Diamond microcrystallites formation through the phase transition graphite→liquid→diamond
The paper presents the results of synthesizing the diamond microparticles (3 to 5 µm) in a spark discharge in hydrogen at the low pressure (100 Torr). The obtained growth rate ~5 µm/s is uniquely high. Our analysis of the nature of particles by using SEM and Raman spectroscopy demonstrates that thes...
Gespeichert in:
| Datum: | 2006 |
|---|---|
| 1. Verfasser: | |
| Format: | Artikel |
| Sprache: | English |
| Veröffentlicht: |
Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України
2006
|
| Schriftenreihe: | Semiconductor Physics Quantum Electronics & Optoelectronics |
| Online Zugang: | http://dspace.nbuv.gov.ua/handle/123456789/121583 |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| Назва журналу: | Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| Zitieren: | Diamond microcrystallites formation through the phase transition graphite→liquid→diamond / T.V. Semikina // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2006. — Т. 8, № 1. — С. 22-28. — Бібліогр.: 19 назв. — англ. |
Institution
Digital Library of Periodicals of National Academy of Sciences of Ukraine| id |
irk-123456789-121583 |
|---|---|
| record_format |
dspace |
| spelling |
irk-123456789-1215832017-06-15T03:04:39Z Diamond microcrystallites formation through the phase transition graphite→liquid→diamond Semikina, T.V. The paper presents the results of synthesizing the diamond microparticles (3 to 5 µm) in a spark discharge in hydrogen at the low pressure (100 Torr). The obtained growth rate ~5 µm/s is uniquely high. Our analysis of the nature of particles by using SEM and Raman spectroscopy demonstrates that these particles are cubic high quality diamond. Using the result of SIM images after cross-cutting of the sample by FIB, it is concluded that diamond does not grow on the substrate and running process is not CVD. Discussing the theory of the spark discharge, it is suggested that the process occurs at high pressures and temperatures. A hypothesis of diamond formation through a diffusion mechanism of the direct phase transition is presented. 2006 Article Diamond microcrystallites formation through the phase transition graphite→liquid→diamond / T.V. Semikina // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2006. — Т. 8, № 1. — С. 22-28. — Бібліогр.: 19 назв. — англ. 1560-8034 PACS 81.05.Uw http://dspace.nbuv.gov.ua/handle/123456789/121583 en Semiconductor Physics Quantum Electronics & Optoelectronics Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України |
| institution |
Digital Library of Periodicals of National Academy of Sciences of Ukraine |
| collection |
DSpace DC |
| language |
English |
| description |
The paper presents the results of synthesizing the diamond microparticles (3 to 5 µm) in a spark discharge in hydrogen at the low pressure (100 Torr). The obtained growth rate ~5 µm/s is uniquely high. Our analysis of the nature of particles by using SEM and Raman spectroscopy demonstrates that these particles are cubic high quality diamond. Using the result of SIM images after cross-cutting of the sample by FIB, it is concluded that diamond does not grow on the substrate and running process is not CVD. Discussing the theory of the spark discharge, it is suggested that the process occurs at high pressures and temperatures. A hypothesis of diamond formation through a diffusion mechanism of the direct phase transition is presented. |
| format |
Article |
| author |
Semikina, T.V. |
| spellingShingle |
Semikina, T.V. Diamond microcrystallites formation through the phase transition graphite→liquid→diamond Semiconductor Physics Quantum Electronics & Optoelectronics |
| author_facet |
Semikina, T.V. |
| author_sort |
Semikina, T.V. |
| title |
Diamond microcrystallites formation through the phase transition graphite→liquid→diamond |
| title_short |
Diamond microcrystallites formation through the phase transition graphite→liquid→diamond |
| title_full |
Diamond microcrystallites formation through the phase transition graphite→liquid→diamond |
| title_fullStr |
Diamond microcrystallites formation through the phase transition graphite→liquid→diamond |
| title_full_unstemmed |
Diamond microcrystallites formation through the phase transition graphite→liquid→diamond |
| title_sort |
diamond microcrystallites formation through the phase transition graphite→liquid→diamond |
| publisher |
Інститут фізики напівпровідників імені В.Є. Лашкарьова НАН України |
| publishDate |
2006 |
| url |
http://dspace.nbuv.gov.ua/handle/123456789/121583 |
| citation_txt |
Diamond microcrystallites formation through the phase transition graphite→liquid→diamond / T.V. Semikina // Semiconductor Physics Quantum Electronics & Optoelectronics. — 2006. — Т. 8, № 1. — С. 22-28. — Бібліогр.: 19 назв. — англ. |
| series |
Semiconductor Physics Quantum Electronics & Optoelectronics |
| work_keys_str_mv |
AT semikinatv diamondmicrocrystallitesformationthroughthephasetransitiongraphiteliquiddiamond |
| first_indexed |
2025-07-08T20:09:53Z |
| last_indexed |
2025-07-08T20:09:53Z |
| _version_ |
1837110813968564224 |
| fulltext |
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2006. V. 9, N 1. P. 22-28.
