Exciton tuning and strain imaging in WS2 supported on PDMS micropillars Cite as: Appl. Phys. Lett. 121, 253101 (2022); https://doi.org/10.1063/5.0130927 Submitted: 16 October 2022 • Accepted: 09 December 2022 • Published Online: 21 December 2022 Published open access through an agreement with Universidad del País Vasco Facultad de Ciencias Químicas M. Sledzinska, P. Xiao, E. Puig Vilardell, et al. COLLECTIONS Paper published as part of the special topic on Phononics of Graphene, Layered Materials, and Heterostructures ARTICLES YOU MAY BE INTERESTED IN Excitonic performance and ultrafast dynamics in defective WSe2 Applied Physics Letters 121, 083102 (2022); https://doi.org/10.1063/5.0098100 Exciton–polaritons of hBN/WS2 heterostructure in cavity observed at room temperature Applied Physics Letters 121, 231106 (2022); https://doi.org/10.1063/5.0118834 Phononics of graphene, layered materials, and heterostructures Applied Physics Letters 122, 070401 (2023); https://doi.org/10.1063/5.0144480 Appl. Phys. Lett. 121, 253101 (2022); https://doi.org/10.1063/5.0130927 121, 253101 © 2022 Author(s). Applied Physics Letters ARTICLE scitation.org/journal/apl Exciton tuning and strain imaging in WS2 supported on PDMS micropillars Cite as: Appl. Phys. Lett. 121, 253101 (2022); doi: 10.1063/5.0130927 Submitted: 16 October 2022 . Accepted: 9 December 2022 . Published Online: 21 December 2022 M. Sledzinska,1,a) P. Xiao,1 E. Puig Vilardell,1,2 E. Cha  vez Angel,1 M. J. Esplandiu,1 and C. M. Sotomayor Torres 1,3 AFFILIATIONS 1 Catalan Institute of Nanoscience and Nanotechnology (ICN2), CSIC and BIST, Campus UAB, Bellaterra, 08193 Barcelona, Spain 2 Departamento de Fısica, Universidad Auto  noma de Barcelona, Bellaterra, 08193 Barcelona, Spain 3 ICREA, Passeig Lluis Companys 23, 08010 Barcelona, Spain Note: This paper is part of the APL Special Collection on Phononics of Graphene, Layered Materials, and Heterostructures. a) Author to whom correspondence should be addressed: marianna.sledzinska@icn2.cat ABSTRACT Since the raise of 2D materials, significant research has been dedicated to their strain-dependent electronic and mechanical properties. In this work, we studied exciton energies and low-frequency phonon modes in CVD-grown mono- and few-layer WS2 transferred on PDMS micro- pillars. The modification of the band structure under strain was investigated by photoluminescence (PL) spectroscopy at room temperature. Machine learning (ML) methods were used to analyze the PL spatial maps and facilitate the spectral deconvolution. For monolayer (1L) WS2, red shift in the exciton energy was detected as a function of the position, which was ascribed to the presence of residual strain. For three-layer (3L) strained WS2, a significant increase in the PL intensity corresponding to direct (K–K) band transition together with a change of exciton energy was observed. From the PL spectra, strain distribution maps were extracted for both studied samples, which strongly resembled the ML clustering results. Finally, the low-frequency Raman modes of WS2 were studied on both Si/SiO2 and PDMS substrates and no significant change of their frequency was observed for the 3L-WS2. C 2022 Author(s). All article content, except where otherwise noted, is licensed under a Creative Commons Attribution (CC BY) license (http:// V creativecommons.org/licenses/by/4.0/). https://doi.org/10.1063/5.0130927 One of the most exciting properties of 2D transition metal dichal- increasing number of layers, the bandgap becomes indirect and, conse- cogenides (TMDs) is that their electronic structure1–3 and mechanical quently, the PL quantum yield decreases significantly. Another draw- properties4,5 can be tuned by external strain. At the same time, it is rel- back of the previous works is that the presented spectra were taken atively easy to experimentally induce