In an extensive review published in AAPS Open, researchers Rupam Deka, Shrutie Shivam, Debarupa Dutta Chakraborty, and Prithviraj Chakraborty synthesize the fundamental physical principles, diverse microfabrication strategies, and biomedical utilities of microfluidic technologies. The authors report that conventional drug delivery, pharmacokinetic modeling, and biochemical analysis systems face severe constraints, including imprecise targeting, irregular release profiles, unpredictable throughputs, and high dosage requirements that cause systemic toxicity. By manipulating fluid handling at the microscale, engineered microfluidic architectures resolve these bottlenecks through the creation of scalable drug delivery devices, microparticles, microneedles, lab-on-a-chip systems, and organ-on-a-chip platforms. The foundational operational mechanisms rely predominantly on low Reynolds number fluid dynamics where viscous forces completely dominate inertial forces, supplemented by electrokinetic fluid handling, capillary action, and centrifugal force. These microfluidic platforms operate with liquid volumes spanning nanoliters to femtoliters inside channels with micrometer cross-sections, enabling continuous synthesis, reproducible testing environments, and high-precision analytical pipelines across oncology, single-cell analysis, disease diagnosis, and personalized therapeutics.
Microscale fluid physics differs substantially from macroscopic pipe flows because minute channel dimensions enforce predictable and observable laminar flow without turbulent mixing or formation disruption. For non-circular geometries, the calculation incorporates the hydraulic diameter defined as four times the cross-sectional area divided by the wetted perimeter. While macroscopic conduits sustain laminar conditions up to Reynolds numbers between 2000 and 3000, microfluidic channels measuring between 1 and 100 micrometers yield extremely low Reynolds numbers, typically ranging from 10 to the negative sixth power up to 10, and practically remaining well below 1. Under these conditions, the linear Stokes equation governs flow behavior because inertial effects are negligible. This fluid regime ensures diffusion-dominated mixing where molecular diffusion replaces chaotic advection, facilitates hydrodynamic focusing, and allows multiple distinct fluid streams to flow side-by-side within a single channel without convective mixing to generate parallel reaction zones.
Fluid handling strategies exploit multiple physical principles to manipulate microscale volumes without relying on cumbersome external mechanical infrastructure. Electrokinetic handling uses applied electric fields to control fluids and charged solutes through microchannels via electroosmosis and electrophoresis, proving vital for solute separation and precise sample transport. Capillary microfluidics operates as a passive transport mechanism governed by surface tension and wettability across porous substrates like paper and nitrocellulose without external pumping, as exemplified by paper-based analytical devices detecting sulfites in wine. Capillary and thermocapillary Marangoni forces also regulate liquid transport in microgravity environments aboard the International Space Station for plant watering systems, biosensors, and specialized drinking cups. In contrast, centrifugal microfluidics employs rotational forces from spinning disk-like platforms to push liquids outward through microchannels. By balancing centrifugal acceleration against channel geometry and surface tension while harnessing secondary Coriolis and Euler forces, centrifugal devices achieve automated valving, metering, mixing, and separation along parabolic flow profiles, providing highly portable fluid handling suited for automated laboratory diagnostics and space applications.
The physical construction of microfluidic devices directly governs experimental performance, fluid flow, biocompatibility, and system integration. Conventional fabrication paradigms originated with photolithography adapted from semiconductor manufacturing. The standard photolithographic process begins by designing a photomask using computer-aided software, spin-coating an SU-8 negative photoresist onto base substrates of silicon, glass, or quartz, and exposing the assembly to ultraviolet radiation. Because negative photoresist hardens upon ultraviolet exposure, unexposed regions are dissolved during solvent development, leaving raised structures that function directly as microchannels or as master molds. However, pure photolithography suffers from substrate inflexibility, elevated costs associated with repetitive silicon mold fabrication, and time-intensive processing. Consequently, soft lithography using polydimethylsiloxane has become the dominant method for academic prototyping and elastomeric chip assembly. In soft lithography, a mixture of polydimethylsiloxane prepolymer and curing agent in a 10-to-1 ratio is degassed, poured onto the patterned master mold, and cured at 65 degrees Celsius. Inlets and outlets are mechanically punched, and substrate surfaces undergo oxygen plasma bombardment to replace surface methyl groups with reactive hydroxyl groups, which form irreversible, watertight covalent silicon-oxygen-silicon bonds upon contact.
