Water-penetration-assisted mechanical transfer of large-scale molybdenum disulfide onto arbitrary substrates

Abstract

The transfer of two-dimensional (2D) material layers to arbitrary substrates from growth substrates is critical for many applications. Although several studies of transfer processes have been reported, a transfer method that does not degrade 2D layers and damage growth substrates is still required. In this paper, we report a method that with the assistance of water penetration enables the mechanical transfer of MoS₂, one of the most widely studied 2D materials, from the growth substrate to a target substrate without any etching of the growth substrate or leaving any polymer residue on the MoS₂ layer. The difference between the adhesion forces of the MoS₂ and carrier films and the difference between the hydrophobicities of MoS₂ and the growth substrate means that water can easily penetrate the interspace at the MoS₂/growth substrate interface generated by the peeling off process. We also experimentally confirmed the usefulness of Cu carrier films as a contact material for MoS₂ that enables its clean separation. Our transfer method protects the original quality and morphology of large area MoS₂ without leaving any polymer residue, and enables the reuse of the growth substrate. This clean transfer approach is expected to facilitate the realization of industrial applications of MoS₂ and other 2D materials.

---

Introduction

Recently, two-dimensional (2D) materials have emerged as an important research topic. Single layer molybdenum disulfide (MoS₂) is one of the most widely investigated 2D materials, and possesses outstanding properties with applications in band gap engineering, optical and electrical devices, and spin–orbit research.^1,2 It has been reported that large-area and uniform single layer MoS₂ can be synthesized by performing chemical vapor deposition (CVD) on various substrates with atomically smooth surfaces such as SiO₂ on Si, sapphire, and mica.³ However, the full exploitation of the potential of CVD MoS₂ requires a method for the clean transfer of the as-grown MoS₂ from the growth substrate to the target substrate without degeneration of its properties. The current approach to the transfer of CVD MoS₂ usually involves coating a polymer carrier film on top of the CVD MoS₂ as a supporting layer and the chemical etching of the growth substrate to separate MoS₂ from its surface. For example, Lin et al. reported the transfer of CVD MoS₂ via the introduction of a PMMA polymer carrier film coating on top of the MoS₂ layer on the growth substrate. They floated PMMA/MoS₂ by etching the SiO₂ surface in hot NaOH, transferring PMMA/MoS₂ onto a target substrate, and then removing PMMA in acetone.⁴ However, this etching process causes serious damage to the 2D material due to the harsh corrosivity of the etchant, usually hydrogen fluoride (HF) or a strong base (NaOH or KOH), and the removal of PMMA leaves a polymer residue (see ESI Fig. S1 and S2†).⁵ Thus the development of a gentle transfer approach that is highly efficient, repeatable, and environmentally friendly is required. The mechanical transfer of CVD 2D materials is a good approach to the preparation of clean and high quality films because of the absence of a chemical etchant, but physical damage due to the abrupt force applied during the peeling off process is inevitable.⁶ For example, Lin et al. assessed the mechanical transfer by exfoliation with thermal tape of CVD MoS₂ grown on SiO₂ under ambient conditions. Cracks and holes due to such physical damage were clearly observed in their transferred sample.⁷

In this paper, we report the water-penetration-assisted mechanical transfer of CVD MoS₂ onto an arbitrary substrate. This approach is aided by the fast penetration of water into the interspace at the MoS₂/growth substrate interface generated by peeling off the PDMS/PMMA/Cu carrier film and MoS₂.⁷ The penetrated water provides effective mechanical support for the MoS₂ layer and accelerates the delamination, and thus prevents the physical damage that always occurs in typical mechanical transfer processes.⁸ We also experimentally established the usefulness of Cu as a metal carrier film that enables the separation of the underlying MoS₂ layer.