strain in those materials, either only from one point on the sample without considering the strain dis- by atomic force microscopy tip,6–8 in bending1,9–12 or stretching13 tribution in the whole flake.3,9–11 The importance of strain distribution setup, by creating wrinkles14,15 or by interactions with pre-patterned mapping was recently discussed by Wang et al., who showed that WS2 substrates.16,17 follows continuum mechanics on the microscale and strain generates a Among the 2D TMDs family, 1L-WS2 has attracted considerable non-uniform bandgap distribution even in a single WS2 flake placed interest due to its unique properties such as PL quantum yield (up to on flexible polymer substrates.22 6%) higher than other 2D semiconductors (1% for MoS2)18 and Raman and PL spectroscopies are fast, non-destructive techni- large exciton–trion binding energy.19 Furthermore, due to the break- ques to study the impact of the strain on the optical properties. ing of inversion symmetry and strong spin–orbit coupling in 1L-WS2, Moreover, they require little or no sample preparation, and can pro- the electronic states of the two valleys have different chiralities, which vide qualitative and quantitative maps of the chemical structure and have recently been applied to a nanophotonic circuit that can unidirec- composition of the sample. However, the analysis of each individual tionally route the valley indices of 1 L-WS2.20 spectrum in large spectroscopical maps can be time consuming. Tuning of the optical and electronic properties of WS2 and other Machine learning is a rapidly growing field that has shown great TMDs by strain has been investigated extensively both theoretically potential in several research areas due to its ability to detect patterns in and experimentally, but to date, most of the studies have been realized complex or very similar datasets that humans would likely miss.23 on TMDs monolayers due to their direct bandgap.9–11,21 With an Additionally, rapid spectral clustering can be performed in seconds, Appl. Phys. Lett. 121, 253101 (2022); doi: 10.1063/5.0130927 121, 253101-1 C Author(s) 2022 V Applied Physics Letters ARTICLE scitation.org/journal/apl providing valuable features not present in the intensity maps that The Raman spectrum of 1L-WS2 shows the characteristic peaks cor- default spectroscopy setup software gives to the users. responding to in-plane vibrational modes E2g (C-point in the Brillouin In this work, we studied room temperature PL and Raman spec- zone) and 2LA (M), in plane A1g (C) and their combinations tra of single- and few-layer CVD-grown WS2, transferred on PDMS [Fig. 1(c)].25 High PL intensity of the as-grown samples together with micro-pillars. We applied machine learning techniques, such as data the Raman spectrum confirms their high quality. clustering (k-means), to simplify processing of the datasets and fit Subsequently, the flakes were water-transferred to plain and them more accurately. We measured spatial distributions of the PL micropatterned PDMS substrates.26 After transfer to plain PDMS, the spectra and observed significant changes in the exciton peak intensity PL intensity decreased considerably, and the energy of both XA and and energy as a function of the position. These maps were subse- X excitons increased to 2.03 and 1.99 eV, respectively, due to the quently converted into strain maps. We show that such a procedure charge transfer induced by interaction with PDMS [Fig. 1(b)].27 With can be straightforwardly applied to detect strain in both single and the increasing number of layers, we observe a significant decrease in the few-layer TMDs. Finally, the low-frequency Raman mode of 3L-WS2 PL intensity, together with a redshift to 1.93 eV, corresponding to the was studied as a function of strain. direct (K–K)–indirect (C–K) bandgap transition.15,18 The PL signal for WS2 was synthesised on Si/SiO2 substrates using CVD from solid samples thicker than the monolayer is strongly suppressed [Fig. 1(d)], precursors at 900 C under constant Ar flow. The resulting samples and PL peaks from the PDMS substrate become visible. This is due to consisted of large area monolayer and few layer triangles, as shown in the fact that the PL quantum yield with respect to the monolayer is Fig. 1(a). 