-
This content has been downloaded from IOPscience. Please scroll
down to see the full text.
Download details:
IP Address: 115.145.196.174
This content was downloaded on 22/06/2017 at 06:40
Please note that terms and conditions apply.
Highly uniform wafer-scale synthesis of -MoO3 by plasma enhanced
chemical vapor
deposition
View the table of contents for this issue, or go to the journal
homepage for more
2017 Nanotechnology 28 175601
(http://iopscience.iop.org/0957-4484/28/17/175601)
Home Search Collections Journals About Contact us My
IOPscience
You may also be interested in:
Wafer-scale monolayer MoS2 grown by chemical vapor deposition
using a reaction of MoO3 and H2S
Youngchan Kim, Hunyoung Bark, Gyeong Hee Ryu et al.
The important role of water in growth of monolayer transition
metal dichalcogenides
Christoph Kastl, Christopher T Chen, Tevye Kuykendall et al.
Controlled MoS2 layer etching using CF4 plasma
Min Hwan Jeon, Chisung Ahn, HyeongU Kim et al.
Wafer-scale production of highly uniform two-dimensional MoS2 by
metal-organic chemical vapor
deposition
TaeWan Kim, Jihun Mun, Hyeji Park et al.
Direct synthesis of large-area continuous ReS2 films on a
flexible glass at low temperature
Youngchan Kim, Byunggil Kang, Yongsuk Choi et al.
Large-area, high-quality monolayer graphene from polystyrene at
atmospheric pressure
Junqi Xu, Can Fu, Haibin Sun et al.
Synthesis of -MoO3 nano-flakes by dry oxidation of RF sputtered
Mo thin films and their
application in gas sensing
Priyanka Dwivedi, Saakshi Dhanekar and Samaresh Das
Synthesis of wafer-scale hexagonal boron nitride monolayers free
of aminoborane nanoparticles by
chemical vapor deposition
Jaehyun Han, Jun-Young Lee, Heemin Kwon et al.
http://iopscience.iop.org/page/termshttp://iopscience.iop.org/0957-4484/28/17http://iopscience.iop.org/0957-4484http://iopscience.iop.org/http://iopscience.iop.org/searchhttp://iopscience.iop.org/collectionshttp://iopscience.iop.org/journalshttp://iopscience.iop.org/page/aboutioppublishinghttp://iopscience.iop.org/contacthttp://iopscience.iop.org/myiopsciencehttp://iopscience.iop.org/article/10.1088/0953-8984/28/18/184002http://iopscience.iop.org/article/10.1088/2053-1583/aa5f4dhttp://iopscience.iop.org/article/10.1088/0957-4484/26/35/355706http://iopscience.iop.org/article/10.1088/1361-6528/aa6958http://iopscience.iop.org/article/10.1088/1361-6528/aa6958http://iopscience.iop.org/article/10.1088/2053-1583/aa6533http://iopscience.iop.org/article/10.1088/1361-6528/aa6235http://iopscience.iop.org/article/10.1088/0268-1242/31/11/115010http://iopscience.iop.org/article/10.1088/0268-1242/31/11/115010http://iopscience.iop.org/article/10.1088/0957-4484/25/14/145604http://iopscience.iop.org/article/10.1088/0957-4484/25/14/145604
-
Highly uniform wafer-scale synthesis ofα-MoO3 by plasma enhanced
chemical vapordeposition
Hyeong-U Kim1,5, Juhyun Son2,5, Atul Kulkarni2, Chisung Ahn3,Ki
Seok Kim4, Dongjoo Shin2, Geun Yong Yeom1,4 and Taesung Kim1,2
1 SKKU Advanced Institute of Nanotechnology (SAINT),
Sungkyunkwan University, Suwon, Republic ofKorea2 School of
Mechanical Engineering, Sungkyunkwan University, Suwon, Republic of
Korea3 Institute of Advanced Machinery and Technology, Sungkyunkwan
University, Suwon, Republic of Korea4 School of Advanced Materials
Science and Engineering, Sungkyunkwan University, Suwon, Republic
ofKorea
E-mail: [email protected]
Received 4 December 2016, revised 24 February 2017Accepted for
publication 15 March 2017Published 3 April 2017
AbstractMolybdenum oxide (MoO3) has gained immense attention
because of its high electron mobility,wide band gap, and excellent
optical and catalytic properties. However, the synthesis of
uniformand large-area MoO3 is challenging. Here, we report the
synthesis of wafer-scale α-MoO3 byplasma oxidation of Mo deposited
on Si/SiO2. Mo was oxidized by O2 plasma in a plasmaenhanced
chemical vapor deposition (PECVD) system at 150 °C. It was found
that thesynthesized α-MoO3 had a highly uniform crystalline
structure. For the as-synthesized α-MoO3sensor, we observed a
current change when the relative humidity was increased from 11%
to95%. The sensor was exposed to different humidity levels with
fast recovery time of about 8 s.Hence this feasibility study shows
that MoO3 synthesized at low temperature can be utilized forgas
sensing applications by adopting flexible device technology.
