WS2 Lateral Heterostructures

Subscriber access provided by the University of Exeter

Communication 2

2

Diffusion-Mediated Synthesis of MoS/WS Lateral Heterostructures Kevin Bogaert, Song Liu, Jordan Chesin, Denis Titow, Silvija Gradecak, and Slaven Garaj Nano Lett., Just Accepted Manuscript • DOI: 10.1021/acs.nanolett.6b02057 • Publication Date (Web): 20 Jul 2016 Downloaded from http://pubs.acs.org on July 20, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Nano Letters is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

Diffusion-Mediated Synthesis of MoS2/WS2 Lateral Heterostructures

1 2

3

Kevin Bogaert1, 2†, Song Liu1†, Jordan Chesin2, Denis Titow2, 3, Silvija Gradečak2*,

4

Slaven Garaj1, 4, 5*

5

1

6

of Singapore, 6 Science Drive 2, Singapore, 117546

7

2

8

Technology, Cambridge, MA 02139

9

3

Department of Chemistry, Justus-Liebig-University, Giessen, Germany, 35394

10

4

Department of Physics, National University of Singapore, 2 Science Drive 3, Singapore,

11

117542

12

5

13

Singapore, 117575

Centre for Advanced 2D Materials and Graphene Research Centre, National University

Department of Materials Science and Engineering, Massachusetts Institute of

Department of Bioengineering, National University of Singapore, 9 Engineering Drive 1,

14 15

†

These authors contributed equally to this work.

16 17 18 19

Abstract: Controlled growth of two-dimensional transition metal dichalcogenide (TMD)

20

lateral heterostructures would enable on-demand tuning of electronic and optoelectronic

21

properties in this new class of materials. Prior to this work, compositional modulations in

22

lateral TMD heterostructures have been considered to depend solely on the growth

23

chronology. We show that in-plane diffusion can play a significant role in the chemical

24

vapor deposition of MoS2/WS2 lateral heterostructures leading to a variety of non-trivial

25

structures whose composition does not necessary follow the growth order. Optical,

26

structural, and compositional studies of TMD crystals captured at different growth

1 ACS Paragon Plus Environment

Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 2 of 22

1

temperatures and in different diffusion stages suggest that compositional mixing vs.

2

segregation are favored at high and low growth temperatures, respectively. The observed

3

diffusion mechanism will expand the realm of possible lateral heterostructures,

4

particularly ones that cannot be synthesized using traditional methods.

5

6

Keywords: Transition metal dichalcogenides, heterostructures, diffusion, chemical vapor deposition

7 8 9 10

Table of Contents Graphic

11 12

13


Page 3 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

Two-dimensional (2D) transition metal dichalcogenides (TMDs) are atomically thin

2

direct bandgap semiconductors with high carrier mobilities, visible light direct bandgaps,

3

high on/off current ratios, and unique excitonic, valleytronic, and spintronic properties.1–4

4

TMDs have a hexagonal graphene-like structure consisting of transition metal atoms (M

5

= Mo, W, etc.) on non-neighboring sites covalently bonded to a stacked pair of chalcogen

6

atoms (X = S, Se, or Te) on the alternate sites resulting in the MX2 chemical formula.

7

Because of the similarity in their crystal structure and lattice constants, alloyed TMD

8

materials across the entire compositional range have been theoretically predicted5,6 – and

9

several of these have been synthesized5,7–13 – allowing for continuous bandgap

10

modulations.

11

In analogy to conventional semiconductors, more complex heterostructures based on 2D

12

TMDs would expand the realm of possible device architectures, particularly because

13

different TMD materials could have opposing levels of intrinsic doping. Both vertical and

14

lateral heterostructures have been suggested and recently demonstrated.14–22 The vertical

15

heterostructures, consisting of stacked layers of different materials, can be made via

16

growth or via mechanical exfoliation and stacking. The lateral heterostructures,

17

consisting of in-plane domains of different materials, must be grown to ensure proper

18

chemical and structural matching at the interface. Despite these additional requirements,

19

lattice continuity of lateral heterostructures yields superior in-plane electronic properties,

20

including excellent current rectification behavior and photocurrent generation

21

characteristics.16,18 Furthermore, lateral in-plane heterostructures can act as intrinsic p-n

22

junctions and have arbitrarily tuned bandgaps via alloying. Several chalcogen-changing

23

(e.g. MoS2/MoSe2 or WS2/WSe2), metal-changing (e.g. MoS2/WS2 or MoSe2/WSe2), and

24

dual-changing (e.g. MoS2/WSe2) lateral heterostructures grown by chemical vapor

25

deposition (CVD) have been recently reported.14–20

26

Understanding the mechanisms that govern the growth of pure and multi-phase TMDs

27

will be critical for engineering more advanced heterostructures with controlled properties.


