Rate‐Dependent Decohesion Modes in Graphene‐Sandwiched Interfaces

Abstract

Practical industrial mass production of macrosized graphene has been realized with the development of the chemical vapor deposition (CVD) method, envisioning a wide range of potential applications in microelectronic devices.[1–3] However, as a key step in the fabrication process, the successful transfer of highquality macrosized graphene to a specific target substrate of these devices continues to be a challenge. Compared with the typical transfer method of graphene, which often involves wet chemical etching, the etching-free mechanical dry-transfer process is fast, renewable, cost-competitive, and environmentally friendly.[4,5] The adhesion energy of graphene to various substrate materials, as a key parameter characterizing the mechanical resistance to delamination of the graphene/substrate interface, has been shown to be a critical factor determining the quality of the transferred graphene and the performance of graphene-based devices.[5] Therefore, the development of an appropriate method for measuring the adhesion energy of the graphene/substrate interface is essential for large-scale fabrication and device applications of graphene. In the last decade, significant progress has been made in experimental investigations characterizing the interfacial properties of graphene using the shear-lag method,[6–10] the blister tests,[11–16] the double cantilever beam (DCB) fracture tests,[5,17,18] and nanoindentation methods.[19,20] Shearlag methods are often employed to study sliding and shear interactions of the graphene/substrate interfaces, whereas the adhesion energy is typically measured using the DCB and blister tests. For experimental investigation of the adhesive interactions, Koenig et al.[11] performed a pressurized blister test by creating a pressure difference across the graphene membrane to directly measure the adhesion energy between graphene and a silicon-oxide substrate and obtained a value of Γg/Si = 0.45 ± 0.02 J m−2, whereas Wang et al.[12] used the same method to obtain a smaller value of Γg/Si = 0.19 ± 0.02 J m−2. Recently, Xin et al.[13] measured the adhesion energy of the as-grown graphene on copper foil of different roughness using a blister test and obtained values Mechanical dry transfer of large-area graphene is increasingly applied in fabrication of graphene-based electronic devices, and adhesion energy of graphene/substrate interface is a key factor affecting reliability and performance of these devices. Herein, the adhesion energy of a graphene/poly(ethylene terephthalate) (PET) interface is measured by widely adopted double cantilever beam (DCB) fracture tests. Results show that the apparent adhesion energy of sandwiched interface is highly rate-dependent. When separation rate increases from 20 to 150 μm s−1, apparent adhesion energy increases by an order of magnitude. By examining fractured interfaces after DCB tests with micro-Raman spectroscopy, the graphene is found to be fractured and transferred in fragments, with residual tensile strain up to 3% for high separation rates. The results are contrary to earlier reports, where higher separation rate in dry-transfer process would typically enhance the dry transfer of graphene, resulting in better integrity and performance. Based on Raman spectroscopy measurements, three distinct decohesion modes are identified for PET-/ graphene-/adhesive-sandwiched interface, which consistently explain the rate-dependent apparent adhesion energy. The complicated decohesion modes also suggest that an optimal separation rate should be used to properly measure the adhesion energy and improve the dry-transfer technique of graphene with minimum damage and residual strain.