Supplementary material for this article is available online
Keywords: MoO3, plasma enhanced chemical vapor deposition
(PECVD), wafer-scale, Ramanspectroscopy, humidity sensor
(Some figures may appear in colour only in the online
journal)
1. Introduction
Two-dimensional (2D) materials have attracted immenseattention
owing to their unique chemical and physical prop-erties [1]. For
example, graphene, the first 2D atomic crystal,possesses excellent
material properties such as mechanicalstiffness, strength,
elasticity, high electrical and thermalconductivities [2]. However,
graphene lacks semiconductingcharacteristics because of its zero
band gap. In the past fewyears, it has been revealed that
transition metal dichalcogen-ides can replace graphene owing to
their tunable band gap.
Especially, MoS2 has a tunable band gap, which can be tunedfrom
an indirect (1.29 eV) to a direct (1.90 eV) band gapdepending on
the number of layers. However, it has a lowcarrier mobility despite
its tunable band gap [3]. On the otherhand, 2D semiconducting metal
oxides (e.g., molybdenumoxide (MoO3)) have high dielectric constant
(high-k), highelectron mobility [4], and a wide band gap (3.2 eV).
More-over, MoO3 is relatively abundant in nature and shows
opticaland catalytic properties.
2D material synthesis methods are classified into physicaland
chemical methods. Mechanical exfoliation is a typicalphysical
method, which yields 2D materials with high qualityand small size.
Chemical vapor deposition (CVD) is a popular
Nanotechnology
Nanotechnology 28 (2017) 175601 (6pp)
https://doi.org/10.1088/1361-6528/aa67d1
5 These authors contributed equally.
0957-4484/17/175601+06$33.00 © 2017 IOP Publishing Ltd Printed
in the UK1
mailto:[email protected]://doi.org/10.1088/1361-6528/aa67d1https://doi.org/10.1088/1361-6528/aa67d1http://crossmark.crossref.org/dialog/?doi=10.1088/1361-6528/aa67d1&domain=pdf&date_stamp=2017-04-03http://crossmark.crossref.org/dialog/?doi=10.1088/1361-6528/aa67d1&domain=pdf&date_stamp=2017-04-03
-
chemical method for the synthesis of uniform and large-areathin
films. Graphene can be introduced a band gap bynanoribbons but the
required complex synthesis processeswith significant loss in
carrier mobility [5]. Therefore, layeredmaterials (such as MoO3 and
MoS2) synthesized by varioustypes of CVD using different types of
sources and precursorshave been reported to be potential
alternatives to graphene[6]. However, the synthesis of these
layered materials is atime-consuming process and requires high
temperatures(>500 °C). To overcome this problem, the
atmosphericpressure CVD technology was reported for the preparation
oftransition metal oxide thin films at low temperature [7] andalso
metal organic CVD has been developed for the synthesisof MoS2 thin
films. In this method, two source gases [8] and aprecursor film
with one source gas are used for reducing theprocess temperature
[9]. On the other hand, plasma enhancedCVD (PECVD) is used for the
synthesis of MoS2 thin films atlow temperatures (
-
performance of the MoO3-based sensors were determined
bymeasuring current using semiconductor characterization sys-tem
(KEITHLEY 4200-SCS) in real time. Water vapor wasinjected using an
atomizer into the chamber. The RH levelwas controlled with a two
channel mass flow controller at1 l min−1, one for water vapor and
one for dry air. The lowcapacity vacuum pump was connected to
outlet of thechamber at 1 l min−1. The RH level in chamber was
measuredby hygrometer, which is placed near the sensor device
undertest. When the sensor response reached a saturated value
forset RH level, the inlet was closed for removing the humidityin
the chamber. All measurements were carried out at roomtemperature
(18 °C–21 °C) at different level of RH. Theresponse time and
recovery time are defined on the basis ofthe time required for the
sensor to achieve 90% of the totalcurrent change.
