U N I V E R S I T À D I P I S A
DIPARTIMENTO DI INGEGNERIA DELL’INFORMAZIONE Dottorato di Ricerca in Ingegneria dell’Informazione
PhD Course Activity Report by the Student Marta Perucchini PhD Program in Information Engineering, cycle XXXII
Tutors: Prof. Gianluca Fiori, Prof. Giuseppe Iannaccone
1. Research Activity
During the course of the Ph.D. course, I have worked on several topics concerning the modeling of 2D-material based devices for next generation electronics.
With the scaling of transistor dimensions come the so-called short channel effects, which hinder the devices performance. In order to overcome this limitation, it is of great interest to find alterna-tives to the traditional silicon-based technologies, one of them being the use of low-dimensional materials and new architectures.
In the first year of the Ph.D. program, while fulfilling most of my formation activities, I have
worked on a model of an ultra-scaled MoS2-based device gated with a 1-nm cylindrical gate.
More-over, I analyzed the band structure of graphene nanoribbons as well as preliminarily studied a par-ticular class of 2D materials: Noble-TMDs.
During the second year of PhD, I have continued with three main research lines. In the first one, I have dealt with the literature review of simulations techniques available for the two-dimensional materials class. The main topic has however been the development of a new model for transport in non-homogeneous materials, able to support the experimental realization of ink-jet printed mi-cro/nano-electronic devices. We have been able to define the material structure and to study the electrostatic of the system while setting the base for the development of a transport model which was dealt with in the third year of my Ph.D. course. Lastly, I continued to explore the potential ap-plications of Noble-TMDs on lateral heterostructures.
Finally, in the last year of the Ph.D. program, my main focus has been the development of a transport model for 2D-material based networks. In addition, I have finalized the paper and projects of the previous two years concerning the modeling of a system with a carbon nanotube gate, and lateral heterostructure FETs based on Noble transition-metal dichalcogenides.
This topic was developed during the course of the first and second years of the Ph.D. program.
In the scaling perspective, reducing the channel length is not the only bottle-neck. Technological problems such as lithography patterning limit also the shaping of the gate. The group of Prof. Javey at the University of California Berkeley has recently proposed and realized a device gated with a metallic carbon nanotube. This bottom-up approach used for the fabrication of single-walled CNTs allows one to easily reach diameters as small as 1 nm. We thus modeled a system inspired by the experimentally realized one, aiming to investigate the ultimate performance of an ultra-scaled MoS2-based device gated with a 1 nm cylindrical gate. Moreover, we evaluated the
impact of the finite density-of-state of a metallic carbon nanotube by comparing its performance to that of a traditional metal gate.With this aim, we performed quantum transport simulations by self-consistently solving Poisson and Schroedinger equations employing the NanoTCAD ViDES simulation environment. We showed that the finite density of states of the nanostructure
negatively affects the electrostatic control over the MoS2 channel with respect to the metallic one,
which can be seen as an upper-limit case. We also noticed a strong degradation of performance with scaling. The subthreshold swing reaches in fact values of more than 300 mV/decade for channel lengths around 5 nm. We explain this degraded performance by short channel effects and tunneling currents, due to poor electrostatic control. To boost the performance, we proposed a reduction of the oxide thickness. On the contrary, the electrostatic doping through the back gate and the gate diameter did not show relevant improvement of the device performance. Noticeably, for ultra-scaled devices, the gap between a CNT- and metal-gated structures was reduced, with their behavior being dominated by short channel effects. We were thus able to set a framework for the potential applications and limitations of this innovative structure, taking into account the experimental outcomes.
The results of this work have been reported in a letter for the journal Applied Physics Letter [J1].
Project 2: Graphene Nanoribbons
This topic was developed during the course of the first year of the Ph.D. program.
Reducing one dimension of a graphene sheet in order to produce 1D nanoribbons is one of the principal strategies used to open a bandgap in this otherwise semi-metallic material. The nature of GNRs however, depends not only on the width but also on the edge-shape and symmetry. A pre-liminary work has been done in order to confirm the semiconducting behavior of 4-aGNR (arm-chair) and the effects of hydrogen passivation and atoms relaxation by means of a tight binding model.
The results of this work have not been reported in publications.
This topic was developed during the course of the first, second and third years of the Ph.D. program.
In the miniaturization process, the metal/semiconductor contact resistance has been identified as one of the main bottlenecks for the performance of 2D-based electronic devices. A possible way to improve the electrical transport across the interface could be the realization of a single-material device. Noble-TMDs are particularly suitable in this context since their bandgap can be tuned by modulating the thickness of the crystalline phase. This option allows very good heterointerface qual-ity thanks to lattice matching. We studied the band structure of three compounds (NiS2, PdS2 and
PtS2) obtained by first-principle calculations in mono and bilayer form, showing steep transitions in
the density of states (DOS) of the 2D materials. We eventually exploited the aforementioned prop-erties to engineer lateral heterostructures such as Field-Effect Transistors (FET) and a Resonant Tun-neling Diode (RTD). To do so, we adopted a multi-scale simulation approach combining DFT, Wan-nier up to full device simulations based on quantum transport. The results obtained showed that PdS2 could be used to achieve sub-60 mV/decade subthreshold swing in LH-FET thanks to the
en-ergy-filtering source effect. Moreover, our findings also suggested that devices made with PdS2 and
PtS2 can comply with IRDS performance requirements, differently from those realized with NiS2.
Lastly, we evidenced a strong peak-to-valley ratio in a PtS2- based RTD, which is suitable for
experi-mental observation.
The results of this work have been reported in an article for the journal ACS Nano [J2]
Project 4: Review on modeling of 2D materials based on 2D materials
This topic was covered during the course of the second year of the Ph.D. program.
