Recent advancements in nanomaterials and microfabrication have led to significant progress in flexible wearable electronics. However, conventional manufacturing methods often involve complex, multi-step processes that require costly cleanroom facilities and specialized personnel, limiting scalability and affordability. This research addresses these limitations by presenting a new approach to additive nanomanufacturing of functional materials, creating a wireless, multilayered, seamlessly interconnected, and flexible hybrid electronic system. The core innovation lies in employing a functionalized conductive graphene (FCG) that exhibits enhanced biocompatibility, resistance to oxidation, and solderability, thereby enabling the creation of a wireless flexible circuit. This all-printed electronics approach, integrated with machine learning, offers versatile and multi-class human-machine interfaces (HMIs). The high-aspect-ratio FCG provides gel-free, high-fidelity recording of muscle activities, surpassing the performance of traditional metal electrodes. The study demonstrates the efficacy of this system by using real-time EMG signals to control external systems. A key aspect of the study is the use of deep learning for electrophysiology mapping, enabling optimization of electrode placement to accurately capture finger movements. This introduction highlights the need for more efficient and cost-effective manufacturing of wearable electronics and introduces the presented solution as a significant advancement in the field.

Existing literature demonstrates the advantages of thin, stretchable hybrid electronic packages for seamless integration with the human body. However, traditional microfabrication processes used in creating these systems are often expensive, inefficient, and require specialized cleanroom facilities. While advancements have been made in creating flexible, multi-layered electronics capable of conforming to human physiology and incorporating commercial electronic components, these methods still rely on resource-intensive techniques. The exploration of additive manufacturing for stretchable hybrid electronics offers a compelling alternative, promising reduced material waste, faster turnaround times, scalable parallel printing, and simplified equipment requirements. Advances in printing methods and soft materials are shifting wearable electronics from rigid designs toward comfortable, skin-integrated soft form factors. The development of highly conductive nanomaterials such as silver, copper, and carbon nanotubes (CNTs) has led to improvements in electrophysiological signal recording. However, limitations exist, including the cytotoxicity of silver and copper and the relatively low conductivity of CNTs. This work builds upon these prior advancements, aiming to overcome these limitations by employing a novel biocompatible and highly conductive material for the fabrication of a fully printed wearable electronic system.

This research employed an all-printed nanomembrane hybrid electronics (p-NHE) fabrication strategy based on comprehensive studies of nanomaterial preparation, material processing, and printing optimization. The additive nanomanufacturing process, utilizing aerosol-jet-based printing (AJP), allowed for high-precision alignment in multilayer printing. The thin, flexible structures enabled natural integration and deformation with elastomers. The key material, functionalized conductive graphene (FCG), was prepared with a method that conserved its electrical and morphological properties while improving biocompatibility and oxidation resistance. The high-aspect-ratio FCG served as an oxidation barrier for the silver (Ag) conductive traces and also as the sensing electrode material. Polyimide (PI) was utilized as the insulating and structural support layer. The AJP method enabled direct deposition of inks with a wide range of viscosities (1–1000 cP) without masks or screens. The FCG ink, formulated with carboxylic and hydroxyl groups, allowed for easy dispersion in aqueous solvents without the need for dispersing agents. The printing process involved sequential deposition of layers of PI, FCG, and Ag to create both the sensing electrodes and the wireless circuit. The completed devices integrated functional chip components via soldering and were encapsulated with a low-modulus silicone elastomer. Characterization techniques included atomic force microscopy (AFM), transmission electron microscopy (TEM), scanning electron microscopy (SEM), X-ray diffraction (XRD), X-ray photoelectron spectroscopy (XPS), and finite element analysis (FEA) to assess the structural, electrical, and mechanical properties of the printed materials and devices. Cell viability assays were conducted using human keratinocyte cells to assess the biocompatibility of the FCG electrodes. The EMG data acquisition and analysis involved signal processing techniques and machine learning algorithms (Convolutional Neural Networks (CNN) and k-nearest neighbors (KNN)) to classify hand gestures and control external devices. The experimental protocol for the HMI included the use of single and three-device configurations to control drones, RC cars, presentation software, and a robotic hand.

The study successfully demonstrated the fabrication and functionality of an all-printed, flexible, wireless bioelectronic system using functionalized conductive graphene (FCG). AFM and TEM analysis confirmed the high-aspect-ratio structure of the FCG, with lateral dimensions between 1 and 5 µm and a thickness of ~3.1 nm. The printed FCG electrodes exhibited high performance in EMG recording, demonstrating signal-to-noise ratios comparable to gel-based electrodes without the associated drawbacks of skin irritation and impedance fluctuations. The FCG effectively served as an oxidation barrier for the silver (Ag) circuit traces, preventing degradation during soldering, as confirmed by XRD and XPS analyses. FEA and experimental bending/stretching tests verified the exceptional mechanical flexibility and reliability of both the electrodes and the circuit, with negligible resistance changes even under extreme deformation. The system achieved stable, long-range wireless communication via Bluetooth, with signal strength comparable to rigid circuit boards. In the HMI experiments, deep learning was crucial; anatomical study with deep learning-embedded electrophysiology mapping allowed for an optimal selection of three channels to capture all finger motions. Using a single p-NHE, real-time control of a drone, RC car, and presentation software was achieved with over 99% accuracy across six classes. With three p-NHEs, the system classified seven different hand gestures (including individual finger movements) with 98.6% accuracy, enabling real-time control of a robotic hand. The p-NHE proved exceptionally lightweight and thin, facilitating comfortable and seamless lamination to the skin.

The findings of this research demonstrate a significant advancement in the field of wearable bioelectronics. The all-printed nanomembrane hybrid electronics (p-NHE) system offers a highly efficient and cost-effective alternative to conventional manufacturing methods. The use of biocompatible, solderable FCG addresses the limitations of previously used conductive materials, enhancing both biocompatibility and circuit reliability. The integration of deep learning algorithms optimizes electrode placement and significantly improves the accuracy of EMG signal classification for controlling multiple external devices. The exceptional mechanical flexibility and wireless communication capabilities of the p-NHE system broaden the potential applications in areas such as prosthetic control, rehabilitation training, and other human-machine interface applications. The high accuracy and versatility of the multi-class HMIs demonstrated in this work highlight the potential for seamless integration of advanced electronic systems with the human body for a variety of applications.

This paper presents a groundbreaking approach to the creation of all-printed, flexible, and wireless bioelectronic systems using biocompatible and solderable functionalized conductive graphene (FCG). The demonstrated capabilities—high-fidelity EMG recording, reliable solderability, exceptional mechanical flexibility, and accurate multi-class human-machine interfaces—represent a significant advancement in the field of wearable electronics. Future research will focus on translating these findings into clinically relevant applications, such as developing biofeedback-enabled prosthetics and enhancing rehabilitation training. Further optimization of the manufacturing process and exploration of additional applications are also planned.

While the study demonstrates the significant potential of the p-NHE system, there are some limitations. The current integration process still involves manual assembly of electronic components, which could be further automated to improve yield. Further research is needed to thoroughly assess the long-term stability and reliability of the devices in real-world conditions. The study focused on a limited set of hand gestures, and broader testing with diverse user populations would strengthen the generalizability of the findings. While the biocompatibility studies demonstrate good cell viability, longer-term in vivo studies are needed to fully characterize the biocompatibility and potential long-term effects of the materials.