Centrifugal microfluidic flow and mixing with applications in cell culture and cell lysing

Pao Yue-kong Library Electronic Theses Database

Centrifugal microfluidic flow and mixing with applications in cell culture and cell lysing


Author: Ren, Yong
Title: Centrifugal microfluidic flow and mixing with applications in cell culture and cell lysing
Degree: Ph.D.
Year: 2013
Subject: Microfluidics.
Hong Kong Polytechnic University -- Dissertations
Department: Dept. of Mechanical Engineering
Pages: xxiii, 223 p. : ill. ; 30 cm.
Language: English
InnoPac Record: http://library.polyu.edu.hk/record=b2639238
URI: http://theses.lib.polyu.edu.hk/handle/200/7100
Abstract: With the recent advances in microfabrication and lab-on-a-chip (LoC), significant development in centrifugal microfluidics (CM) with a wide variety of applications in bio-MEMS has been made. As an indispensable step in all LoC systems, mixing is no-doubt an important process in CM. Despite numerous approaches to achieve efficient mixing either in batch (i.e. mixing in a rotating chamber containing a fixed amount of sample) or continuous flow mode (i.e. mixing of samples and reagents in flow through a rotating channel) have been demonstrated on CM platform, there still remains a number of issues requiring in-depth investigations. These form the basis of our research with results presented in this thesis. In batch mixing using CM, there is inadequate understanding on (a) the complicated vortical flow pattern generated by inertial acceleration of the CM chamber containing the samples to be mixed; (b) geometric and size effect on flow and mixing; and (c) amplitude of angular acceleration/deceleration on flow and mixing. In continuous mixing of fluids in flow through microchannels, the quantitative results of aspect ratio and rotation speed of the microchannel on mixing are not known; therefore the scale-up, design and operation of microchannel for such application is by trial-and-error. To-date, the investigations on CM mixing are primarily on Coriolis acceleration developed from the throughflow in the radial channel under high rotating speed. However, the rotation speed could have been reduced (with benefits of power savings, smaller drive motor, and reducing wear) to only moderate speed level by deploying additional innovative configurations (i.e., zigzag channel, patterning obstruction followed by width constriction, just to name two) so that the Coriolis effect can be further enhanced and other effects can help with mixing. Additional accelerations and effects are brought into play (e.g. centrifugal acceleration from flow around corners of obstruction, splitting and combining flow leading to folding and stretching/compressing the fluid layers) interacting with Coriolis effect in a complex manner, which is the subject of the present study. The knowledge gained can be furthermore adopted in important biochemical technologies and applications such as cell culture, detection and lysis, and mixing in micro-reactors. First, mixing of sample in a rotating chamber was studied. A numerical model has been developed to study the circulatory/vortical flow induced from continuously transient angular acceleration-deceleration of the rotating chamber. A primary vortex, responsible for mixing fluids in the radial-circumferential planes of the chamber, is generated from the inertial effect with the temporal change of vorticity directed opposite to the rotation direction. This dominant vortex obtained from the numerical simulation has been confirmed by two independent means: (a) analytical model developed by Leung assuming quasi-equilibrium under constant acceleration-deceleration, (b) flow visualization by monitoring the motion of the interface between two miscible liquid dyes, and by tracking the movement of neutrally buoyant particles in the flow. In addition to the primary vortex, a pair of toroidal vortices, responsible for mixing fluids in the radial-axial planes of the chamber, is generated from the Coriolis acceleration acting on the primary vortex. The resultant three-dimensional spiral toroidal vortex flow provides effective momentum as well as mass transfer (mixing of different species) in the chamber. Different schemes have been used to realize the continuous acceleration-deceleration, and the scheme involving linear change of angular speed over time has been found to be most effective. The mixing time is under 11 s for milli-chambers with volumes ranging between 1 and 16 μl, and varies from 6 to 14 s for a micro-chamber with volume of 0.07 μl for which higher acceleration/deceleration has been used to compensate for the increased viscous effect and reduced vorticity. The time for mixing per unit volume, specific mixing time (SMT), which measures the effectiveness of mixing, is 0.09 - 2.7 s/μl for milli-chambers, and 87 - 204 s/μl for micro-chambers. More effective mixing with smaller SMT, attributed to higher vorticity and lower viscous friction, can be obtained from higher acceleration/deceleration and with larger chamber (longer radial extent, increased height, and wider angular span). These conclusions have been verified experimentally with our CM platform wherein mixing is slower for smaller angular span (i.e., 5o) chamber for which increasing acceleration from 17 to 34 rad/s2 can compensate the mixing, whereas for larger angular span chamber (i.e. 20o), the improvement is only marginal, as mixing is already quite effective. Experimental measurements on SMT agree well with those of the numerical model. Second, the flow and mixing in unobstructed rotating radial microchannel were studied with objective to improve mixing. The aspect ratio of channel height/width, and rotation speed were correlated to the mixing quality. The numerical flow velocity profiles were in agreement with the analytical model developed by Leung. Too small, or too large, in channel aspect ratio leads to low mixing due to low throughflow and thus low induced crossflow velocity, as well as increased viscous effect. There exists an optimal channel aspect ratio which is a function of the rotation speed.