© 2006, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
22
PACS 81.05.Uw
Diamond microcrystallites formation through
the phase transition graphite→liquid→diamond
T.V. Semikina
Department of Environmental and Materials Science, Teikyo University of Science & Technology,
2525 Yatsusawa, Uenohara-machi, Kitatsuru-gun, Yamanashi-pref., 409-0193, Japan
E-mail: semikina@edd.ntu-kpi.kiev.ua
Abstract. The paper presents the results of synthesizing the diamond microparticles (3 to
5 µm) in a spark discharge in hydrogen at the low pressure (100 Torr). The obtained
growth rate ~5 µm/s is uniquely high. Our analysis of the nature of particles by using
SEM and Raman spectroscopy demonstrates that these particles are cubic high quality
diamond. Using the result of SIM images after cross-cutting of the sample by FIB, it is
concluded that diamond does not grow on the substrate and running process is not CVD.
Discussing the theory of the spark discharge, it is suggested that the process occurs at
high pressures and temperatures. A hypothesis of diamond formation through a diffusion
mechanism of the direct phase transition is presented.
Keywords: diamond microparticle, spark discharge, direct phase transition, threefold
point of the phase diagram.
Manuscript received 24.10.05; accepted for publication 15.12.05.
1. Introduction
Synthesis of diamond has attracted attention ever since it
was established in 1797 that diamond is a crystalline
form of carbon. From the time of announcement in 1955
of a process for artificial diamond synthesis with a
molten transition metal solvent-catalyst at pressure
where diamond is the thermodynamically stable phase,
the topic of diamond and diamond films (DF) growth
and application has been on the top of research activity
in the whole world still now. In spite of the fact that
many technologies of diamond and DF deposition are
developed, structural and physical properties of
synthesis products are well studied, the unexpected
preparation of fullerenes in 1991 year and then
nanotubes, success in deposition of nanodiamonds [1]
show that the issues in the field of diamond and diamond
like materials have a lot of unknown up to date.
As known, the methods of diamond synthesis at low
pressures, where it is metastable, are divided into
physical and chemical vapor deposition (CVD) [2]. The
physical methods are based on the diamond growth from
supersaturating carbon vapor. The CVD methods are
characterized by the chain of vapor transport reactions in
hydrocarbon or hydrocarbon-hydrogen mixtures. The
gas activation is provided by thermal, electrical,
chemical, photochemical, or combined methods [3].
Diamond growth takes place through formation of
chemical bonds between carbon atoms.
Because in our experiment the diamond synthesis
runs by the way characterized as physical (but new
physical), it is useful to remind some aspects of the
known physical methods. In one of his numerous
reviews [4], B.V. Spitsyn mentioned that Brinkman et al.
was among the first to suggest a diamond growth from
carbon vapor obtained by the high temperature graphite
sublimation. Carbon transferring from a graphite source
occurs in vacuum or gas. Diamond can be grown on a
silicon substrate, quartz glass, stainless steel and other
materials. However, the best results are obtained when
diamond grows on the facet of natural diamond. The
growth rate depends on the technology conditions. The
growth rate increase is the most actual problem up to
now. In Table, there are some examples of diamond film
physical deposition from carbon ions of medium energy,
which was characterized by the high growth rate and big
coated area [4].