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 22

1

In one-step growth processes, two metallic precursors are simultaneously present in the

2

reaction chamber during the growth; after the growth the material with higher vapor

3

pressure and nucleation rate is found in the core of the 2D crystal.10,11,14,15,19,22,23 In

4

two-step growths, the core domain of the first material is grown first, followed by the

5

exchange of precursors and the growth of the ring of the second material.16–18,20,24 It has

6

been suggested that the growth chronology is the determining factor in radial domain

7

distribution and that each subsequent growth step simply adds material onto the existing

8

crystal edge. In this work, however, we show that the growth chronology is not the sole –

9

and sometimes not the governing – factor that determines the resulting structure. We

10

report a two-step growth process in which MoS2 is found in the core and it is surrounded

11

by a WS2 ring, opposite from the order in which the precursors were introduced into the

12

CVD chamber. We propose a new thermodynamically-driven heterostructure synthesis

13

model that utilizes in-plane diffusion to minimize crystal energy at the growth

14

temperature. This micro-scale diffusion is unique to synthesis of 2D materials regardless

15

of the growth method that can enable new opportunities (e.g. synthesis of new

16

heterostructures) and/or should be considered in optimizing growth of 2D structures that

17

require abrupt interfaces.

18 19

Figure 1. Proposed growth stages of MoS2/Mo1-xWxS2 heterostructures (W atoms are

20

shown in green, Mo atoms are red, S atoms are black). WS2 crystals are first grown at

21

1100°C (Stage I). MoS2 is then formed at the crystal edge after the growth temperature is


Page 5 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

reduced and MoO3 is introduced into the growth chamber (Stage II). Depending on the

2

substrate growth temperature during the MoS2 growth step, the crystal forms either a

3

phase-segregated heterostructure with a Mo-core at low temperatures (Stage III-L) or a

4

near-uniform Mo1-xWxS2 alloy at high temperatures (Stage III-H).

5 6

Our two-step atmospheric pressure CVD heterostructure growth procedure employs a

7

succession of the growth processes developed for the individual growth of high-quality

8

MoS2 and WS2 2D crystals (experimental details are provided in Supporting Information).

9

We first synthesized monolayer and few-layer WS2 crystals at the growth temperature (!)

10

𝑇!

= 1100 °C using WO3 and sulfur powders (Figure 1, Stage I). The presence of WS2

11

crystals and their purity were confirmed by Raman spectroscopy and PL and the samples

12

were then reintroduced to the CVD reactor (Figures S2, S3, and S4a-c in Supporting

13

Information). In the second step, WO3 was replaced by MoO3 and the growth processes

14

were performed at temperatures between 𝑇!

15

optimal for single-phase MoS2 growths (Figure 1, Stage II). As we discuss later on, the

16

structure of the resulting heterostructures consistently show a strong dependence on the

17

growth temperature: heterostructures grown at 𝑇!

18

MoS2 core surrounded by an Mo1-xWxS2 ring (Figure 1, Stage III-L), whereas higher

19

growth temperatures between 𝑇!

20

Mo1-xWxS2 alloy throughout the entire crystal (Figure 1, Stage III-H).

(!)

= 650 − 710 °C , which would be

(!)

(!)

= 650 °C consisted of a distinct

= 680 − 710 °C yielded a more evenly distributed

21 22

An optical image of a representative Stage III-L heterostructure crystal grown at 650 °C

23

(Figure 2a) shows two domains forming a triangular core-ring structure. The thickness of

24

the crystal measured by atomic force microscopy (AFM) was determined to be

25

approximately 2.5 nm, corresponding to 3 TMD layers, with no significant topological

26

variations between the core and the ring. To investigate the source of the optical contrast,


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 22

1

we performed Raman spectroscopy measurements along the radial direction of the

2

heterostructure (Figure 2b). Throughout the paper, we will use A and E to refer to the

3

! 𝐴′! and 𝐸′ vibrational modes in odd layer number crystals and 𝐴!! and 𝐸!!

