Electrically induced negative differential resistance states mediated by oxygen octahedra coupling in manganites for neuronal dynamics
2025
Read Electrically induced negative differential resistance states mediated by oxygen octahedra coupling in manganites for neuronal dynamics (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2502-00137.txt.
Introduction
Azminul Jaman and co-authors coax a manganite crystal into behaving like a neuron. Their 2025 paper addresses the power hunger of modern computing by turning to neuromorphic hardware built from correlated oxides. The catalog's creator field is empty; the head names seven authors. The material, LSMO, was grown on a twinned substrate that strains it unevenly, and driven out of equilibrium electrically.
Under those conditions the device shows two distinct states of negative differential resistance, where current falls as voltage rises. Harnessed in a circuit, the states produce oscillations at different frequencies and a leaky integrate-and-fire pattern, the basic rhythm of an artificial neuron. A material model based on bond-angle distortions between neighboring crystal units explains both states. The work positions LSMO for low-power brain-like chips.
Catalog record
| YEAR | 2025 |
|---|---|
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-00137
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Electrically induced negative differential resistance states mediated by oxygen octahedra coupling in manganites for neuronal dynamics Abstract The precipitous rise of consumer network applications reiterates the urgency to redefine computing hardware with low power footprint. Neuromorphic computing utilizing correlated oxides offers an energy-efficient solution. By designing anisotropic functional properties in LSMO on a twinned LAO substrate and driving it out of thermodynamic equilibrium, we demonstrate two distinct negative differential resistance states in such volatile memristors. These were harnessed to exhibit oscillatory dynamics in LSMO at different frequencies and an artificial neuron with leaky integrate-and-fire dynamics. A material based modelling incorporating bond angle distortions in neighboring perovskites and capturing the inhomogeneity of domain distribution and propagation explains both the NDR regimes. Our findings establish LSMO as an important material for neuromorphic computing hardware. keywords Negative differential resistance (NDR), Self-Oscillation, Leaky-integrate and fire(LIF) Neuron Azminul Jaman* Lorenzo Fratino Majid Ahmadi Rodolfo Rocco Bart J. Kooi Marcelo Rozenberg Tamalika Banerjee* \dedication Azminul Jaman, Majid Ahmadi, Bart J. Kooi, Tamalika Banerjee Zernike Institute for Advanced Materials, Groningen Cognitive Systems and Materials Centre, University of Groningen, Nijenborgh 4, Groningen, 9747 AG, The Netherlands. Email Address: [email protected], [email protected] Lorenzo Fratino, Rodolfo Rocco, Marcelo Rozenberg CNRS Laboratoire de Physique des Solides, Université Paris-Saclay, 91405, Orsay, France Lorenzo Fratino CNRS Laboratoire de Physique Théorique et Modélisation, CY Cergy Paris Université, 95302 Cergy-Pontoise Cedex, France. 1 Introduction The high volume of consumer network, that precipitously rose, during the world-wide stay-at-home orders, few years ago, brought to light the urgent need to redefine computing hardware with low power footprint. In this, neuromorphic computing has gained steep interest due to the energy-efficient solution capabilities that new materials, such as oxides, offer for in-memory computing. For such non von Neumann computing approaches, correlated properties intrinsic to the oxides are commonly exploited. However, at the hardware level, these properties have been less explored for demonstrating the rich non-linear phenomenon present in biological systems. The emergence of collective order is the cornerstone of the class of correlated oxides [ 1 ] , primarily stemming from strain tailoring at heterointerfaces and the intricate coupling between charge, spin, orbital, and lattice [ 2 , 3 , 4 , 5 , 6 ] . Manganites with their intrinsic chemical inhomogeneities and complex phase diagrams are particularly relevant. Electronic transport in manganites is susceptible to external stimuli and leads to inhomogeneous spatial distribution of electronic phases with varying local properties [ 7 , 8 , 9 , 10 , 11 , 12 ] . These coexisting and competing phases collectively result in non-linear responses of the order parameter, important for realizing nanoelectronic devices [ 13 ] . By driving such systems out of thermodynamic equilibrium, electrically or optically, such non-linear phenomena are realized, widening the repertoire of emergent properties. Here we design