I. Core Research Background and Scientific Questions of the Paper

In the design of modern neuromorphic computing—or “brain‑inspired chips”—conventional printed electronic devices typically can only emulate extremely simple neuronal functions, such as a single threshold‑based integration and firing mechanism.

However, real biological neurons exhibit “multi-order complexity,” as their firing patterns are influenced by local microenvironmental perturbations, short-term and long-term memory effects, and the dynamic regulation of multisynaptic cascade networks.

Mark C. Hersam   The professor’s team has proposed a… A brand-new technological roadmap : through Solution-based method and AJP micro‑nano printing process , will two-dimensional materials   MoS ₂ nanosheet Assembled into a disordered yet highly interconnected Memristive Networks , thereby enabling the seamless emergence of the high-order, complex firing dynamics of biological neurons on a single printed device.

II. The Core Role of AJP Micro‑ and Nano‑Printing and Its Process Breakthroughs

In this study, AJP Printing It is far more than just an ordinary manufacturing tool; it is the means to realize that… The “unique topological structure” of memristive nanosheet networks and Physical properties of the Decisive Core Process

1. Breaking Through the Regulation of Ultrafine Lines and Three-Dimensional Surface Microenvironments

Advantages of the AJP printing process: AJP printing technology can deposit materials containing MoS ₂ The ink of the nanosheets is atomized into aerosol droplets with diameters on the micrometer scale, and a focused sheath gas precisely focuses these droplets into ultra‑fine lines—typically down to the 10‑μm level. This approach delivers resolution and three‑dimensional topography control that far surpasses conventional screen printing or piezoelectric inkjet printing.

Significance for the device: This high-precision aerosol fluid-dynamic deposition enables MoS ₂ Nanosheets form a structure between the electrodes that exhibits high randomness yet statistical uniformity. Network topology 。

2. Ink Compatibility and Shear-Assisted Assembly of the Nanosheet Network

MoS used in this study ₂ Two-dimensional nanosheets belonging to the liquid‑phase exfoliation category. AJP printing technology boasts an exceptionally broad ink viscosity compatibility range—spanning from 1 to 1000 cP—making it ideally suited for printing composite inks containing suspended nanoparticles.

During high-speed jetting and rapid solvent evaporation, the aerosol facilitates MoS ₂ Nanosheets self-assemble within the micro‑regions to form countless “nanoscale contact interfaces.” These interfaces give rise to tens of thousands of miniature “ion–electron hybrid conductive memristive junctions.”

III. Core Innovations and Emergent Effects of the Paper

MoS printed using AJP ₂ Memristive networks have demonstrated remarkable… Biological neuromorphic simulation capability ：

1. Emergent Dynamics of Multi-Stage Systems

Within this nanosheet network, the migration of sulfur vacancies under an electric field, combined with localized thermal effects, gives rise to multiple parallel current pathways (percolation pathways) inside the device. As a result, the printed spiking neuron not only mimics basic potential integration but also exhibits multi‑order temporal dynamics—such as dynamic adaptation and discharge‑frequency modulation conditioned by its prior state—that are extremely difficult to achieve in conventional silicon‑based devices.

More importantly, the system’s complex pulse dynamics do not arise from explicit algorithmic programming in conventional digital circuits; instead, they emerge from the collective coupling and self-organized evolution of numerous memristive junctions within a two-dimensional nanosheet network, exhibiting typical emergent behavior.

2. Multiphysics and Microenvironment Simulation:

This being   AJP printing technology The precisely engineered, intricate three-dimensional network architecture exhibits a sensitive, nonlinear response to localized electrical, thermal, and electrochemical perturbations. Its dynamical characteristics closely resemble the spatially organized modulation of neuronal firing patterns by local neurotransmitter release or microenvironmental changes in biological tissues.

IV. High Degree of Alignment with the Development of the “Micro‑Nano Manufacturing—Intelligent Analysis” System

The cross-disciplinary integration of printed electronics and biological models: This study demonstrates how to use Industrial-grade micro- and nano-printing (AJP) To replicate the complex dynamics of biological systems. This demonstrates that AJP is capable of constructing Highly integrated, highly flexible three-dimensional bio-microfluidic chips and organoid multi-array electrode systems , or In-situ Intelligent Sensing and Analytics Network In this regard, it boasts an unparalleled technological ceiling.

A new hardware platform for AI for Science: This low-cost, highly complex pulse‑neuron network fabricated using printing technologies can serve as the hardware substrate for “physical reservoir computing,” enabling direct analysis of extremely intricate, nonlinear biological signals—such as long‑term, multi‑channel trends in metabolic reactive oxygen species (ROS) levels.

V. Summary

This study by Professor Mark C. Hersam’s team represents a quintessential effort to advance the physical construction of cutting-edge intelligent hardware through advanced micro‑ and nanofabrication techniques—specifically, advanced jet printing (AJP). The research demonstrates that by precisely tuning the microenvironment governing the assembly of two-dimensional nanomaterials during the printing process, multi‑scale, hierarchical complex dynamical behaviors reminiscent of biological systems can emerge within micron‑scale rigid material networks.

More importantly, this complex pulse‑based behavior does not rely on explicit algorithmic programming in conventional digital circuits; instead, it arises from the collective coupling and self‑organized dynamics among the vast array of memristive junctions within a two‑dimensional nanosheet network. This implies that future computational capabilities may no longer depend solely on software algorithms, but could instead emerge directly from the intricate physical dynamics inherent in the material network itself.

From a technical perspective, this study has in fact established a highly representative cutting-edge development pathway:

AJP micro- and nanoscale printing → Random two-dimensional material networks → Multi-scale complex dynamics → Neuromorphic computing → AI for Science

From a deeper perspective, this work embodies not merely “advanced manufacturing” itself, but rather the notion that “manufacturing is computation.”
A new technological paradigm—“Manufacturing-as-Computing.” Within this framework, AJP micro‑ and nanofabrication is no longer confined to device fabrication; instead, it actively constructs network topologies, modulates material heterogeneity, shapes complex dynamics, and ultimately endows the system itself with computational capabilities.

In a sense, this work goes beyond merely “printing electronic devices”; it leverages advanced micro- and nanofabrication techniques to construct physical intelligence systems endowed with emergent intelligent properties, thereby paving the way for novel hardware implementations in cutting-edge fields such as neuromorphic computing, reservoir computing, physical AI, and AI for science.

Image and Text Guide

Figure 1. Printed threshold-switch memristor with negative differential resistance (NDR) characteristics.

a. Schematic diagram of a biological neuron, illustrating its typical firing threshold and all-or-none response characteristic. b, Schematic diagram of the aerosol jet printing (AJP) process, along with an optical micrograph of a printed threshold-switching memristor array. c, Schematic diagram of the device structure of a single printed memristor. d. The current–voltage (I–V) characteristic curve of the device exhibits typical negative differential resistance (NDR) behavior and threshold-switching characteristics. e, Schematic diagram of a spiking neuron circuit built on printed memristors. f, The spiking behavior of the device under various input conditions exhibits dynamic response characteristics analogous to those of biological neurons.

Figure 2. Evolution of the thermal spatial distribution during the negative differential resistance (NDR) process in a printed threshold-switching memristor.

Figure 3. Artificial neuron circuit based on a printed threshold-switching memristor

Figure 4. First-, second-, and third-order complex dynamics in a spiking neuron circuit

Figure 5. Pulse characteristics related to physiology and in vitro cell stimulation experiments