4

vibrational modes in even layer number crystals, respectively. Interestingly, the MoS2 E

5

(382 cm-1) and A (407 cm-1) Raman peaks dominate the spectrum in the core and no WS2

6

peaks were detected in this area despite the fact that WS2 was grown first. Two sets of

7

in-plane E and out-of-plane A peaks, which can be ascribed to WS2 and MoS2-derived

8

vibrational modes, coexist in the outer ring area, indicating that this region is a

9

Mo1-xWxS2 alloyed structure. Moving away from the crystal center, the relative intensity

10

of WS2 E mode (352 cm-1) decreases, whereas WS2 A mode (417 cm-1) increases,

11

indicating a compositional gradient with decreasing x (decreasing W composition). In

12

addition to the intensity variations, the MoS2 E and A mode peaks down shift to lower

13

wavenumbers (373 cm-1 and 395 cm-1, respectively) and broaden in the alloyed

14

Mo1-xWxS2 outer ring. This behavior can be attributed to localized strain and disorder in

15

an alloy of two domains with different lattice constants, consistent with those observed in

16

other Mo1-xWxS2 alloy crystals.18,22

17


Page 7 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

(!)

1

Figure 2. (a) Optical image of a typical heterostructure grown at 𝑇!

= 650 °C (Stage

2

III-L) reveals a distinct core-ring structure. AFM cross-sectional contour shows the

3

height of the crystal to be 2.5nm. Scale bar is 5 µm. (b) Raman spectra collected at the

4

points marked in Figure 2a. The dashed lines show the wavenumber shift of the A and E

5

modes of WS2 and MoS2, respectively.

6 7

To understand the spatial compositional variations in the core-ring structure, we

8

performed Raman surface mapping (Figure 3). In agreement with our spot Raman

9

measurements, the intensity of E and A vibrational modes (Figures 3a and b, respectively)

10

corresponding to pure MoS2 are localized in the core, whereas E and A vibrational modes

11

of WS2 (Figures 3c and d, respectively) are localized in the ring. The observed radial

12

decrease of the WS2 E peak intensity towards the edge of the crystal can be attributed to

13

the corresponding decrease in W composition. However, the opposite trend observed for

14

the WS2 A is likely due to a peak convolution between A vibration modes of MoS2 and

15

WS2, as observed previously.7 This observation is supported by the fact that the intensity

16

of the E and A vibrational modes of the alloy-shifted MoS2 (Figures 3e and f,

17

respectively) increase near the outer edge of the crystal, indicating a higher composition

18

of Mo in this region of the Mo1-xWxS2 ring. Figure 3g summarizes these findings in a

19

composite Raman map, which can be directly correlated with the contrast features

20

observed in the scanning electron microscope (SEM) secondary electron image of the

21

same crystal (Figure 3h). The secondary electron image indicates that the variation in the

22

work function – and therefore composition change – exists between the core and outer

23

ring, corroborating the Raman results. Finally, the compositional segregation was

24

confirmed by energy-dispersive x-ray spectroscopy (EDS) measurements (Supporting

25

Information, Figure S5).


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Page 8 of 22

2

Figure 3. Raman and secondary electron images of the same crystal shown in Figure 2

3

(Stage III-L). MoS2 (a) E (382 cm-1) and (b) A (407 cm-1) mode intensity maps showing

4

pure MoS2 composition in the core of the crystal, despite the fact that MoS2 was grown in

5

the second step. WS2 (c) E (352 cm-1) and (d) A (417 cm-1) mode intensity maps showing

6

strong localization of WS2 in the crystal ring. Alloy-shifted MoS2 (e) E (373 cm-1) and (f)

7

A (395 cm-1) mode intensity maps showing that the ring is composed of an Mo1-xWxS2

8

alloy. (g) Raman composite map calculated from (a-f) where green color intensity

9

corresponds to the sum of intensities of MoS2 A and alloyed MoS2 E peaks, and red

10

corresponds to the intensity of the WS2 E peak. The dark yellow (ocher) color of the

11

crystal edge identifies an alloy of MoS2 (red) and WS2 (green). (h) SEM secondary

12

electron image of the same crystal. Scale bar is 5 µm.


