Temperature-Dependent Circuit Modeling and Performance Evaluation Due to Crosstalk in Capacitively Coupled Interconnects of Intercalation-Doped Multilayer Graphene Nanoribbon
Tajinder Kaur,Mayank Kumar Rai,Rajesh Khanna
DOI: https://doi.org/10.1007/s13369-020-04892-x
IF: 2.807
2020-09-04
Arabian Journal for Science and Engineering
Abstract:The present work explores the temperature-dependent performance evaluation due to crosstalk in coupled interconnects of intercalation-doped multilayer graphene nanoribbons (ID-MLGNR). A temperature-dependent equivalent single conductor (ESC) model of arsenic pentafluoride (AsF<sub>5</sub>) ID-MLGNR is considered and is employed to analyze its crosstalk performance (i.e., crosstalk-induced noise and crosstalk-induced delay). Other metallic interconnect materials viz. mixed carbon nanotube bundles (MCB) and copper (Cu) are also analyzed in a similar manner. The temperature-dependent crosstalk performance of ID-MLGNR is compared with that of MCB and Cu, at different temperatures from 300 to 500 K. For the specified temperature range, in case of crosstalk-induced noise, the AsF<sub>5</sub> ID-MLGNR-based long interconnects demonstrate superior performance to its MCBs and Cu counterparts. Moreover, the temperature-dependent crosstalk-induced delay is less with ID-MLGNR in comparison with both MCB and Cu interconnects. The large values of the mean free path (MFP) and conductivity of AsF<sub>5</sub> ID-MLGNR contribute to its improved performance. In addition, the impact of variation in MFP, interlayer spacing and Fermi level is studied on the crosstalk delay performance of ID-MLGNR. From the obtained results, it is revealed that variation in these parameters results in the significant variation of 62 ps, 51 ps and 169 ps, respectively, in the odd mode (OM) crosstalk delay. Hence, for emerging nano-technology nodes (e.g., 14 nm), ID-MLGNR interconnects with its improved crosstalk performance are proposed as a potential alternative to its MCBs and Cu counterparts.
multidisciplinary sciences