Thinning of n-layer MoS₂ by annealing a palladium film under vacuum

Abstract

When Pd is thermally evaporated onto n-layer MoS₂, a uniform film is formed on the surface. If Pd/n-layer MoS₂ is annealed at high temperature under vacuum, thinning of n-layer MoS₂ is observed. By controlling the thickness of the Pd film and annealing temperature, one or two top layers of n-layer MoS₂ can be removed during the annealing process. The thinning method presented in this work benefits the future design and fabrication of MoS₂-based devices.

---

Compared to graphene,^1–4 which is a zero-gap material and limits on/off ratios of graphene-based field effect transistors, semiconducting two-dimensional materials have attracted great interest in electrical and photoelectric devices. As a typical transition-metal dichalcogenide semiconductor, molybdenum disulfide (MoS₂) exhibits unique physical, optical and electrical properties,^5–10 which have attracted more attention recently. One remarkable property of MoS₂ is the indirect-to-direct band gap transition for bulk to monolayer MoS₂.⁵ Therefore, it is important to prepare and control the thickness of n-layer MoS₂.

At present, n-layer MoS₂ can be grown by chemical vapor deposition (CVD)^11–13 or be thinned to ultrathin flakes by the micromechanical cleavage method like that used for preparing graphenes on different substrates.¹ The n-layer MoS₂ prepared by the micromechanical cleavage method has the advantage of less defects and higher on/off ratios for MoS₂ devices,⁷ indicating higher quality comparing to that grown by CVD. In this work, n-layer MoS₂ were mechanically exfoliated from a piece of natural crystalline MoS₂ sample (SPI Supplies) with Scotch transparent tape 600 (3M). Then the MoS₂ were transferred to 300 nm SiO₂/Si substrates by adhering and taking off the tape. The position and layers number were determined by optical microscope (Leica DM4000) and micro-Raman spectroscope (Renishaw in Via Raman Spectroscope). A film of Pd was evaporated onto n-layer MoS₂. Before and after annealing treatment, scanning electron microscope (SEM) and micro-Raman spectroscope are used to study the morphologies and properties of Pd on n-layer MoS₂. An interesting result is that by controlling the thickness of Pd film and annealing temperature, one or two top layers of n-layer MoS₂ can be removed during the annealing process. The corresponding mechanism was proposed and discussed.

The n-layer MoS₂ were first selected in an optical microscope based on their colour contrast as shown in Fig. 1a. The layer number is estimated according to the color contrast. And their coordinates were recorded according to a mark in these optical images. Then the Raman spectra of these n-layer MoS₂ were measured, which has proved to be a powerful technology to identify the properties of ultrathin MoS₂ flakes,^14,15 especially the determination of the layer number of n-layer MoS₂. The layer identification was based on the difference of Raman spectra among these n-layer MoS₂.¹⁴ Raman spectra of n-layer MoS₂ excited by 514 nm laser lines are measured as shown in Fig. 1b; thus, the layer number of n-layer MoS₂ is determined accurately. More details of the experiment are given in the ESI.†

Fig. 1 (a) Typical photograph of n-layer MoS₂ under normal white light. The layer number is estimated according to the color contrast, which is confirmed by the corresponding Raman spectra. (b) Raman spectra of pristine n-layer MoS₂, which can be used to confirm the layer number. The Raman peaks correspond to E¹_2g and A_1g bands of n-layer MoS₂ as indicated in the Raman spectra, respectively. The dotted vertical lines in (b) represent the positions of E¹_2g, A_1g peaks of pristine monolayer MoS₂. The wavelength of Raman excitation laser is 514 nm.

A Pd film with controllable thickness was evaporated onto MoS₂ layers via a thermal evaporator at vacuum of 10⁻⁴ Pa. The morphologies of the Pd film were studied in detail by SEM. As shown in Fig. 2a and b, after thermal deposition of a Pd (1.6 nm) onto n-layer MoS₂, the Pd forms a uniform films on the surface of MoS₂ and substrate. The morphologies of Pd on the surface of MoS₂ and substrate are different and the boundary between the substrate and n-layer MoS₂ can be seen as shown in Fig. 2. Meanwhile, the morphologies of Pd on n-layer MoS₂ with different layer numbers, such as n = 1, 3 and bulk, are not so clear as shown in Fig. 2a and b.

