Phaseguides - a paradigm shift in microfluidic priming and emptying

Reference Link

Here lie my notes on the paper titled the above. Essentially this expands on previous works on phaseguides and shows the basic methods for making them and physical properties of them (as well as various applications which can be viewed in the linked video somewhere in the notes below).

##Paper Notes:


  • Microfluidics haven’t reached their full potential due to issues such as:
    • “chip priming, sample recovery, evaporation issues, large-to- small volume interfacing and bulky control instrumentation”
    • Priming behaviour
      • Constraints on the design results in concession between its envisioned functionality and the sheer need of filling the channel network.
      • Determined by the pressure gradient over the liquid and results from system geometry, applied pressure, and wettability
      • With curved channels, high likelihood of trapping air bubbles
        • also complications from dirt contamination, compound absorption, or multiple material usage
    • Recovery of material that has been separated or manipulated in the channel network is difficult
    • Whereas flow-through systems still facilitate sample recovery through continuous elution of sample, becomes harder to recover from static liquid volumes
      • where emptying of the network is already an issue, selective recovery of compounds from a larger bulk solution is almost impossible
    • Some techniques for specific applications have been used to circumvent these issues, but are not integral:
      • Lateral flow tests
        • capillary pressure control of liquid advancement has been used for this
        • More complex use of capillary pressure is to use it for both passive valving as well as driving force.
      • Centrifugal platforms
        • use combination of pressure driven flow with passive valving components based on capillarity (also used for non-centrifugal platforms)
      • Electrowetting platforms
        • represent advanced control over capillary pressure by electrically controlling the surface of wettability

Working toward a solution

  • Introduced a technique to stepwise control the progress of a liquid-air interface (previous paper that they cite).
    • Done by patterning stripes of metrial acting as a capillary pressure barrier, perpendicular to the advancement direction of the liquid-air meniscus.
      • Stripes force meniscus to align itself with the stripe before jumping over
    • coined the term phaseguides for this
    • used to fill and empty a static liquid dielectrophoresis microchamber with sample and buffer
    • flow profile controlled
      • buffer & sample could be inserted and recovered in parallel, without disturbing the two lane profile
  • continued this work in another paper
    • selectively filled a microfluidic chamber with gel in a reproducible square shape, inserting a sample in a square microchamber, and completely recovering the sample without introducing bubbles or leaving behind remains

This paper

  • Further show phaseguides for filling and emptying of microfluidic structures, extend their functionality and categorize them accordingly
  • also define design principles for phaseguides?

Phaseguiding principle

  • “A phaseguide is a line of material or a change in geometry that spans the complete length of a moving liquid–air boundary.”
  • Essentially the change in materials causes a meniscus that perfectly aligns and then jumps to the next PG
    • Obv. this allows for controlled filling/emptying
  • See below:

Meniscus pinning

  • Meniscus pinning, I guess, is what causes the liquid not to keep going and has this timestep approach
  • Here use patterned bumps on the bottom substrate as phaseguides (you can see the bump on Fig 1c)
  • There are some equations regarding this that govern a complete meniscus pinning
    • contact angles with the top ($\theta_1$ in Fig1c, above) and with the phaseguide ($\theta_2$)
    • $\alpha$ is the angle between the vertical side of the bump and the horziontal top surface
    • Meniscus formed when
$$ \theta_1 + \theta_2 \ge 180 - \alpha$$
  • $\alpha_{crit}$ for when the liquid meniscus is fully stretched is defined by Concus-Finn criterion:
$$\alpha_{crit} = 180 - (\theta_1 + \theta_2)$$
  • For bumps with a vertical sidewall, $\theta_1 + \theta_2$ should be larger than 90 deg. to obtain complete pinning.

Phaseguide overflow

  • When pressure builds up in the pinned meniscus, overflow will occur
    • happens at weakest point along the phaseguide (least amount of energy required for meniscus to advance)
      • point where minimal meniscus stretching is required and maximal surface wetting is achieved
      • typically design a v-shaped solid surface (V-groove). Fig 2 below
        • sharp angle between phaseguide and channel wall (Fig 2a)
        • a sharp bend of the phaseguide (Fig2b)
        • a branch of the phaseguide resulting in a sharp V-shaped groove (Fig 2c)
  • The angle of $\alpha_2$ in all of the above (the angle of the V) determines the stability of the phaseguide.
  • Smaller angle means more likelihood over overflow occuring there?
  • Cricical angle determines when external pressure is required to evoke overflow and the phaseguide typically only consists of one material $\theta_1 = \theta_2$ and thus if all V-groove angles are greater, the thing is stable and allows for complex systems with controlled overflow

Meniscus alignment and geometry

  • Need to make sure the meniscus is completely aligned for accuracy
  • Other dependencies as seen in fig 3:
  • Channel connected to chamber in fig.
    • Phaseguide acts as virtual wall so flow profile in the chamber remains similar to that of the channel
    • Overflow expected to occur at phaseguide-wall interface
      • can measure pressure at this point ($P_2$) by the circuit in Fig. 3 and the following equation
$$ P_2 = \frac{R_2}{R_1 + R_2} (P_{appl} + P_{cap}) - P_{cap}$$ Here, $P_2$ = pressure at expected overflow point
$P_{appl}$ = applied pressure (from circuit??)
$P_{cap}$ = capillary pressure of advancing meniscus
$R_1$ = hydrodynamic resistance until that point
$R_2$ = hydrodynamic resistance from that point till the advancing meniscus
  • Ah, this equation shows that the geometry of the phaseguide pattern is of influence to the overflow behaviour and position
    • BUT, the eq. does not hold for most designs… Instead, need to simulate breakthrough pressure at every point along the phaseguide?


  • Used similar experimental procedure to prev. papers, so they seem to leave a lot out….
  • Ordyl SY 330 for phaseguides
    • 29 μm high and 30 μm wide
  • Heigh of channel struct. 116 μm
  • pipette used as a pump, and access hole was made to achieve seal with pipette tip


This video from the supplementary actually helped a lot in undertanding what they were talking about.

  • Basically, you can do a bunch of cool stuff with phaseguides, you can either control the passing of the phaseguide by making the “weak” angle $\gt 43^{\circ}$
  • For the V indent, an interface angle of $40^{\circ}$ was sufficient for repeatable overflow at that position

Dead angle filling

  • This is an important feature of phaseguides, you can fill angles that normally wouldn’t be filled during normal filling process
  • Do this by placing a phaseguide breaking point at that corner (see butterfly example in the video around the 35s mark)

Meniscus rotation

  • You can actually change the advancement of direction of the meniscus by using the structure shown in the relevant (180 flip) part of the video (around 41 second mark)

Passive valving through sequntial overflow

  • You can actually use phaseguides to selectively choose one of several phaseguides to overflow (based on angles, etc.) to accomplish tasks such as half filling a chamber (around 60s mark of video)
    • They actually have several phaseguides, one in the middle and several to fill dead angles and get rid of airbubbles.
      • The latter are less stable because they have at least one smaller angle, the big divider has a phaseguide of $90^{\circ} so is more stable and thus does not break until the other phaseguides have broken


  • Basically, lots of potential for phaseguides all of which is unproven…
Written on October 9, 2014