In this paper, the diamond microparticles are
obtained as a result of d.c. discharge between two
graphite rods at low pressure 100 Torr in a chamber. The
issues concerning with determination of the growth
mechanism type, elucidation of the discharge type,
discharge mechanism and characteristics of the pressure
and temperature in the area of discharge are discussed.
The diamond particles were characterized using
micro-Raman spectroscopy fulfilled on the JASCO
spectrometer of NPS-1000 type. In the course of measu-
rements, a laser beam with the wavelength 532.07 nm
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2006. V. 9, N 1. P. 22-28.
© 2006, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
23
Fig. 1. SEM images of two different diamond particles.
Table. Some deposition parameters and basic properties of DF prepared from carbon atoms and ions.
N Deposition
method
Starting substances
(sputtering or plasma-
forming gas)
Temperature,
K
Substrate
material
Linear rate,
nm/s
Maximum
thickness of
film, µm
Diameter of
substrate,
mm
1 Arc in inert gas Graphite (Ar) 300 Si, stainless
steel, glass
5 10 –
2 Arc in inert gas Graphite 300 Si, Ni, NaCl,
KCl
– 0.5 –
3 Double beam Graphite (Ar) 300 NaCl, quartz
glass, hard alloy
– – –
4 Double beam Graphite, diamond
(Ar)
453 NaCl, KCl 2×10-3 3×10-2 10
5 Duoplasmo-
tron
Graphite 298-323 Quartz glass, Si,
Ni, KCl
0.5 0.1 50
and 1 μm spot size was focused onto the crystallites. The
images of diamond particles were obtained using JEOL
field-emission scanning electron microscope (SEM).
Scanning ion microscopy (SIM) was fulfilled after cross
cutting the sample by a focused ion beam (FIB) of
gallium. Experimental results were obtained in
Department of Environmental and Materials Science,
Teikyo University of Science & Technology, Japan.
2. Technology of preparation
Synthesis of diamond particles were performed using the
equipment for carbon deposition JEOL, JEE-5B
modified with additional pipes to inlet gases to reaction
chamber. The experiments were carried out without
hydrogen (in vacuum), with hydrogen (100 Torr) and in
the mixture Ar/H2 (100 Torr). The graphite rods were
fixed in position end-to-end in special holders. The
graphite rods were given by Mitsubishi Pensil CO.,
LTD. The shape of graphite rods were cylindrical,
diameter 3 mm with sharp tip (diameter 1–2 mm). Four
shapes of sharp tips were used in all the set of
experiments. We used silicon, Al2O3 ceramics and non-
alkaline glass substrates. Distance between graphite rods
and substrate was 5 and 10 mm. The position of
substrates was also changed (up-and-down) relatively to
graphite rods in various experiments. Before the process,
the substrates and graphite rods were ultrasonically
treated in liquid acetone.
Approaching d.c. discharge the applied voltage has
been increased to 7–8 V for 30 s. Within this time
interval, the graphite rods were heated and began to
emit. After the moment of discharge appearance, the
applied voltage was maintained for 10 s in the first group
of experiments and 1 s in the second one. The current
reached 25 A. The discharge was accompanied with
bright emission and crackling sound. Then the voltage
was removed and the process stopped. The temperature
in discharge area was observed with two-colored
pyrometer (CHINO, IR-AQ and IR-GAG). In the
illumination area, the temperature reached 2400 °C. The
substrate was heated due to illumination of discharge
and rods. The substrate temperature was controlled by a
K-type thermo-couple attached to the back of the
substrate and reached 300 °C in its maximum.
As a product of synthesis, the graphite flake,
amorphous carbon films and diamond microparticles in
quantity 2 to 4 particles per the rectangular 1.5×1.0 cm2
were obtained on all the types of substrates.
