Effects of interfacial structure and polarity on charge transfer between carbonaceous nanomaterials and rutile (110) surface

Electronic Structure and Charge Transfer in the TiO2 Rutile (110)/Graphene Composite Using Hybrid DFT Calculations

“…Note that this alignment of Ti and carbon states is different from several recently published studies of TiO 2 –graphene interfaces, where the graphene conduction band edge is usually found to lie near, just below, or just above the TiO 2 band edge. ,,,, There is, however, no agreement on the alignment of C- and Ti-based conduction band states in the published studies, and there are examples of C states being lower in energy than Ti states, similar to the results of this work, notably in graphene interfaces with anatase (001) , and rutile (110) . This difference is likely to be caused by differences in the electronic properties of anatase and rutile polymorphs of TiO 2 (indeed, the CB of anatase is believed to be 0.2–0.4 eV below that of rutile , ).…”

“…Although the Dirac point has not been captured in our DoS plots (Figure 6 and Figure S3), it is clearly present in the band structure. The apparent band gap in the DoS, similar to the small 44,46,47 or very small 50 band gaps observed in some of the previous studies of TiO 2 /graphene interfaces, has likely been caused by the use of insufficiently dense k-point grids, similar to the early studies of strained isolated graphene alone. 86,94 The absence of a band gap means that electrons can be easily promoted to the unoccupied graphene states.…”

“…41,44,47,49,50 There is, however, no agreement on the alignment of C-and Ti-based conduction band states in the published studies, and there are examples of C states being lower in energy than Ti states, similar to the results of this work, notably in graphene interfaces with anatase (001) 45,48 and rutile (110). 46 This difference is likely to be caused by differences in the electronic properties of anatase and rutile polymorphs of TiO 2 (indeed, the CB of anatase is believed to be 0.2−0.4 eV below that of rutile 113,114 ). It is also likely that the TiO 2 band positions in the DFT+U calculations 41,[44][45][46]48 are affected by the choice of the "U" parameter which is applied to describe the on-site Coulomb interaction of Ti 3d electrons.…”

“…These key studies were followed by a number of further studies of the interfaces of graphene with rutile 46 and anatase 44,45,47−50 polymorphs of TiO 2 and other oxides, such as CeO 2 51 and SrTiO 3 , 52 as well as several studies of small TiO 2 clusters on graphene 53−58 and graphene-like flakes on TiO 2 surfaces, 59,60 and larger TiO 2 /graphene/dye heterostructures. 61 Notably, several of these works confirm the role of graphene as a photosensitizer, leading to photoinduced charge transfer from graphene to TiO 2 , 44,50 similar to the first computational studies 41,43 and in agreement with experiments.…”

“…In particular, for this system it has been common for studies to employ pure DFT with a Hubbard`U' correction, 65 which can correctly reproduce the band gaps and band positions of TiO 2 , but where the choice of the parameter U can be quite ambiguous (values ranging from 4.0 to 5.5 eV 41,43,45,46,56,60 and even as high as 9.0 eV 44,48 have been used in studies of TiO 2 -graphene interfaces). Hybrid functionals would be a more reliable choice, 66 68 which allow consistent comparison of TiO 2 -based systems.…”