Semiconductor Science and Technology

ACCEPTED MANUSCRIPT

Large-area, thermally-sulfurized WS₂ thin films: Control of

growth direction and use as a substrate for GaN epitaxy
To cite this article before publication: Emroj Hossain et al 2020 Semicond. Sci. Technol. in press https://doi.org/10.1088/1361-6641/ab6bb3

Manuscript version: Accepted Manuscript

Accepted Manuscript is “the version of the article accepted for publication including all changes made as a result of the peer review process,
and which may also include the addition to the article by IOP Publishing of a header, an article ID, a cover sheet and/or an ‘Accepted
Manuscript’ watermark, but excluding any other editing, typesetting or other changes made by IOP Publishing and/or its licensors”

This Accepted Manuscript is © 2020 IOP Publishing Ltd.

During the embargo period (the 12 month period from the publication of the Version of Record of this article), the Accepted Manuscript is fully
protected by copyright and cannot be reused or reposted elsewhere.
As the Version of Record of this article is going to be / has been published on a subscription basis, this Accepted Manuscript is available for reuse
under a CC BY-NC-ND 3.0 licence after the 12 month embargo period.

After the embargo period, everyone is permitted to use copy and redistribute this article for non-commercial purposes only, provided that they
adhere to all the terms of the licence https://creativecommons.org/licences/by-nc-nd/3.0

Although reasonable endeavours have been taken to obtain all necessary permissions from third parties to include their copyrighted content
within this article, their full citation and copyright line may not be present in this Accepted Manuscript version. Before using any content from this
article, please refer to the Version of Record on IOPscience once published for full citation and copyright details, as permissions will likely be
required. All third party content is fully copyright protected, unless specifically stated otherwise in the figure caption in the Version of Record.

View the article online for updates and enhancements.

This content was downloaded from IP address 128.197.229.194 on 18/01/2020 at 06:17

Page 1 of 10 AUTHOR SUBMITTED MANUSCRIPT - SST-106086.R2

1
2
3
4

pt
5
6
7
8
9
Large-area, thermally-sulfurized WS2 thin films:

cri
10
11 Control of growth direction and use as a substrate
12
13 for GaN epitaxy
14
15
16 Emroj Hossain, A Azizur Rahman, Amit P Shah, Bhagyashree

us
17 A Chalke, Arnab Bhattacharya
18
19 Tata Institute of Fundamental Research, Homi Bhabha Road, Colaba, Mumbai
20 400005, India
21 E-mail: arnab@tifr.res.in
22
23
24
25
26
27
28
3 December 2019

an
Abstract. We present a detailed study of the influence of metal seed thickness,
amount of sulfur, sulfurization time and temperature on the morphology of large-area
WS2 thin films prepared by sulfurization of electron-beam evaporated tungsten layers.
dM
29 We show how preferably horizontally or vertically oriented films can be achieved. The
30 WS2 films with horizontal morphology were used as nearly lattice-matched substrates
31 for the growth of GaN layers by MOVPE. While epitaxial films of GaN could be
32 demonstrated, the thin-film WS2 layers show a enhanced propensity for degradation
33 in the strongly reducing and high-temperature MOVPE growth environment.
34
35
36
37
1. Introduction
38
pte

39
40 The limited availability and high cost of bulk gallium nitride (GaN) has led to most GaN
41 layers being deposited on substrates like sapphire [1] or silicon [2] which have a large
42
lattice and thermal mismatch [3]. This results in epilayers with relatively high defect
43
44 densities [4] which consequently impacts device performance [5, 6]. Recently it was
45 shown that several layered transition metal dichalcogenides (TMDCs) have in-plane
46
lattice parameters very close to that of GaN and were proposed as potential lattice-
ce

47
48 matched substrates for III-nitride growth [7]. Preliminary results on the growth of GaN
49 on exfoliated flakes of TMDCs such as MoS2 and WS2 were demonstrated [7, 8]. While
50
such exfoliated flakes of 2D materials have been extensively studied for their novel
51
52 electronic properties [9, 10] their size (typically 5-10 µm) severely limits their utility
Ac

