Automated sample collection

What makes pressure-driven flow suitable here?

Faster response times come with precise control, yet stability matters most when handling tiny volumes during separation steps. One channel runs at one microliter per minute, another at 2, and a third at 10 – each managed separately by an OB1 MK4 unit on the same device. Samples move forward in parallel, guided steadily through without overlap. Collection shifts to fresh tubes rhythmically, timed to match output cycles.

Which components are present in the tested configuration?

A single pressure controller – model OB1 MK4, equipped with three channels covering 0 to 2000 mbar – forms the core. Flow is monitored using inline sensors: two MFS2 units handle lower rates, while an MFS3 manages higher volumes. Routing relies on three 12-port MUX distribution valves, enabling access to 36 distinct pathways during a single automated run. Narrow tubing, with an inner diameter of just 1/32 inch, links components with minimal dead volume. Precise fluid resistance comes from calibrated capillaries: 65 micrometers wide, set at 25 centimeters for lines running at 1 to 2 microliters per minute, and at 15 centimeters for the 10 microliters-per-minute path. All streams feed into a multi-channel microfluidic chip where manipulation occurs.

What shape did the chip have, yet why is that significant?

A single chip, model Fluidic 138, holds four long channels running side by side. Each measures 58.5 millimeters in length, one full millimeter across, and 2-tenths of a millimeter deep, built in Topas material. Because these paths are broad and deep relative to typical designs, introducing cells is easier while avoiding high stress at gentle flow rates. Although this version uses four lanes, similar chips can support more – up to sixteen – if needed. However, when pushing liquid containing cells through any such device, confirming the vertical spacing within it matters most.

What does the setup of the collection process look like?

Starting at one microliter per minute, the controller maintains a steady flow, then increases to 2 microliters per minute, then to 10 microliters per minute. Every two hours, the MUX valves shift position, guiding fluid into new channels. A full day of operation produces twelve collected samples per flow rate. That makes thirty-six separate fractions overall. The volume per segment closely matches the intended value. Twelve hundred microliters gather at the highest rate during each interval. At the lowest setting, around 120 microliters collect. Two hundred forty is typical when running at twice the base speed.

How did the living systems respond? What patterns emerged over time?

TNF-α release from microglia increased with flow rate, consistent with mechanosensitive activation (Piezo1). Peak amounts (per 2-hour fraction) occurred around 6 hours for the two faster conditions: roughly 1000 pg at 10 µL/min and ~700 pg at 2 µL/min; at 1 µL/min, the maximum was ~400 pg after about 12 hours. Concentrations were normalized to the actual volume collected for each fraction.

Could there be parts of the control loop still useful to retain?

Flow control P, along with I, reached equilibrium at 0.001, promoting gentle conditions suitable for delicate cells. Verification of the actual versus intended flow is performed through inline sensors. When smoother output is required, applying a brief moving average (5-10 seconds) to the sensor data reduces minor fluctuations while preserving genuine changes in signal shape.

How little effort on the bench is required to ensure proper sterile seeding?

Starting with hydrophilic surfaces makes things easier, though standard coatings work too. A cell suspension at 0.5-1 × 10^6 cells/mL should be fully dispersed before loading. Flow it slowly into the channels using Mini Luer ports to avoid damage. For proper adhesion, incubate for 4 to 16 hours. Throughout this time, maintain fluid levels at both ends with warmed medium. Small reservoirs or short tubes can stop the openings from drying out.

What method reduces leftover parts when working with fractions?

To keep things clean, fit short, inflexible tubes leading straight to the collector. Position the nozzle directly over every tube axis. After each advance, add a quick priming burst that clears fluid trapped in the tip. Well-organized piping helps reduce leftover traces down to just one or two percent of the dispensed amount. Confirm results using fluorescein in a stepwise fashion, then record how quickly the signal moves from 10% to 90%.

How much output should one expect when detailing procedures?

Measurements show the channel size to be 58.5 by 1.0 by 0.2 millimeters. Flow rates were fixed at 1, 2, and then 10 microliters per minute. Each sample remained in place for 2 hours before moving on. Collected volume reached 120, 240, up to 1200 microliters across runs. Controller settings used a proportional gain of 0.001 and an integral term of 0.001. Two MFS2 sensors plus one MFS3 monitored flow during tests. Tubing resistance came from capillaries with an inner diameter of 65 microns and lengths of 25 and 15 centimeters; twelve fractions are formed along each output line within a full day. In testing, TNF-alpha levels rose within each collection window. Peak concentration times followed earlier observations.

What are the main day-1 failure modes – and the fixes?

-A gentle current shift? Try cutting down flexible piping, then go over obstruction distances carefully. Finally, verify the control unit settings match expected values.

-Bubble formation during incubation can be reduced by warming the fluid beforehand. The liquid should also undergo degassing to minimize gas release later. A barrier made of water-repelling material, placed before entry into the chamber, helps capture bubbles early.

-When mis-indexing occurs after extended idle periods, begin by returning the MUX units to their original positions before starting operations. Realign them at regular intervals throughout the process flow.

-A slight deviation in fraction volume – more than 5%- calls for weighing several vials while timing their collection. When discrepancies appear, changes to flow calibration may be necessary. Adjusting dwell time can improve delivery accuracy. Precision improves when measured responses follow observed variance.

Where does the Microfluidics Innovation Center fit for grants and prototypes?

One way to look at the Microfluidics Innovation Center is through its role in grant applications and prototype development. Starting from initial concepts, MIC builds complete fractionation systems – handling chip design, fluid routing, automation logic, and measurement integration – before moving into fabrication of tailored microfluidic devices and delivering functional prototypes. As a specialized R&D company participating in European research collaborations, we often see significantly higher approval rates than the average submission outcome. This effect comes from joint drafting of experimental approaches, verification strategies, and commercialization frameworks, along with full responsibility taken for the microsystem engineering components.

Can the same approach scale to multi-condition or higher throughput screens?

Yes. Using chips that support multiple channels – anywhere from four to sixteen – and coordinated MUX banks allows for switching between different flow speeds, signals, or types of cells. When running at a two-hour interval and a rate of 10 microliters per minute, a single channel produces twelve times 1200 microliters daily. By running several channels together or reducing the time spent on each step, for instance, cutting it down to half or one hour, the total output decreases while gaining finer detail over time.

Small quality-of-life tips that help downstream analysis

Start each label with the exact time, followed by the fraction number. Record target values together with actual flow rates observed. Save the space between the chip exit and the collection tube – this gap affects timing accuracy later. When adjusting the media mix or heat during operation, factor in the altered fluid thickness. Then update control settings to match those shifts smoothly.