Fig. 2 (a and b) Typical SEM images of the morphologies of 1.6 nm Pd film on n-layer MoS₂. The layer numbers are 1, 3 and bulk, in (a) and (b), respectively. The scale bars are 100 nm. (c and d) Raman spectra of n-layer MoS₂ deposited with (c) 1.6 nm or (d) 3.2 nm Pd film. The dotted vertical lines in (c) and (d) represent the positions of E¹_2g, A_1g peaks of pristine monolayer MoS₂ for comparing. The wavelength of Raman excitation laser is 514 nm.

After deposition of Pd film on n-layer MoS₂, their typical Raman spectra are shown in Fig. 2c and d for film thickness of 1.6 nm and 3.2 nm, respectively. For pristine n-layer MoS₂, the two peaks at 384 cm⁻¹ and 404 cm⁻¹ which are ascribed to be E¹_2g and A_1g for laser line of 514 nm excitation. The E¹_2g mode is in-plane which results from opposite vibration of the two S atoms with respect to the Mo atom while the A_1g mode is associated with the out-of-plane vibration of only S atoms in opposite directions.^16–18 These two Raman modes, E¹_2g and A_1g, exhibited sensitive to the thickness of MoS₂. Because of the coulombic interactions and possible stacking-induced changes of the intra-layer bonding, the frequency of the E¹_2g peak decreases and that of the A_1g peak increases with increasing layer number of n-layer MoS₂ as shown in Fig. 1b.

Comparing to the Raman spectra of pristine MoS₂, the Raman spectra of n-layer MoS₂ decorated Pd film have several significant changes. Firstly, as shown in Fig. 2c, the frequencies of E¹_2g peak of Pd decorated n-layer MoS₂ show obvious red shifts. And the shift increases with the increasing of the layer number of n-layer MoS₂, which means that the Raman frequency shift of E¹_2g peak is dependent on the layer number “n” of n-layer MoS₂. For monolayer MoS₂, the E¹_2g peak has the largest frequency shift 4.0 cm⁻¹. For bulk MoS₂, the frequency shift of E¹_2g peak almost disappears. Secondly, the A_1g peak frequency of Pd decorated n-layer MoS₂ shows similar increasing tendency as that in pristine MoS₂ and the magnitudes of peak shifts of the A_1g peaks are reduced obviously. For Pd decorated monolayer MoS₂, the new shoulder peaks of the A_1g peak and E¹_2g peak can be found in the spectra. Thirdly, the intensity ratios of the Raman peaks (E¹_2g and A_1g) of Pd decorated n-layer MoS₂ are different from those of pristine n-layer MoS₂. The deposition of Pd film change the intensity ratios of E¹_2g and A_1g peaks, especially for bilayer and trilayer MoS₂. For example, for pristine bilayer MoS₂, the intensity of E¹_2g peak is larger than that of A_1g peak. But for Pd decorated bilayer MoS₂, the intensity of E¹_2g peak is weaker than that of A_1g peak.

Meanwhile, it should be noted that the thickness of Pd has little effect on the modification of the Raman spectra of n-layer MoS₂. As shown in Fig. 2d, the Raman spectra of bilayer and trilayer MoS₂ decorated with 3.2 nm Pd film are almost the same as those of 1.6 nm Pd film decorated bilayer and trilayer MoS₂. For monolayer MoS₂, the Raman spectra of 3.2 nm Pd film decorated MoS₂ show slight differences in the intensity of the characteristic peaks as shown the green component peaks in Fig. 2c and d.

An important and interesting question here is why Raman spectra of n-layer MoS₂ changes after the deposition of Pd film. The changes of Raman spectra for metals on graphenes^19–21 have been attributed to two main reasons: one is the doping effect, another is the stress effect. The changes of Raman spectra of n-layer MoS₂ after the deposition of Pd film in this work are believed to the lattice mismatch between MoS₂ and Pd, which is supported by the experimental result (the result will be shown in Fig. 3) that the changes of Raman spectra disappears after annealing.

Fig. 3 (a and b) SEM images of the morphologies of 1.6 nm Pd film on n-layer MoS₂ after 800 °C vacuum annealing. Scale bars are 1 μm. (c) Raman spectra of n-layer MoS₂ deposited with 1.6 nm Pd film after annealed under 750 °C for 4 min. (d) The peak position of E¹_2g and A_1g bands for pristine MoS₂, 1.6 nm Pd film decorated MoS₂ and after annealing at 750 °C. (e) Raman spectra of n-layer MoS₂ deposited with 3.2 nm Pd film after annealed under 750 °C for 4 min. (f) The frequency difference of the E¹_2g and A_1g peak for pristine MoS₂, 1.6 nm or 3.2 nm Pd decorated MoS₂ with different annealing temperature. The dotted vertical lines in (c) and (e) represent the positions of E¹_2g, A_1g peaks of pristine monolayer MoS₂ for comparing. The wavelength of Raman excitation laser is 514 nm.

