The in situ tensile experiment in the transmission electron microscope (TEM) for two-dimensional (2D) materials is a very challenging experiment in the area of nanoscience. Many groups have attempted to perform a tensile test in TEM in order to study the mechanical strength and the deformation at nanoscale by using different techniques.
Bimetal technique  is a tensile technique to generate the straining by using a heating holder. This technique is excellent to image deformation during straining but it cannot obtain the mechanical strength. Wei et al. performed tensile test on graphene and Boron Nitride (BN) by using TEM-AFM holder to obtain the mechanical strength  but they could not obtain the atomic deformation during tensile. However, they fabricated the 2D nanomaterials from nanotube using electron beam irradiation inside TEM. Push-to-pull (PTP) device and a pico indenter are often used to obtain the strength of the materials but it could not image the atomic deformation during tensile. Exfoliating 2D materials and preparing the tensile test have become major challenges. Recently, Kim et al. has successfully conducted the tensile test on few layers of graphene on PTP by using gel method . Their report described very well the challenges and difficulties in performing in situ tensile test for 2D materials.
We performed tensile test for 2D materials using our homemade tensile stage and the main goal was to verify whether the 2D materials will fail in a brittle or in a ductile mode depending on the orientation of the hexagonal structure with respect to tensile direction. These 2D materials possess a hexagonal crystal structure which has a very strong covalent bond among the atoms in the basal plane and has a relatively weak Van der Waals interaction between the layers. We did tensile tests on graphene, graphene oxide, boron nitride (BN), ball milled BN, WS2 and MoS2. There were some theoretical and simulation studies which predicted that 2D materials could fail in ductile or in brittle mode depending on the orientation of the hexagonal structure of the material with respect to the tensile direction [4, 5].
We would like to test this prediction with our tensile experiment.
Method and Materials
We have developed a homemade tensile stage that can be used for a straining holder in TEM as shown in figure 1a. This tensile stage enabled us to pull the sample in certain direction relative to the hexagonal structure of the material. There are two main parts of the tensile stage. The first part was made of 100 μm sheet of copper. It was cut into a rectangular shape with three holes in it to fit the straining holder. The second part is a smaller copper bridge made from a commercial TEM grid that has a 50 μm hole at the center.
Initially, a thin holey pre-patterned aluminum foil was attached by using an epoxy to the commercial grid and to cover its hole as shown in the inset of figure 1a. After sonication using a method described by Zhou et al. , the solution of 2D material was dropped on the foil. We did not use Platinum deposition to attach the flakes on the foil. The flakes would be attached relatively strong to the surface of the foil. Some very thin flakes of 2D material could be easily found covering the holes of the foil. Initial examination for the structure and orientation of the flakes was done by using TEM. After the desired orientation of the thin 2D nanomaterial was determined then the grid was cut into a rectangular piece with the longer side is parallel to the tensile direction and it was attached to the bigger copper piece by using an epoxy. A focused ion beam (FIB) was used to cut the thin flake into a dog bone shape parallel to tensile direction and also to mill the flake (fig. 1b). Some slits on the foil nearby the sample were also made by using FIB. This was to ensure the stress will be concentrated near the sample when the tensile stress is applied. Once finished with FIB, the tensile test is ready to be performed in TEM.
