Carbon 214 (2023) 118332

Contents lists available at ScienceDirect

Carbon
journal homepage: www.elsevier.com/locate/carbon

A comprehensive study of simultaneous RF-PECVD synthesis of n-diamond

and carbon nanowalls hybrid structure over a large area
A.M. Mumlyakov a, *, E.A. Pershina a, Ju.V. Bondareva a, b, P.A. Nekludova a, A.A. Shibalova a,
M.V. Shibalov a, Yu.V. Anufriev a, A.M. Tagachenkov a, M.A. Tarkhov a
a
Institute of Nanotechnology of Microelectronics of Russian Academy of Science, Leninsky Prospekt 32A, Moscow, 119991, Russia
b
Center for Materials Technologies, Skolkovo Institute of Science and Technology, 121205, Russia

A B S T R A C T

In this study, n-diamond–carbon nanowalls hybrid structure was synthesized on a 4-inch diameter silicon wafer by RF-PECVD technique (type CCP). A comprehensive
analysis of the structure was carried out at the beginning of growth and after a long period of synthesis using electron microscopy, transmission microscopy, Raman
spectroscopy, and X-ray diffraction techniques. As the analysis showed, feeding carbon monoxide (CO) to a gas mixture of methane (CH4), argon (Ar), and hydrogen
(H2) induces the formation of n-diamond particles in a basal layer. By using plasma-enhanced etching in an oxygen-containing medium, carbon nanowalls (CNWs)
were successfully removed, and the underlying layer containing n-diamond was analyzed. For the first time, X-ray diffraction results of a separate n-diamond phase
were obtained for this layer.

1. Introduction
Scientific advances in applied research and technology have accel­
erated the discovery of hybrid high-tech materials with unrivaled
physical and chemical properties. Scientists and researchers in this field
are paying special attention to a unique element called carbon [1–7].
The ability of carbon to create different coexisting allotropic structures
with completely different properties is the main characteristic of this
element [8]. Recently, the research community has been primarily
focusing on CNWs, i.e., one of the carbon allotropic forms.
The exceptional morphological and structural properties of CNWs,
including their vertical orientation, open and ultra-thin edges of gra­
phene sheets, large specific surface area and open areas between gra­
phene sheets, have attracted significant interest [9,10]. Due to their
important properties, such as high electrical conductivity, thermal
conductivity, and chemical stability, CNWs have been successfully
applied in nanoelectronics, photoelectronics, and electrochemistry
[11–14].
CNWs are mainly grown using Plasma Enhanced Chemical Vapor
Deposition (PECVD) from various types of hydrocarbons with different
plasma sources such as radio-frequency (RF) [15], microwave (MW)
[16], and direct-current (DC) [17]. The interfacial layer formed between
a substrate and CNWs, i.e., the basal layer, is an important phase of
CNWs synthesis. The basal layer ensures a uniform heat and current
distribution within the substrate plane and remains significant for
interface resistance and CNW formation [18]. Studying the basal layer
has demonstrated that it can be represented by a layer of graphite sheets
growing parallel to a substrate [19], a layer of amorphous carbon [20],
or various carbides formed during interaction with the substrate mate­
rial [17]. The type of the formed basal layer depends on both the syn­
thesis technique and gas precursors applied during the growth.
Throughout our research, we found that feeding an additional car­
bon source in the form of carbon monoxide (CO) to a gas mixture (Ar/
CH4/H2) induces the formation of a unique interface layer consisting of -
new-diamond (n-diamond, FCC-carbon). A new diamond crystalline
phase (carbon γ phase) was first produced during the low-temperature
annealing of a carbon film exposed to Ar ions during deposition [21].
Various methods have been developed to generate nano-sized n-dia­
mond crystallites, including hydrothermal synthesis [22], hydrogen
plasma treatment of a diamond surface [23], microwave plasma
enhanced chemical vapor deposition (MW-PECVD) [24], and
ultrasonic-assisted alcoholic-alkaline treatment [25]. The n-diamond
phase can also be a byproduct of the synthesis of other carbon structures
using the High Temperatures – High Pressures (HTHP) technique [26],
carbon shock compression [27], and carbon detonation [28]. However,
these methods produce n-diamond as nanoparticles or mixtures with
other carbon-containing phases [29,30].
There is no detailed study of the n-diamond structure using X-ray
diffraction in the scientific papers due to the small size of nanoscale
crystallites and by-products arising during synthesis. Therefore, most of

* Corresponding author.
E-mail address: mumlyakov.a@inme-ras.ru (A.M. Mumlyakov).

