GNR Interconnects Research Articles

Variable Length and Temperature-Dependent Analysis of MLGNR at Nano-Scale Regime

In this paper, a thermally aware equivalent single conductor is proposed, along with analytical modeling to evaluate the parasitic parameters of multilayer graphene nanoribbon (MLGNR) as interconnect, and its performance is analyzed in terms of delay and power delay product (PDP) for 32[Formula: see text]nm, 22[Formula: see text]nm and 16[Formula: see text]nm technology nodes at variable global interconnect lengths (500–2000[Formula: see text]m). It was examined that with rising temperature, there is a strident decrease in the mean free path (MFP) of GNR interconnect, which further influences its own resistance at global length (2000[Formula: see text][Formula: see text]m) for all three technology nodes. The simulation tool Simulation Program with Integrated Circuit Emphasis (SPICE) is used to estimate and compare MLGNR performance in terms of signal delay and PDP for three different technology nodes. It is revealed from the outcomes that the propagation delay and PDP increase at long interconnects (500–2000[Formula: see text][Formula: see text]m) over a temperature range of 200–500[Formula: see text]K for deep submicron technology nodes (16[Formula: see text]nm, 22[Formula: see text]nm and 32[Formula: see text]nm). A similar investigation was performed on the copper interconnect, and it was discovered that the MLGNR performs better in terms of delay and PDP at global levels with a temperature range of 200–500[Formula: see text]K for nano-scaled technology nodes (32[Formula: see text]nm, 22[Formula: see text]nm and 16[Formula: see text]nm).

Signal Integrity Assessment of GNRFET-Based Ternary Logic for Multi Layered GNR Interconnects with Dielectric Insertion

In this paper, signal integrity assessment is carried out for graphene nanoribbon field effect transistor (GNRFET) based ternary logic with dielectric inserted multi-layered GNR (MLGNR) interconnects. The analyses are carried out for crosstalk effects and eye diagrams with and without shielding lines. In this paper firstly, it is observed that the dielectric inserted MLGNR interconnects show better than copper (Cu) and multiwall carbon nanotube (MWCNT) interconnects. Secondly, active shield technique is adopted, and observed that it exhibits better performance than without shield and passive shield techniques. Also, the power-delay product performance parameter is evaluated and envisaged that active shield technique outperforms passive technique. Further, the eye diagram analysis is carried out for different bit rates. The different performance analyses in the paper have been carried out for 10 nm technology node.

Comparative Analysis of Crosstalk Effects in Dielectric Inserted Horizontal andVertical Multi-layer GNR Interconnects for Ternary Logic System

In this work, the performance of copper (Cu), dielectric inserted horizontal graphene nanoribbon (Di-HGNR) interconnect and dielectric inserted vertical graphene nanoribbon (Di-VGNR) interconnects are investigated using active shielding and passive shielding techniques. However, the analysis is carried out by adapting driver-interconnect-load system. This analysis considers the interconnect length from 500 to 2000 μm for 10 nm technology node. Further, the crosstalk induced effects on various interconnect structures are examined. It is envisaged that Di-VGNR exhibits lowest propagation delay compared to Cu and Di-HGNR. Further, the in-phase and out-phase crosstalk delay among the coupled interconnect lines is determined. It is investigated that active shielded Di-VGNR has least crosstalk induced delay compared to other interconnect structures considered in this study. Therefore, Di-VGNR interconnects outperforms Cu and Di-HGNR and are best suited for future VLSI interconnects.

Performance and signal integrity analysis of intercalation‐doped MLVGNR interconnects

In this work, signal integrity, power consumption and stability of multilayer vertical graphene nanoribbon (MLVGNR) interconnects are determined and compared with multilayer horizontal GNR (MLHGNR) and conventional copper interconnects for different intercalation dopants and specularity indices. Lithium-intercalated MLVGNR outperforms these conventional interconnects in terms of delay. However, it is not very much immune to noise and also consumes more power but overall noise delay product and power delay product are least among all. As far as stability is concerned, MLVGNR interconnects have a comparable gain and phase margin and a steep transient response without an undershoot as compared with other interconnect configurations, which experience undershoot. It is worth to point out that MLVGNR interconnects are more amenable to scaling compared with other alternatives. This study for the first time analyses and compares the performance, signal integrity and stability of MLVGNR, MLHGNR and copper interconnects.