Experimental section

The whole transfer process is schematically illustrated in Fig. 1a–h. The CVD process for the growth of high quality single layer MoS₂ on SiO₂ substrates is the same as in previous studies.⁹ Before synthesis, growth substrates (SiO₂/Si) were treated with O₂ plasma. The growth of the MoS₂ films was carried out in a LPCVD furnace system. MoO₃ powder was placed close to the heating zone center in an Al₂O₃ boat, and another boat containing sulfur powder was placed with 30 cm distance from MoO₃ powder upstream relative to the gas flow direction. Growth substrates were placed 10 cm from MoO₃ powder. After purging the furnace with 100 sccm Ar for more than 10 minutes, the furnace was heated to 850 °C at a rate of 13.75 °C min⁻¹. MoS₂ was then synthesized under 850 °C for 20 minutes. After synthesis, the furnace system was cooled down to room temperature. After the growth of MoS₂, a thin Cu carrier film (50–100 nm) is prepared with e-beam deposition on top of the MoS₂ layer supported by the growth substrate (Fig. 1a and b). After a layer of polymethyl methacrylate (PMMA) (150–400 nm) is spin-coated onto Cu/MoS₂/SiO₂, the baking process is performed, and a layer of polydimethylsiloxane (PDMS) is attached onto PMMA/Cu/MoS₂/SiO₂ (Fig. 1c). Next, PDMS/PMMA/Cu/MoS₂/SiO₂ is immersed in deionized (DI) water (Fig. 1d). Even though water has a natural tendency to penetrate the interface between MoS₂ and the growth substrate due to the differences between their surface energies, penetration cannot actually start spontaneously.¹⁰ However, once the peeling off process starts, water will immediately flood into the space generated by the separation of PDMS/PMMA/Cu/MoS₂ because of this natural tendency and the hydraulic pressure.¹¹ The penetrated water mechanically supports the upper PDMS/PMMA/Cu/MoS₂ with buoyancy force, which prevents physical damage. The key strategy for the clean peeling off of the carrier film and MoS₂ is to find a suitable contact material that has a stronger adhesion force with MoS₂ than the MoS₂/growth substrate interface.⁷ We experimentally demonstrated that Cu is an ideal candidate, as discussed below. After separation of PDMS/PMMA/Cu/MoS₂, it is dried under ambient conditions, and then PDMS/PMMA/Cu/MoS₂ is placed and pressured onto a target substrate (a SiO₂/Si wafer in this study) (Fig. 1e). The whole sample is heated at 130 °C on a hotplate. Once PDMS has lost adhesion, it can easily be peeled off leaving PMMA/Cu/MoS₂ on top of the target substrate (Fig. 1f).¹² In the next step, the PMMA/Cu/MoS₂/target substrate is dipped in acetone to remove PMMA (Fig. 1g) and then in Cu etchant (FeCl₃ solution) to remove Cu (Fig. 1h). In contrast to the residue left on the MoS₂ layer after the removal of the polymer contact material in previous studies,⁵ the Cu layer is completely removed by the Cu etchant due to the de-wetting of metals on 2D materials.¹³ Finally, perfectly uniform, large-area CVD MoS₂ with a clean surface is obtained on the target substrate with our transfer process.

Fig. 1 Schematic diagram of the transfer process (a–k), OM images (l and m), and optical images (n–p) of samples after this process. Scale bars in (n–p): 1 cm.

Results and discussion

To clearly demonstrate the mechanical support provided by water penetration in our transfer approach, we performed a contrast experiment. After preparing two identical PDMS/PMMA/Cu/MoS₂/SiO₂ samples (Fig. 1c), we carried out peeling off under water (Fig. 1d) and out of water under ambient conditions (Fig. 1i) as in other mechanical transfer processes.⁷ After peeling off, the PDMS/PMMA/Cu/MoS₂ of the sample processed under water was found to be successfully separated (Fig. 1n). In contrast, PDMS/PMMA/Cu/MoS₂ was only partially separated for the sample processed out of water (Fig. 1o), and many cracks are evident (Fig. 1m) that are not present in the sample processed under water (Fig. 1l). These effects are due to the absence of the mechanical support provided by the water buoyancy force, which prevents damage to the thin PMMA/Cu/MoS₂ sandwich due to the abrupt peeling off.⁶ We also found that the use of the Cu carrier film as a contact material is crucial to the successful separation of MoS₂ from the growth substrate. As shown in Fig. 1j and k, another contrast experiment was conducted. After peeling off MoS₂ and the Cu carrier film (Fig. 1d), the growth substrate was found to be quite clean without any residue. Further, the shape and dimension of the separated PMMA/Cu/MoS₂ sandwich on PDMS are identical to those of the corresponding growth substrate (Fig. 1n), which indicates a perfect peeling off and release process. However, without Cu, the separation of MoS₂ fails (Fig. 1k). As shown in Fig. 1p, the PMMA carrier film directly attached to MoS₂ cannot be peeled from the growth substrate. This observation reveals the important role of Cu in the whole transfer process, but a full understanding of the exact mechanism will only be obtained with further study.⁷