https://doi.org/10.1016/j.carbon.2023.118332
Received 10 April 2023; Received in revised form 25 July 2023; Accepted 26 July 2023
Available online 27 July 2023
0008-6223/© 2023 Elsevier Ltd. All rights reserved.
A.M. Mumlyakov et al.
the data on the n-diamond structure has been obtained based on electron
diffraction results from TEM studies [22,24,25,31], which is insufficient
to determine the exact structure of the n-diamond. Thus, the n-diamond
crystalline structure remains uncertain and requires further scientific
investigation [32–34].
The study of n-diamond-CNWs hybrid structure formation is
important fundamental research for producing applied micro- and
nanodevices, optoelectronic systems, energy storage and conversion
systems, various sensors and other tools. Understanding the factors
affecting the development of hybrid structures facilitates the control of
the synthesis and growth processes, as well as the phase composition.
In this study, the n-diamond-CNWs structure was manufactured on a
four-inch silicon wafer using Plasma Enhanced Chemical Vapor Depo­
sition (PECVD). A comparative analysis of the structure was carried out
using electron microscopy, transmission microscopy, Raman spectros­
copy, and X-ray diffraction with and without CO. As a result of applying
the plasma-based surface treatment methods, CNWs were selectively
removed, allowing the X-ray diffraction pattern of a nanocrystalline n-
diamond phase to be obtained. The structure formation was also studied
at an early stage of synthesis with CO feeding, which allowed to collect
experimental data on the evolution of the n-diamond-CNWs hybrid
structure.
2. Experimental section
2.1. Structure formation procedure
P-type silicon wafers (100) with a diameter of 4 inches and a re­
sistivity of 12 Ohm*cm were used as blank wafers. The wafers were
preliminarily cleaned in a piranha solution (a mixture of concentrated
sulfuric acid with hydrogen peroxide), washed in de-ionized water, and
then processed in a centrifugal drier with hot nitrogen purging.
The n-diamond-CNWs hybrid structure was synthesized in the Ox­
ford Instruments Nanofab 1200 Agile reactor. At room temperature, the
wafers were placed at a grounded graphite base electrode, which is used
with an upper electrode to generate high-frequency capacitively coupled
plasma (CCP). The heating to 750 ◦ C was performed in a hydrogen flow
at a pressure of 1 Torr, and an average heating rate of 14 ◦ C/min. The
wafer was then cleaned in hydrogen-argon plasma at 1 Torr and 200 W
and 150 W power at radio frequency (13.56 MHz) and low frequency
(100 kHz) sources, respectively, for 10 min. The depositing phase
comprised the feeding of Ar, CH4, CO, and H2 process gases (ratio
200:10:7.5:2.5), at a pressure of 250 mTorr and the RF source power of
300 W. It is important to note that a graphite electrode can also serve as
the carbon source for the synthesis of carbon hybrid structure [35–37].
Once the deposition was complete, the graphite electrode was cooled
under a helium flow until it reached a temperature of 100 ◦ C, and the
wafer was then removed from the system.
To analyze the formation of the n-diamond-CNWs hybrid structure,
the wafers with a synthesis period of 38; 75; 150; 300; 450, and 3600 s
were prepared. Synthesis was also performed on a silicon wafer for 3600
s without feeding carbon monoxide (CO) to a gas mixture.
CNWs etching in a hybrid structure with a synthesis period of 3600 s
was carried out in oxygen plasma using the Diener TETRA 30 device.
The pressure was 375 mTorr, the heater temperature was 100 ◦ C, and
the RF source power was 500 W.
2.2. Characterization
2.2.1. Raman spectroscopy
Raman-scattering spectra were obtained using the Renishaw inVia
Qontor confocal Raman dispersive spectrometer, with an excitation
wavelength of 532 nm focused on the sample using the 50× magnifi­
cation lens. Spectra were recorded Over a range of 1000–2000 cm− 1
during 10–20 s and 2–5 accumulations. The laser output power was
maintained below 1 mW to avoid overheating and any associated the
2
Carbon 214 (2023) 118332
Raman shift.
2.2.2. Scanning electron microscopy
Scanning electron microscopy (SEM) was performed using FEI He­
lios, a dual beam FIB-SEM (focused ion beam) system. Images (patterns,
micrographs) were acquired using the secondary electron imaging mode
while scanning the surface of analyzed objects with an electron beam.
Imaging parameters were selected experimentally according to the
criteria of sufficient resolution for visualizing relevant characteristics
and an acceptable signal-to-noise ratio. Scanning electron microscopy
enabled taking micrographs of enlarged areas of the sample surface,
their edges on cleaves perpendicular to the surface of analyzed objects.
2.2.3. Transmission electron microscopy
The microstructure of the samples was studied by transmission
electron microscopy with high resolution (TEM and HRTEM) using a
JEOL JEM-2100 Plus microscope at an accelerating voltage of 200 kV.
The two-dimensional Fourier transform technique (FFT) was used to
obtain diffraction patterns of separate nanoparticles and minimize the
noise of high-resolution TEM images.
A thin fragment of a sample (membrane, lamellae) was prepared
using the FIB dual-beam system by producing the cross-section,
extracting it from the volume, transferring it onto the selected grid
using a micromanipulator, and subsequent reducing the extracted object
thickness for the TEM analysis (less than 100 nm). Upon extracting,
transferring, and securing the relevant sample fragment on a specially
selected grid using the FIB, the object was polished to decrease (to carry
out thinning) its thickness to a value sufficient for acquiring the needed
data by the TEM technique (less than 100 nm). The area of interest was
polished by a focused ion beam at a current of 100–5000 pA, which is
lower than the current applied to remove material around the fragment
before its extraction from the sample. A niobium layer of 50–100 nm was
applied to protect the sample surface from exposure to an ion beam, and
was not subject to structural analyses.
2.2.4. X-ray diffraction
The samples were analyzed on an X-ray diffractometer using the X-
ray diffraction (XRD) technique at a grazing incidence of X-ray radiation
to a sample (incident angle ≈ 0.5◦ , selected based on the maximum
intensity of peaks from CNWs and n-diamond). X-ray diffraction patterns
were taken in the scanning mode at an angle of 2θ using Cu Kα radiation
(λ = 1.540605 Å) under the X-ray source parameters of 45 kV and 40
mA. A paraboloidal reflector with a divergence slit of ¼ and an anti-
scatter slit of ¼ was set on a primary beam. A parallel plate collimator
with a divergence amounting to 0.19◦ was set on a diffracted beam. The
value of Soller slit divergence was equal to 0.04◦ . A point-linear detector
was used as a detector.
3. Results
3.1. Section I The effect of CO feeding
SEM micrographs (Fig. 1 a, b) provide the morphology comparison of
synthesized CNWs without CO and with CO feeding after 3600 s of
synthesis. The SEM images show a top view of the samples, revealing a
petal-like structure characteristic of CNWs [38,39] for both samples. It
can be observed that the addition of CO to the reaction mixture increases
the lateral wall size at the same synthesis time.
The Raman spectra for samples synthesized with and without CO
feeding for a synthesis duration of 3600 s are given in Fig. 1 (c, d). The
Raman spectra display typical features of CNWs, including specific
graphite peaks of D, G, and D’. The Raman G band at ~1583 cm− 1 arises
from the in-plane vibration mode of neighboring carbon atoms [40]. The
wide peak at ~1348 cm− 1 belongs to a disorder-induced D band acti­
vated during the first-order scattering. This band may indicate the
presence of substitutional heteroatoms in the plane, vacancies, grain
A.M. Mumlyakov et al. Carbon 214 (2023) 118332

