Microfluidic systems enable controlled manipulation of small fluid volumes and have become essential tools in biomedical applications, including point-of-care diagnostics1,2, organ-on-chip platforms3,4, and sensor technologies5,6. Despite the diversity of their applications, these devices share core engineering demands: appropriate material selection, reproducible fabrication strategies, and sustained biocompatibility7,8. These requirements become more stringent in blood-contacting systems, where materials and device architectures must ensure hemocompatibility, form robust fluid sealing, and maintain physiologic flow parameters such as shear stress and pressure gradients9,10,11.
Hybrid microfluidic systems integrate thermoplastics with rigid substrates such as silicon and glass, expanding the functionality of biomedical devices (Supplemental Figure S1)12,13,14,15,16,17. This approach is particularly intriguing for blood-contacting applications, where mechanically fragile yet highly precise silicon or glass components are coupled to low-cost polymer housings to enable scalable, biocompatible designs18,19,20,21,22,23,24. While PDMS and thermoplastic microdevices remain prevalent due to their fabrication simplicity, these materials limit chemical resistance, mechanical durability, and incorporation of rigid functional materials, motivating the development of hybrid architectures25,26,27,28. Despite the advantages of hybrid devices, assembling dissimilar materials presents inherent challenges. Differences in thermal expansion and mechanical properties can undermine seal integrity, and fabrication conditions must be tightly controlled to preserve hemocompatibility, maintain channel geometry, and prevent damage to brittle silicon or glass structures. Successful implementation of hybrid systems, therefore, depends on deliberate choices of material properties with appropriate fabrication strategies.
Thermoplastic materials, including polycarbonate and poly(methyl methacrylate) (PMMA), are widely used in microfluidic systems because they provide mechanical strength, biocompatibility, cost efficiency, and are compatible with scalable fabrication strategies such as solvent bonding, thermocompression, and adhesives12,29,30,31,32,33,34,35. In contrast, silicon and glass substrates offer enhanced chemical resistance and high-precision geometries, yet their integration with thermoplastics remains technically challenging due to disparities in surface chemistry and mechanical properties36,37,38,39. Conventional thermocompression bonding can subject brittle silicon or glass components to elevated pressures, increasing the risk of cracking. Alternatively, adhesive-based approaches may reduce mechanical stress but can introduce variability in bond uniformity or raise concerns regarding long-term biocompatibility40.
Our lab is developing blood-contacting devices intended for organ replacement functions, including oxygenators and hemofilters17,21. These systems integrate silicon nanopore membranes as active functional elements with polycarbonate housings that establish the blood flow pathway and provide mechanical support18,19,21. Achieving robust and scalable fabrication of these platforms depends on reliable bonding between these dissimilar materials.
This work evaluates three bonding approaches developed to address the constraints of conventional high-temperature and high-pressure assembly methods used for hybrid polycarbonate–silicon microfluidic systems. The platform design is derived from ongoing efforts to develop modular components for a microfluidic extracorporeal membrane oxygenator17. The device incorporates a rigid polycarbonate housing that provides mechanical stability and defines a machined blood flow channel, integrated with a semi-permeable silicon membrane intended to facilitate gas exchange. To isolate bonding performance during testing, a non-porous silicon substrate was used in place of the functional membrane.
The three bonding approaches evaluated were: 1) a solvent-based method that avoids high-temperature and significant mechanical load with a protective masking strategy to preserve channel geometry, 2) a low-temperature thermal bonding technique incorporating a 100 µm low-density polyethylene (LDPE) interlayer between polycarbonate plates, and 3) a mechanically sealed configuration using an O-ring gasket cord positioned around the fluidic channel. These strategies were systematically compared to determine their feasibility and reliability for rapid prototyping of hybrid microfluidic devices in a laboratory research setting. All methods were designed to enable low-temperature assembly while supporting moderate operating pressures (10–700 mmHg) and flow conditions relevant to blood-contacting microfluidic systems. The following article builds upon a chapter from the dissertation of the first author41.