Thermal Stability Analysis of Graphene Nano-ribbon Interconnect and Applicability for Terahertz Frequency

In this paper, we have studied the thermal stability of GNR and Cu interconnects for terahertz frequency range. The top-contact (TC) and side-contact (SC) GNR interconnects are analyzed for stability, RC-delay, and power consumption at different interconnect lengths ($$10\,{\upmu }\hbox {m}$$, $$50\,{\upmu }\hbox {m}$$ and $$100\,{\upmu }\hbox {m}$$) and different chip operating temperatures (233 K, 300 K and 378 K). The RC-delay and power consumption are analyzed with the help of transmission line model (TLM) using 16nm PTM-HP standard CMOS model. In our analysis, we have shown that the RC-time constant for SC-GNR is $$\sim (3{-}7)\times $$ and $$\sim (2{-}6)\times $$ less than that of TC-GNR and Cu interconnects. Similarly, we have shown that the power consumption for SC-GNR is $$\sim (4{-}7)\times $$ and $$\sim (2{-}5)\times $$ less than that of TC-GNR and Cu. In terms of stability, SC-GNR interconnect is more stable at higher frequency range ($$>10^{12}$$ Hz) compared with TC-GNR and Cu for $$10\,{\upmu }\hbox {m}$$ interconnect length and three different chip operating temperatures.

Thermally Aware Modeling and Performance Analysis of MLGNR as On-Chip VLSI Interconnect Material

In this study, a thermally aware electrical equivalent single conductor along with an analytical model to evaluate the parasitic parameters of multilayer graphene nanoribbon (MLGNR) as an on-chip very-large-scale integration interconnect is proposed and its performance is analyzed in terms of delay, power dissipation, and power delay product (PDP) with variable temperatures ranging from 200 K to 500 K for 32 nm, 22 nm, and 16 nm technology nodes. It was observed that with a rise in the temperature, there is a sharp decrease in the mean free path of the GNR interconnect, which further dominates its own resistance at variable global lengths (500–2000 μm) for all three technology nodes. The Simulation Program with Integrated Circuit Emphasis (SPICE) simulation tool is used to estimate and compare the performance of MLGNR in terms of signal delay, power dissipation, and PDP for three different technology nodes. It is revealed from the results that the propagation delay and PDP increase with increasing temperature. A similar analysis was also done for the copper interconnect and it was found that the MLGNR gives better performance in terms of delay, power dissipation, and PDP, at long interconnects (2000 μm) over a temperature range of 200–500 K for deep submicron technology nodes (32 nm, 22 nm, and 16 nm).

Analysis of a temperature-dependent delay optimization model for GNR interconnects using a wire sizing method

A temperature-dependent delay optimization model for a multilayered graphene nanoribbon (MLGNR) with top contact (TC-GNR), side contact (SC-GNR), and Cu-based nano-interconnects using a wire sizing method was applied to determine the delay for different interconnects widths (11 nm, 16 nm, and 22 nm) and lengths (10 μm, 50 μm, and 100 μm), being the first such model for TC-GNR, SC-GNR, and Cu interconnects applied at three different chip operating temperatures (233 K, 300 K, and 378 K). The results reveal that the SC-GNR requires ~ 3–6× and ~ 2–3× fewer repeaters w.r.t. the TC-GNR or Cu interconnect, and that the SC-GNR and Cu interconnects can achieve ~ 4–5× and ~ 2–2.5× reduction in repeater dimension compared with the TC-GNR interconnect. Meanwhile, the SC-GNR interconnect can achieve 73× less propagation delay w.r.t. the TC-GNR interconnect for interconnect width of 22 nm, interconnect length of 10 μm, and two different chip operating temperatures of 233 K and 300 K. Similarly, the Cu interconnect can achieve 6× less propagation delay w.r.t. the TC-GNR interconnect at interconnect width of 22 nm and 16 nm, interconnect length of 10 μm, and 300 K.

Modeling and Performance Analysis of MLGNR Interconnects

In this work, we have presented the temperature-dependent analytical time domain model for top-contact multilayer graphene nanoribbon (TC-MLGNR) and side-contact multilayer graphene nanoribbon (SC-MLGNR) interconnects for 16[Formula: see text]nm technology node. Using this analytical model, the effective mean free path (MFP) is calculated for different temperatures and then the resistance of GNR interconnect is calculated. The lower resistance of MLGNR is one of the important factors to reduce interconnect delay. The equivalent capacitance for TC-MLGNR is also calculated. It is observed that the performance of graphene interconnects seriously deteriorates due to the presence of the interlayer capacitance. The presence of this interlayer capacitance increases the equivalent capacitance which is the dominant factor that inhibits the performance of TC-MLGNR interconnects. Further, the delay ratio between copper and TC-MLGNR for different interconnect lengths and for three different temperatures (233[Formula: see text]K, 300[Formula: see text]K, 378[Formula: see text]K) is calculated. It is observed that for longer interconnect lengths, the improvement in delay in TC-MLGNR is less as compared to traditional copper-based interconnect at low temperature. Further, power delay product (PDP) of copper and TC-MLGNR for different interconnect lengths and for three different temperatures is also calculated. It is shown that TC-MLGNR interconnects have better PDP than copper interconnects. The crosstalk analysis is performed to estimate the noise and overshoot/undershoot in TC-MLGNR and SC-MLGNR interconnects. It is shown that SC-MLGNR interconnect has better performance as far as the crosstalk is concerned as compared to that of Cu and TC-MLGNR interconnects.