3. Result and discussion
The wafer-scale MoO3 obtained by PECVD of the Modeposited
Si/SiO2 wafer was characterized by Ramanspectroscopy, XPS, HR-TEM,
and EDS.
Figure 2(a) shows the Raman spectrum of the synthe-sized
wafer-scale MoO3 at an excitation wavelength of532 nm and in the
100–1100 cm−1 range. The obtained peakscorrespond to the vibration
of the ordered structure of MoO3.The peaks observed at 300, 671,
821, and 999 cm−1 corre-spond to the characteristic peaks of MoO3.
The band at300 cm−1 (B2g, B3g) is attributed to the Mo3–O
bendingmode. The band at 671 cm−1 (B2g, B1g) is attributed to
theasymmetric stretching of the triply connected bridge-oxygenMo3–O
bridge entity along the c-axis. This bridge-oxygenMo3–O bridge
entity was formed by edge sharing of oxygen(Mo3–O) with the three
adjacent octahedra. The 821 cm
−1
(Ag, B1g) band was observed only under the 30 nm conditionand
corresponded to the symmetric stretching of the terminaloxygen
atoms or the doubly connected bridge-oxygen Mo–O–Mo entity. Mo–O–Mo
was formed by corner sharing ofoxygen with two octahedra. The 999
cm−1 (Ag, B1g) bandcorresponds to the symmetric stretching of the
terminaloxygen atoms (Mo6+=O), which are responsible for thelayered
structure of α-MoO3 [12, 13]. β-MoO3 shows twomajor vibrational
bands at 775 and 850 cm−1 [14]. However,these peaks were not
observed in the PECVD-synthesizedMoO3, as seen from figure 2(a).
Further, the wafer-scaleuniformity of the synthesized MoO3 was
investigated byobtaining the Raman spectra at three different
locations(figure 2(b)). Moreover, for Raman mapping three peaks
wereselected 671 cm−1 (Mo3–O), 821 cm
−1 (Mo2–O) and999 cm−1 (Mo=O) as shown in figure S4. It is
observed thatall the three peaks show uniformity in the 20×20
μm2
mapping area. The Raman spectra and mapping clearly revealthat
MoO3 was deposited uniformly on the wafer. The che-mical bonding
state and surface composition of the synthe-sized MoO3 were
investigated by XPS. The XPS scans forMo and O binding energies of
the MoO3 layer are shown infigure 2(c). The Mo 3d5/2 and 3d3/2
peaks were observed at
234.0 and 237.2 eV, respectively. The energy differencebetween
the 3d5/2 and 3d3/2 peaks was 3.2 eV. The lowbinding energy
component located 533.2 eV (O 1s)(figure 2(d)) originated from the
lattice oxygen of MoO3. Theatomic ratio of Mo to O was 1:3.10 and
it is very similar withstoichio-metrically expected ratio value of
3. Additionally,Mo and O binding energy were compared for their
actualintensity as depicted in online supplementary figure S5.
Thus,the XPS results confirmed that MoO3 was successfully
syn-thesized [15].
The structure of the synthesized MoO3 was furtherinvestigated by
HR-TEM. Figure 2(e) shows the cross-sectionimage obtained by
exposing the MoO3 sample to a focusedion beam (FIB). The inset
shows the EDS spectrum of MoO3(30 nm Mo condition). The EDS
analysis confirmed the pre-sence of O and Mo atoms (denoted by
yellow circles). Thestoichiometric ratio (Mo:O) estimated from the
EDS spectrumwas close to 3. The result further confirmed the
formation ofMoO3. The preferential (100) orientations of MoO3 shown
infigure 2(f) were used to determine the lattice spacing of
thesynthesized MoO3 with bright field mode of the scanningTEM
(STEM) mode. The lattice spacing was similar to thatreported for
MoO3 in previous reports. A lattice spacing of0.38 nm was obtained,
corresponding to the (100) plane of theorthorhombic phase of MoO3
[12].