Fig. 2a–c show optical microscopy (OM) images of MoS₂ before and after transfer under or out of water. The sample transferred out of water was prepared following steps a–c to i and e–h in Fig. 1. In Fig. 2a, high quality, uniform MoS₂ with a large area can be seen, and the appearance is similar at a higher resolution see the inset in (Fig. 2a). In Fig. 2b, the MoS₂ transferred under water also appears to be uniform, continuous, and clean without any wrinkles, cracks or polymer residue. This remarkable uniformity and cleanness is similar to that found in as-grown MoS₂. Note that wrinkles, cracks, and other defects, if any, can easily be detected with optical microscopy because of the strong contrast,¹⁰ as was the case for the sample transferred out of water (Fig. 2c). The holes in the sample transferred out of water were generated by the abrupt forces generated by the peeling off process; the absence of support by water means that the MoS₂ layer, which is just one atomic layer thick, can easily be damaged. The scanning electron microscopy (SEM) images in Fig. 2d–f show the high quality and uniformity of MoS₂ after transfer under water compared with those of as-grown MoS₂, but high density microcosmic defects are present in the sample transferred out of water. The difference in quality between the MoS₂ samples transferred under and out of water clearly shows the importance of the mechanical support effects of water. The cleanness and uniformity of the transferred MoS₂ was confirmed with atomic force microscopy (AFM) mapping measurements (Fig. 2g–i). The surface roughnesses (RMS) of our transferred single layer MoS₂ samples (both under water and out of water) were determined to be approximately 0.49 nm, which is comparable to that of as-grown MoS₂ (0.45 nm). This low roughness clearly demonstrates that no Cu residue is present after transfer. In contrast to polymer carrier films, the Cu carrier film deposited on the MoS₂ surface can easily be etched in Cu etchant due to the de-wetting of metals on 2D material.¹³ This fundamental advantage of Cu makes it an ideal mechanical supporting carrier film for 2D materials in transfer processes as a replacement for traditional polymer carrier films. Even though the roughness of the MoS₂ sample transferred out of water is similar to that of the sample transferred under water, its continuity is totally different. As shown in Fig. 2i, there are many holes in the MoS₂ sample transferred out of water. This observation is consistent with our OM and SEM results, and reinforces the important role of water penetration in our mechanical transfer process.

Fig. 2 OM images (400 μm × 400 μm, insets: 30 μm × 30 μm) of MoS₂ before (a) and after transfer under (b) or out of (c) water. SEM images of MoS₂ before (d) and after transfer under (e) or out of (f) water. AFM images of MoS₂ before (g) and after transfer under (h) or out of (i) water. Scale bars in (g–i): 1 μm.

Our transfer approach also preserves the layer number, crystallinity, and other physical properties of the MoS₂ layer. Fig. 3a shows the Raman spectra collected from MoS₂ before and after transfer under or out of water. The peak positions of the two characteristic peaks A_1g and E¹_2g, and the frequency difference Δk between them, which is widely used as an indicator of the number of layers of MoS₂,¹⁴ are not changed at all by either transfer process. The Δk value, ∼20 cm⁻¹, confirms the successful transfer of single layer MoS₂ from the growth substrate to the target substrate; this transfer is more challenging than the transfer of multiple layer MoS₂.¹⁰ The Raman peak positions of the samples transferred under and out of water are the same because physical damage does not influence the atom vibration mode. The Raman mapping image of the sample transferred under water (Fig. 3b) clearly demonstrates the preservation of single layer, high quality, and uniform MoS₂. For instance, the Δk value of MoS₂ in the mapping region always lies in a narrow range (19.5–20.5 cm⁻¹) after the transfer process. Fig. 3c shows the photoluminescence (PL) spectra of single layer MoS₂ before and after transfer under or out of water. Typical PL peaks are present in all the MoS₂ samples at ∼665 nm due to the direct band structure at the K point.¹⁵ These PL spectra of MoS₂ are consistent with the theoretical band structure and are also evidence for the high crystal quality of our MoS₂ samples before and after both transfer processes. The PL peak positions of both transferred samples are similar, because any defects in the sample transferred out of water will not change the MoS₂ band structure. However, the peak intensity of the MoS₂ sample transferred out of water is much weaker than in the other two curves due to the physical damage generated by the transfer process. The PL peak centers in the mapping image of the sample transferred under water (Fig. 3d) are uniformly distributed within the narrow range 660–670 nm over the whole mapping region. This result confirms that uniform large area single layer MoS₂ has been successfully transferred under water without quality degradation. The Raman mapping images in Fig. 3e–g were collected from MoS₂ after the out-of-water transfer process. The E¹_2g and A_1g Raman peaks and the PL peak intensity mapping clearly demonstrate the presence of defected holes in the MoS₂ sample, as is consistent with our AFM and other microscopy results.

Fig. 3 Raman (a) and PL (c) spectra of MoS₂ before and after transfer under or out of water. Raman E¹_2g and A_1g peak distance (b) and PL peak center position (d) mapping of MoS₂ after transfer under water. Raman (e) E¹_2g peak (380–395 cm⁻¹) and (f) A_1g peak (400–415 cm⁻¹) and (g) PL peak (650–670 nm) mapping of MoS₂ after transfer out of water. Laser: 532 nm and all mapping image size: 20 μm × 20 μm.