Modeling and Performance Analysis of Graphene Nanoribbon Interconnects

Attempts have been made to propose a compact resistive model for graphene nanoribbon interconnects. Top-contact (TC-GNR) and side-contact (SC-GNR) GNR interconnects have been modeled. The number of conduction channels, mean free path, and resistance are calculated for different Fermi energies and interconnect widths. Present study showed that the resistance of SC-GNR is ~10× less and delay is ~2–16× less than that of TC-GNR.

Intercalation Doped Multilayer-Graphene-Nanoribbons for Next-Generation Interconnects.

Copper-based interconnects employed in a wide range of integrated circuit (IC) products are fast approaching a dead-end due to their increasing resistivity and diminishing current carrying capacity with scaling, which severely degrades both performance and reliability. Here we demonstrate chemical vapor deposition-synthesized and intercalation-doped multilayer-graphene-nanoribbons (ML-GNRs) with better performance (more than 20% improvement in estimated delay per unit length), 25%/72% energy efficiency improvement at local/global level, and superior reliability w.r.t. Cu for the first time, for dimensions (down to 20 nm width and thickness of 12 nm) suitable for IC interconnects. This is achieved through a combination of GNR interconnect design optimization, high-quality ML-GNR synthesis with precisely controlled number of layers, and effective FeCl3 intercalation doping. We also demonstrate that our intercalation doping is stable at room temperature and that the doped ML-GNRs exhibit a unique width-dependent doping effect due to increasingly efficient FeCl3 diffusion in scaled ML-GNRs, thereby indicating that our doped ML-GNRs will outperform Cu even for sub-20 nm widths. Finally, reliability assessment conducted under accelerated stress conditions (temperature and current density) established that highly scaled intercalated ML-GNRs can carry over 2 × 108 A/cm2 of current densities, whereas Cu interconnects suffer from immediate breakdown under the same stress conditions and thereby addresses the key criterion of current carrying capacity necessary for an alternative interconnect material. Our comprehensive demonstration of highly reliable intercalation-doped ML-GNRs paves the way for graphene as the next-generation interconnect material for a variety of semiconductor technologies and applications.

Modeling and analysis of crosstalk induced overshoot/undershoot effects in multilayer graphene nanoribbon interconnects and its impact on gate oxide reliability

Crosstalk induced overshoot/undershoot effects in multilayer graphene nano ribbon interconnects (MLGNRs) are investigated with the help of ABCD parameter matrix approach for intermediate level interconnects at both 11nm and 8nm technology node. The worst case crosstalk induced peak overshoot voltage for perfectly specular, doped multilayer zigzag GNR interconnects is comparable to that of copper interconnects. The performance of neutral GNR interconnects is better than that of its doped counterpart with respect to peak crosstalk overshoot. But from the perspective of overall overshoot width and overshoot area contribution, perfectly specular, doped MLGNR interconnects outperform all other alternatives. As far as the effective electric field across the gate oxide is concerned, the doped MLGNR interconnects outperform neutral ones and copper interconnects for all the cases. It is estimated that the doped perfectly specular multilayer GNR interconnects have gate oxide failure rates (AFR) of ~240× and ~790× lesser than copper interconnects for 11nm and 8nm technology node respectively. So, from the gate oxide reliability perspective, perfectly specular, doped multilayer zigzag GNR interconnects are great advantageous to copper interconnects for the future integrated circuit technology generations.

Stability and delay analysis of multi-layered GNR and multi-walled CNT interconnects

This paper analyzes and compares the propagation delay and relative stability of multi-layered GNR (MLGNR) and multi-walled CNT (MWCNT) for different interconnect lengths. The transfer function of a driver-interconnect-load system is obtained by representing the interconnect line with equivalent single conductor model of either a MLGNR or a MWCNT. Using the transfer function of distributed transmission line, an analytical expression of delay is obtained that exhibits significant accuracy. It is observed that a similar delay performance is obtained using fewer number of MWCNT shells than MLGNR layers. On an average, a MWCNT requires 54.5 % lesser area compared to MLGNR interconnects for the similar delay performance. Additionally, using Nyquist criterion, the relative stability of MLGNR and MWCNT is compared for local, semi-global and global interconnect lengths. It is observed that the relative stability of MLGNR is lesser in comparison to MWCNT at local interconnect lengths, whereas, encouragingly for semi-global and global interconnect lengths, both MLGNR and MWCNT demonstrate better stability characteristics.