Subtractive and replication techniques provide additional manufacturing avenues for rigid substrates, polymers, and thermoplastic composites. Subtractive etching selectively removes substrate materials through chemical wet etching, reactive ion etching, or deep reactive ion etching. Wet etching utilizes isotropic chemical etchants such as hydrofluoric acid on patterned silicon or glass. Reactive ion etching provides anisotropic profiles by exposing substrates to ionized reactive gases within vacuum chambers, combining physical ion bombardment with chemical reactivity. Deep reactive ion etching achieves deep vertical features through alternating cycles: a fluorocarbon gas like octafluorocyclobutane deposits a thin passivation layer across all exposed surfaces, after which directional direct-current plasma ions strip the passivation exclusively from the trench floor, allowing sulfur hexafluoride radicals to etch silicon vertically while sidewalls remain protected. Conversely, hot embossing replicates microscale and nanoscale structures into thermoplastic polymers such as polymethyl methacrylate and cyclic olefin copolymer. In batch embossing, the thermoplastic substrate is heated above its glass transition temperature between 120 and 240 degrees Celsius and compressed against a mold under forces reaching 30 kilonewtons, whereas continuous roller embossing feeds polymer sheets between heated cylinders.
Additive manufacturing overcomes the geometric constraints of planar two-dimensional lithography by producing complex, assembly-free three-dimensional microchannel networks directly from digital computer models. Stereolithography and digital light processing construct devices layer-by-layer by photopolymerizing photosensitive resins using focused ultraviolet beams or projected digital images, yielding internal channels below 100 micrometers without manual bonding. Extrusion-based methods include fused deposition modeling, which extrudes melted thermoplastic filaments like acrylonitrile butadiene styrene, polylactic acid, polyethylene terephthalate, thermoplastic polyurethane, polypropylene, or polyvinyl alcohol through heated nozzles to fabricate molds or macroscale fluidic channels. Direct ink writing dispenses shear-thinning viscoelastic inks, hydrogels, fugitive waxes, or cell-laden pastes under mechanical pressure without thermal melting, facilitating sacrificial channel templating and biological tissue engineering. Inkjet techniques like PolyJet jet photopolymer droplets alongside sacrificial support matrices to combine rigid and flexible elements in single builds. Furthermore, hybrid manufacturing integrates additive manufacturing with soft lithography by embedding 3D-printed sacrificial structures of wax or polyvinyl alcohol into cast polydimethylsiloxane, which are subsequently dissolved to generate hollow, multi-layered three-dimensional microvascular architectures.
Advanced subtractive and patterning technologies include laser micromachining and nanoimprint lithography for high-throughput and submicron channel production. Laser micromachining employs continuous-wave carbon dioxide infrared lasers operating at 10.6 micrometers and 45 watts, as well as pulsed lasers that vaporize materials when energy exceeds local ablation thresholds. While nanosecond pulses cause thermal ablation, ultrafast femtosecond lasers operating around 800 nanometers with 100-femtosecond pulses and picosecond lasers operating at 515 nanometers with 6-picosecond pulses utilize non-linear multiphoton absorption, photo-ionization, and avalanche ionization to remove material mechanically via plasma formation with minimal collateral heat damage, achieving surface roughness down to 0.4 micrometers. Picosecond lasers at 1030 nanometers also perform interface microwelding to seal glass microchannels without external adhesives. Nanoimprint lithography provides scalable replication across microscale and nanoscale domains. In thermal nanoimprint lithography, rigid stamps press heated thermoplastic films, while ultraviolet nanoimprint lithography crosslinks liquid resists or inorganic precursor solutions of silicon dioxide, titanium dioxide, or zinc oxide under transparent molds. Soft nanoimprinting utilizes flexible polydimethylsiloxane stamps at ambient conditions, and continuous roll-to-roll processing enables high-volume manufacturing of flexible microfluidic devices.