53 for practical commercial applications. There have been several methods proposed for
54 the growth of large-area (cm size or larger) films of TMDCs such as CVD [10, 11]
55
56 MOCVD [12, 13] etc, of which chalcogenization of the corresponding metal film [14] is
57 a relatively easy method. Very few results of the growth of III-nitrides on large-area
58 TMDCs have been reported. Low-temperature ALD of AlN on WS2 was attempted by
59
60
 AUTHOR SUBMITTED MANUSCRIPT - SST-106086.R2 Page 2 of 10

1
2
3
4

pt
5 Chung et al. [15] and recently Yin et al. deposited crack-free AlN films by MOCVD ON
6 WS2 layers formed in-situ from WO3 [16].
7
This work is motivated by the desire to explore the growth of GaN via metal-
8
9 organic vapour phase epitaxy (MOVPE) on large-area WS2 films. A pre-requisite for

cri
10 this is the synthesis of smooth, continuous, oriented WS2 films that could serve as
11 substrates. This process is however complicated by the fact that during the sulfurization
12
13 of a tungsten thin film the WS2 flakes can grow in a horizontal (basal plane parallel
14 to the substrate) or vertical direction or in certain cases even form nanotubes [17].
15 While vertically standing WS2 flakes are candidate materials for efficient and scalable
16

us
17 hydrogen-evolving cathodes [18], for use as a growth substrate the horizontal morphology
18 is required. Previously it was shown that the growth direction of WS2 thin films could be
19 controlled only by metal sheet layer thickness [14]. We first show that direction control
20
21 of the growth of WS2 films can also be independently achieved for different metal layer
22 thicknesses through the appropriate combination of various growth parameters such
23
24
25
26
27
28
an
as temperature, sulfurization time and amount of sulfur. Following this we examine
the stability of WS2 layers in the typical MOVPE growth environment and develop a
procedure for growing GaN on WS2 , and compare the structural and optical properties
of GaN grown on exfoliated and thin-film WS2 . While epitaxial films of GaN could
dM
29 be demonstrated, the degradation of the underlying WS2 is a significant issue that will
30 need to be solved before such substrates can find widespread use.
31
32
33 2. Experimental Details
34
35 WS2 thin film samples were prepared by sulfurization of tungsten layers of different
36
thicknesses deposited by electron beam evaporation (vacuum ∼10-6 mbar, substrate
37
38 temperature ∼150 o C, evaporation rate ∼ 6 nm/min) on cleaned (0001)-oriented
pte

39 sapphire substrates. The deposited tungsten samples along with sulfur (typically in
40
excess of the stoichometrically required amount) were sealed in a HF-cleaned quartz tube
41
42 evacuated to a pressure of ∼ 10-5 mbar. An extensive pump-purge sequence and argon
43 flush were used to remove background impurities prior to sealing. The sulfurization of
44
the tungsten layers was carried out in a single zone furnace. In a series of experiments,
45
46 the tungsten thickness, sulfurization temperature and time, and the amount of sulfur
ce

47 were varied from 6-20 nm, 650-950 o C, 1 min - 2 hr, and 12-54 mg, respectively. After the
48
formation of WS2 , the films were structurally characterized using electron microscopy,
49
50 x-ray diffraction (XRD), Raman and optical absorption spectroscopy.
51 In the next part, WS2 films with a horizontal growth morphology were used as
52 substrates for the epitaxy of GaN in a closed-coupled showerhead MOVPE system.
Ac

53
54 Along with the large-area sulfurized tungsten WS2 films, bulk WS2 was exfoliated on
55 to SiO2 -coated silicon wafers to serve as a control substrate. Trimethylgallium and
56 ammonia were used as group III and group V precursors, respectively. The precursors
57
58 were transported to the reactor using H2 and/or N2 carrier gas. The growth of III-
59 nitrides was monitored in-situ using optical interferometry. The observation of Fabry-
60
2
Page 3 of 10 AUTHOR SUBMITTED MANUSCRIPT - SST-106086.R2