Fig. 1. Comparison of CNWs for a synthesis duration of 3600 s with and without CO feeding. SEM images in secondary electrons are presented in the same scale in
Fig. 1(a), without CO feeding, and in Fig. 1(b), with CO feeding. The Raman spectra for the samples without CO and with CO feeding are provided in Fig. 1(c) and (d),
respectively. (A colour version of this figure can be viewed online.)

boundaries, or other defects that reduce the crystal symmetry [41]. The
D′ band peak at 1617 cm− 1 corresponds to phonon vibrations on the
edges of graphene planes and is typical for the CNWs structure [42]. The
high intensity of the D peak and the narrow width of the G peak are also
characteristic features of CNWs [43]. The ID/IG value is a key parameter
for determining the degree of defectiveness of carbon materials. An in­
crease in the ID/IG ratio may indicate a disorder degree, alteration of
structural orientation or length of CNWs formed. The phase composition
may also affect the ratio of the D and G peaks; therefore, the ID/IG ratio
may be higher due to the presence of other allotropic carbon structures
[44]. Furthermore, ID/IG increases with an increase in sp2 content,
indicating the formation of small graphite clusters. Thus, the value of
ID/IG for the CNWs sample synthesized with СО feeding is 2.5 ± 0.1,
which is lower than the value for the sample synthesized without СО
feeding, i.e., 3.0 ± 0.2.
The phase composition of samples synthesized with and without CO
was determined using the X-ray diffraction technique. To analyze the
surface layer of the samples (CNWs + interface), the Grazing Incidence
X-ray Diffraction (GIXRD) method was applied. The X-ray diffraction
pattern of the sample synthesized without CO (see Fig. 2, blue chart)
displays peaks at 2θ angles equal to 25.4 , 43.0 , 45–46 , and 78.5 ,
◦ ◦ ◦ ◦

which refer to CNWs [45]. The lack of (004) reflex at 2θ ≈ 53.3 on the
о

diffraction pattern can be explained by a low number of graphene layers
in nanowalls along [0001] or their aperiodicity. However, the diffrac­
tion pattern of the sample synthesized with CO (Fig. 2, dark-red chart)
displays additional peaks at 2θ angles equal to 44.6 , 52.1 , 76.7 , 93.2 ,
◦ ◦ ◦ ◦
Fig. 2. The Grazing Incidence X-ray Diffraction patterns for the samples syn­
thesized without CO (blue chart) and with CO feeding (red chart). The location
of peaks for the CNWs phase and n-diamond phase are marked on a line dia­
gram. (A colour version of this figure can be viewed online.)

and 98.8 , which were identified as the n-diamond phase according to

◦
for the range of 2θ angles from 25 to 105 distinctively displays peaks at
◦ ◦

scientific literature [28,46].

2θ angles equal to 44.6 , 51.8 , 76.4 , 93.0 , and 98.5 for the Kα1 po­
◦ ◦ ◦ ◦ ◦

Several peaks in the diffraction pattern overlap with each other.

Since experimentally obtained X-ray diffraction patterns for the n-dia­ sition. The diffraction pattern background was removed due to its higher
mond phase without overlapping peaks from accompanying phases are depth at narrow angles, which is associated with the GIXRD measure­
not presented in the literature, we removed CNWs from the hybrid ment geometry. The fundamental difference between the diffraction
structure by treatment in oxygen plasma. The diffraction pattern (Fig. 3) pattern of n-diamond and common diamond is the presence of the HKL