Four-dimensional printing represents a dynamic manufacturing frontier that incorporates time as an active fourth dimension through stimulus-responsive smart materials. Unlike static three-dimensional geometries, four-dimensional microfluidic structures alter their physical configuration, porosity, or mechanical function when exposed to external environmental triggers such as temperature changes, pH variations, light irradiation, or magnetic fields. Common functional matrices include shape-memory polymers that undergo pre-programmed conformational folding or channel occlusion during thermal transitions, as well as stimuli-responsive hydrogels that swell or contract reversibly. Standard additive platforms such as stereolithography, fused deposition modeling, selective laser sintering, and direct ink writing are modified to print multi-material systems capable of responsive actuation. Despite significant promise in soft robotics and adaptive bio-interfaces, four-dimensional microfluidics confronts notable hurdles, including material fatigue, low response sharpness, elevated hysteresis, and structural sagging within long unsupported channels. Furthermore, complex time-dependent fluid-structure interactions remain challenging to model accurately because environmental responses depend on non-linear deformation histories, requiring advanced computational simulations and predictive design frameworks to realize stable adaptive devices.
Biomedical implementations of microfluidic systems span analytical point-of-care chips, biomimetic physiological models, and high-throughput microdroplet reactors. Lab-on-a-chip platforms integrate microchannels with microelectrodes, actuators, and biochemical sensors to automate multi-step diagnostic assays from minimal sample volumes. Exemplary devices include potentiometric lab-on-a-chip monitors utilizing ion-selective membranes to measure ammonium concentrations in tap water, microneedle-integrated devices extracting interstitial fluids, and microfluidic liquid biopsy cartridges that isolate circulating tumor cells and exosomal biomarkers directly from whole blood. Organ-on-a-chip platforms mimic complex organ physiology and pharmacokinetic processes by culturing living parenchymal and endothelial cells within microfluidic channels under continuous fluid perfusion, recreating mechanical shear stress, transepithelial electrical resistance, and tissue-tissue interfaces. Specific models include bone-on-a-chip devices incorporating osteoblasts and osteocytes within three-dimensional scaffolds to simulate bone remodeling, osteoporosis, and periodontitis, as well as human cancer-on-a-chip platforms that model metastatic intravasation, extravasation, and vascular transit under physiological flow. Meanwhile, droplet microfluidics uses hydrodynamic flow-focusing geometries to generate highly monodisperse emulsion droplets with coefficient of variation values around 2 percent, achieving active substance loadings up to 67 weight percent and encapsulation efficiencies exceeding 70 percent.
The convergence of microfluidics with single-cell sequencing, wearable interfaces, and artificial intelligence establishes highly automated platforms for molecular analysis and decentralized healthcare. Single-cell microfluidic workflows isolate individual cells into nanoliter droplets, valve-actuated reaction chambers, or nanowell arrays, attaching molecular barcodes to process transcriptomic and genomic material simultaneously across thousands of individual cells. Wearable microfluidic patches adhere directly to skin surfaces, using capillary wicking and sweat pressure to channel perspiration, tears, or interstitial fluid across colorimetric dyes and electrochemical electrodes to monitor sweat rates, pH, metabolites, and electrolytes for diabetes or cystic fibrosis management. Concurrently, artificial intelligence algorithms process massive microfluidic imaging datasets and signal streams. Three-step analytical pipelines employ generative adversarial networks for bubble artifact removal, convolutional neural networks for automated cell classification, and recurrent networks for temporal signal decoding. Nevertheless, technical limitations persist across the discipline, including channel clogging by biological particulates, batch-to-batch variation, high fabrication dependencies on polydimethylsiloxane, and stringent regulatory requirements that continue to govern translation into routine clinical workflows.
Subject of Research: Microfluidic technologies, microfabrication methods, and biomedical applications
Article Title: Microfluidic technologies: from fundamental principles to advanced fabrication and applications
Article References: Deka, R., Shivam, S., Chakraborty, D. D., & Chakraborty, P. (2026). Microfluidic technologies: from fundamental principles to advanced fabrication and applications. AAPS Open, 12(1), Article 45. https://doi.org/10.1186/s41120-026-00175-1
Image Credits: AI Generated
DOI: 10.1186/s41120-026-00175-1
Keywords: Microfluidics, Microfabrication, Lab-on-a-Chip, Organ-on-a-Chip, Droplet Microfluidics, Nanotechnology, Biomedical Engineering, Artificial Intelligence, Microfluidic, technologies, fundamental, principles
News Source: Ophelia Keating. (November 8, 2026). Review Details Advanced Microfluidic Fabrication Systems and Biomedical Applications. Scienmag.