1
2
3
4

pt
5 Perot oscillations due to reflection from growth surface provided real-time information
6 on the growth rate and quality of growth. Several different experiments for the growth
7
of GaN on WS2 were carried out, varying the heat-up conditions, growth temperature,
8
9 growth rate, NH3 flow etc. Various characterization techniques were used to understand

cri
10 the structural and optical properties of the GaN on WS2 .
11
12
13 3. Results and Discussion
14
15
16 W: 6 nm W: 20 nm

us
a W: 14 nm b c
17 Time: 45 min Time: 5 min Time: 60 min
18 T: 950 oC T: 900 oC T: 800 oC
19 S: 0.03 g S: 0.03 g S: 0.03g
20
21
22
23
24
25
26
27
28
100 nm G1 100 nm
an G2 100 nm G3
dM
W: 16 nm d W: 16 nm e W: 16 nm f
29 Time: 30 min Time: 5 min Time: 5 min
30 T: 800 oC T: 950 oC T: 900 oC
31 S: 0.03 g S: 0.03 g S: 0.012 g
32
33
34
35
36
37
38
pte

39 100 nm G4 100 nm G5 100 nm G6

40
41 Figure 1: SEM images of WS2 grown under different conditions and with different
42 metal seed layer thickness. In each image “W” represents the thickness of the tungsten
43 layer, “Time” represents the sulfurization time at the growth temperature (“T”). “S”
44
45 represents the mass of sulfur used. The growth ID is given in the bottom right corners
46 of each image.
ce

47
48
49 Figure 1 shows representative electron micrographs of WS2 films grown under
50 different conditions. As is evident, different choices of growth parameters and metal seed
51
52 layer thickness lead to remarkably different morphologies of WS2 , with predominantly
Ac

53 horizontally or vertically-oriented crystallites being preferentially obtained. For very

54 thin tungsten layers, isolated flakes of WS2 of ∼ 100-200 nm size were observed
55
56
(Fig.1(a)). With increasing thickness of the metal seed layer the flakes tend to join
57 together (Fig. 1(b)) forming semi-continuous horizontal WS2 films. An increase in the
58 metal seed layer usually leads to vertical growth of WS2 due to the compression and
59
60
extrusion of the flakes out of plane [14]. Figure 1(c) shows such a dense arrangement
3
 AUTHOR SUBMITTED MANUSCRIPT - SST-106086.R2 Page 4 of 10

1
2
3
4

pt
5 Table 1: Growth details of WS2 thin films
6 ID Tungsten Sulfurization Growth Sulfur
7
8
thickness (nm) time (min) temp. (o C) (mg)
9 G1 6 45 950 30

cri
10 G2 14 5 900 30
11
12 G3 20 60 800 30
13 G4 16 30 800 30
14 G5 16 5 950 30
15
16 G6 16 5 900 12

us
17
18 nm
19
e
20 Sapphire a
21
22
23

an
Intensity (a.u)

24 b
25
26
27 c
28
dM
WS2(002)
29 WS2(004) WS2(006) WS2(008)
30 d
31 500 nm
32 10 20 30 40 50 60 70 80

33 2q (degree)
34
35
Figure 2: XRD data of (a) sapphire substrate (b) W/sapphire (c) WS2 /sapphire Tg =900
o
36 C, sulfur 0.012 g (sample ID: G6), (d) WS2 /sapphire Tg=950 o C, sulfur 0.03 g (sample
37 ID: G5), (e) AFM image of horizontal WS2 thin film (sample ID: G6)
38
pte

39
40
41
of vertically-oriented flakes. However, for a given tungsten thickness, a careful control
42 of the sulfurization time and the amount of sulfur provided can be used to change the
43 morphology, as shown in the SEM images in Fig. 1(d-f). A significant reduction of
44
45
vertical growth can be achieved by reducing the sulfurization time and amount of excess
46 sulfur used in the process. As seen in Fig. 1(f) reducing the growth time to 5 min. and
ce

47 amount of sulfur to 0.012 g allows one to obtain fully horizontal WS2 thin films for a
48
49
metal seed layer thickness of 16 nm. The transition in morphology from mostly vertical
50 to almost fully horizontal for the same metal seed layer thickness at different growth
51 conditions, as shown in sequence of Figs. 1(d-f), shows that not only the metal seed layer
52
thickness but the sulfurization parameters also play an important role in determining
Ac

53
54 the growth direction of WS2 thin films.
55 Since our interest was in using the WS2 films as a substrate for GaN growth,
56
57
we mostly focus on the characterization of the layers with a predominantly horizontal
58 morphology in the rest of this work. Figure 2(a-d) shows the powder XRD pattern of
59 the underlying sapphire substrate, the evaporated tungsten layer, and WS2 grown under
60
4
Page 5 of 10 AUTHOR SUBMITTED MANUSCRIPT - SST-106086.R2