Results and Discussion
Our technique has successfully performed tensile tests on 2D materials including graphene, graphene oxide, BN, ball milled BN, WS2 and MoS2. Figure 2a showed the initial condition of the sample. The thickness of the sample was estimated around 20 nm by using thickness map. The sample was pulled with the strain rate around 10-4/s. Upon straining, the sample experienced several surface deformations. This surface deformation was due to some curled or wavy profiles on the surface when they are stretched under tensile straining. Prior to the failure, these surface deformations occurred more rapidly. We observed that these 2D materials would always fail in brittle mode and it was independent to the orientation of the hexagonal structure with respect to the tensile direction. We did not observe any necking process and the brittle fracture occurred in a very short time (less than 0.05 s). Figure 2b showed the sample after failure. We have performed more than 30 tensile experiments and all of them ended with a brittle fracture. These experimental results did not seem to agree with some theoretical and simulation studies which suggested the materials could fail in ductile mode when the tensile direction is parallel to the armchair or zig-zag direction of the sheet . Figure 3a showed the broken layers of the sample after failure and the fracture edge was shown in figure 3b. Interestingly, we observed that the zig-zag edge occurred more often compared to the armchair edge. The fracture in certain layer could also occur at relatively far from the other layers. When this occurred, one layer could slide on top of the others and the Moiré pattern was created as the layers slide. After failure in a brittle mode, usually it was followed by buckling on the edge of fracture and eventually the edge showed a turbostratic structure. We also performed tensile tests on BN nanosheets obtained by using ball-milling process. Ball-milled BN usually has multilayer structure and some structural defects that could potentially behave differently as the tensile is performed . However, we observed that the ball milled BN also failed in brittle mode experimentally. Nevertheless, in our early experiments with BN, we did observe a few of ductile fractures. Figure 3c showed the ductile tip of the sample after tensile test. The ductile fracture was actually originated from the aluminum layer deposited on the BN surface as we used FIB to cut the foil and the dog bone shape. During galium ion bombardment, the ejected aluminum would redeposit on the surface of BN and formed a polycrystal aluminum layer with the grain size around 10 to 50 nm. When aluminum and BN became in contact, they created a strong interlayer bonding and became a BN-Al nanocomposite. Therefore, it is always important to mill the sample using FIB to clean the 2D material from other redeposited materials and to minimize the contamination. After milling and cleaning the 2D material, the 2D materials failed in brittle mode due to the tensile. It is actually surprising to learn from the discovery of graphene up to present that in the area of tensile test, we have not reached a satisfying condition for 2D materials due to its very difficult preparation. We still could not pull a single layer of 2D material in TEM. Nevertheless, we have learned that 2D material would fail in a brittle mode by pulling few layers of it. However, this field is still widely open and every milestone will always give important contribution for future efforts.
We have performed the in situ tensile test on 2D nanomaterials in the TEM. In general, these materials would fail in brittle mode experimentally and it is independent to the orientation of the hexagonal structure to the tensile direction.
The tensile experiments were performed at the Australian Centre for Microscopy and Microanalysis (ACMM) at the University of Sydney, NSW, Australia. The author would like to thank Prof. Xiaozhou Liao, Dr. Yanbo Wang and Dr. Hongwei Liu for the supports. Prof. Ying Chen from Institute for Frontier Materials (IFM) at Deakin University (Victoria, Australia), Dr. Shaikh Nayeem Faisal and Dr. David Mitchell from the University of Wollongong (NSW, Australia) were gratefully acknowledged for providing samples and STEM operation.
1School of AMME, The University of Sydney, NSW, Australia.
2Macquarie University International College, NSW, Australia.
Dr. Ade Kismarahardja
Macquarie University International College
 Y. Yue, P. Liu, Q. Deng, E. Ma, Z. Zhang, X. Han, Quantitative evidence of crossover toward partial dislocation mediated plasticity in copper single crystalline nanowires, Nano Lett 12(8) 4045-9 (2012) doi 10.1021/nl3014132
 X. Wei, S. Xiao, F. Li, D.M. Tang, Q. Chen, Y. Bando, D. Golberg, Comparative fracture toughness of multilayer graphenes and boronitrenes, Nano Lett 15(1) 689-94 (2015) doi 10.1021/nl5042066
 K. Kim, J.C. Yoon, J. Kim, J.H. Kim, S.W. Lee, A. Yoon, Z. Lee, Dedicated Preparation for in situ transmission electron microscope tensile testing of exfoliated graphene, Applied Microscopy 49(3) (2019) doi org/10.1007/s42649-019-0005-5
 Y.I. Jhon, Y.M. Jhon, G.Y. Yeom, M.S. Jhon, Orientation dependence of the fracture behavior of graphene, Carbon 66, 619-628 (2014) doi 10.1016/j.carbon.2013.09.051
 X. Qi-lin, L. Zhen-huan, T. Xiao-geng, The defect-induced fracture behaviors of hexagonal boron-nitride monolayer nanosheets under uniaxial tension, Journal of Physics D: Applied Physics 48(37) 375502 (2015) doi 10.1088/0022-3727/48/37/375502
 K.G. Zhou, N.N. Mao, H.X. Wang, Y. Peng, H.L. Zhang, A mixed-solvent strategy for efficient exfoliation of inorganic graphene analogues, Angew Chem Int Ed Engl 50(46) 10839-42 (2011) doi 10.1002/anie.201105364
 W. Xia, L. Ruiz, N.M. Pugno, S. Keten, Critical length scales and strain localization govern the mechanical performance of multi-layer graphene assemblies, Nanoscale 8(12) 6456-62 (2016) doi 10.1039/C5NR08488A
 D. Lahiri, V. Singh, L.H. Li, T. Xing, S. Seal, Y. Chen, A. Agarwal, Insight into reactions and interface between boron nitride nanotube and aluminum, Journal of Materials Research 27(21) 2760-2770 (2012) doi org/10.1557/jmr.2012.294
More information on TEM
More about tensile testing