In Fig. 3a and b, typical SEM images of the Pd on n-layer MoS₂ after annealing at 800 °C for 4 min are shown. It can be seen that after annealing treatment, the Pd film has become nanoparticles on the substrate. On n-layer MoS₂, the morphologies of Pd metal are quite different from those on the substrate. On monolayer and trilayer MoS₂, the metal Pd has aggregated into irregular nanostrips. At the boundary between monolayer and trilayer MoS₂, there are no Pd as shown in Fig. 3a. The difference of the morphologies of Pd on monolayer and trilayer MoS₂ is small. For example, as shown in the Fig. 3a, the size of Pd strip on trilayer MoS₂ is little larger than that on monolayer MoS₂. And the size of the Pd nanostrips on bulk MoS₂ is larger than those on monolayer and trilayer MoS₂ as can be seen from Fig. 3b, which results in a low density on bulk MoS₂. Therefore, it can be concluded that the size of the nanostrips becomes larger and the density becomes smaller with the increase of MoS₂ layer number.

Why the Pd on n-layer MoS₂ show the thickness-dependent morphologies? Thickness-dependent morphologies of metals on graphenes have been reported and the mechanism is attributed to thermodynamic (e.g., energetics and stability) and kinetic (e.g., surface diffusion) factors.^22,23 Since similar systems (Pd on n-layer MoS₂ versus metals on n-layer graphenes) are studied, the same mechanism is believed to be applicable. This is to say, the density of nanostrips (N) can be obtained using the following equation: N ∝ exp(E_n/3KT), where E_n is the diffusion barriers of Pd on n-layer MoS₂, K is the Boltzmann constant, T is the temperature. Therefore, the different diffusion barrier E_n of Pd on substrate and n-layer MoS₂ lead to the different nanostrips (N) and the thickness-dependent morphologies on Pd on n-layer MoS₂.

The Raman spectra of Pd nanostrips on n-layer MoS₂ have been measured. Typical Raman spectra of Pd nanostrips on n-layer MoS₂ are shown in Fig. 3c after the Pd films of 1.6 nm on n-layer MoS₂ have been treated at 750 °C for 4 minutes in vacuum. Comparing to those of pristine and Pd film coated n-layer MoS₂, the Raman spectra after annealing treatments show interesting characteristics: firstly, the red shift of E¹_2g peak in monolayer, bilayer MoS₂ and its layer dependence in Pd film coated n-layer MoS₂ almost disappear. Secondly, the positions of E¹_2g peak and its layer dependence are quite similar to that of pristine n-layer MoS₂ as shown in Fig. 3d, which is red-shift behaviour with the increasing layer. Thirdly, the position of the A_1g peak do not change much among those of pristine, Pd decorated and heating treated n-layer MoS₂. Fourthly, the intensity ratios between A_1g and E¹_2g after annealing have similar values as those of pristine n-layer MoS₂.

The above results indicate that after annealing at 750 °C, the Raman spectra of 1.6 nm Pd film decorated n-layer MoS₂ are quite similar to those of the pristine n-layer MoS₂, suggesting the recovery of n-layer MoS₂. In order to study the effect of the thickness of Pd metal, the thickness of Pd film is increased to 3.2 nm. As shown in Fig. 2c and d, the Raman spectra of n-layer MoS₂ decorated with Pd films of 1.6 or 3.2 nm are almost the same, suggesting that the thickness of Pd films has little effect on the Raman spectra of Pd decorated MoS₂.