1
2
3
4

pt
9x103
5 E2g(G) 70 b
a
6 8x103

2LA(M)
A
7 7x103
60

8 50

Tansmission (%)
6x103
B
9
Intensity (a.u)

A1g(M)+LA(M)

cri
5x103 A1g(M)-LA(M) 40

A1g(G)
10
11 4x103 30
LA(M)

4LA(M)
12 3x103 20 C
13 2x103
10
14 3
1x10
15 0

16 0

us
100 200 300 400 500 600 700 800 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5
17 -1 Energy (nm)
Raman shift (cm )
18
19 Figure 3: (a) Room temperature Raman spectrum of the horizontal WS2 sample with
20 532 nm excitation showing LA(M), E2g I , A1g I vibrational modes at 176 cm-1 , 356 cm-1 ,
21
22 418 cm-1 respectively (b) Room temperature optical transmission spectrum of WS2
23
24
25
26
27
28
deposited on double-side-polished sapphire, showing the A, B, C excitons at 2.03, 2.4,
2.75 eV respectively (sample ID: G6)
.
an
dM
29 two different conditions. For the perfectly horizontal WS2 sample, the intensity of the
30 (002) reflection is quite low (Fig. 2(c)) while, for a mixture of horizontal or vertical
31 growth shows (Fig. 2(d)) the intensity of XRD peaks for the (002) reflection and its
32
33 multiples is much higher. This may be just because of the volume of material interacting
34 with the x-ray beam is much lower in the horizontal case as the film is extremely thin.
35 AFM measurements were done on a WS2 thin film with mostly horizontal morphology
36
37 using a Park XE AFM over 2×2 µm2 area in tapping mode to evaluate the surface
38 roughness and morphology of the sample. From the AFM image (Fig. 2(e)) the rms
pte

39 roughness was found to be ∼2.5 nm.

40
41 Raman scattering spectra taken with a WITec α-300 spectrometer using 532 nm
42 excitation show the vibrational modes of the WS2 film. In Fig. 3(a) the LA(M), E2g I ,
43 and A1g I vibrational modes can be seen at 176 cm-1 , 356 cm-1 , and 418 cm-1 respectively
44
45 along with some linear combination of these excitations [19]. No significant differences
46 in the Raman signal of horizontal and vertical WS2 samples were observed. For optical
ce

47 absorption measurements the horizontal WS2 films were grown on double-side-polished

48
49
sapphire to minimize scattering from unpolished back surface. Three excitonic features
50 at energy 2.03, 2.4, and 2.75 eV were observed (figure 3(b)) corresponding to the A,
51 B, and C transitions as reported earlier [20]. Along with the XRD results, the optical
52
Ac

53
characterization of the layers suggests that the WS2 films are of high crystal quality.
54 GaN was grown using MOVPE on the horizontal WS2 large-area films (sample
55 ID: G6) along with exfoliated WS2 flakes on SiO2 -coated Si as control substrates. A
56
57
summary of the conditions used in different attempts is shown in the Table 1 in the
58 supporting information. Initially, the GaN growth was carried out using the conditions
59 proposed by Gupta et al. [7] in their first demonstration of GaN growth on several
60
5
 AUTHOR SUBMITTED MANUSCRIPT - SST-106086.R2 Page 6 of 10

1
2
3
4

pt
5 transition metal dichalcogenides. A two step growth process was used, with heat up
6 under N2 to prevent reduction of the WS2 layer, an initial GaN layer grown at 900
7 o
C for 40 s followed by a second layer grown at 1040 o C for 300 s, which is the usual
8
9 temperature for ”high-temperature” GaN grown on sapphire. (All temperature values

cri
10 mentioned are thermocouple set-point temperatures, as the WS2 sample is kept on a
11 sapphire carrier in the wafer pocket, the actual surface temperature of the sample is
12
13 likely to be ∼40o C lower than the set-point).
14
15 5
16

us
a
17
18 4 GaN
Raman intensity (a.u)

19
20
3
21
22
23
24
25
26
27
28
2

300 400 500

an600 700
GaN

800 900
dM
29 Raman shift (cm-1)
30
8
31
b
32 7 SiO2 c d e
33
Raman intensity (a.u)