Previous studies, PECVD are used to deposit by theadsorption and
reaction of dissociated gases with precursorson the substrate [16].
In this study, a Mo film was depositedon the Si/SiO2 substrate and
the resulting Mo-depositedSi/SiO2 was used as the precursor. The
precursor was thenoxidized by O2 for the synthesis of MoO3. We have
used thistechnique in our previous studies for the synthesis of
MoS2thin films [10, 17]. The mechanism for the synthesis of
wafer-scale MoO3 is illustrated in figure 3. During the
plasmaactivation, the Ar gas present in the chamber was
ionized.Herein, the specific reason to use Ar is to produce
moreoxygen ions in the sheath area of plasma, this can further
helpto penetrate O2 plasma into the 30 nm thick Mo film toachieve
MoO3.
This ionization of Ar gas led to a charge transfer
reactionbetween the resulting Ar ions and O2 because of
collision.This collision resulted in the conversion of O2 to
+O ,2 and asheath layer was formed above the Mo-deposited wafer.
AsO2+ carried a positive charge and the Mo wafer was chargeless
(i.e. it was grounded), an electric field was generated and
+O2reacted with Mo on the Si/SiO2 substrate. Because of
thisreaction, it was possible to oxidize Mo at a low
temperature(i.e. 150 °C). Further, this reaction continued for
approxi-mately 90 min to finally yield MoO3 on the Si/SiO2wafer
[18].
The dependence of current versus RH is shown in figure 4.The
sensing characteristics of the optimized MoO3 film devicewere
measured current change by increasing or decreasing RHat the room
temperature. It is reported that, the water moleculeacts as an
electron donor (n-type doping), hence with theexposure of RH, the
Fermi level of MoO3 shifted toward theconduction band by decreasing
the resistance of MoO3,resulting in increase in the current [19].
It is obvious that the
3
Nanotechnology 28 (2017) 175601 H-U Kim et al
-
Figure 2. (a) Raman spectra of MoO3 (for 30 nm Mo condition).
(b) Raman spectra of three points on the synthesized MoO3 (for 30
nm Mocondition). (c) XPS spectra of Mo 3d. (d) O 1s spectrum (for
30 nm Mo condition). (e) TEM cross-section image of the sample
obtained byexposing the sample to a FIB; the inset shows the EDS
spectrum of MoO3. (f) STEM mode image of MoO3.
4
Nanotechnology 28 (2017) 175601 H-U Kim et al
-
response and recovery behavior is one of the most
importantcharacteristics for evaluating the performance of
humiditysensors. The stability of the MoO3 gas sensor was
checkedwith two cycles. The stability of the sensor was performed
inconjugative cycles between two RH levels, by increasing theRH
from 11% to 95%, as shown in figure 4(a). It is observedthat, for
RH 95%, the response and recovery time are ∼195 and∼15 s,
respectively. Figure 4(b) presents the current–time plotobtained by
exposing the MoO3 sensor to the RH level such as95%, 71%, and 45%
and returned to low humidity level of11%. At low humidity (11% RH),
only a few water moleculeswill be present. At 71% RH, the response
time and recoverytime were ∼185 s and ∼11 s, respectively. At a
lower RH of45%, the response time and recovery time were ∼175 s
and∼8 s, respectively. As observed, the sample had a quickrecovery
to the RH change. Table 1 shows the comparativeresults for various
kinds of metal oxide nanostructures withdifferent morphology, in
terms of response time and recoverytime, for comparing the
performance of the humidity sensorfabricated in this study [20–22].
For the MoO3 sensor fabricatedin this study, the recovery time was
very low because theadsorbed water molecules desorbed quickly
during humiditydecrease from high to low levels. Hence, low
temperature
Figure 3. Schematic of the synthesis of MoO3 on Si/SiO2, showing
the film formation mechanism.
Figure 4. A typical current–time plot of the MoO3-based RH
sensor.(a) Response of the sensor during the RH switching between
11%and 95% for repeated cycles. (b) RH sensing response at
differenthumidity levels at room temperature.