3
A.M. Mumlyakov et al.
Fig. 3. Experimental X-ray diffraction pattern of a nanocrystalline n-diamond
phase indicating the fractional intensity of each peak. The diffraction pattern
background was removed. (A colour version of this figure can be
viewed online.)
(200) diffraction peak at the ~51.8◦ angle position by 2θ. The diffrac­
tion pattern without background removal is shown in Fig. S1. The value
of integral FWHM (integral breadth) for all peaks is ≈ 1.3, which cor­
responds to an average size of crystallites equal to ≈10 nm and calcu­
lated according to the Scherrer formula [47]. The intensity ratio of each
peak is given in the insert to Fig. 3. The number of reflexes, their relative
position, and intensity ratio substantiate that the n-diamond structure
can be described by a crystallographic set Fm3m, as it was suggested in a
series of scientific papers [48,49].
Fig. 4. Cross-sectional a) SEM and c) TEM images of CNWs grown on Si substrate for 3600 s without CO. Cross-sectional b) SEM and d) TEM images of CNWs grown
on Si substrate for 3600 s with CO. (A colour version of this figure can be viewed online.)
4
Carbon 214 (2023) 118332
According to the diffraction pattern interpretation using the Fm3m
structure, the n-diamond lattice parameter is equal to а = 3.52 Å, which
is in agreement with the literature data [48]. This lattice parameter
exceeds those appropriate for a crystalline structure consisting solely of
carbon atoms. Scientific papers contain theoretical evaluation assuming
the presence of hydrogen atoms in a crystallite lattice [34], which can
explain the high value of the lattice parameter for n-diamond.
A general cross-section view of the sample containing the hybrid
structure grown with and without CO for the same synthesis duration of
3600 s is shown in Fig. 4 a, b in SEM images of secondary electrons. It
can be seen that at the same synthesis time, adding CO to the reaction
mixture leads to the growth of higher CNWs. The increase in CNWs
growth rate is associated with a shift in the balance between the process
of etching in the oxygen plasma and the deposition towards deposition,
which occurs when a small amount of CO is added to the gas mixture.
The morphology and structural differences of the obtained samples
samples were also analyzed using the TEM technique. The general view
of the CNWs-Si interface for samples grown without CO and with CO
feeding is provided in Fig. 4 c and Fig. 4 d, respectively. The TEM images
show a distinctive feature of the sample grown with CO, i.e., the pres­
ence of individual nanoparticles formed on the CNWs-Si boundary.
According to TEM images, the average size of the particles is 10 nm. The
presence of graphene layers of CNWs growing around the observed
nanoparticles is also apparent in Fig. 4 d. By contrast, in the sample
grown without CO, only graphene layers with an interplanar spacing of
3.74–3.77 Å can be visible at the CNWs–Si boundary (Fig. 4 c, insert)
corresponding to the interplanar spacing for the (0001) plane of tur­
bostratic graphite [50] and capable of responding to an unformed CNWs
structure.
High-resolution TEM images of the CNWs-Si section were taken in
several areas of the sample grown with CO to study the nanoparticle’s
structure. An example of those areas is given in the insert to Fig. 4 d, i.e.,
an enlarged image of a yellow square. Some particles overlapped or were
A.M. Mumlyakov et al.
composed of several crystallites, while most of them represented indi­
vidual monocrystals and had a shape close to spherical. To identify the
phase, we only used FFT images of monocrystals. Significantly, different
allotropes of carbon may exist simultaneously and possess the same
value of interplanar d-spacings. Therefore, to ensure precise identifica­
tion of a phase of carbon nanoparticles under study, we determined
interplanar d-spacing and angles between directions in the lattice on FFT
images and compared them with values defined for registered allotropes
of carbon according to the ICDD database.
A typical HRTEM image of a particle taken for the evaluation and a
relevant FFT pattern can be found in the insert to Fig. 4 d. The reflection
positions on the displayed FFT pattern indicate that this electron
diffraction pattern corresponds to the [101] cross-section of the Fm3m
lattice structure [51], and the lattice parameters correspond to the
n-diamond phase. By summarizing the electron diffraction data of
various individual nanoparticles in the sample, it can be concluded that
during synthesis with CO feeding, randomly oriented crystals with an
n-diamond structure and Fm3m lattice are formed. The obtained data are
consistent with the results of the investigation by the XRD technique.
3.2. Section II Evolution of the n-diamond-CNWs structure over time
Samples with shorter synthesis times, i.e., 38 s, 75 s, 150 s, 300 s, and
450 s, were prepared to monitor the formation of the n-diamond-CNWs
hybrid structure. SEM, Raman spectroscopy, TEM, and XRD were used
jointly to examine the potential formation of CNWs and the nano­
structured phase for each sample.
Raman-scattering spectra of the n-diamond-CNWs hybrid structure
on a silicon substrate with various synthesis duration are presented in
Fig. 5 a. Raman scattering was measured to describe and monitor
changes in the n-diamond-CNWs hybrid structure of various synthesis
duration. Raman spectra display the typical spectra of a graphite
structure with specific D peaks (1349 cm− 1) and G peaks (1604 cm− 1) at
38 s of synthesis. The shoulder D’ peak (1620 cm− 1) typical of carbon
nanostructures for 75 s of synthesis indicates graphite crystallites of a
limited size and graphene edges [43]. After 150 s of synthesis, the 2D
peak (2680 cm− 1), which is a second-order peak, and the D + G peak
(2942 cm− 1) are distinctively observed on Raman spectra and comprise
a combination of the G and D peaks [52]. Full width at half maximum
(FWHM) of Raman G bands at 75 s and 150 s in Fig. 5 а is sharper
compared with other spectra and equal 28 and 25 cm− 1 respectively.
Fig. 5. The results of studying the n-diamond-CNWs hybrid structure for various synthesis duration by Raman scattering. a) Raman spectrum for various synthesis
duration b) Dependence of ID/IG ratio on synthesis duration c) Dependence of the G peak position ratio on synthesis duration. (A colour version of this figure can be
viewed online.)
5
Carbon 214 (2023) 118332
Such a narrow width of G means a higher degree of graphitization and
the slight decrease in the FWHM of the D peak corresponds to small
crystallite size. It is assumed that at low synthesis times, CNWs are
graphite-like carbon types consisting of small crystallites with a high
degree of graphitization. A significant variation in the intensity of
measured D and G peaks was observed in the samples. The ID/IG ratio for
different synthesis durations is shown in Fig. 5 b. Changes in the ID/IG
ratio are caused by variations in the specific surface area of the material
and the average size of a carbon nanowalls [41], which can be seen in
SEM micrographs (Fig. 6 a-d). Extending the synthesis time to 300 s led
to an increase in specific surface area, which enhanced edge density,
formed defects on the grain boundaries, and increased bond breaking on
the surface of the carbon nanostructures [53]. After 450 s of synthesis,
the specific surface of carbon nanostructures decreases while the wall
length increases (Fig. 6 d). This process results in thinning of the defects
in the carbon nanostructures and lowering of the D peak intensity
Increasing synthesis duration leads to a shift of the G peak into the red
region (Fig. 5 c), which indicates the formation of a graphitized struc­
ture demonstrating a high concentration of carbon nanostructures on
the substrate and implies a potentially high oxygen content in the carbon
nanostructures [54]. It is worth noting that in the Raman spectra shown
in Fig. 5 a, we do not observe the n-diamond peak. This may be because
the n-diamond core is surrounded by a shell of disordered sp3-carbon
and formed graphene sheets of carbon nanowalls, whose signal is much
more intense than that of the crystal phase. The data on the peaks’
location, FWHM, and the ID/IG ratio are given in supporting information
(Table S1).
A significant change in the surface morphology and the corre­
sponding increase in the height of carbon nanostructures as a result of
increasing the synthesis duration (Fig. 6 e) can be distinctively observed
from SEM micrographs (Fig. 6 a-d). The thickness of CNWs increases