34 6
35
5 4 mm 4 mm 4 mm
36
37 4
38 WS2
pte

39 3 GaN
40
2
41
42
43 300 400 500 600 700 800 900
44 Raman shift (cm-1)
45
46 Figure 4: Raman spectra of GaN on (a) thin film WS2 (b) flakes of bulk WS2 exfoliated
ce

47 on to SiO2 -coated Si substrates,(c) Optical microscope image of GaN on exfoliated WS2

48
49
(d) Raman map corresponding to GaN peaks (e) Raman map corresponding to WS2
50 peaks
51
52
Ac

53 The Raman spectra of GaN on thin film WS2 and exfoliated WS2 are shown in
54 Fig. 4A and B. In both cases two intense peaks at 568 cm-1 , 742 cm-1 corresponding
55 to the E2 and E1 (LO) phonon modes of GaN can be clearly seen. However, the two
56
57 characteristic Raman peaks of WS2 at 355 and 420 cm-1 corresponding to the E2g I and
58 A1g I modes are seen only for the sample grown on the exfoliated WS2 . This sample also
59 shows a peak at 520 cm-1 arising from the underlying silicon substrate. Two weaker
60
6
Page 7 of 10 AUTHOR SUBMITTED MANUSCRIPT - SST-106086.R2

1
2
3
4

pt
5x10 4
5 WS2(002) Sapphire(006) g
a
6
7 4x10 4
b
8
Intensity (a.u.)

Intensity (a.u.)
9 3x10 4 Tg=800 oC

cri
c
10
11
d 2x10 4
12
13 GaN(002)
GaN(004)
14 e
1x10 4
Tg=700 oC
15
f
16

us
0
17 10 20 30 40 50 60 70 80 300 400 500 600 700 800
18 2q (deg) Wavelength (nm)
19
20 Figure 5: XRD of (a) thin film WS2 ; GaN on WS2 grown at (b) 550 o C temperature in
21
N2 ambient (c) 700 o C temperature in N2 ambient (d) 800 o C temperature in N2 ambient
22
23
24
25
26
27
28
in H2 ambient an
(e) 700 o C temperature in H2 ambient (f) 800 o C temperature in H2 ambient., (g) PL
spectra of GaN grown on thin film WS2 at temperature 700 o C and 800 o C respectively
dM
29 peaks (Fig. 4(a)) at 535 and 560 cm-1 corresponding to A1 (TO) and E1 (TO) are also
30 observed. These peaks are present in the GaN on exfoliated WS2 samples but were
31
not clearly visible due to the high-intensity of adjacent peaks at 520 and 560 cm-1 .
32
33 This suggests while GaN can be grown on both exfoliated and thin-film WS2 , the thin
34 film WS2 degrades during the growth process. Energy-dispersive X-ray measurements
35
in an SEM showed a much reduced sulfur signal post growth in the thin-film WS2
36
37 samples. The exfoliated flakes from naturally occurring bulk crystals can be expected to
38 be more stable against reduction under hydrogen at high temperature than the sulfurized
pte

39
tungsten films. Compared to an essentially single crystal exfoliated flake that can be
40
41 attacked only from the edges, the thin film WS2 layers have a small grain size with
42 many edges and defects providing a large number of sites that allow for a much quicker
43
reaction with H2 and consequent degradation.
44
45 Raman mapping of GaN on exfoliated WS2 (Fig. 4 c,d and e) of the GaN E2 and
46 WS2 E2g modes showed the spatial correspondence between the substrate and epilayer
ce

47
and confirmed the presence of the WS2 flake in all regions below the GaN.
48
49 Since it was seen that the WS2 films degrade at the typical MOVPE growth
50 temperature for GaN on sapphire - 1040 o C, we next attempted a series of growth
51
runs with the GaN layers grown at a lower temperature range and with different carrier
52
Ac

53 gas to check if we could minimize the WS2 substrate degradation.

54 Fig. 5(a-f) shows a series of XRD profiles of GaN growth on thin-film WS2
55 under different conditions. While the peaks due to the underlying WS2 are seen after
56
57 low-temperature GaN layer grown at 550 o C. there is a swift degradation at higher
58 temperatures even in an N2 ambient. This is most likely due to the reaction of WS2
59 with the much more reactive atomic hydrogen generated by the cracking of NH3 . In
60
7
 AUTHOR SUBMITTED MANUSCRIPT - SST-106086.R2 Page 8 of 10

1
2
3
4

pt
5
6
7 GaN/Al O
8 2 3

GaN/W S
9 1.0
2

cri
10
11 Intensity (a.u.)