3. SEM and Raman spectroscopy results
SEM images of obtained diamond particles (Fig. 1)
demonstrate that synthesized diamonds obviously have a
cubic shape with the upper facets (311) or (711) and
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2006. V. 9, N 1. P. 22-28.
© 2006, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
24
600 800 1000 1200 1400 1600 1800 2000
0
10
20
30
40
50
60
70
Ra
m
an
in
te
ns
ity
, a
. u
.
Wave number, cm-1
Fig. 2. Micro-Raman spectrum of the crystal, which indicates
the peak at 1331.885 cm−1 that is characteristic for diamond.
a
b
Fig. 3. SIM images of the cross-section of the substrate with
diamond particles under small (a) and bigger (b) magnification.
lateral facets (110) or (101). The presented cubic shape
is also typical for spontaneous diamond synthesis. This
result is in a good correlation with the fact [2] that at
high temperature of supersaturating vapor it is observed
the plasma-chemical growth of diamond with dominant
cube faces. From SEM images, it is seen that diamond
particle sizes are 3 to 5 μm. Hence, the obtained growth
rate could reach ~3…5 µm/s proceeding from the
discharge appearance time 1 s.
The analysis of Raman spectra (Fig. 2) confirms the
diamond nature of particles. Spectrum exhibits the first-
order diamond peak centered at 1330 – 1332 cm−1. The
high intensity of the diamond peak, its small full-width
at half-maximum (5.42–6.32 cm−1) and small contri-
bution of the graphite phase demonstrate the high quality
of the diamond crystal particles. The Raman analyses
were also carried out on deposited areas surrounding the
diamond crystals: there are graphite flakes in the center
of the sample (directly under the place of the graphite
tips contact) and then areas that have D and G signals
usually characterizing diamond-like carbon films.
4. Discussion of the results
What do the presented results have new and unusual?
The first question which is necessary to be answered
what synthesis method occurs in this process: plasma-
chemical or physical. Initially it was declared [5] that the
process of growth was chemical vapor deposition. In
usual CVD [4], the hydrocarbon radicals are deposited
on the surface of a heated substrate and form the
diamond and graphite nucleation centers. Following this
fact in our experiments, the chamber was modified
specially to inlet hydrogen with the goal to get hydrogen
ions that would interact with activated carbon. It should
be mentioned that in experiments without hydrogen the
diamond particles were not obtained. Hydrogen inlet
promoted diamond synthesis. But the growth rate
3…5 µm/s obtained in our experiments is unique high.
Thus, the growth rates for diamond deposited using the
chemical vapor deposition method lie within the range
0.1 to 1.4 μm/h for the optimal conditions [4].
To elucidate the process of growth, the cross cutting
of samples in the place of diamond particle position was
carried out by the focused gallium ion beam. The results
of SIM (Fig. 3a, b) fulfilled in the place of cutting deny
the initial version of CVD as a running process. Both
physical and plasma-chemical methods of diamond
particle growth at low pressures are based on the fact
that nucleation and the following growth occur on the
substrate. That is why nature, temperature and treatment
of substrate influence on the deposition result very
strongly. But in the obtained results (Fig. 3), it is evident
that particles do not nucleate on the substrate because
Diamond
particle
Carbon
coating
Void
Si
substrate
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2006. V. 9, N 1. P. 22-28.
© 2006, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
25
Fig. 4. SIM images of the cross-section of the substrate with
CVD diamond particles.
Substrate Graphite rods
Graphite and
amorphous carbon films
area
Diamond
particle area
Fig. 5. Alignment of graphite rods, substrate and location of
products of synthesis.
only corner of particles lies on it. In comparison, the
cross cutting of the substrate with diamond deposited by
hot filament CVD shows that particle lies on the
substrate by a whole face and hence grows on the
substrate (Fig. 4). Possibly in our case the diamond
particle “drops” on the substrate. So, the running process
is not CVD. This conclusion coincide with that reported
in [6], where it was put under doubt the plasma-chemical
nature of the growth of diamond microparticles that were
obtained as a result of a pulse discharge between two
sharpened tips (~0.5 mm in diameter) of graphitic
electrodes. In the experiment [6], the carbon plasma is
formed by an electric pulse discharge (20 – 100 ms) at
the huge current (~800 – 1,000 A) and pressure (0.01 –
0.05 Torr). It was detected ~10 to 50 diamond
microparticles (10 – 100 µm) per square centimeter [6].