Page 9 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1 2

Taken together, optical, Raman, and SEM results show that the final composition

3

distribution in the TMD heterostructure is independent of the growth order. We note that

4

similar results were obtained on a number of crystals from the same substrate, and

5

therefore Figures 2 and 3 are representative of the specific growth conditions. Our results

6

are unexpected, as they are the first observation of a micron-scale diffusion of metals in

7

the TMD crystallites during the heterostructure growth. In all of the previously reported

8

two-step CVD growths, the first deposited TMD forms the core of the lateral

9

heterostructure, and the edge atoms serve as preferential sites for epitaxial growth of the

10

second TMD that forms the ring.16–18,20,24 In our case, however, MoS2 is deposited after

11

WS2 and yet, an inverted heterostructure and alloying are observed.

12

13

To gain insights into the growth mechanism and investigate the role of the growth

14

parameters on the apparent substitution of W atoms in the core with Mo, we performed a

15

series of two-step growths with different MoS2 deposition parameters (Supporting

16

Information, Table S1). Figure 4 shows optical, AFM, and Raman characterization results

17

of a sample grown using a higher growth temperature during the MoS2 deposition

18

(𝑇!

19

characteristic MoS2 and WS2 vibrational modes were obtained (Figure 4a-d), and are

20

summarized in the composite Raman map (Figure 4e). The Raman results indicate higher

21

concentration of WS2 in the outer ring, while MoS2 is present across the entire crystal.

22

The optical image (Figure 4f) shows relatively uniform crystals and the corresponding

23

AFM line-scan indicates that the crystal is a bilayer with the thickness of 1.5 nm (AFM

24

height image is provided in Supporting Information, Figure S6a).

(!)

= 680 °C vs. 650 °C). As in the previous case, Raman intensity maps of the


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Page 10 of 22

2

Figure 4. Raman intensity maps of the (a) E and (b) A modes of MoS2 and (c) E and (d)

3

A modes of WS2 of a triangular heterostructure grown with a second step CVD growth

4

temperature of 680 °C (Stage III-H). (e) Raman composite map calculated from (a-d)

5

where green represents the MoS2 A mode, and red represents the WS2 A mode. (f)

6

Optical image with the height profile from the corresponding AFM image. Scale bar is 5

7

µm.

8

Although Raman studies provide a useful insight into the overall structure of the TMD

9

materials, they offer limited precision in quantifying composition of the Mo1-xWxS2

10

alloys.7 Other techniques, such as photoluminescence (PL) spectroscopy, could refine

11

Raman results regarding the composition of lateral heterostructures. We used two

12

complementary techniques, AFM phase imaging and PL mapping (Figure 5a and b,

13

respectively), to analyze the same Stage III-H crystal shown in Figure 4. Despite the

14

fundamental differences between these two techniques, they reveal qualitatively similar

15

results that are a sensitive measure of the composition variation within the material from


Page 11 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

the center to the edge of the crystal. Both show a high-contrast interface separating the

2

core and ring with different alloy composition, situated approximately 3.5 µm from the

3

crystal core (Figure 5c). The AFM phase image was generated by monitoring the phase

4

of the oscillating AFM tip; as the composition of the crystal changes, the resulting strain

5

in the crystal causes a phase shift. Figure 5b shows the map of the PL emission peak

6

wavelength within the crystal and the extracted radial distribution of the measured PL

7

peak wavelengths is shown in Figure 5c. The PL peaks of pure MoS2 and WS2 are

8

centered at 680 nm and 620 nm, respectively; the deviation from these values can

9

therefore be directly translated into the compositional information (Figure 5c).5 We note

10

that this translation does not account for any additional doping, strain, or other factors

11

that may contribute to shifts in the PL emission peak. Therefore, the computed alloy

12

compositions (Figure 5c) should be interpreted only qualitatively. In agreement with our

13

Raman data, the PL map and composition plot indicate that the entire crystal is a

14

Mo1-xWxS2 alloy with a MoS2-dominant center and a slightly higher concentration of WS2

15

in the outer regions. This claim is further supported by EDS measurements (Figure S7 in

16

Supporting Information). In contrast to the segregated core-ring structure of the Stage

17

III-L crystals (Figures 2 and 3), these Stage III-H crystals are relatively uniform – the

18

composition varies over a range between 43% and 56% and the interface could not be

19

observed in Raman or optical images.

20

21

Figure 5. (a) AFM tapping mode phase image from the crystal shown in Figure 4. (b) PL

22

map of the same crystal, darker color represents MoS2 and lighter color represents WS2.

23

(c) PL emission peak and computed W composition at varying distances from the crystal


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 22

1

core along the yellow arrow shown in Figure 5b. W composition values are calculated

2

from the PL emission peak data.5 Scale bar is 5 µm.

