Graphene is one of the stiffest and thinnest materials in the world. However, the mechanical properties of monolayer graphene pieces bigger than a few micrometers have never been tested, simply because moving such an ultrathin film to a standard testing apparatus has not been possible. Researchers at the Center for Multidimensional Carbon Materials, within the Institute for Basic Science (IBS), have managed to measure the tensile strength of centimeter-scale monolayer graphene films, using camphor - a chemical that easily volatilizes at room temperature - as a temporary support layer. Published in Advanced Materials, this trick could be applied to transfer and study other challenging 2D materials.

Recently, new and impressive characteristics of ultrathin films have been unveiled, with promising applications in microelectromechanics, optoelectronics, and biology. For example, the combination of graphene’s transparency, impermeability, conductivity, and elasticity could be used for flexible electronics, transparent protective coatings, and barrier films.

Measuring the tensile strength of materials involves pulling samples until they break. However, testing 2D materials is a complex task. Traditionally, a layer of another material (substrate) is used as a “tray”, to help with the transfer and to give these ultrathin films some support. Such substrates have pros and cons: there is the risk of damaging the film when the substrate is peeled away, or if the substrate layer is too thick, distinguishing the properties of the substrate from the ones of the 2D material of interest becomes impossible.

In this study, camphor is used as a transient support, and what differentiates it from conventional methods is that it is sublimed away in air at room temperature naturally, or at higher temperatures for faster processing. Thanks to this method, ultrathin films with an area larger than 1 cm x 1 cm are transferred without damage, then the camphor layer disappears in the air without leaving traces. In this way, tensile measurements were made on centimeter-scale 300 nm-thick graphene oxide film specimens, almost ten times thinner than previously reported. It was also possible to work with a graphene oxide film that was only 35 nm thick, and suspend it over a 1 cm x 1 cm hole.

Figure 1: Schematics of the newly developed process of transferring a thin film composed of stacked and overlapped graphene oxide platelets using camphor as transient substrate. Initially, a copper (Cu) foil is used as substrate to support the thin film. (i) Camphor is deposited onto the graphene oxide film to form a solid layer with thickness of around 500 μm. (ii) The sandwich (camphor/graphene oxide film/Cu) is immersed into 0.1 m (NH₄)₂S₂O₈, which etches away the copper layer. (iii) The camphor/graphene oxide film assembly is dried in air and transferred onto a hollow substrate. (iv) The camphor is removed naturally, by exposing the sample to air for 24 hours so that the camphor layer sublimed away leaving the thin graphene oxide film suspended on the hollow substrate. Thanks to this method, IBS scientists could transfer a 35 nanometer thick graphene oxide film to a hollow substrate with a 1 cm x 1 cm large hole.

Figure 2: Transfer and tensile testing. A “dog-bone”- shaped graphene oxide film specimen with dimensions of 12 mm (gauge length) × 2 mm (gauge width), and thickness of around 300 nm, is transferred to a frame with a hole in the center, where tensile tests were carried out.

Moreover, the IBS researchers tested the new camphor-method with one-atom-thick graphene films. In this case, stretching a freestanding monolayer 2D film is not yet possible, and a polycarbonate (PC) film is used as a support layer. Without camphor, the thinnest PC/monolayer graphene assembly tested was around 1 μm thick, and graphene’s mechanical properties could not be obtained because the thick PC film dominates the response. However, IBS scientists managed to distinguish monolayer graphene properties using the camphor method and a 100 nm thick PC substrate.

Figure 3: Comparing stress-strain responses between polycarbonate–monolayer graphene (PC–G) assemblies with PC thickness of 1 μm or 100 nm. a) In the 1 μm thick sample, the PC substrate dominates the response and graphene properties do not emerge. b) In the 100 nm thick sample, a clear difference is visible in the samples with and without graphene, thus graphene properties can be distinguished. Thanks to the camphor technique, it is possible to transfer and test thinner assemblies, and thus uncover more mechanical properties of monolayer graphene.

The tensile tests demonstrated that these large-scale samples do have high stiffness. The team measured graphene mechanical properties in terms of the Young’s modulus, which characterizes the intrinsic stiffness of a material. The study reports that centimeter-scale polycrystal monolayer graphene has a Young’s modulus of 637-793 GPa, but that single-crystal graphene had high-end values at or near 908 GPa. As a matter of comparison, a block of high strength steel has a Young’s modulus of about 200 GPa, but accounting for the higher density of steel, that is, corrected for weight, the intrinsic stiffness of large-scale graphene is about 20 times higher.

“I think this method might become a standard around the world for testing single layer 2D materials, including graphene at length scales meaningful for applications,” notes Rod Ruoff, director of the IBS Center for Multidimensional Carbon Materials.

Now that the team has this technique in hand, their next challenge is to produce graphene sheets that are not only astonishingly stiff, but also incredibly strong. “Strong carbon-carbon bonds confer graphene with an impressive ideal strength, which is however lowered if there are defects in its structure,” explains first author Bin Wang. “The ideal strength of graphene is about 130 GPa, second only to the ideal strength of diamond along a particular crystallographic direction (240 GPa). But these are calculated strengths, and are never obtained for “large” specimens, like a centimeter or larger—because such specimens always have defects, also called flaws, like small cracks. Now that we have this new method for measuring large graphene specimens, we can begin to understand why the strength for our graphene samples, even the single crystal graphene samples, is typically around 8 GPa, and not higher. Understanding this will provide us a “road map” to make graphene that is much stronger—that breaks at a much higher value of percent elongation,” continues Ruoff. “And when we do that—that will change the world.”

Letizia Diamante