thinner, smoother diamond films that could be
of use in fabrication of microelectronic devices or optical
coatings. An investigation is currently underway in which
bias-pretreated SiO surfaces will be more thoroughly char-
2
acterized and process conditions used during the pretreat-
ment will be optimized in order to increase the density of
nuclei produced by biasing.
1
B. R. Stoner, G.-H. M. Ma, S. D. Wolter, and J. T. Glass, Phys. Rev. B 45,
1
1067 ͑1992͒.
2
3
P. Reinke, P. Kania, and P. Oelhafen, Thin Solid Films 270, 124 ͑1995͒.
S. D. Wolter, M. T. McClure, J. T. Glass, and B. R. Stoner, Appl. Phys.
Lett. 66, 2810 ͑1995͒.
S. D. Wolter and J. T. Glass, J. Appl. Phys. 77, 5199 ͑1995͒.
J.-S. Lee, K.-S. Liu, and I.-N. Lin, J. Appl. Phys. 81, 486 ͑1997͒.
S. D. Wolter, T. H. Borst, A. Vescan, and E. Kohn, Appl. Phys. Lett. 68,
4
5
6
3
558 ͑1996͒.
7
C. Wild, P. Koidl, W. M u¨ ller-Sebert, H. Walcher, R. Kohl, N. Herres, R.
Locher, R. Samlenski, and R. Brenn, Diam. Relat. Mater. 2, 158 ͑1993͒.
S. Yugo, T. Kimura, and T. Kanai, Diam. Relat. Mater. 2, 328 ͑1992͒.
S. P. McGinnis, M. A. Kelly, and S. B. Hagstr o¨ m, Appl. Phys. Lett. 66,
FIG. 4. ͑a͒ Silicon 2p and ͑b͒ carbon 1s x-ray photoelectron spectra taken
after a 30 min bias pretreatment.
8
9
3
117 ͑1995͒.
1
1
1
0
1
2
J. Gerber, S. Sattel, K. Jung, H. Ehrhardt, and J. Robertson, Diam. Relat.
Mater. 4, 559 ͑1995͒.
B. R. Stoner, G. H. Ma, S. D. Wolter, W. Zhu, Y.-C. Wang, R. F. Davis,
and J. T. Glass, Diam. Relat. Mater. 2, 142 ͑1993͒.
R. Kohl, C. Wild, N. Herres, P. Koidl, B. R. Stoner, and J. T. Glass, Appl.
Phys. Lett. 63, 1792 ͑1993͒.
Diamond may subsequently nucleate on the silicon carbide
or other carbonaceous species present on this surface layer.
It is generally accepted that one primary effect of the
applied bias is to increase the flux of positively charged radi-
ϩ
ϩ
8–10
cals such as C and CH to the substrate surface.
These
13
J. Gerber, S. Sattel, H. Ehrhardt, J. Robertson, P. Wurzinger, and P. Pon-
gratz, J. Appl. Phys. 79, 4388 ͑1996͒.
radicals may react with, or attach to, the substrate surface at
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: 130.236.82.7
18 Appl. Phys. Lett., Vol. 71, No. 5, 4 August 1997 Irwin et al.
On: Fri, 19 Dec 2014 11:52:56
7