1L-WS2 has a direct bandgap located at the K-point in the more than three orders and four orders of magnitude lower for the Brillouin zone; therefore, it shows a strong A exciton peak (XA) corre- bilayer and trilayer, respectively.18 sponding to the direct excitonic transitions at 1.97 eV. A trion The freshly transferred 1L samples showed only XA and X exci- (charged exciton, X ) is present at 1.95 eV,24 as shown in Fig. 1(b). tons; however, after some time (1 week), an additional peak (XD) FIG. 1. Optical spectroscopy on CVD-grown WS2. (a) Optical image of the WS2 flakes synthesized on Si/SiO2 substrates. Scale bar corresponds to 50 lm. (b) PL spectra of the as-grown and transferred to the PDMS substrate 1L-WS2. (c) Raman spectrum of as-grown 1L-WS2. (d) PL spectra of the 1, 2, and 3L-WS2 on the PDMS substrate. Appl. Phys. Lett. 121, 253101 (2022); doi: 10.1063/5.0130927 121, 253101-2 C Author(s) 2022 V Applied Physics Letters ARTICLE scitation.org/journal/apl FIG. 2. 1L-WS2 on the PDMS micropillar. (a) Optical image of the sample. Scale bar corresponds to 5 lm. (b) Clustering of the spectra, where clusters C1 and C2 correspond to spectra of WS2 in the center of the pillar, C4 and C5 to pillar edges, and C3 to the background. (c) Examples of PL spectra from each cluster. The color outline corresponds to the standard deviation of spectra, and the black curve corresponds to the mean of the spectra in each cluster. (d) Strain distribution in the WS2 layer. located at 1.86 eV emerged in the PL spectrum (Fig. S1). It has been processing (baseline correction and peak area normalization), we fur- previously attributed to the presence of defects;28 however, the origin ther investigated higher number of clusters. In Figs. 2(b) and 2(c), the of the peak in this sample is not fully clear. The defects could arise results for five clusters are shown. C1 and C2 correspond to the high- from the interaction between WS2 and the environment, as it is known intensity spectra taken from the center of the pillar, C4 and C5 to pillar that 2D materials can easily degrade when in contact with air and edges, and C3 to the background. Each cluster groups spectra of differ- moisture. The presence of defects did not affect the XA and X peaks, ent PL energies. A similar analysis was also carried out using principal which remained at the energies of 2.02 and 1.99 eV. component analysis (PCA) as shown in the supplementary material. For strain mapping experiments, the 1L-WS2 was transferred Figure S5 shows the maps of the scores of PC1 and PC2 as a function to PDMS micropillars [5 lm diameter, 10 lm pitch, 5 lm height; of the position in the sample. The PC1 score-map basically shows the Figs. 2(a) and S2]. The PL spatial map was taken from the area in difference between the spectra of WS2, the edge, and the background. Fig. 2(a) with 1 lm resolution. To facilitate the data analysis, we While the PC2 score-map is more sensitive to the strain difference in recurred to ML techniques, in particular, unsupervised learning. We the WS2 sample, other PCs does not show any significant information. used the k-means clustering algorithm to label samples based on fea- With the different clusters identified, we proceeded to fit the tures that group it with similar samples. Silhouette score contrasts spectra. All the spectra except for the ones grouped in C3 (back- average distance to elements in the same cluster with the average