Overview
Microfluidic systems operate at very low flow rates, typically below 1 mL/min. While this reduces sample and reagent consumption and promotes highly ordered laminar flow profiles, it makes flow monitoring with conventional flow sensors challenging. These challenges have driven development of various sensors designed for extremely low flow measurements, including multiple types of flow sensors.
Active Flow Sensors
According to MEMS Consulting, flow sensors can be classified as active or passive. Active flow sensors use external energy to perform measurements. Common commercial active sensing techniques include thermal flow sensors and Coriolis flow sensors. Thermal flow sensors are based on convective heating and typically include a heater and two temperature sensors. Heat injected by the heater increases the temperature of the nearby fluid, which is measured by the temperature sensors. As flow rate increases, more heat is carried away, and the temperature difference between the two sensors changes with flow. The flow rate can be inferred from the relationship between temperature difference and fluid velocity.
Recent work has produced single-element thermal flow sensors using transient thermal deflection methods, where a metal structure functions as both heater and temperature sensor. Thermal flow sensors have no moving parts and therefore do not introduce additional mechanical stress, such as shear stress, into nearby fluids, which is important when handling mechanically sensitive samples like human blood. However, added heat can denature thermally sensitive proteins in samples. Thermal sensors can measure dynamic flows, but response times may be limited by the high specific heat of aqueous solutions.
Coriolis flow sensors operate on the Coriolis effect, in which a moving mass in a rotating reference frame experiences a force perpendicular to both its motion and the rotation axis. Commercial Coriolis sensors typically use a U-shaped vibrating tube driven at resonance by magnetic coils. When fluid moves through the oscillating tube, it experiences Coriolis forces proportional to mass flow. These forces produce opposite torques in the tube inlet and outlet arms, causing tube deformation that is measured by sensitive motion sensors attached to the tube. Coriolis sensors provide high accuracy and are insensitive to fluid properties, but the vibrating tube actuators and sensors can make them relatively expensive. Tube vibration can also impose unwanted mechanical stress on mechanically sensitive cells.
Passive Flow Sensors
Passive flow sensors do not require external energy to measure flow. Common passive approaches include gravity-based and cantilever-based sensors. Gravity flow sensors determine flow by measuring changes in collected liquid mass over time. They can be very accurate but require large equipment and have low data acquisition rates.
Cantilever flow sensors infer flow from deflection of a cantilever embedded in a fluidic channel. Cantilever deflection results from viscous drag proportional to flow and is often measured using a piezoresistor at the cantilever base. Cantilever sensors have fast response but require complex, time-consuming fabrication and can be prone to fouling.
Strain-based sensors have also been developed by integrating flexible polyolefin membranes into microfluidic channels. Pressure from fluid motion strains the membrane and changes its electrical resistance, enabling flow measurement based on resistance change.
Particle-Based Optical Methods
For fluids containing high-density suspended particles or cells, particle image velocimetry (PIV) and laser Doppler velocimetry (LDV) have been used to monitor flow. PIV tracks particles within microfluidic structures to measure velocity and is widely used for flow visualization. High-speed cameras combined with continuous-wave lasers can capture high transient flows. Although PIV offers high precision and the ability to map 3D flow fields, it requires expensive optical equipment, including an inverted microscope with a camera and pulsed laser source, and data analysis can be time-consuming.
LDV uses coherent, polarized laser beams and photoelectric detectors to measure frequency shifts in scattered light from moving particles. The Doppler-induced frequency shift is proportional to particle velocity. LDV can also measure flows without particles by detecting diffraction gratings produced by flowing bubbles or droplets. LDV provides localized, instantaneous velocity measurements, but even miniaturized implementations still require costly optical hardware.
Role of 3D Printing
Advances in 3D printing have improved resolution, material variety, and print quality while reducing cost, enabling rapid prototyping of customized microfluidic components. 3D printing offers high precision and three-dimensional freedom for constructing complex structures, facilitating integration of commercially available sensors for flow sensing. These capabilities support broader application of flow sensors in microfluidics.
Design of a 3D-Printed Flow Sensor
Researchers have developed 3D-printed flow sensors that combine: 1) a 3D-printed microfluidic channel; 2) two miniature commercial pressure sensors; 3) a 3D-printed enclosure; and 4) a printed circuit board. The sensor measures flow from the viscous pressure drop between the two pressure sensors mounted on the 3D-printed channel. Changing the channel diameter adjusts the sensor's operating range. The developers demonstrated the sensor for both steady and dynamic flow driven by syringes, piezoelectric pumps, and pressure pumps, and characterized performance with water, water-glycerol solutions, and human blood. Using the sensor's sensitivity, they measured blood viscosity at physiological and room temperatures. They also demonstrated monitoring of transient flows generated by manual pipetting. Tests showed the sensor is compact, low-cost, fast-responding, and free of moving parts, and it can be easily customized, connected, and operated for microfluidic applications.
Manufacturing Process and Characterization
Typical manufacturing steps for the 3D-printed flow sensor are: (a) PCB fabrication; (b) 3D printing of the enclosure and microfluidic structure; (c) patterning semicircular microgrooves to facilitate bubble removal; (d) assembly; and (e) the completed 3D-printed flow sensor.
The sensor was characterized under constant flow driven by a syringe pump.