12
13
14
15 0.5

16

us
17
18
19
20
0.0
21 -0.4 -0.2 0.0 0.2 0.4

22
23
24
25
26
27
28
an
(arcsec)

Figure 6: Rocking curves for GaN grown on WS2 and sapphire

dM
29 an N2 ambient WS2 films were able to survive only up to 700 o C. As a result, high-
30 temperature GaN growth was not possible. EDX quantification confirmed that there is
31
no sulfur in the sample after GaN growth at 800 o C for 250 seconds. To check if the
32
33 degradation of WS2 layer could be eliminated or reduced by minimizing the NH3 flow,
34 we grew GaN at 800 o C using an NH3 flow of 0.5 SLM instead of 2 SLM, however, in this
35 case too, it was not possible to avoid the degradation of the underlying WS2 . In an H2
36
37 ambient, growth of GaN is much faster, as a result, for the same GaN thickness, there is
38 less degradation of WS2 . Moreover, the quality of GaN layer is better when grown with
pte

39 H2 carrier gas than with N2 . The XRD profiles show sharper peaks of higher intensity
40
41 for GaN on WS2 grown in H2 ambient. But even in an H2 ambient the WS2 underlayer
42 degrades completely at 800 o C after 250 s of GaN growth. The degradation results were
43 also corroborated with EDX and Raman measurement. Hence it seems difficult to be
44
45 able to grow high-quality GaN on thin-film WS2 keeping the substrate intact. To check
46 the quality of GaN layers at these unusually low growth temperatures, room temperature
ce

47 PL measurements were performed, and are shown in Fig.5(g). The PL spectra for GaN
48
49 grown at temperatures of 700 o C and 800 o C in H2 ambient show remarkable differences.
50 The GaN grown at 700 o C has very weak band-edge luminescence, and predominantly
51 yellow-defect emission, whereas the sample grown at 800 o C temperature shows high
52
Ac

53 band-edge and low defect luminescence (even though the underlying WS2 has degraded
54 at that temperature).
55 Clearly, the material quality is not as good as that for state-of-the-art GaN
56
57 on sapphire, which has been optimized over several decades. Fig. 6 compares the
58 (normalized) rocking curve for a standard 1µm-thick GaN layer grown on sapphire by
59 the standard two-step process with that for the GaN grown on WS2 . The FWHM
60
8
Page 9 of 10 AUTHOR SUBMITTED MANUSCRIPT - SST-106086.R2

1
2
3
4

pt
5 of the rocking curve GaN on sapphire is ∼300” and GaN on WS2 is ∼550” with a
6 broad background. While we cannot perform a detailed analysis of the underlying
7
microstructure, this may arise from a combination of local regions of good quality GaN
8
9 where the WS2 is relatively intact and other regions of poor crystallinity. However, this

cri
10 shows the GaN on WS2 is not yet comparable with state-of-the-art GaN on sapphire.
11 In conclusion, we demonstrated that by suitable choice of growth parameters and
12
13 metal seed layer thickness one can control the growth direction of WS2 allowing one
14 to synthesize large-area films with a horizontal morphology. While GaN can be grown
15 by MOVPE on such WS2 films, the substrates are extremely sensitive to degradation
16

us
17 during the GaN growth at high temperature due to hydrogen from decomposition of
18 ammonia, and are not as stable as exfoliated single-crystal WS2 flakes. However, even if
19 the WS2 layer eventually degrades it still seems to be able to allow the nucleation of GaN
20
21 on it allowing for the growth of epitaxial films. However, for potential applications of
22 WS2 /GaN heterostructures, where presence of the underlying layer would be essential,
23
24
25
26
27
28
Acknowlegements
an
these films would have limited utility, and further optimization of the growth process
for single-crystal, large-area WS2 would be required.
dM
29
30 The authors thank the Prof. Sandip Ghosh for discussion in optical measurement,
31 TIFR glass blowing section for the help in quartz tube sealing, Nilesh Kulkarni and V
32
M Chopde and Vilas Mhatre for XRD and R D Bapat for some SEM measurements.
33
34 The work was supported by internal grant 12P0165.
35
36
37 References
38
[1] Tendille F, Hugues M, Vennéguès P, Teisseire M and Mierry P D 2015 Semiconductor Science and
pte