anisotropic functional properties in strained thin films of La 0.7 Sr 0.3 MnO 3 (LSMO), a well-established correlated oxide in the manganite family, by uniquely accommodating variations in the local octahedral tilts, across a twinned substrate of LaAlO 3 (LAO), with bond angle variations ranging from 180 ∘ to 150 ∘ , at the adjacent atomic sites. Recent advancements in complex oxide engineering demonstrate an efficient control of oxygen octahedra coupling (OOC) at the film-substrate interface, at the atomic scale [ 14 , 15 ] , however their cooperative impact on the electronic properties remains unexplored. We utilize this to demonstrate a non-linear response to current, in thin film devices of LSMO on LAO, driving the system to two distinct negative differential resistance (NDR) states, close to the metal-to-insulator (MIT) transition in LSMO, not reported for other types of volatile memristors. Electronic instabilities such as NDR are a key indicator of spatially distributed inhomogeneous phases, close to a phase transition, in a material [ 16 ] . The electrical current ( I ) associated with these phase inhomogeneities manifest as a non-ohmic and negative slope for the electrode potential ( V ) [ 17 ] and has been commonly observed in Mott insulators when driven out of thermodynamic equilibrium by electric field or current-induced Joule heating [ 18 , 19 , 20 ] . We harness these NDR states to exhibit oscillatory dynamics in LSMO at different frequencies in a Pearson-Anson circuit with a capacitor and resistor in parallel and in series respectively, with the volatile memristor. We demonstrate the dynamic response of the entire volatile memristive network as a leaky-integrate-fire (LIF) neuron when the external stimuli are large enough to trigger an output, transiting the network to a low resistance state. Our material based modelling incorporates the distribution of Curie temperature ( T_{C} ) in the network, that corroborates with the bond angle distortions, as found from Scanning transmission electron microscopic (STEM) studies. The model captures the inhomogeneity of the domain distribution with increasing temperature that amplifies close to the phase transition temperature, where high resistance domains dominate. Joule heating driven temperature rise nucleates a high resistive barrier, that expands until it blocks charge migration and eventually snaps, equilibrating the sample and substrate temperature in the paramagnetic insulating state, explaining the observation of the low bias NDR. The unique observation of the second NDR is explained by the model, considering the enhancement of the barrier with increasing voltage, suppressing the ferromagnetic state and transiting to the paramagnetic metallic state, characterizing the phase transition beyond T_{C} in LSMO. These findings, well explained by the rich correlated physics in LSMO widens the ambit of this material class for applications in low power computing, specifically within the realm of neuromorphic computing, opening new avenues towards more biologically plausible brain functionalities. 2 Results and Discussion 2.1 Strongly coupled phases and HAADF-STEM studies of the interface We characterize the coupled phase transition in a 10 nm strained thin LSMO film, on Lanthanum Aluminate (LAO) [ 21 , 22 ] . The double exchange interaction governs the functional properties in LSMO, where, in the parallel alignment, the adjacent oxygen octahedra ( MnO_{6} ) with the manganese (Mn) spins, allows the itinerant electron to hop from one Mn^{3+} to the next Mn^{4+} site. This alignment plays a crucial role in the strong coupling between electronic and magnetic phases, with a metal-to-insulator transition temperature (T_{MIT}) and Curie temperature (T_{C}) coinciding at 325 K shown in Figure 1 a. Below 325 K, the film is a ferromagnetic metal and above 325 K, the film transforms into a paramagnetic insulating state. The conduction mechanism is hopping transport dominated where the probability of an electron moving from Mn^{3+} to Mn^{4+} depends on the bond angle between them shown in Figure 1 b, top panel. This hopping probability is maximized if the bond angle between them is 180 ∘ , and decreases with decreasing bond angle (See, Figure 1 b, bottom panel). A 10 nm thick LSMO film grown on a textured LAO substrate shows that the film is compressed along the in-plane direction and thus elongated in the out-of-plane direction, as revealed from reciprocal space map studies shown in Figure (Supporting Information). Additionally, a slight smearing of the film peak has been observed, signifying an inhomogeneous strain in the film. The inhomogeneities in the strain arise from the structural twins boundaries present on the surface of