dis- ground) were fitted with a Lorentzian curve and a Gaussian curve for tance to elements in other clusters. As a rule, the higher the silhouette XA and X peaks, respectively. The energy distribution of XA as a score, the value indicates that a given point is well matched to its own function of the position is shown in Fig. S6. It was previously reported cluster and well separated from the neighbor clusters. For the raw that XA energy scales linearly with strain as 46 meV/% strain for 1L- spectra, the highest silhouette score was obtained for two clusters (Fig. WS29. From this dependence, the strain map was calculated, as shown S3). From the PL intensity map (Fig. S4) and cluster map (Fig. S3), it is in Fig. 2(d). A small variation of strain between 0% and 0.3% was clear that C1 and C2 roughly correspond to the background, and WS2 detected in the WS2 layer. The highest strain was measured in the cen- supported on the pillar. After performing simple spectra pre- ter of the pillar, at the edge of the WS2 film, which is ascribed to the Appl. Phys. Lett. 121, 253101 (2022); doi: 10.1063/5.0130927 121, 253101-3 C Author(s) 2022 V Applied Physics Letters ARTICLE scitation.org/journal/apl FIG. 3. 3L-WS2 on the PDMS micropillar. (a) Optical image of the sample. Scale bar corresponds to 5 lm. (b) Clustering of the PL spectra. (c) Representative PL spectra from each cluster and corresponding peak deconvolution. (d) Strain distribution in the WS2 layer. residual strain coming from the transfer process. The maps in We applied the k-means algorithm to cluster the spectra in the Figs. 2(b) and 2(d) strongly resemble each other; therefore, we can PL map, and the result is shown in Fig. 3(c). The highest silhouette conclude that clustering algorithms could discern between spectra cor- score corresponds to the division into two clusters: background and responding to different strain levels. WS2 spectra, similarly as to the case of 1L-WS2 discussed previously. We further investigated few-layer WS2 transferred to the same The second highest score corresponds to four and five clusters, which type of pillars [Fig. 3(a)]. By the optical contrast, we estimated the differ only by a few points. Clusters C1–C3 correspond roughly to the thickness to be 3L; however, the exact thickness of the sample is rather left and right pillar and suspended (folded) WS2 spectra, respectively. difficult to evaluate with optical methods. Contrary to other 2D mate- Representative spectra of each cluster are compared in Fig. 3(b). The rials, such as graphene,29,30 when the thickness of TMDs is larger than spectrum corresponding to the left pillar can be fitted by two 2L, Raman spectra closely resemble each other and cannot be used to Lorentzian curves at 1.91 and 1.84 eV. The peaks can be assigned to confirm the exact thicknesses. The comparison between PL spectra of direct (XA) and indirect (XI) band transitions. There is a clear trend of few-layer samples is typically difficult due to their low intensity. redshift in the XA peak between the left and right pillars from 1.91 to However, this sample showed enhanced PL signal from WS2, which 1.87 eV, respectively, a consequence of increasing strain. Furthermore, was much stronger with respect to the background PDMS signal than for the folded region, the peak shifts further to 1.85 eV. for 3L-WS2 placed on plain PDMS [Figs. 3(b) and 1(d)]. The increase Using clustering results, the spectra were fitted either by one or in the signal intensity corresponding to XA excitons could be explained two curves, and the shift of XA (50 meV/% strain15) was used to calcu- by the bandgap renormalization due to strain. According to density late the strain distribution, as shown in Fig. 3(d). The highest strain functional theory calculations, the tensile strain lowered the energy level (>2%) is present in the folded