After annealing, quite interesting and unexpected phenomena are observed and the results are shown in Fig. 3e. It shows the Raman spectra of the 3.2 nm Pd film decorated MoS₂ after 750 °C annealing treatment in vacuum for 4 minutes. It can be seen that after annealing treatment, there is no any peaks observed for monolayer MoS₂, indicating that the original monolayer MoS₂ has disappeared. The optical images and Raman mapping of the monolayer MoS₂ have been added as Fig. S4 in the ESI† to show the uniformity of the thinning behavior. The Raman spectrum of bilayer MoS₂ has similar characteristics as that of the Raman spectrum of pristine monolayer MoS₂ as shown in Fig. 1b and 3e. Meanwhile, the frequency difference of A_1g and E¹_2g is 20.2 cm⁻¹ in bilayer MoS₂, which is consistent with that of 19.5 cm⁻¹ in pristine monolayer MoS₂ (see Fig. 3f). For original trilayer MoS₂, its Raman spectrum is similar to that of the Raman spectrum of pristine bilayer MoS₂ (see Fig. 3e). And the frequency difference between A_1g and E¹_2g of 22.2 cm⁻¹ is consistent with that of 21.7 cm⁻¹ in pristine bilayer MoS₂ (see Fig. 3f). These results indicate that for Pd film of 3.2 nm the top layer of n-layer MoS₂ is removed after annealing at 750 °C in vacuum. Thus, after this treatment monolayer MoS₂ disappears, bilayer MoS₂ becomes monolayer MoS₂ and so on. It should be noted that when few-layer MoS₂ flakes are annealed in vacuum as long as one hour, one layer of MoS₂ can be removed.²⁴ If the vacuum annealing time is shorted to several minutes (Fig. S3 in ESI†), no removing behavior is observed, suggesting the important role of palladium film in the thinning behavior of n-layer MoS₂.

These data suggest that the removal of the top layer of n-layer MoS₂ depends on both the thickness of Pd metal and the temperature. For example, when the annealing temperature is increased to 800 °C for Pd film of 3.2 nm, both monolayer and bilayer MoS₂ disappeared after annealing treatment as shown in Fig. 3f. The original trilayer MoS₂ becomes a monolayer MoS₂ after the annealing treatment. The frequency difference of A_1g and E¹_2g is 20.4 cm⁻¹ in the original trilayer MoS₂ as shown in Fig. 3f. These results indicate that the top two layers of the MoS₂ can be successfully removed at the same time. This means that, if we control the annealing temperature and the thickness of Pd film, the number of layers removed can also be controlled.

The effects of annealing temperature and thickness of Pd on the removal of n-layer MoS₂ are systematically studied and the results are summarized in Table 1. It can be seen that the thickness of Pd film plays an important role for the removal of the top layers of n-layer MoS₂ after vacuum annealing. When the thickness of Pd is 1.3, 1.6 or 2.7 nm, no removal of the top layers of n-layer MoS₂ is observed for annealing temperature of 750 °C/800 °C. When the thickness of Pd is 3.2 nm, one layer at the top of n-layer MoS₂ is removed at annealing temperature of 750 °C. If the annealing temperature is raised to 800 °C, two layers at the top of n-layer MoS₂ are removed. This means that after annealing monolayer and bilayer MoS₂ disappear and trilayer MoS₂ becomes a monolayer MoS₂. Meanwhile, the photoluminescence (PL) spectra of these n-layer MoS₂ were measured. The conclusion is shown in Fig. S1 and S2 (ESI†) which consistent with the result of Raman spectra. Since the thinning occurs only at the annealing process, the mechanism of thinning of n-layer MoS₂ is attributed to a dissolving-like process of MoS₂ in the metal Pd, which can be expressed by the following reaction:

where δ_Pd is the thickness of the Pd film and T is the annealing temperature.

Table 1 Effects of annealing temperature and thickness of Pd on the removal of n-layer MoS₂ after vacuum annealing

Thickness of Pd film	Annealing temperature	Results
3.2 nm	800 °C	1 & 2L removed, 3L → 1L
3.2 nm	750 °C	1L removed, 2L → 1L; 3L → 2L
2.7 nm	750 °C/800 °C	No removal
1.6 nm	750 °C/800 °C	No removal
1.3 nm	750 °C/800 °C	No removal

---

Conclusions

In summary, when Pd is thermally evaporated onto n-layer MoS₂, a uniform film is formed on the surface. If Pd/n-layer MoS₂ is annealed at high temperature in vacuum, Pd nanostrips are observed on n-layer MoS₂. The size of Pd nanostrips becomes larger and the density becomes smaller with the increase of layer number of n-layer MoS₂, respectively. By studying the Raman and PL spectra, the thinning behavior of n-layer MoS₂ has been observed. By controlling the thickness of Pd film and annealing temperature, one or two top layers of n-layer MoS₂ can be removed during the annealing process. The mechanism of the thinning behavior is proposed and the thinning method presented in this work benefits the future design and fabrication of MoS₂-based device.