3 4

In addition to the radial changes observed in AFM and PL images (Figures 5a and b),

5

straight, discontinuous lines can be seen extending from the center to the three corners of

6

the triangular crystal (darker lines in AFM image, lighter in PL image). These lines are

7

attributed to strain-related micro-tearing, caused by the difference in the coefficients of

8

thermal expansion between the crystal and its substrate. Intra-grain boundaries are

9

formed during the cooling stage, which affect optical and electronic properties of the

10

crystal observed as a drop in PL amplitude in pure MoS2 and WS2 crystals25–27. In the

11

case of heterostructures, we observe not only the drop in PL intensity (Figure S6b in

12

Supporting Information), but also as a noticeable shift in PL peak wavelength consistent

13

with higher W composition (Figure 5b). These results suggest that W preferentially

14

diffuses to these intra-grain boundaries, further supporting our observation of the W

15

micro-scale in-plane diffusion toward the crystal edges (internal or external).

16 17

To understand the interplay between the segregation and mixing observed for different

18

MoS2 growth temperatures, we performed a score of crystal growth experiments with

19

varying MoS2 growth parameters to capture different stages of the Mo atomic diffusion

20

process. From these experiments detailed in Figures S4 and S8 in Supporting Information,

21

we stipulate the following growth model: WS2 crystals form on the substrate during the

22

first growth step (Stage I in Figure 1). In the second growth step, MoS2 nucleates directly

23

at the activated WS2 edge but the overall size and thickness of the crystal do not change

24

(Stage II in Figure 1). Then, as inter-diffusion occurs, the Mo atoms substitute some or

25

effectively all of the W atoms at the core of the crystal (Stage III-H or III-L in Figure 1,

26

respectively). Because the final morphology is determined by the initial WS2 crystal,


Page 13 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

regardless of its thickness, it can be concluded that the Mo atoms participate in intra-layer

2

diffusion, from the edge to the core, substituting W atoms at the M sites of the MX2

3

lattice to form a tunable alloy structure.

4 5

The substantial effect of the inter-diffusion is sensitive to the growth temperature. Two

6

distinct heterostructures observed at different growth temperatures, as discussed above,

7

indicate that thermodynamic and kinetic effects compete in determining the final

8

structure and composition of the TMD lateral heterostructures. At higher temperatures

9

(e.g. 𝑇!

(!)

= 680 °C or 710 °C), the alloy has a more even distribution of Mo and W

10

atoms without domain separation (Stage III-H in Figure 1). The entropic contribution to

11

the Gibbs free energy becomes increasingly important with the increased temperature. As

12

such, mixing is favored and the crystal approaches a homogeneous Mo0.5W0.5S2

13

stoichiometry.28–30 At lower temperatures (e.g. 𝑇!

14

with a sharp interface separating it from the alloyed ring (Stage III-L in Figure 1). In that

15

regime, the enthalpic contribution causes segregation into separate domains. Because

16

MoS2 has a lower Gibbs energy of formation than WS2, the Mo coalesces into the core to

17

minimize its interfacial energy.31 We note that similar thermodynamically-driven effects

18

leading to alloying vs. phase segregation may play a role in chalcogen-modified

19

systems.32

(!)

= 650 °C), MoS2 forms the core

20 21

Others have alluded to possible temperature-dependent domain separation in the

22

Mo1-xWxS2 heterostructure system,22 but previous experiments involved a one-step

23

growth, thus making it impossible to decouple the WS2 and MoS2 growth processes. As

24

such, two-domain lateral heterostructures with a distinct interface were attributed to the

25

different lifetimes of the metal precursors (MoO3/MoO2 vs WO3) present during different

26

stages of the sulfurization step, in agreement with the commonly accepted growth

27

mechanism.14,15,22 Our two-step growth reveals another important mechanism in TMD


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 22

1

heterostructure growth in which the final structure is a thermodynamically-driven process

2

and independent of the growth chronology. We note that a two-step WS2/MoS2 growth

3

with WS2 localized in the core was recently reported by Heo et al.20 However, their

4

growth was performed at low pressure, whereas our growth was performed at

5

atmospheric pressure, a condition that may affect the nucleation and growth kinetics of

6

the crystal edge, as well as the relative availability of metal and sulfur precursors. For

7

example, the growth rate observed by Heo et al. was approximately ten times larger than

8

in our case, allowing the outer ring of MoS2 to reach a critical size to stabilize before Mo

9

atoms could diffuse inward.