part of the sample, in agreement level of the conduction band valley and valence band hill at the K point with previously reported studies.15 WS2 supported on the left pillar of the Brillouin zone. This resulted in the conduction band at the K shows almost no strain (<1%), and WS2 on the right pillar displays point aligned below the conduction band at the K point.15 Therefore, considerable amount of strain (1.5%–2%). Such non-uniform strain we observed the indirect-direct bandgap transition for the strained distribution is a result of the fabrication process, where the interaction samples and the intensity of the direct transition corresponding to XA between the 2D material and the transfer substrate is difficult to peak increased significantly. Interestingly, a significant increase in the control. For industrial applications, when transferring large-area 2D PL intensity from the parts of the folded WS2 layer was also measured materials for optical or optoelectronic devices, the presence of strain [for example, in the region suspended between two pillars in and the corresponding change in the electronic bandgap will be crucial Fig. 3(b)]. A similar increase was observed for wrinkled 2D materials for their correct operation. Noteworthy, the strain levels are adequately and is an experimental evidence for the funneling of photogenerated represented by clusters in Fig. 3(c). Therefore, we can use ML methods excitons toward regions of a higher strain at the top of the wrinkle for detecting high-strain points in few-layer TMDs. geometry.14,15 For comparison, in the case of few-layer WS2 suspended Finally, we investigated the low-frequency (<100 cm 1) phonon on pillars where no strain was present, such an increase in the PL modes of this sample. These modes originate from motions of atomic intensities was not observed (Fig. S7). layers instead of individual atoms and, therefore, are susceptible to Appl. Phys. Lett. 121, 253101 (2022); doi: 10.1063/5.0130927 121, 253101-4 C Author(s) 2022 V Applied Physics Letters ARTICLE scitation.org/journal/apl correlate the clusters with the strain maps calculated from the shift in the PL energies. In the case of 1L-WS2 the algorithm was able to dis- tinguish between the spectra which differed by only 0.3% strain and in the case of highly strained 3 L-WS2, the clustering results strongly resembled the strain map. Finally, we studied the strain dependence of the low-frequency Raman mode for the multilayer WS2 and did not observe any fre- quency shift up to approximately 2.5% of strain, only a significant intensity increase for the highly strained regions. Our results confirm machine learning techniques together with PL and Raman spectros- copy can provide a useful, fast, and efficient tool to analyze 2D materials. See the supplementary material for the details of the effect of aging of the sample on the PL signal and SEM images of the patterned PDMS substrates. For the 1L-WS2 on the pillar, we show the PL inten- sity and energy distribution of the XA exciton, clustering, and PCA analysis of the spectra. Finally, we show the PL spectra analysis for few-layer WS2 suspended on pillars when no strain is present. This work has been supported by the Severo Ochoa Program (No. SEV-2017-0706 funded by MCIN/AEI/10.13039/501100011033), by the FIG. 4. Low frequency Raman spectra on the original Si/SiO2 growth substrate and Spanish Ministry of Economy and Competitiveness (MINECO) under after transfer to PDMS for (a) 1L-WS2 and (b) 2L-WS2. (c) Low frequency Raman spectra of 3L-WS2 corresponding clusters from Fig. 3(b) (from the bottom: C1: left Contract Nos. PGC2018-095032-B-I00 and PID2021-124568NB-I00, pillar; C2: right pillar, and C3: suspended folded area). and by the CERCA Programme/Generalitat de Catalunya. The authors acknowledge the European Union’s H2020 FET Proactive Project