Toward a Theoretical Roadmap for Organic Memristive Materials

TL;DR

Organic memristive materials offer tunability and scalable processing but require a unified theory to translate molecular design into device performance. The authors propose a multiscale computational roadmap that integrates quantum chemistry, molecular dynamics, coarse-grained simulations, and device-level modeling to connect molecular structure to memristive function. They detail three representative mechanisms—ionic migration, redox switching, and magnetic–chirality conduction—and outline modeling and screening strategies to optimize switching energy, retention, and multistate behavior. This framework aims to accelerate the discovery and reliable implementation of chemically engineered synaptic materials for neuromorphic hardware.

Abstract

Neuromorphic computing aspires to overcome the intrinsic inefficiencies of von Neumann architectures by co-locating memory and computation in physical devices that emulate biological neurons and synapses. Memristive materials stand at the core of this paradigm, enabling non-volatile, history-dependent electronic responses. While inorganic oxides currently dominate the field, molecular and polymeric systems can offer untapped advantages in terms of chemical tunability, structural flexibility, low-cost processing, and biocompatibility. However, progress has been hindered by the absence of a theoretical framework able to rationalize how molecular structure translates into memristive function. Here, a multiscale computational perspective is presented, outlining how quantum chemistry and molecular dynamics, among other approaches, can be integrated into a coherent methodology to design next-generation organic memristors. Three mechanisms, ionic migration, redox-driven switching, and conduction interplay in chiral molecules are examined as representative routes toward molecular neuromorphic hardware. The opportunities and challenges associated with each mechanism are discussed, together with a view on how a theoretically guided roadmap can accelerate the emergence of chemically engineered synaptic materials.

Figures (7)

• Figure 1: Schematic depiction of the organic materials described in this work as proposals for achieving high-efficient memritive properties
• Figure 2: Schematic 1D representation of ion migration mechanisms. (a) Electrodynamical model (injection-limited regime, no effective polymer doping exists) (b), Electrochemical model (ohmic injection regime, the polymer conductivity is modified by electronic doping). Reproduced from Ref. PradoSocorro2022
• Figure 3: (a) Schematic of a single synaptic device, where the memristive material shows homogeneous presence of salt ions and composition of the memristive layer: (b) lithium triflate (LiCF3SO3), (c) Super Yellow (semiconductive polymer), and (d) $Hybrane^{\copyright}$ DEO750 8500 (ion-transport polymer). Reproduced from Ref. PradoSocorro2022
• Figure 4: Redox-activity of Fe^3+/2+(phen)_3 complex
• Figure 5: (left) Structure of the LBTC peptide complexed with Tb^3+ (green sphere). In this LBT variant the two C-terminal aminoacids have been replaced with a cysteine residue (right) Schematic functioning of a spintronic multistate memristive device based on the CISS effect in a paramagnetic molecular material. Let us start the experiment with the voltage in the ON position for an unspecified but long enough time to guarantee that the magnetic polarization of the metal ions is maximal due to the interaction with the CISS-polarized current, and the resistance is correspondingly minimal. A sudden switch in the voltage to the OFF position, e.g. a change in sign has two consequences: it starts a relatively slow process of relaxation of the magnetic polarization, and it instantly increases the memristance to its maximum value. This sudden jump in the memristance is due to the ‘wrong’ polarization of the paramagnetic ions, which now opposes the spin polarization of the CISS current. As the magnetic polarization evolves towards its new equilibrium situation, the memristance decreases too, accessing a continuum of values, i.e. behaving as a multistate memristance. Depending on the timing of the subsequent changes in the voltage between the ON and OFF values, the time evolution of the magnetic polarization of the system and of the memristance of the device will present different shapes; a fine timecontrol of the voltage allows achieving any desired memristance between the ON and OFF limit values. Reproduced from Ref. Cardona-Serra2021 with permission from the PCCP Owner Societies.