linearly with synthesis time, confirming their existence within the pro­
cess window. The average growth rate was 10.8 nm/min. SEM images
(Fig. S2) show graphite nanoislands with immature carbon nanowalls at
38 s of synthesis, which is confirmed by Raman spectroscopy (Fig. 5 a).
At 75 s (Fig. 6 a), the development of randomly oriented carbon nano­
walls started. Alongside the formation of separate graphene walls, car­
bon nanofibers (CNFs) significantly exceeding their height began to
form (insert, Fig. 6 e). A similar development of carbon nanowalls and
carbon nanotubes (CNTs) was observed in the direct current glow-
discharge plasma [17]. However, such a growth model in the
A.M. Mumlyakov et al.
Fig. 6. The results of studying the n-diamond-CNWs hybrid structure for various synthesis duration by electron microscopy. SEM image (top view and a cleave) for
synthesis during а) 75 b) 150 c) 300 d) 450 s. e) Dependence of carbon nanowalls height on synthesis duration. (A colour version of this figure can be viewed online.)
RF-PECVD system (CCP type) has not been reported in scientific papers.
At 150 s of synthesis (Fig. 6 b), carbon nanowalls primarily change their
growth direction to vertical. This phenomenon is likely caused by in­
ternal strain, due to the discrepancy of lattices [55,56] between
new-diamond particles and graphene sheets. At this stage, the CNFs
reach their peak density and demonstrate noticeable variations in height
and thickness (Fig. S3). At 300 s of synthesis (Fig. 6 c), the specific
surface reaches its maximum while the height of carbon nanowalls
measures up to 50 nm. At this point, a large portion of CNFs are
removed, presumably due to the redistribution of the electric field in the
plasma layer during the change in growth direction of CNWs [19,57].
This effect is amplified only in thin and long fibers, where the concen­
tration of the electric field is maximum due to the high aspect ratio. As a
result, only fibers of larger diameter and shorter height remain, with
significantly lower density. These remaining CNFs are covered with
carbon nanowalls during the 3600 s of synthesis, as confirmed by SEM
images (Fig. S4). At 450 s of synthesis (Fig. 6 c), the specific surface
6
Carbon 214 (2023) 118332
starts to decrease, the height of carbon nanowalls reaches approximately
80 nm and their structure is highly developed.
Microstructures of the prepared lamellae cross-cut for synthesis
duration of 150, 300, and 450 s are provided in Fig. 7 a, c, e. For these
synthesis durations, non-agglomerated, almost spherical nanocrystal
particles along the entire lamellae length, surrounded by carbon nano­
walls, are displayed in TEM images (Fig. 7 b, d, e). To determine the
average size of particles, we analyzed over 50 nanocrystals for each
duration. The general size of particles for 150, 300, and 450 s amounts to
7, 9, and 8 nm, respectively. Generalized dependence of an average
nanocrystal size including the synthesis duration of 3600 s is shown in
Fig. 7 g. We did not encounter any indication of nanoparticle formation
in HRTEM images for samples at 38 and 75 s of synthesis.
To interpret the n-diamond phase evolution, we analyzed the sample
FFT images with a synthesis duration of 150 s. The results proved that
some of the observed nanoparticles were polycrystalline, composed of
several particles, while the majority were monocrystals that we used for
A.M. Mumlyakov et al.
Fig. 7. The results of studying the n-diamond-CNWs hybrid structure for various synthesis duration by transmission microscopy. TEM image of a cross-section for
synthesis during а) 150 b) 300 c) 450 s. e), d), f) Enlarged view of a selected area. g) Dependence of nanocrystal size on synthesis duration. (A colour version of this
figure can be viewed online.)
phase identification. Using FFT images, we determined interplanar d-
spacing and angles between directions in the lattice for each nano­
particle and compared them with values defined for registered allo­
tropes of carbon according to the ICDD database. Based on this analysis,
we identified three types of carbon modifications in the CNWs-Si
boundary, i.e., n-diamond (JCPDS 43-1104), hexagonal diamond
(JCPDS 19-0269), and cubic diamond (JCPDS 6-0675). Typical electron-
diffraction patterns of the three modifications can be found in Fig. 8.
It is also significant that we examined samples with short synthesis
duration by the GIXRD technique. During the analysis, no distinctive
peaks were observed in the diffraction pattern, which may be due to the
small size of the crystallites. Consequently, our research into the for­
mation of the n-diamond-CNWs hybrid structure has revealed the
presence of other allotropes of carbon in earlier synthesis periods. Other
forms of carbon are most likely to remain thermodynamically unstable
under selected synthesis conditions. Hence, hexagonal diamond and
cubic diamond forms of carbon eventually transform into the n-diamond
as a more thermodynamically stable crystal structure under the selected
synthesis conditions.
4. Discussion
In order to better understand the mechanism of n-diamond nucle­
ation and growth, we conducted a comparative analysis of literature
data related to the plasma-chemical synthesis of carbon materials spe­
cifically focusing on cases where n-diamond or nanodiamond crystallites
are also formed. The obtained results are presented in the summary
Table 1.
As one can see from the Table T1, that the prevalent approach for
producing carbon materials containing n-diamond is MW-PECVD, which
is one of the classical methods for bulk diamond growth. However, a
different technique CCP-PECVD operates with lower energy stored in the
electrodes, leading to a plasma with significantly lower density.
Consequently, a direct comparison between these two methods becomes
challenging due to their inherent differences in plasma characteristics.
In Ref. [58], a CCP PECVD technique employing DC bias was used to
fabricate films consisted of crystalline grains. This method stands out
due to the incorporation of an extra electrode bias, which generates
7
Carbon 214 (2023) 118332
high-energy Ar ions, enabling the sputtering of side carbon phases. The
use of low substrate temperatures is thought to encourage nucleation of
n-diamond by “quenching” the hot ions. Such synthesis conditions
significantly differ from those given in our research, consequently
leading to different mechanisms governing the growth of n-diamond. In
Ref. [60], it is highlighted that the presence of hydrogen gas plays a
crucial role in the formation of n-diamond. Additionally, the study
suggests the existence of intermediate crystalline phases involving CHx
as essential contributors before the actual formation of diamond nuclei.
This finding offers valuable insights into the nucleation pathway of
diamond. In the works [61,62], diamond nucleation starts with the
formation of a-C on the Si surface, followed by the nucleation of abun­
dant spherical diamond clusters, which is attributed to the presence of
large amounts of C2 and CN species in the N2/CH4 plasma. As can be
seen from these works, it is rather difficult to discuss the mechanism of
n-diamond formation due to the differences in the synthesis conditions.
It is also worth noting that the authors of the papers do not use carbon
monoxide in the gas mixture, which we think play key role in the for­
mation of n-diamond.
The closest to us explanation of the nucleation process is presented in
Ref. [63], where the authors provide a comprehensive explanation of the
synthesis of carbon materials in different gas environments using
MW-PECVD. We believe that the introduction of carbon monoxide into
the CH4-Ar-H2 gas mixture accelerate a reaction between excited carbon
monoxide and hydrogen molecules, resulting in the formation of atomic
carbon. This atomic carbon then reacts with hydrogen molecules or
radicals, leading to the generation of various radicals like methyl,
methylene, and methine. These radicals can adsorb on the substrate
surface and thermally decompose into diamond, amorphous carbon and
graphite.
5. Conclusion
The n-diamond-CNWs hybrid structure was successfully synthesized
in capacitively coupled plasma discharge using a mixture of carbon
monoxide, methane, hydrogen, and argon. The study has demonstrated
that feeding CO to the gas mixture is an essential factor in the n-diamond
formation. The addition of carbon monoxide promotes the formation of
A.M. Mumlyakov et al. Carbon 214 (2023) 118332