10 11

In conclusion, we report a thermodynamically-driven TMD lateral heterostructure

12

synthesis via two-step CVD growth. The resulting core-ring crystal structure does not

13

reflect the growth sequence, but is sensitive to temperature, favoring mixing at high

14

temperatures and compositional segregation at lower temperatures. The Mo atoms are

15

absorbed at the edge of a WS2 crystal and diffused throughout the crystal. The resulting

16

crystals consist of either distinctive MoS2 core at low growth temperatures, or a rather

17

uniform alloy at higher growth temperatures. The observed atom diffusion from the edge

18

to the center in the structure offers an opportunity to synthesize the advanced

19

heterostructure architectures with controlled properties. It could also provide an effective

20

method to incorporate new transition metals into TMDs that have been theorized to exist,

21

but impossible to synthesize using current techniques. These materials in turn can

22

introduce unique physical properties to the field of 2D materials and open up exciting

23

opportunities for investigating the contact interaction and carrier dynamics to create

24

high-performance functional devices.

25 26

Associated Content

27

Supporting Information Available: Details of the chemical vapor deposition growth


Page 15 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

1

process, characterization and data analysis methods, additional WS2 and heterostructures

2

growth results. This material is available free of charge via the Internet at

3

http://pubs.acs.org.

4 5

Author Information

6

Corresponding Authors

7

*Email: [email protected], [email protected]

8

Notes

9

The authors declare no competing financial interest.

10 11

Acknowledgements

12

The work in Singapore was supported by National Research Foundation, Prime Ministers

13

Office, Singapore, under its Fellowship Program (Award No. NRF-NRFF2012-09). Work

14

at MIT was supported by the Center for Excitonics, an Energy Frontier Research Center

15

funded by the US Department of Energy, Office of Basic Energy Sciences under Award

16

Number DE-SC0001088. This work made use of Shared Experimental Facilities provided

17

by the MIT Center for Materials Science and Engineering (CMSE), supported in part by

18

the MRSEC Program of the National Science Foundation under award number

19

DMR-0213282. Any opinions, findings and conclusions are those of the authors and do

20

not necessarily reflect those of NSF or DOE.

21 22 23

References:

24

(1)

Mak, K. F.; Lee, C.; Hone, J.; Shan, J.; Heinz, T. F. Phys. Rev. Lett. 2010, 105, 136805.

26

(2)

Tan, C.; Zhang, H. J. Am. Chem. Soc. 2015, 137, 12162-12174.

27

(3)

Wang, Q. H.; Kalantar-Zadeh, K.; Kis, A.; Coleman, J. N.; Strano, M. S. Nat. Nanotechnol. 2012, 7, 699–712.

25

28


Nano Letters

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 22

(4)

Bhimanapati, G. R.; Lin, Z.; Meunier, V.; Jung, Y.; Cha, J. J.; Das, S.; Xiao, D.; Son, Y.; Strano, M. S.; Cooper, V. R.; Liang, L.; Louie, S. G.; Ringe, E.; Zhou, W.; Sumpter, B. G.; Terrones, H.; Xia, F.; Wang, Y.; Zhu, J.; Akinwande, D.; Alem, N.; Schuller, J. A.; Schaak, R. E.; Terrones, M.; Robinson, J. A. ACS Nano 2015, 9, 11509–11539.

(5)

Chen, Y.; Xi, J.; Dumcenco, D. O.; Liu, Z.; Suenaga, K.; Wang, D.; Shuai, Z.; Huang, Y.-S.; Xie, L. ACS Nano 2013, 7, 4610–4616.

8

(6)

Kang, J.; Tongay, S.; Li, J.; Wu, J. J. Appl. Phys. 2013, 113, 143703.

9

(7)

Chen, Y.; Dumcenco, D. O.; Zhu, Y.; Zhang, X.; Mao, N.; Feng, Q.; Zhang, M.; Zhang, J.; Tan, P.-H.; Huang, Y.-S.; Xie, L. Nanoscale 2014, 6, 2833–2839.

(8)

Mann, J.; Ma, Q.; Odenthal, P. M.; Isarraraz, M.; Le, D.; Preciado, E.; Barroso, D.; Yamaguchi, K.; Von Son Palacio, G.; Nguyen, A.; Tran, T.; Wurch, M.; Nguyen, A.; Klee, V.; Bobek, S.; Sun, D.; Heinz, T. F.; Rahman, T. S.; Kawakami, R.; Bartels, L. Adv. Mater. 2014, 26, 1399–1404.