Because these microparticles do not stick to the substrate
and are easy removed, it was concluded that they
dropped on the substrate. But authors did not explain
how and where microparticles form in this case.
The second issue of doubts about plasma-chemical
nature of obtained diamond is the extremely high growth
rate (~3…5 µm/s) that is higher than that in known
publications [7-14]. The only rate 500–100 µm/s
calculated in [6] exceeds that of our results.
The additional fact opposing the plasma-chemical
process is independence of nature, temperature and
position of the substrates, which is in contrast to our
results and those of the work [6]. And finally, the
arrangement of microparticles is also strange from the
viewpoint of CVD. The matter is that microparticles are
located not in the area directly under a place of discharge
but further (Fig. 5). In the beginning, there are areas like
amorphous carbon films, which are formed at
temperatures considerably less than the diamond growth
temperature. Then, there are located areas with a small
amount of diamond particles (2 to 4 particles per area of
1.5×1.0 сm2). Hence, from the direction of the
temperature gradient, the assumption of diamond growth
on the substrate is unreal.
It is possible to conclude that diamond particles in
our experiments are formed by some physical
mechanism. The known physical mechanism assumes
growth from supersaturated carbon vapor on the
substrate. In our experiment, supersaturated carbon
vapor could be obtained as a result of heating of graphite
rods at big direct current running and reaching in its
maximum 25 A. However, the particles do not nucleate
on the substrate. We assume that diamond particles
process formation occurs in the place of contact of
graphite rods [15]. The results with different shape of
graphite tips prove this idea. Diamond microparticles
could not be formed in experiments with cylindrical rods
(diameter 3 mm) and were synthesized only in the case
of sharpened tips (1-2 mm). Sharpening of the end leads
to the electric field strength increase within the discharge
range. Hence, the electric characteristics determine the
synthesis. That is why, the next question concerns with
the process in the discharge place. Because the process is
accompanied of crackling that is peculiarity of spark
discharge [16], in the beginning the discharge was
determined as a spark [5, 15]. However, it is known [16]
that the spark discharge appears at relatively high gas
pressure what is close to the atmospheric one or higher.
In our experiments, the pressure is only 100 Torr. Strong
heating of graphite rods at high current resulting in
graphite evaporation despite of the fact that graphite is
refractory material. As graphite vapor is ionized, ions
have weak mobility, so the pressure in discharge place
increases and could reach the atmospheric one.
Assuming that the process is the spark discharge, the gas
temperature in the discharge channel reaches the value
of the order of 10,000 °C [16]. Such sharp temperature
increase for a short time of spark being generates very
strong and instant (exposure type) increase in the gas
pressure that leads to characteristic crackling. In this
case, the process of diamond formation does not run at
low pressure and should be characterized as the process
occurring at a high temperature and high pressure. In this
Diamond
particle
AlN
Substrate
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2006. V. 9, N 1. P. 22-28.
© 2006, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
26
situation, it should be taken into account another
mechanism of diamond synthesis. That is why, these
papers suggest the hypothesis of diamond formation
through the mechanism of the direct phase transition of
the first order at the high pressure. From the data [17]
the phase transition graphite→diamond at shock
compression takes place within the pressure range 200 to
400 kbar, the volume effect of transformation composes
18 %. According to our estimating calculation, the
temperature at 200 kbar is 470 K, at 400 kbar – 770 K, at
600 kbar – 1300 K [17]. Diamond synthesized under
shock compression of pure graphite without special
introduction of additives – coolers and not dissipated in
cooled matrix has a cubic structure in stored after
compression samples. In work [17], it is mentioned the
results of Bandy who got transformation
graphite→diamond for 6 – 8 s at the temperature 4000 K
due to application of pulse heating the pressed graphite
samples using a discharge from a powerful condenser
battery. For cubic diamond synthesis at the pressure
140 kbar and temperature 3000 K, it was necessary 4 ms.