Fig. 8. The time scale for alteration of the nanodiamonds phase composition. TEM images of a cross-section for a synthesis duration of 150 s and enlarged images of
separate monocrystals with identified allotropic modification. (A colour version of this figure can be viewed online.)

Table 1
An overview table on plasma-chemical synthesis of carbon materials containing n-diamond.
Process parameters

Method Gas Temperature ( C)

◦
Pressure (Torr) Flow rate (sccm) Power (W) The size of diamond crystallites Ref.

CCP PECVD with DC bias CH4/Ar 20 0.5 and 0.3 20:80 and 5:95 250–375 10–40 nm and 200 nm [58]
MW PECVD CH2Cl2/O2/Ar - 20 120:100 300 50 nm, 200 nm [59]
MW PECVD CH4/H2 950 50 2-4:200 1100 10 nm [60]
MW PECVD with DC bias CH4/N2 450 50 Flow ratio <1 1200 5 nm [61]
MW PECVD CH4/Ar 460 120 1:99 1200 >5 nm [24]
MW PECVD CH4, 650 50 4, 1200 8–10 nm [62]
N2, 0-10,
H2 86–96
CCP PECVD CH4, 750 0.25 10, 300 7–10 nm Our work
CO, 7.5,
H2, 2.5,
Ar 200