interactions with surroundings such as substrate interactions or inter- TOCHA (Grant No. 824140) and the ERC-AdG Project LEIT (Grant layer coupling in 2D materials.31 For bulk WS2, shear (S) and layer No. 885689). breathing (LB) modes are present at 27.5 and 47.8 cm 1.32 These modes are not expected for 1L samples, as they require the interaction AUTHOR DECLARATIONS of at least two layers. Accordingly, the S and LB modes were not Conflict of Interest detected for 1L TMDs with an exception of WS2, where the low- The authors have no conflicts to disclose. frequency phonon mode at 27.0 cm 1 was measured.33 It has been tentatively assigned to LA (M); however, its origin in still not fully Author Contributions understood. In the study by Xiang et al., the mode was recently attrib- uted to the out-of-plane motion of tungsten and sulfur atoms, and it Marianna Sledzinska: Conceptualization (equal); Data curation was explained as an interaction between WS2 and the substrate in the (equal); Funding acquisition (equal); Investigation (equal); CVD-grown WS2.34 Contrary to the work of Xiang et al., we observed Visualization (equal); Writing – original draft (equal). Peng Xiao: clear signature of the LA (M) mode for 1L-WS2 for the as-grown sam- Data curation (equal); Investigation (equal); Visualization (equal). ple on Si/SiO2 and after the transfer to PDMS. This may suggest strong Eulalia Puig Vilardell: Data curation (supporting); Formal analysis interaction between WS2 and PDMS and could enable novel studies of (supporting); Investigation (supporting); Visualization (supporting). low frequency modes in monolayers on flexible substrates.  Emigdio Chavez-Angel: Formal analysis (equal); Methodology For 2L-WS2, we observed both S and LB modes on the growth (equal); Visualization (equal); Writing – original draft (supporting). substrate and after transfer to PDMS [Fig. 4(b)]. In the case of Maria Jose Esplandiu: Investigation (equal); Supervision (equal); 3L-WS2, we measured the S mode for both PDMS-supported and Writing – review & editing (equal). Clivia M. Sotomayor Torres: suspended samples. Spectra corresponding to different clusters, as in Conceptualization (equal); Funding acquisition (equal); Supervision Fig. 3(b), are compared in Fig. 4(c). No detectable difference in the (equal). peak frequency was observed in the strain range of 0%–2.5% from which we can conclude that for few-layer WS2, the S mode is not sen- DATA AVAILABILITY sitive to strain. In the folded regions, we observed a strong increase in the peak intensity in comparison with the central laser peak (Fig. 4, The data that support the findings of this study are available C3). Similar to the PL signal enhancement, we attribute it to electron from the corresponding author upon reasonable request. funneling into the wrinkles. In summary, we studied exciton energies and corresponding REFERENCES strain distributions for mono- and few layer CVD-grown WS2 trans- 1 F. Carrascoso, H. Li, R. Frisenda, and A. Castellanos-Gomez, Nano Res. 14, ferred on PDMS micropillars. We used clustering algorithms to group 1698 (2021). similar spectra with two main objectives. First, clustering facilitated 2 C. D. Giorgio, E. Blundo, G. Pettinari, M. Felici, F. Bobba, and A. Polimeni, the spectral deconvolution of the PL signal. Second, we were able to Adv. Mater. Inter. 9, 2102220 (2022). Appl. Phys. Lett. 121, 253101 (2022); doi: 10.1063/5.0130927 121, 253101-5 C Author(s) 2022 V Applied Physics Letters ARTICLE scitation.org/journal/apl 3 19 A. Chaves, J. G. Azadani, H. Alsalman, D. R. da Costa, R. Frisenda, A. J. Z. Ye, T. Cao, K. O’Brien, H. Zhu, X. Yin, Y. Wang, S. G. Louie, and X. Zhang, Chaves, S. H. Song, Y. D. Kim, D. He, J. Zhou, A. Castellanos-Gomez, F. M. Nature 513, 214 (2014). 20 Peeters, Z. Liu, C. L. Hinkle, S.