39
40 Technology 30 065001
41 [2] Zhang L, Lee K H, Riko I M, Huang C C, Kadir A, Lee K E, Chua S J and Fitzgerald E A 2017
42 Semiconductor Science and Technology 32 065001
43 [3] Jiang Q, Allsopp D W E, Bowen C R and Wang W N 2013 Semiconductor Science and Technology
44 28 094010
45 [4] Sorokin L M, Myasoedov A V, Kalmykov A E, Kirilenko D A, Bessolov V N and Kukushkin S A
46 2015 Semiconductor Science and Technology 30 114002
ce

47 [5] Liliental-Weber Z 2014 Japanese Journal of Applied Physics 53 100205

48
[6] Shen X Q, Matsuhata H and Okumura H 2005 Applied Physics Letters 86 021912
49
[7] Gupta P, Rahman A A, Subramanian S, Gupta S, Thamizhavel A, Orlova T, Rouvimov S,
50
51 Vishwanath S, Protasenko V, Laskar M R, Xing H G, Jena D and Bhattacharya A 2016 Scientific
52 Reports 6 23708
Ac

53 [8] Susanto I, Tsai C Y, Fachruddin, Rahmiati T, Ho Y T, Tsai P Y and Yu I S 2019 Applied Surface
54 Science 496 143616
55 [9] Wang Q H, Kalantar-Zadeh K, Kis A, Coleman J N and Strano M S 2012 Nature Nanotechnology
56 7 699–712
57 [10] McCreary K M, Hanbicki A T, Jernigan G G, Culbertson J C and Jonker B T 2016 Scientific
58 Reports 6 19159
59
60
9
 AUTHOR SUBMITTED MANUSCRIPT - SST-106086.R2 Page 10 of 10

1
2
3
4

pt
5 [11] Gao Y, Liu Z, Sun D M, Huang L, Ma L P, Yin L C, Ma T, Zhang Z, Ma X L, Peng L M, Cheng
6 H M and Ren W 2015 Nature Communications 6 8569
7 [12] Marx M, Nordmann S, Knoch J, Franzen C, Stampfer C, Andrzejewski D, Kümmell T, Bacher G,
8 Heuken M, Kalisch H and Vescan A 2017 Journal of Crystal Growth 464 100–104
9 [13] Lin Y C, Jariwala B, Bersch B M, Xu K, Nie Y, Wang B, Eichfeld S M, Zhang X, Choudhury T H,

cri
10 Pan Y, Addou R, Smyth C M, Li J, Zhang K, Haque M A, Fölsch S, Feenstra R M, Wallace
11 R M, Cho K, Fullerton-Shirey S K, Redwing J M and Robinson J A 2018 ACS Nano 12 965–975
12 [14] Jung Y, Shen J, Liu Y, Woods J M, Sun Y and Cha J J 2014 Nano Letters 14 6842–6849
13 [15] Chung J W and Ohuchi F 1996 MRS Proceedings 449 379
14
[16] Yin Y, Ren F, Wang Y, Liu Z, Ao J, Liang M, Wei T, Yuan G, Ou H, Yan J, Yi X, Wang J and
15
Li J 2018 Materials 11
16

us
17 [17] Hossain E, Rahman A A, Bapat R D, Parmar J B, Shah A P, Arora A, Bratschitsch R and
18 Bhattacharya A 2018 Nanoscale 10 16683–16691
19 [18] Yan Y, Xia B, Li N, Xu Z, Fisher A and Wang X 2015 J. Mater. Chem. A 3 131–135
20 [19] Berkdemir A, Gutiérrez H R, Botello-Méndez A R, Perea-López N, Elı́as A L, Chia C I, Wang B,
21 Crespi V H, López-Urı́as F, Charlier J C, Terrones H and Terrones M 2013 Scientific Reports 3
22
23
24
25
26
27
28
1755

an
[20] Zhu B, Chen X and Cui X 2015 Scientific Reports 5 9218
dM
29
30
31
32
33
34
35
36
37
38
pte

39
40
41
42
43
44
45
46
ce

47
48
49
50
51
52
Ac

53
54
55
56
57
58
59
60
10