8
A.M. Mumlyakov et al.
methyl radicals, which are necessary for the growth of diamond crystals.
Selective removal of CNWs from the hybrid structure has allowed to
experimentally obtain X-ray diffraction patterns for the n-diamond
phase, free of overlapping peaks of accompanying phases. The average
size of the n-diamond particles was approximately 10 nm at 3600 s of
synthesis. Systematic experimental studies of the hybrid structure evo­
lution revealed two main features of n-diamond formation. Firstly, an
incubation period when graphite layers and carbon nanowalls form
precedes the formation of nanocrystals and the n-diamond. Secondly,
the n-diamond formation is accompanied by phase transformations from
other allotropic forms of carbon. Having grown the n-diamond-CNWs
hybrid structure on 4-inch silicon wafers offers opportunities for large-
scale integration and wide practical application.
CRediT authorship contribution statement
Mumlyakov A.M: Supervised the research, developed and formatted
the idea of manuscript. Provided TEM analysis of the samples. Pershina
E.A.: Provided XRD analysis of the samples and formatted the body of
the manuscript. Bondareva Ju.V. Provided Raman analysis of the
samples and formatted the body of the manuscript. Nekludova P.A.
Provided SEM analysis of the samples. Shibalova A.A. Provided XRD
analysis Shibalov M.V. Discussion the results and preparing the body of
manuscript. Anufriev Yu.V. Discussion the results and preparing the
body of manuscript. Tagachenkov A.M.: Discussion the results and
preparing the body of manuscript. Tarkhov M.A.: Developed and
formatted the idea of manuscript, supervised the research, prepared and
formatted the body of the manuscript.
Declaration of competing interest
The authors declare that they have no known competing financial
interests or personal relationships that could have appeared to influence
the work reported in this paper.
Acknowledgments
The authors are grateful to Grant No. FFFG-2022-0003 in support of
the Ministry of Science and Higher Education of the Russian Federation.
We thank E. M. Eganova for preparing samples in TEM. Fabrication and
characterization were carried out at large-scale facility complexes for
heterogeneous integration technologies and silicon + carbon
nanotechnologies.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.carbon.2023.118332.
References
[1] Liming Dai (Ed.), Carbon Nanotechnology: Recent Developments in Chemistry,
Physics, Materials Science and Device Applications, 2006.
[2] Anke Krueger, Carbon Materials and Nanotechnology, John Wiley & Sons, 2010.
[3] Lingyu Li, et al., Carbon nanotube induced polymer crystallization: the formation
of nanohybrid shish–kebabs, Polymer 50 (4) (2009) 953–965.
[4] Bohua Wu, et al., Noble metal nanoparticles/carbon nanotubes nanohybrids:
synthesis and applications, Nano Today 6 (1) (2011) 75–90.
[5] Bibekananda De, Niranjan Karak, Recent progress in carbon dot–metal based
nanohybrids for photochemical and electrochemical applications, J. Mater. Chem.
5 (5) (2017) 1826–1859.
[6] Eun Roh, et al., Stretchable, transparent, ultrasensitive, and patchable strain sensor
for human–machine interfaces comprising a nanohybrid of carbon nanotubes and
conductive elastomers, ACS Nano 9 (6) (2015) 6252–6261.
[7] Maosheng Cao, et al., Graphene nanohybrids: excellent electromagnetic properties
for the absorbing and shielding of electromagnetic waves, J. Mater. Chem. C 6 (17)
(2018) 4586–4602.
[8] Jana Vejpravová, Mixed sp2–sp3 nanocarbon materials: a status quo review,
Nanomaterials 11 (10) (2021) 2469.
9
Carbon 214 (2023) 118332
[9] Junhong Chen, Bo Zheng, Ganhua Lu, Vertically-oriented Graphene, 10, Springer
International Publishing Switzerland, 2015, p. 978, 3.
[10] Mineo Hiramatsu, Masaru Hori, Carbon Nanowalls: Synthesis and Emerging
Applications, Springer Science & Business Media, 2010.
[11] Quan Zhou, et al., The controlled growth of graphene nanowalls on Si for Schottky
photodetector, AIP Adv. 7 (12) (2017), 125317.
[12] A.M. Mumlyakov, et al., Fabrication and characterization of quasi-three-
dimensional capacitor structure based on carbon nanowalls, Carbon 184 (2021)
698–705.
[13] E. Stratakis, et al., Three-dimensional carbon nanowall field emission arrays, Appl.
Phys. Lett. 96 (4) (2010), 043110.
[14] Stanislav A. Evlashin, et al., N-doped carbon nanowalls for power sources, Sci. Rep.
9 (1) (2019) 1–7.
[15] Jianjun Wang, et al., Synthesis of carbon nanosheets by inductively coupled radio-
frequency plasma enhanced chemical vapor deposition, Carbon 42 (14) (2004)
2867–2872.
[16] Zhipeng Wang, Mao Shoji, Hironori Ogata, Carbon nanosheets by microwave
plasma enhanced chemical vapor deposition in CH4–Ar system, Appl. Surf. Sci. 257
(21) (2011) 9082–9085.
[17] V.A. Krivchenko, et al., Evolution of carbon film structure during its catalyst-free
growth in the plasma of direct current glow discharge, Carbon 50 (4) (2012)
1477–1487.
[18] Wei Zheng, Xin Zhao, Wenjie Fu, Review of vertical graphene and its applications,
ACS Appl. Mater. Interfaces 13 (8) (2021) 9561–9579.
[19] Mingyao Zhu, et al., A mechanism for carbon nanosheet formation, Carbon 45 (11)
(2007) 2229–2234.
[20] Shingo Kondo, et al., Initial growth process of carbon nanowalls synthesized by
radical injection plasma-enhanced chemical vapor deposition, J. Appl. Phys. 106
(9) (2009), 094302.
[21] L.S. Palatnik, et al., y carbon, Zh. Eksp. Teor. Fiz. 87 (9) (1984) 14–19.
[22] Xuanxuan Ma, et al., Influence of crystallization temperature on fluorescence of n-
diamond quantum dots, Nanotechnology 31 (50) (2020), 505712.
[23] M. Fatow, et al., Carbon modification with the fcc crystal structure, Vacuum 68 (1)
(2002) 75–78.
[24] Joseph P. Thomas, et al., Preferentially grown ultranano c-diamond and n-diamond
grains on silicon nanoneedles from energetic species with enhanced field-emission
properties, ACS Appl. Mater. Interfaces 4 (10) (2012) 5103–5108.
[25] Dejian Dai, Yuanyuan Li, Jiyang Fan, Room-temperature synthesis of various