15

(9)

Liu, H.; Antwi, K.; Chua, S.; Chi, D. Nanoscale 2014, 6, 624–629.

16

(10)

Lin, Z.; Thee, M. T.; Elías, A. L.; Feng, S.; Zhou, C.; Fujisawa, K.; Perea-López, N.; Carozo, V.; Terrones, H.; Terrones, M. APL Mater. 2014, 2, 092514.

1 2 3 4 5 6 7

10 11 12 13 14

17 19

(11) Zheng, S.; Sun, L.; Yin, T.; Dubrovkin, A. M.; Liu, F.; Liu, Z.; Shen, Z. X.; Fan, H. J. Appl. Phys. Lett. 2015, 106, 063113.

20

(12)

Fu, Q.; Yang, L.; Wang, W.; Han, A.; Huang, J.; Du, P.; Fan, Z.; Zhang, J.; Xiang, B. Adv. Mater. 2015, 2, 4732-4738.

(13)

Li, H.; Zhang, Q.; Duan, X.; Wu, X.; Fan, X.; Zhu, X.; Zhuang, X.; Hu, W.; Zhou, H.; Pan, A.; Duan, X. J. Am. Chem. Soc. 2015, 137, 5284–5287.

(14)

Gong, Y.; Lin, J.; Wang, X.; Shi, G.; Lei, S.; Lin, Z.; Zou, X.; Ye, G.; Vajtai, R.; Yakobson, B. I.; Terrones, H.; Terrones, M.; Tay, B. K. K.; Lou, J.; Pantelides, S. T.; Liu, Z.; Zhou, W.; Ajayan, P. M. Nat. Mater. 2014, 13, 1135–1142.

(15)

Huang, C.; Wu, S.; Sanchez, A. M.; Peters, J. J. P.; Beanland, R.; Ross, J. S.; Rivera, P.; Yao, W.; Cobden, D. H.; Xu, X. Nat. Mater. 2014, 13, 1096–1101.

(16)

Duan, X.; Wang, C.; Shaw, J. C.; Cheng, R.; Chen, Y.; Li, H.; Wu, X.; Tang, Y.; Zhang, Q.; Pan, A.; Jiang, J.; Yu, R.; Huang, Y.; Duan, X. Nat. Nanotechnol. 2014, 9, 1024–1030.

(17)

Gong, Y.; Lei, S.; Ye, G.; Li, B.; He, Y.; Keyshar, K.; Zhang, X.; Wang, Q.; Lou, J.; Liu, Z.; Vajtai, R.; Zhou, W.; Ajayan, P. M. Nano Lett. 2015, 15, 6135–6141.

(18)

Li, M.-Y.; Shi, Y.; Cheng, C.-C.; Lu, L.-S.; Lin, Y.-C.; Tang, H.-L.; Tsai, M.-L.; Chu, C.-W.; Wei, K.-H.; He, J.-H.; Chang, W.-H.; Suenaga, K.; Li, L.-J. Science.

18

21 22 23 24 25 26 27 28 29 30 31 32 33 34 35


Page 17 of 22

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Nano Letters

2015, 349, 524–528.

1 2

(19)

3 4

(20) Heo, H.; Sung, J. H.; Jin, G.; Ahn, J.-H.; Kim, K.; Lee, M.-J.; Cha, S.; Choi, H.; Jo, M.-H. Adv. Mater. 2015, 27, 3803–3810.

5

(21)

Kang, J.; Tongay, S.; Zhou, J.; Li, J.; Wu, J. Appl. Phys. Lett. 2013, 102, 012111.

6

(22)

Kobayashi, Y.; Mori, S.; Maniwa, Y.; Miyata, Y. Nano Res. 2015, 8, 3261–3271.

7

(23)

Chen, K.; Wan, X.; Xie, W.; Wen, J.; Kang, Z.; Zeng, X.; Chen, H.; Xu, J. Adv. Mater. 2015, 27, 6431–6437.

(24)

Chen, K.; Wan, X.; Wen, J.; Xie, W.; Kang, Z.; Zeng, X.; Chen, H.; Xu, J.-B. ACS Nano 2015, 9, 9868–9876.