Table 1. The comparative data of various kinds of
nanomaterialswith MoO3 film related to response/recovery time.
Material TypeResponsetime
Recoverytime References
CuO NW 120 120 [19]SnO2 NW 120–170 20–60 [20]TiO2 Nanotube 100
190 [21]MoO3 Film 175 8 Present work
5
Nanotechnology 28 (2017) 175601 H-U Kim et al
-
synthesized MoO3 can be suitable candidate for the
sensitivehumidity/gas sensor application and can be applied
flexibledevice technology.
4. Conclusions
In the present study, wafer-scale α-MoO3 was
successfullysynthesized on a Si/SiO2 wafer using a PECVD system
at150 °C. To the best of our knowledge, this is the first
attemptmade till date to synthesize wafer-scale MoO3 at such a
lowtemperature. The synthesized α-MoO3 characterization byRaman
spectroscopy, HR-TEM, XPS, and EDS confirms thesuccessful synthesis
which are comparable to other synthesismethods. The Raman
spectroscopy results revealed that the α-MoO3 was deposited
uniformly on the Si/SiO2 wafer. Thesynthesized α-MoO3 responded
well for the RH from 11%–95%. Hence this feasibility study shows
that MoO3 synthe-sized at low temperature can be utilized for the
gas sensingapplications by adopting flexible device technology.
Acknowledgments
This work was supported by the International S&T
CooperationProgram of China (ISTCP) (2014DFG52760) and
NRF-2013K1A3A1A20046951. This work was also supported by aNational
Research Foundation of Korea (NRF) grant funded bythe Korean
government (NRF-2015R1D1A1A01057861).
References
[1] Bonaccorso F et al 2010 Graphene photonics
andoptoelectronics Nat. Photon. 4 611–22
[2] Geim A K and Grigorieva I V 2013 Van der
Waalsheterostructures Nature 499 419–25
[3] Lee C et al 2010 Anomalous lattice vibrations of
single-andfew-layer MoS2 ACS Nano 4 2695–700
[4] Balendhran S et al 2013 Enhanced charge carrier mobility
intwo‐dimensional high dielectric molybdenum oxide Adv.Mater. 25
109–14
[5] Jiao L et al 2010 Facile synthesis of high-quality
graphenenanoribbons Nat. Nanotechnol. 5 321–5
[6] Jeon J et al 2015 Layer-controlled CVD growth of
large-areatwo-dimensional MoS2 films Nanoscale 7 1688–95
[7] Gesheva K et al 2014 APCVD transition metal
oxides—functional layers in ‘smart windows’ J. Phys.: Conf. Ser.559
012002
[8] Mun J et al 2016 Low-temperature growth of layeredmolybdenum
disulphide with controlled clusters Sci. Rep. 621854
[9] Lee Y et al 2014 Synthesis of wafer-scale uniformmolybdenum
disulfide films with control over the layernumber using a gas phase
sulfur precursor Nanoscale 62821–6
[10] Ahn C et al 2015 Low‐temperature synthesis of
large‐scalemolybdenum disulfide thin films directly on a
plasticsubstrate using plasma‐enhanced chemical vapor
depositionAdv. Mater. 27 5223–9
[11] Galatsis K et al 2001 Semiconductor MoO3–TiO2 thin film
gassensors Sensors Actuators B 77 472–7
[12] Kalantar-Zadeh K et al 2010 Synthesis of
nanometre-thickMoO3 sheets Nanoscale 2 429–33
[13] Lupan O et al 2014 Versatile growth of
freestandingorthorhombic α-molybdenum trioxide
nano-andmicrostructures by rapid thermal processing for
gasnanosensors J. Phys. Chem. C 118 15068–78
[14] Haro-Poniatowski E et al 1998 Laser-induced
structuraltransformations in MoO3 investigated by Ramanspectroscopy
J. Mater. Res. 13 1033–7
[15] Lee Y J et al 2009 Chemical vapour transport synthesis