Since the discovery of graphene 14 years ago, two-dimensional, layered materials have been a trending topic with the number of synthetized or predicted materials constantly growing as it is the number of published papers. To approach the subject of 2D-materias from a theoretical point of view there are various modeling and simulation techniques ranging from atomistic, first-principle calculations to compact models. Needless to say, the accuracy and computational time are most of the times inversely proportional and the right approach to use is strictly dependent on the parame-ters of interest. In the review paper we were invited to write, we classified the main methods to study electron transport, we examined how they could be combined in a multiscale approach and finally we showed some examples of their application in the various 2D -materials families.
The results of this work have been reported in an article for the journal IEEE Transactions on
Elec-tron Devices [J3]
Project 5: Modeling of electronic transport in 2D material networks for printable electronics
Devices made on transparent or flexible substrates are expected to be leading actors in the next electronic revolution. On the other hand, the possibility of product customization and rapid and low-cost prototyping is nowadays appealing to both industries and consumers. Ink-jet printing is an emerging technology that is believed to be able to satisfy these two market needs, making it possi-ble to exploit unconventional but promising materials (such as low-dimensional materials) and com-plex structures. Layered materials in fact provide interesting electronic properties combined with extreme flexibility and mechanical strength, attributes that make them ideal candidates for printed nanoelectronics devices. Being the ink-jet printing process still in its infancy, the printed structures still require extensive baseline studies and physical models in order to achieve the expected device performance and to gain profit from material properties. To support the experimental work in the research group and to fill a literature gap on the subject, the goal of this work was to develop a simulation tool for printed networks of two-dimensional materials flakes such as graphene, MoS2
and hBN. We started by defining a method for the structure generation, which had to resemble the statistical and random experimental process.
In order to provide fabrication guidelines, we have thus developed a novel transport model able to capture the intrinsic complexity of heterogeneous-material based devices. As a first step, we defined MoS2 and graphene network structures in the 3D space, tuning geometric parameters such as flake
density, size, distribution and shape. This is a non-trivial step since the flakes connectivity strongly determines the conductivity of the structure. In order to test the electrostatic behavior of the ma-terial, we performed the whole device simulations (by means of the NanoTCAD ViDES simulator) choosing a uniform dielectric constant. This has been done not only to account for the non-homogeneous channel material, but also considering the possible water residues from processing. Later, we have implemented a drift-diffusion equation in the software, discretizing it in order to account for the empty points in the structure as well as the different in- and out-of-plane mobilities, which have been extracted from ab-initio simulations (PWCOND from the Quantum Espresso Suite). Finally, we have defined an algorithm able to return the output current at the drain once the input parameters are set together with the desired gate and source-to-drain voltage. This is based on a 3-step self-consistent loop, which solves for the potential, the charge and the quasi-Fermi level in the domain for both electrons and holes in the material. In order to provide statistical relevance to the results, we have carried out sets of 30 simulations with different network configurations, obtaining the system variability for the quantities of interest such as the effective mobility and sheet re-sistance as function of network density and Vds applied.
The results of this work have been reported in an article submitted [J4]
2. Formation Activity
First year
• Thierry Djenizian, Center of Microelectronics in Provence – France, “Electrochemical Energy Storage for Flexible Microelectronics. Principle and applications”, 10-13 April 2017
• Marco Morelli, STMicroelectronics – Italy, “Semiconductor trip: from a simple idea to a com-plex manufacturing”, 19-23 June 2017 (27 hours, 6 credits)
• Eric Pop, Stanford University –USA, “Energy, Thermal and Thermoelectric Effects in Na-noscale Devices”, 12-16 June 2017 (16 hours, 4 credits)
• University of Pisa – Italy, “PhD+: Research valorization, innovation, entrepreneurial mind-set”, April-May 2017 (27 hours, 6 credits)
• Massimo Macucci, University of Pisa – Italy, “Nanoelettronica, corso di Laurea Magistrale”, March-May 2017 (90 hours, 9 credits)
Second year
• Joanne Spataro – CLI, “Academic English C1+” (5 credits)
• Massimo Macucci – University of Pisa, “Quantum computing” (4 credits) • TUDresden – “Summer School Materials4.0” (6 credits)
Third year -
Total number of credits 45
Publications
International Journals
[J1] M. Perucchini, E. G. Marin, D. Marian, G. Iannaccone, G. Fiori, “Physical insights into the opera-tion of a 1-nm gate length transistor based on MoS2 with metallic carbon nanotube gate”, in Applied
Physics Letters, vol. 113, no. 18, pp. 183507, 2018. doi: 10.1063/1.5054281
[J2] E.G. Marin, D. Marian, M. Perucchini, G. Fiori, and G. Iannaccone, “Lateral Heterostructure Field-Effect Transistors based on Two-Dimensional material stacks with varying thickness and energy fil-tering source. In: ACS Nano vol. 14, no. 2 (2020), p. 1982-1989
[J3] E. G. Marin, M. Perucchini, D. Marian, G. Iannaccone and G. Fiori, "Modeling of Electron Devices Based on 2-D Materials," in IEEE Transactions on Electron Devices, vol. 65, no. 10, pp. 4167-4179, Oct. 2018. doi: 10.1109/TED.2018.2854902
[J4] M. Perucchini, D. Marian, E. G. Marin, T. Cusati, G. Iannaccone, G. Fiori, “Electronic transport in 2D-based printed FETs from a multiscale perspective” submitted
Conferences
[C1] M. Perucchini, E. G. Marin, D.Marian, G. Fiori, G. Iannaccone, “Noble-TMDs heterostructure based devices”, Poster contribution at international conference, Graphene, Rome, June 2019 [Award for best PhD poster contribution]