-H. Oh, P. D. Ye, S. J. Koester, Y. H. Lee, P. Y. Chen, S. Qian, K. Wang, X. Xing, A. Wee, K. P. Loh, B. Wang, D. Wu, J. Avouris, X. Wang, and T. Low, Npj 2D Mater. Appl. 4, 29 (2020). Chu, A. Alu, P. Lu, and C.-W. Qiu, Nat. Nanotechnol. 1, 1178–1182 (2022). 4 21 B. Graczykowski, M. Sledzinska, M. Placidi, D. Saleta Reig, M. Kasprzak, F. H. Shi, H. Pan, Y.-W. Zhang, and B. I. Yakobson, Phys. Rev. B 87, 155304 (2013). 22 Alzina, and C. M. Sotomayor Torres, Nano Lett. 17, 7647 (2017). F. Wang, S. Li, M. A. Bissett, I. A. Kinloch, Z. Li, and R. J. Young, 2D Mater. 7, 5 Y. Yoo, J.-H. Yang, and J.-H. Lee, Curr. Appl. Phys. 18, 799 (2018). 045022 (2020). 6 23 A. Castellanos-Gomez, M. Poot, G. A. Steele, H. S. van der Zant, N. Agra€ıt, E. Chavez-Angel, B. Puertas, M. Kreuzer, R. Soliva Fortuny, R. C. Ng, A. and G. Rubio-Bollinger, Nanoscale Res. Lett. 7, 233 (2012). Castro-Alvarez, and C. M. Sotomayor Torres, Foods 11, 1304 (2022). 7 24 S. Bertolazzi, J. Brivio, and A. Kis, ACS Nano 5, 9703 (2011). K. F. Mak, K. He, C. Lee, G. H. Lee, J. Hone, T. F. Heinz, and J. Shan, Nat. 8 K. Bertoldi, V. Vitelli, J. Christensen, and M. van Hecke, Nat. Rev. Mater. 2, Mater. 12, 207 (2013). 25 17066 (2017). H. Zobeiri, S. Xu, Y. Yue, Q. Zhang, Y. Xie, and X. Wang, Nanoscale 12, 6064 9 X. He, H. Li, Z. Zhu, Z. Dai, Y. Yang, P. Yang, Q. Zhang, P. Li, U. (2020). 26 Schwingenschlogl, and X. Zhang, Appl. Phys. Lett. 109, 173105 (2016). M. Sledzinska, B. Graczykowski, M. Placidi, D. S. Reig, A. E. Sachat, J. S. 10 H. J. Conley, B. Wang, J. I. Ziegler, R. F. Haglund, S. T. Pantelides, and K. I. Reparaz, F. Alzina, B. Mortazavi, R. Quey, L. Colombo, S. Roche, and C. M. S. Bolotin, Nano Lett. 13, 3626 (2013). Torres, 2D Mater. 3, 035016 (2016). 11 27 K. He, C. Poole, K. F. Mak, and J. Shan, Nano Lett. 13, 2931 (2013). B. Liu, W. Zhao, Z. Ding, I. Verzhbitskiy, L. Li, J. Lu, J. Chen, G. Eda, and K. P. 12 A. McCreary, R. Ghosh, M. Amani, J. Wang, K.-A. N. Duerloo, A. Sharma, K. Loh, Adv. Mater. 28, 6457 (2016). 28 Jarvis, E. J. Reed, A. M. Dongare, S. K. Banerjee, M. Terrones, R. R. Namburu, P. K. Chow, R. B. Jacobs-Gedrim, J. Gao, T.-M. Lu, B. Yu, H. Terrones, and N. and M. Dubey, ACS Nano 10, 3186 (2016). Koratkar, ACS Nano 9, 1520 (2015). 13 29 M. Sledzinska, G. Jumbert, M. Placidi, A. Arrighi, P. Xiao, F. Alzina, and C. M. A. C. Ferrari, J. C. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri, F. Mauri, S. Sotomayor Torres, ACS Appl. Electron. Mater. 2, 1169 (2020). Piscanec, D. Jiang, K. S. Novoselov, S. Roth, and A. K. Geim, Phys. Rev. Lett. 14 S. Deng, S. Che, R. Debbarma, and V. Berry, Nanoscale 11, 504 (2019). 97, 187401 (2006). 15 30 K. P. Dhakal, S. Roy, H. Jang, X. Chen, W. S. Yun, H. Kim, J. Lee, J. Kim, and S. Sudhindra, F. Rashvand, D. Wright, Z. Barani, A. D. Drozdov, S. Baraghani, J.-H. Ahn, Chem. Mater. 29, 5124 (2017). C. Backes, F. Kargar, and A. A. Balandin, ACS Appl. Mater. Interfaces 13, 16 J. Chaste, A. Missaoui, S. Huang, H. Henck, Z. Ben Aziza, L. Ferlazzo, C. 53073 (2021). 31 Naylor, A. Balan, A. T. C. Johnson, R. Braive, and A. Ouerghi, ACS Nano 12, X. Cong, X.-L. Liu, M.-L. Lin, and P.-H. Tan, Npj 2D Mater. Appl. 4(1), 13 (2020). 32 3235 (2018). L. Liang, J. Zhang, B. G. Sumpter, Q.-H. Tan, P.-H. Tan, and V. Meunier, ACS 17 K. Wang, A. A. Puretzky, Z. Hu, B. R. Srijanto, X. Li, N. Gupta, H. Yu, M. Tian, Nano 11, 11777 (2017). 33 M. Mahjouri-Samani, X. Gao, A. Oyedele, C. M. Rouleau, G. Eres, B. I. M. O’Brien, N. McEvoy, D. Hanlon, T. Hallam, J. N. Coleman, and G. S. Yakobson, M. Yoon, K. Xiao, and D. B. Geohegan, Sci. Adv. 5, eaav4028 Duesberg, Sci. Rep. 6, 19476 (2016). 34 (2019). Q. Xiang, X. Yue, Y. Wang, B. Du, J. Chen, S. Zhang, G. Li, C. Cong, T. Yu, Q. 18 L. Yuan and L. Huang, Nanoscale 7, 7402 (2015). Li, and Y. Jin, Nano Res. 14, 4314 (2021). Appl. Phys. Lett. 121, 253101 (2022); doi: 10.1063/5.0130927 121, 253101-6 C Author(s) 2022 V