(25)

Ogletree, D. F.; Schuck, P. J.; Weber-Bargioni, A. F.; Borys, N. J.; Aloni, S.; Bao, W.; Barja, S.; Lee, J.; Melli, M.; Munechika, K.; Whitelam, S.; Wickenburg, S. Adv. Mater. 2015, 27, 5693–5719.

8 9 10 11 12 13

Zhang, X.; Lin, C.; Tseng, Y.; Huang, K.; Lee, Y. Nano Lett. 2015, 15, 410–415.

17

(26) Bao, W.; Borys, N. J.; Ko, C.; Suh, J.; Fan, W.; Thron, A.; Zhang, Y.; Buyanin, A.; Zhang, J.; Cabrini, S.; Ashby, P. D.; Weber-Bargioni, A.; Tongay, S.; Aloni, S.; Ogletree, D. F.; Wu, J.; Salmeron, M. B.; Schuck, P. J. Nat. Commun. 2015, 6, 7993.

18

(27)

Kobayashi, Y.; Sasaki, S.; Mori, S.; Hibino, H.; Liu, Z.; Watanabe, K. ACS Nano 2015, 9, 4056–4063.

(28)

Gan, L.-Y.; Zhang, Q.; Zhao, Y.-J.; Cheng, Y.; Schwingenschlögl, U. Sci. Rep. 2014, 4, 6691.

(29)

Song, J.-G.; Ryu, G. H.; Lee, S. J.; Sim, S.; Lee, C. W.; Choi, T.; Jung, H.; Kim, Y.; Lee, Z.; Myoung, J.-M.; Dussarrat, C.; Lansalot-Matras, C.; Park, J.; Choi, H.; Kim, H. Nat. Commun. 2015, 6, 7817.

(30)

Wei, X.-L.; Zhang, H.; Guo, G.-C.; Li, X.-B.; Lau, W.-M.; Liu, L.-M. J. Mater. Chem. A 2014, 2, 2101–2109.

(31)

Schaefer, S. C. Electrochemical Determination of Gibbs Energies of Formation of MoS2 and WS2; Report of Investigation 8405, U.S. Department of the Interior, Bureau of Mines, U.S. Government Printing Office: Washington, DC, 1980.

(32)

Su, S.-H.; Hsu, W.-T.; Hsu, C.-L.; Chen, C.-H.; Chiu, M.-H.; Lin, Y.-C.; Chang, W.-H.; Suenaga, K.; He, J.-H.; Li, L.-J. Front. Energy Res. 2014, 2, 27.

14 15 16

19 20 21 22 23 24 25 26 27 28 29 30 31 32

33


Nano Letters

1 W Mo S S 2 3 4 5 6 7 ACS Paragon Plus Environment 8 Stage I Stage II 9 WS2 WS2/MoS2 10 1100oC

Page 18 of 22 Stage III-L MoS2/WS2 650oC

Stage III-H Mo1-xWxS2 ≥680oC

(a)

Page 19 of Nano 22 Letters 4

3

2

Intensity (×103) [a.u.]

1 1 2.5 nm 2 3 4 5 6 7 (b)8 WS2 WS2 E’ A’ 9120 MoS2 MoS2 E’ 10 4 A’ 1180 12 1340 3 14 2 ACS Plus Environment 15 Paragon 0 1 16 350 400 450 17300 Raman Shift [cm-1]

(a)

MoS2 E’ (b)

MoS2 A’

Nano Letters Page 20 of 22

1 2 3 (c) WS2 E’ (d) WS2 A’ 4 5 6 7 8 9 (e)10 Alloyed MoS E’ (f) Alloyed MoS2 A’ 2 11 12 13 14 15 16 (g) (h) 17 18 19 ACS Paragon Plus Environment 20 21 22

(a)

MoS2 E’

(b)

MoS2 A’

Page 21 ofNano 22 Letters

1 2 3 WS2 E’ WS2 A’ (d) (c)4 5 6 7 8 9 10 (f) (e)11 12 13 14 ACS Paragon Plus Environment 1.5 nm 15 16 17

679

Nano Letters

(c)

675

Page 22 of0.55 22

672

0.50

669

0.45

ACS Paragon Plus Environment -3.6°

668

666

0

1

2 3 4 Distance [μm]

5

W Composition [x]

(b)

PL Emission Peak [nm]

1 2 3 4

2.5°

Wavelength [nm]

(a)

WS2 Lateral Heterostructures

Recommend Documents