allotropes of carbon nanostructures (graphene, graphene polyhedra, carbon
nanotubes and nano-onions, n-diamond nanocrystals) with aid of ultrasonic shock
using ethanol and potassium hydroxide, Carbon 179 (2021) 133–141.
[26] Shoichi Endo, et al., X-ray diffraction and transmission-electron microscopy of
natural polycrystalline graphite recovered from high pressure, Phys. Rev. B 49 (1)
(1994) 22.
[27] Kenjiro Yamada, Yasuhiro Tanabe, Akira B. Sawaoka, Allotropes of carbon shock
synthesized at pressures up to 15 GPa, Philos. Mag. A 80 (8) (2000) 1811–1828.
[28] K. Yamada, A.B. Sawaoka, Very small spherical crystals of distorted diamond found
in a detonation product of explosive/graphite mixtures and their formation
mechanism, Carbon 32 (4) (1994) 665–673.
[29] Da Li, et al., Ab initio structure determination of n-diamond, Sci. Rep. 5 (1) (2015)
1–8.
[30] Vadym N. Mochalin, et al., The properties and applications of nanodiamonds, Nat.
Nanotechnol. 7 (1) (2012) 11–23.
[31] J. Xiao, et al., A new phase transformation path from nanodiamond to new-
diamond via an intermediate carbon onion, Nanoscale 6 (24) (2014)
15098–15106.
[32] M.V. Kondrin, et al., Diamond subhydride: unraveling the mystery of “n-diamond”,
CrystEngComm 24 (38) (2022) 6724–6729.
[33] Gael Baldissin, Daniel J. Bull, n-Diamond: dynamical stability of proposed
structures, Diam. Relat. Mater. 34 (2013) 60–64.
[34] Tejinder Singh, et al., First-principles theoretical analysis of pure and
hydrogenated crystalline carbon phases and nanostructures, Chem. Phys. Lett. 474
(1–3) (2009) 168–174.
[35] M.C. Salvadori, et al., Diamond synthesis by microwave plasma chemical vapor
deposition using graphite as the carbon source, Appl. Phys. Lett. 59 (19) (1991)
2386–2388.
[36] Zhipeng Wang, et al., Microwave plasma-induced growth of vertical graphene from
fullerene soot, Carbon 172 (2021) 26–30.
[37] Wenjie Fu, Xin Zhao, Wei Zheng, Growth of vertical graphene materials by an
inductively coupled plasma with solid-state carbon sources, Carbon 173 (2021)
91–96.
[38] Y. Ando, X. Zhao, M. ohkohchi, Production of petal-like graphite sheets by
hydrogen arc discharge, Carbon 35 (1) (1997) 153–158.
[39] Keivan Davami, et al., Synthesis and characterization of carbon nanowalls on
different substrates by radio frequency plasma enhanced chemical vapor
deposition, Carbon 72 (2014) 372–380.
[40] Robert John Nemanich, S.A. Solin, First-and second-order Raman scattering from
finite-size crystals of graphite, Phys. Rev. B 20 (2) (1979) 392.
[41] S. Kurita, et al., Raman spectra of carbon nanowalls grown by plasma-enhanced
chemical vapor deposition, J. Appl. Phys. 97 (10) (2005), 104320.
[42] Andrea C. Ferrari, Denis M. Basko, Raman spectroscopy as a versatile tool for
studying the properties of graphene, Nat. Nanotechnol. 8 (4) (2013) 235–246.
[43] Mineo Hiramatsu, et al., Fabrication of vertically aligned carbon nanowalls using
capacitively coupled plasma-enhanced chemical vapor deposition assisted by
hydrogen radical injection, Appl. Phys. Lett. 84 (23) (2004) 4708–4710.
A.M. Mumlyakov et al.
[44] C.Y. Cheng, K. Teii, Control of the growth regimes of nanodiamond and
nanographite in microwave plasmas, IEEE Trans. Plasma Sci. 40 (7) (2012)
1783–1788.
[45] Hirofumi Yoshimura, et al., Grazing incidence X-ray diffraction study on carbon
nanowalls, Chem. Phys. Lett. 482 (1–3) (2009) 125–128.
[46] J.L. Peng, et al., Nano-crystals of c-diamond, n-diamond and i-carbon grown in
carbon-ion implanted fused quartz, Int. J. Mod. Phys. B 15 (23) (2001) 3107–3123.
[47] A.L. Patterson, The Scherrer formula for X-ray particle size determination, Phys.
Rev. 56 (10) (1939) 978.
[48] L.S. Palatnik, et al., y carbon Sov. Phys." JETP 60 (1984) 520–522.
[49] M. Fatow, et al., Carbon modification with the fcc crystal structure, Vacuum 68 (1)
(2002) 75–78.
[50] Rosalind E. Franklin, The interpretation of diffuse X-ray diagrams of carbon, Acta
Crystallogr. 3 (2) (1950) 107–121.
[51] A. Oberlin, Electron microscopy of thin crystals by PB hirsch, A. Howie, RB
nicholson, DW pashley and MJ whelan, Acta Crystallogr. 21 (3) (1966) 454, 454.
[52] Z.H. Ni, et al., High temperature Raman spectroscopy studies of carbon nanowalls,
J. Raman Spectrosc.: An International Journal for Original Work in all Aspects of
Raman Spectroscopy, Including Higher Order Processes, and also Brillouin and
Rayleigh Scattering 38 (11) (2007) 1449–1453.
[53] André Giese, et al., Synthesis of carbon nanowalls from a single-source metal-
organic precursor, Beilstein J. Nanotechnol. 9 (1) (2018) 1895–1905.
10
Carbon 214 (2023) 118332
[54] Kehan Yu, et al., Growth of carbon nanowalls at atmospheric pressure for one-step
gas sensor fabrication, Nanoscale Res. Lett. 6 (1) (2011) 1–9.
[55] Jiong Zhao, et al., A growth mechanism for free-standing vertical graphene, Nano
Lett. 14 (6) (2014) 3064–3071.
[56] Chia-Hao Tu, et al., Heteroepitaxial nucleation and growth of graphene nanowalls
on silicon, Carbon 54 (2013) 234–240.
[57] Ken Ostrikov, E.C. Neyts, M. Meyyappan, Plasma nanoscience: from nano-solids in
plasmas to nano-plasmas in solids, Adv. Phys. 62 (2) (2013) 113–224.
[58] Gehan Amaratunga, et al., Crystalline diamond growth in thin films deposited from
a CH4/Ar rf plasma, Appl. Phys. Lett. 55 (7) (1989) 634–635.
[59] M.R. Frenklach, W. D, E. K, et al., Homogeneous nucleation of diamond powder in
the gas phase, J. Appl. Phys. 66 (1) (1989) 395–399.
[60] John M. Cowley, et al., Structures of carbon nanocrystals, Chem. Mater. 16 (24)
(2004) 4905–4911.
[61] Kamatchi Jothiramalingam Sankaran, et al., Nitrogen incorporated
ultrananocrystalline diamond microstructures from bias-enhanced microwave N2/
CH4-plasma chemical vapor deposition, Plasma Process. Polym. 13 (4) (2016)
419–428.
[62] K.J. Sankaran, N.H. Tai, I.N. Lin, Microstructural evolution of diamond films from
CH4/H2/N2 plasma and their enhanced electrical properties, J. Appl. Phys. 117 (7)
(2015).
[63] Yukio Saito, et al., Diamond synthesis from CO-H2 mixed gas plasma, J. Mater. Sci.
25 (1990) 1246–1250.