In our experiments, the registered temperature in the
discharge area reached 2400 °C, process runs 1 and 10 s,
diamond has the cubic structure, which is close to the
mentioned above data [17].
The next issue is elucidation of the mechanism of
phase transition: martensite or diffusion. It is known [17]
that structure of forming phase and degree of crystal
perfection depends on the mechanism of the direct phase
transition. As to make the X-ray structural analysis of
obtained particles is not obviously possible because of
their small sizes and difficulties to find them on the
substrate with account of their small amount (2 to 4
particles on the sample), so making the conclusion of the
mechanism responsible for the phase transition could be
done by some indirect way. The transformation observed
by us is very fast (~1 s), which is characteristic of the
martensite mechanism. But the temperature in our
process is high, while the martensite transformation
occurs at low temperatures. That is why, most likely the
character of the transformation observed in our
experiments is diffusion. If we return to SEM images of
the obtained particles (Fig. 1), we could see that the
diamond single crystal has a “tail”. This tail could be
stiffened fused part of melting graphite. The formation
of separated single crystals grown in the whisker shape
is characteristic for crystals prepared from melt [17] as
well as in our case. Moreover, through melting only
diffusion growth of a new phase is possible [17]. But if
diamond formation goes by the diffusion way from
melting graphite, it is possible that we discovered
experimental proof of existence of threefold point of
transition graphite→liquid→diamond on the known
carbon diagram of state [18], theoretically predicted but
earlier not observed. It was mentioned [18] that from the
Bandy estimation the threefold point
graphite→liquid→diamond takes place at the pressure
12 GPa and temperature 4000 K on the carbon diagram
of state.
The described mechanism of diamond formation
could explain the results of our experiments and those of
the work [6]. The difference between our results and [6]
includes the size and amount of diamond microparticles.
Apparently larger size and bigger amount of
microdiamonds in [6] is based on the usage of graphite
of very high purity (99.999 %). While in our
experiments, we use graphite of the mark P-41
consisting of graphite and amorphous carbon and hence
possessing more defects and impurities. The next
divergence with the data [6] concerns with hydrogen
influence. In the work [6], hydrogen was introduced to
the chamber in the latter set of experiments specially to
check the hypothesis of plasma-chemical reaction.
However, its presence does not influence on the results.
Hydrogen changes neither particle size nor their shape.
By contrast, in our experiments the diamond particles
are obtained only with hydrogen. This fact was the initial
reason to support that this process was CVD, because as
is known [2] hydrogen availability is the necessary
component of diamond plasma-chemical deposition.
Apparently, in our process hydrogen acts as a cooling
medium for crystallizing diamond particles. Changing of
gas in the chamber from pure H2 to mixture H2/Ar in
relation 2:98 at the total pressure 100 Torr leads to the
result that pure cubic diamond is not discovered. It
means that the size of gas molecules, their amount and
composition has a big importance and influence on the
process kinetics.
It should be marked that our conclusion is made on
assumption that the discharge is of the spark type. But
the discharge mechanism is not completely clear. In the
work [19], it is approved that at the distance d between
cathode and anode less than 5 cm the atmospheric air is
breakdowned by not streamer but Townsend mechanism
of duplication of avalanches. In our case, the position of
electrode is end-to-end and d is very short. From the
abovementioned information, it means that mechanism
is not streamer like and hence the discharge is not spark.
To elucidate the breakdown mechanism, we should note
that one from the tips of graphite electrodes is sharp
what creates the increasing of electric field strength.
Yu.P. Rayzer writes [19] that due to sharp dependence
of the coefficient of electron duplication α on the electric
field strength E it is enough to get an overvoltage
approximately 10 % for creation of the streamer even in
this interval, while at a static voltage the breakdown is
Townsend. Competition between mechanisms of
avalanche duplication and streamer takes place, and
result could be declined in one or another side. For
example, the streamer mechanism would be dominant at
the overvoltage increase or due to introduction of
another additional gas that decreases the coefficient of
secondary electron emission, which suppresses the
process of avalanche duplication. So, the spark discharge
is possible in our case.