Acknowledgements

We thank the financial support from National Science Foundation of China (Grant No. 11174062, 51472057). W. G. Chu acknowledges financial support from the Strategic Priority Research Program of the Chinese Academy of Sciences (GrantXDA09040101).

Notes and references

1. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov, Science, 2004, 306, 666.
2. A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191.
3. K. Novoselov, A. K. Geim, S. Morozov, D. Jiang, M. Katsnelson, I. Grigorieva, S. Dubonos and A. Firsov, Nature, 2005, 438, 197.
4. A. K. Geim, Science, 2009, 324, 1530.
5. Q. H. Wang, K. Kalantar-Zadeh, A. Kis, J. N. Coleman and M. S. Strano, Nat. Nanotechnol., 2012, 7, 699.
6. K. F. Mak, C. Lee, J. Hone, J. Shan and T. F. Heinz, Phys. Rev. Lett., 2010, 105, 474.
7. B. Radisavljevic, A. Radenovic, J. Brivio, V. Giacometti and A. Kis, Nat. Nanotechnol., 2011, 6, 147.
8. I. Song, C. Park and H. C. Choi, RSC Adv., 2015, 5, 7495.
9. H. Li, Z. Y. Yin, Q. Y. He, H. Li, X. Huang, G. Lu, D. W. H. Fam, A. I. Y. Tok, Q. Zhang and H. Zhang, Small, 2012, 8, 63.
10. X. Zhao, P. Chen, C. Xia, T. Wang and X. Dai, RSC Adv., 2016, 6, 16772.
11. Y. H. Lee, X. Q. Zhang, W. Zhang, M. T. Chang, C. T. Lin, K. D. Chang, Y. C. Yu, J. T. W. Wang, C. S. Chang and L. J. Li, Adv. Mater., 2012, 24, 2320.
12. Y. Zhan, Z. Liu, S. Najmaei, P. M. Ajayan and J. Lou, Small, 2012, 8, 966.
13. Z. Lin, Y. Zhao, C. Zhou, R. Zhong, X. Wang, Y. H. Tsang and Y. Chai, Sci. Rep., 2015, 5, 18596.
14. H. Li, Q. Zhang, C. C. R. Yap, B. K. Tay, T. H. T. Edwin, A. Olivier and D. Baillargeat, Adv. Funct. Mater., 2012, 22, 1385.
15. S.-L. Li, H. Miyazaki, H. Song, H. Kuramochi, S. Nakaharai and K. Tsukagoshi, ACS Nano, 2012, 6, 7381.
16. P. A. Bertrand, Phys. Rev. B: Condens. Matter Mater. Phys., 1991, 44, 5745.
17. D. Bae, K. Kim, Y. Park, E. Suh, K. An, J.-M. Moon, S. Lim, S. Park, Y. Jeong and Y. Lee, Phys. Rev. B: Condens. Matter Mater. Phys., 2001, 64, 233401.
18. C. Lee, H. Yan, L. E. Brus, T. F. Heinz, J. Hone and S. Ryu, ACS Nano, 2010, 4, 2695.
19. A. Das, S. Pisana, B. Chakraborty, S. Piscanec, S. Saha, U. Waghmare, K. Novoselov, H. Krishnamurthy, A. Geim and A. Ferrari, Nat. Nanotechnol., 2008, 3, 210.
20. R. Muszynski, B. Seger and P. V. Kamat, J. Phys. Chem. C, 2008, 112, 5263.
21. Z. H. Ni, T. Yu, Y. H. Lu, Y. Y. Wang, Y. P. Feng and Z. X. Shen, ACS Nano, 2008, 2, 2301.
22. Y. W. Mo, J. Kleiner, M. B. Webb and M. G. Lagally, Phys. Rev. Lett., 1991, 66, 1998.
23. L. Y. Ma, L. Tang, Z.-L. Guan, K. He, K. An, X.-C. Ma, J.-F. Jia, Q. K. Xue, Y. Han, S. Huang and F. Liu, Phys. Rev. Lett., 2006, 97, 266102.
24. X. Lu, M. I. B. Utama, J. Zhang, Y. Zhao and Q. Xiong, Nanoscale, 2013, 5, 8904–8908.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental conditions of Raman measurements and other data. See DOI: 10.1039/c6ra07714e

---