andoptical characterization of MoO3 thin films J. Phys. D:
Appl.Phys. 42 115419
[16] Kim H et al 2013 Synthesis of MoS2 atomic layer usingPECVD
ECS Trans. 58 47–50
[17] Kim H-U et al 2015 In situ synthesis of MoS2 on a
polymerbased gold electrode platform and its application
inelectrochemical biosensing RSC Adv. 5 10134–8
[18] Floquet N et al 1992 Structural and morphological studies
ofthe growth of MoO3 scales during high-temperatureoxidation of
molybdenum Oxid. Met. 37 253–80
[19] Anderson J H Jr and Parks G A 1968 Electrical conductivity
ofsilica gel in the presence of adsorbed water J. Phys. Chem.72
3662–8
[20] Wang S-B et al 2012 CuO nanowire-based humidity sensorIEEE
Sens. J. 12 1884–8
[21] Kuang Q et al 2007 High-sensitivity humidity sensor based
ona single SnO2 nanowire J. Am. Chem. Soc. 129 6070–1
[22] Zhang Y et al 2008 Synthesis and characterization of
TiO2nanotubes for humidity sensing Appl. Surf. Sci. 254 5545–7
6
Nanotechnology 28 (2017) 175601 H-U Kim et al
https://doi.org/10.1038/nphoton.2010.186https://doi.org/10.1038/nphoton.2010.186https://doi.org/10.1038/nphoton.2010.186https://doi.org/10.1038/nature12385https://doi.org/10.1038/nature12385https://doi.org/10.1038/nature12385https://doi.org/10.1021/nn1003937https://doi.org/10.1021/nn1003937https://doi.org/10.1021/nn1003937https://doi.org/10.1002/adma.201203346https://doi.org/10.1002/adma.201203346https://doi.org/10.1002/adma.201203346https://doi.org/10.1038/nnano.2010.54https://doi.org/10.1038/nnano.2010.54https://doi.org/10.1038/nnano.2010.54https://doi.org/10.1039/C4NR04532Ghttps://doi.org/10.1039/C4NR04532Ghttps://doi.org/10.1039/C4NR04532Ghttps://doi.org/10.1088/1742-6596/559/1/012002https://doi.org/10.1038/srep21854https://doi.org/10.1038/srep21854https://doi.org/10.1039/c3nr05993fhttps://doi.org/10.1039/c3nr05993fhttps://doi.org/10.1039/c3nr05993fhttps://doi.org/10.1039/c3nr05993fhttps://doi.org/10.1002/adma.201501678https://doi.org/10.1002/adma.201501678https://doi.org/10.1002/adma.201501678https://doi.org/10.1016/S0925-4005(01)00737-7https://doi.org/10.1016/S0925-4005(01)00737-7https://doi.org/10.1016/S0925-4005(01)00737-7https://doi.org/10.1039/B9NR00320Ghttps://doi.org/10.1039/B9NR00320Ghttps://doi.org/10.1039/B9NR00320Ghttps://doi.org/10.1021/jp5038415https://doi.org/10.1021/jp5038415https://doi.org/10.1021/jp5038415https://doi.org/10.1557/JMR.1998.0144https://doi.org/10.1557/JMR.1998.0144https://doi.org/10.1557/JMR.1998.0144https://doi.org/10.1088/0022-3727/42/11/115419https://doi.org/10.1149/05808.0047ecsthttps://doi.org/10.1149/05808.0047ecsthttps://doi.org/10.1149/05808.0047ecsthttps://doi.org/10.1039/C4RA14839Hhttps://doi.org/10.1039/C4RA14839Hhttps://doi.org/10.1039/C4RA14839Hhttps://doi.org/10.1007/BF00665191https://doi.org/10.1007/BF00665191https://doi.org/10.1007/BF00665191https://doi.org/10.1021/j100856a051https://doi.org/10.1021/j100856a051https://doi.org/10.1021/j100856a051https://doi.org/10.1109/JSEN.2011.2180375https://doi.org/10.1109/JSEN.2011.2180375https://doi.org/10.1109/JSEN.2011.2180375https://doi.org/10.1021/ja070788mhttps://doi.org/10.1021/ja070788mhttps://doi.org/10.1021/ja070788mhttps://doi.org/10.1016/j.apsusc.2008.02.106https://doi.org/10.1016/j.apsusc.2008.02.106https://doi.org/10.1016/j.apsusc.2008.02.106
1. Introduction2. Experimental3. Result and discussion4.
ConclusionsAcknowledgmentsReferences