If, nevertheless, to start with the fact that the pressure
in the place of contact of graphite rods is unknown and
the initial pressure is 100 Torr, so from the
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2006. V. 9, N 1. P. 22-28.
© 2006, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
27
determination [19] of the process running at
pd < 1000 Torr·cm, where p is the pressure and d is the
distance between the anode and cathode, the discharge is
characterized as the decaying one passing into the
process of a current increase up to the arc discharge. In
this case, explaining the result of the phase transition
graphite→diamond is possible via the process of
explosion emission described in the work [19]. At the
breakdown of vacuum interval with strong sharpening
electrode (cathode or anode) the sharp current increasing
after some time leads to explosion of the tip and plasma
cot emission from it. Explosion of electrode materials
(graphite in our case) occurs as a result of emission on
the tip of plenty joule heat from current of autoelectronic
emission and then in the process of heating from the
current of thermoautoelectronic emission. The current
density from tip surrounded by plasma reaches 108 –
109 A/cm2. While the electrons take off from the surface,
the new electrons come to this place from the bulk of
material. I.e., in the graphite near the emitting surface,
the current of a huge density proceeds. This current heats
graphite till explosion-like evaporation of microledge. It
is known [19] that at the currents 15 to 20 A the arch
becomes sibilant. At the further increase of the current,
the torch of heated vapor goes off from anode spot. The
temperature in the anode spot reaches 4200 K. In our
experiment, the current increase goes on up to 25 A.
Hence, it is possible to assume that the process of
evaporation and explosion-like evaporation of graphite
take place without the molten phase. But if only the
evaporation process is running, the diamond formation is
going through the known physical mechanism of growth
on the substrate. However, the presence of “tail” in
Fig. 1 connected with the cubic diamond insists on the
version of melt. So, a most likely the breakdown
mechanism is streamer and discharge is spark.
5. Conclusion
As a result of the spark discharge between two graphite
rods in hydrogen at the low pressure 100 Torr in the
chamber, the diamond microparticles 3 to 5 µm are
obtained on the silicon, Al2O3 ceramics and non-alkaline
glass substrates. The reached growth rate 3 to 5 µm/s is
very high. Diamond nature of particles is confirmed by
the results of SEM and micro-Raman analyses. Cross
cutting the diamond particles by FIB demonstrates that
synthesis does not occur on the substrate, which denies
the known CVD mechanism and physical mechanism of
formation at low pressures. The hypothesis of diamond
synthesis through direct phase transition of the first order
graphite→liquid→diamond via the diffusion mechanism
at high temperatures and pressures has been proposed.
The high temperatures and pressures are generated by
the process in the spark discharge. In this case, hydrogen
plays a role of cooling ambient. The obtained results
could be experimental proofs of existence of threefold
point on the phase diagram of carbon and demonstrates
the new way to create diamond microparticles.
Acknowledgement
The author wish to thank Dr Y. Takagi for opening up
the opportunities to work in the research group in the
Department of Environmental and Materials Science,
Teikyo University of Science & Technology, Japan as
well as for technical supporting the experiments. The
author thanks students T. Hirai and T. Kawai for their
assistance and collaboration in the common
experimental work for growing the diamonds. The
author also thanks Dr A.I. Kutsay (Institute of Superhard
Materials, NANU, Kiev), M.G. Dusheyko and
Yu.V. Yasievich (National Technical University of
Ukraine “KPI”, Kiev) and Professor V.G. Litovchenko
(Institute of Semiconductor Physics, NANU, Kiev) for
the discussion of some parts of this work.
References
1. D.M. Gruen, Nanocrystalline diamond films // Annu.
Rev. Mater. Sci. US Government, 29, p. 211-259
(1999).
2. John C. Angus, and Cliff G. Hayman, Low-pressure,
metastable growth of diamond and “diamond like”
phases // J. Science 241, p. 913-921 (1988).