To demonstrate that this method can be applied to other arbitrary substrates, we transferred MoS₂ films to PET substrate (Fig. S3†), which is widely used as flexible substrate for various applications, by simply replacing SiO₂ target substrate with PET, following the same process in Fig. 1. As shown in Fig. S3c,† the transferred MoS₂ film onto PET substrate also showed excellent uniformity and cleanness.

Conclusions

We have demonstrated a novel water-penetration-assisted transfer process that can perfectly transfer large-scale MoS₂ onto arbitrary substrates without any physical damage or polymer residue. The growth substrate can also be reused after transfer. Although our study focused mainly on the transfer of MoS₂, this process can be applied to the transfer of any other 2D material. We believe that the utilization of the penetration of water between synthesized films and their growth substrates can completely prevent physical damage to 2D films during mechanical transfer processes. This clean, effective, and environmentally friendly transfer approach is a significant step forward in the realization of industry applications of MoS₂ and other 2D materials.

Acknowledgements

This research was supported by the Global Frontier Program through the Global Frontier Hybrid Interface Materials (GFHIM) (2013M3A6B1078873) of the National Research Foundation of Korea (NRF) and the Pioneer Research Center Program through the National Research Foundation of Korea funded by the Ministry of Science, ICT & Future Planning (2014M3C1A3053024), and Basic Science Research Program through the National Research Foundation of Korea funded by the Korean government (MSIP) (grant numbers: 2015R1D1A1A09057297 and 2015M3A7B7045496).

Notes and references

1. Y. X. Huang, J. H. Guo, Y. J. Kang, Y. Ai and C. M. Li, Nanoscale, 2015, 7, 19358–19376.
2. M. Yamamoto, Y. Shimazaki, I. V. Borzenets and S. Tarucha, J. Phys. Soc. Jpn., 2015, 84, 121006.
3. H. F. Liu, S. L. Wong and D. Z. Chi, Chem. Vap. Deposition, 2015, 21, 241–259.
4. Y. C. Lin, W. J. Zhang, J. K. Huang, K. K. Liu, Y. H. Lee, C. T. Liang, C. W. Chu and L. J. Li, Nanoscale, 2012, 4, 6637–6641.
5. D. L. Ma, J. P. Shi, Q. Q. Ji, K. Chen, J. B. Yin, Y. W. Lin, Y. Zhang, M. X. Liu, Q. L. Feng, X. J. Song, X. F. Guo, J. Zhang, Y. F. Zhang and Z. F. Liu, Nano Res., 2015, 8, 3662–3672.
6. S. R. Na, J. W. Suk, L. Tao, D. Akinwande, R. S. Ruoff, R. Huang and K. M. Liechti, ACS Nano, 2015, 9, 1325–1335.
7. Z. Y. Lin, Y. D. Zhao, C. J. Zhou, R. Zhong, X. S. Wang, Y. H. Tsang and Y. Chai, Sci. Rep., 2015, 5, 18596.
8. K. K. Liu, W. J. Zhang, Y. H. Lee, Y. C. Lin, M. T. Chang, C. Su, C. S. Chang, H. Li, Y. M. Shi, H. Zhang, C. S. Lai and L. J. Li, Nano Lett., 2012, 12, 1538–1544.
9. J. Jeon, S. K. Jang, S. M. Jeon, G. Yoo, Y. H. Jang, J. H. Park and S. Lee, Nanoscale, 2015, 7, 1688–1695.
10. A. Gurarslan, Y. F. Yu, L. Q. Su, Y. L. Yu, F. Suarez, S. Yao, Y. Zhu, M. Ozturk, Y. Zhang and L. Y. Cao, ACS Nano, 2014, 8, 11522–11528.
11. P. Gupta, P. D. Dongare, S. Grover, S. Dubey, H. Mamgain, A. Bhattacharya and M. M. Deshmukh, Sci. Rep., 2014, 4, 3882.
12. L. Wang, I. Meric, P. Y. Huang, Q. Gao, Y. Gao, H. Tran, T. Taniguchi, K. Watanabe, L. M. Campos, D. A. Muller, J. Guo, P. Kim, J. Hone, K. L. Shepard and C. R. Dean, Science, 2013, 342, 614–617.
13. R. Addou, A. Dahal and M. Batzill, Nat. Nanotechnol., 2013, 8, 41–45.
14. C. Lee, H. Yan, L. E. Brus, T. F. Heinz, J. Hone and S. Ryu, ACS Nano, 2010, 4, 2695–2700.
15. K. F. Mak, C. Lee, J. Hone, J. Shan and T. F. Heinz, Phys. Rev. Lett., 2010, 105, 136805.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra09681f

---