3. B.V. Spitsyn, and A.E. Alexenko, Origin, currently
abilities and some perspectives of development of
diamond synthesis from gas phase // Proceedings of
5th Intern. Symposium diamond films, April, Kharkov,
Ukraine, p. 122-149 (2002) (in Russian).
4. B.V. Spitsyn, L.L. Bouilov, and B.V. Derjaguin,
Diamond and diamond-like films: deposition from
the vapour phase, structure and properties // Progr.
Crystal Growth and Charact. 17, p. 79-170 (1988).
5. T. Hirai, Y. Takagi, O. Shimizu, Y. Suda, and T.
Semikina, Five micron diamond particles synthesized
in ten seconds // Abstract Book of 15-th European
Conference on diamond, diamond-like materials,
carbon nanotubes, nitrides and silicon carbide, Riva
Del Garda, Trentino, Italy, September (2004).
6. A.V. Palnichenko, A.M. Jonas, J.-C. Charlier,
A.C. Aronin, and J.-P. Issl, Diamond formation by
thermal activation of graphite // J. Nature 402, p.
162-165 (1999).
7. M. Yoshikawa, N. Ohtake, and Zukai Kisou Gousei,
Diamond. Ohmsha, (1995) (in Japanese).
8. A. Chayahara, Y. Kino, Y. Horino, and N. Fujimori,
CVD diamond synthesis with high growth rate //
New Diamond 20(4), p. 26-27 (2004) (in Japanese).
9. R. Velazquer, B. R. Weiner, and G. Morell, Diamond
film synthesis at low temperatures // Abstract Book,
Elsevier, 15th European Conference on diamond,
diamond-like materials, carbon nanotubes, nitrides
and silicon carbide “Diamond 2004”, 12-17 Sept,
2004, Italy.
10. K. Subramanian, W.P. Kang, J.L. Davidson, and
W.H. Hofmeister, The effect of growth rate control
on the morphology of nanocrystalline diamond //
Ibid.
Semiconductor Physics, Quantum Electronics & Optoelectronics, 2006. V. 9, N 1. P. 22-28.
© 2006, V. Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
28
11. T. Bauer, M. Schreck, H. Sternschulte, and
B. Stritzker, High growth rate homoepitaxial
diamond deposition // Ibid.
12. T. Teraji, M. Hamada, H. Wada, M. Yamamoto,
K. Arima, and T. Ito, High-rate growth of high-
quality homoepitaxial diamond films by means of
high-power microwave plasma chemical vapour
deposition // Ibid.
13. N. Fujimori, A. Chayahara, Y. Mokuno, Y. Horino,
Y. Takasu, H. Kato, and H. Yoshikawa, Characte-
ristics of single crystal diamonds under large growth
rate obtained by microwave plasma CVD // Ibid.
14. R. Spitzl, and H. Sung-Spitzl, Large area-high
growth diamond deposition with uniform microwave
plasma // Ibid.
15. T.V. Semikina, Y. Takagi, T. Hirai, T. Kawai,
O. Shimizu, and Y. Suda, New mechanism of “spark
diamond” formation in spark plasma physical
process // Abstract Book, Elsevier, 16th European
Conference on diamond, diamond-like materials,
carbon nanotubes, nitrides and silicon carbide
“Diamond 2005”, Sept, 2005, France.
16. T.A. Voronchev and V.D. Sobolev, Physical basies
of electrovacuum devices. Vysshaya shkola, Moscow
(1967) (in Russian).
17. A.V. Kurdyumov and A.N. Pilyankevich, Phase
transformations in carbon and boron nitride.
Naukova Dumka, Kiev (1979) (in Russian).
18. A.V. Kurdyumov, V.G. Malogolovets, N.V. Novikov,
A.N. Pilyankevich, L.A. Shul’man, Polymorphic
modifications of carbon and boron nitride. Handbook.
Metallurgiya, Moscow (1994) (in Russian).
19. Yu.P. Rayzer, Physics of gas discharge. Nauka,
Moscow (1987) (in Russian).
|