Third, the flow and mixing in rotating zigzag microchannel were investigated. Crossflow can be intensified from the interaction among (a) channel bend induced centrifugal acceleration which generates Gortler vortices, (b) inclined channel segments induced centrifugal acceleration component in the cross-sectional plane, and (c) rotation induced Coriolis acceleration. These accelerations can lead to improved mixing. The effect of bend angle on mixing quality in zigzag channel was also investigated. Fourth, the investigation on crossflow and mixing in rotating radial obstructed microchannel reveals for the first time that crossflow in the obstructed channel is highly intensified from a combination of (a) local centrifugal acceleration, and (b) Coriolis acceleration induced by rotation. When the accelerations are aligned, they can intensify the crossflow. When they are oppositely directed, Coriolis acceleration competes with the local centrifugal acceleration producing a complex crossflow pattern. Experimental results on mixing quality compares well with prediction from the numerical model, and demonstrate that rotating obstructed channel provides higher mixing quality at moderate rotation speed than both the stationary obstructed channel and the rotating unobstructed channel. Higher mixing quality can be obtained from small ratio of obstruction length to spacing (or width-constriction). The best mixing quality realized is 0.952 (30% better compared to other bench marks) in the obstructed and width-constricted channel at rotation speed of 73 rad/s. Fifth, P. pastoris (strain KM 71) cells were used as a demonstration for cell culture followed by cell lysing using the CM platform. Larger angular span chamber and higher acceleration that were previously found to improve mixing have been confirmed to benefit cell culture as well. Based on improved mixing, cell culture using CM reduces the "time duration" of the initial lag phase of cell culture by at least 1/3 when compared to conventional cell culture. This important finding shows that the improved mass and momentum transfer from the 3-D toroidal vortex flow in the culture chamber can affect the physiology of cells especially at the cell preparation phase prior to the growth phase to follow the lag phase. A novel detection method based on the new findings has been developed wherein the time is monitored for cell growth to reach certain target values. For the entire span of cell concentration tested (8.1x10²/ml to 8.1x10⁵/ml), the time to reach the target OD600 (optical density at 600 nm wave length) =0.25 is less than 12 hr and it can be as short as 3 hr for higher feed cell concentration. Both are much lowered when compared to the conventional shake flask method. Also, the conventional approach needs much longer detection time (as the dynamic response range is too narrow for lower OD600 targets) to allow one to distinguish between two different samples despite that they might be 10-fold apart in concentration. The detection time to reach the target will be further trimmed using CM platform with higher acceleration. Furthermore, multiple chambers, each with different nutrients for cell growth in identifying specific cells, can all be installed in a single rotating disk, rendering diagnostics portable and efficient for point-of-care applications. The 3-D vortical flow induced mixing was also used in cell lysing. The DNA concentration upon cell lysis increases with acceleration. This is due to more effective mixing between cells and lysis buffer, and more intensive cell-bead interactions. Glass beads are commonly added in the mixing devices to provide mechanical cell attrition. Increasing amount of glass beads can further improve the CM lysis yield, because this leads to more collision and friction of the beads with the cells. Extension of lysis duration does not make a significant contribution, as all the genomic and proteomic materials will be released once the cell wall (or membrane) is disrupted. Our optimal CM lysis yield by pure mechanical method (i.e. glass beads) relative to conventional lysis protocol is approximately 56%, and the yield can be improved to 78% by incorporating the chemical method (i.e. lysis buffer). The relatively lower lysis capacity of CM approach is attributed to the maximum rotation speed of 1800 rpm that can be attained with the current test rig, versus 2400 rpm adopted by conventional approach. Despite that, our CM approach incorporates both cell culture and lysis procedures on the same platform with inexpensive setup and simple operation, rendering it readily to be incorporated in a LoC system. In summary, the results of the present study would help to design and operate an effective CM platform (batch chamber and/or continuous microchannel) and optimize cell culture, lysis and detection using a CM platform for automated sample preparation and nucleic acid analysis.

Files in this item

Files Size Format
b26392380.pdf 10.26Mb PDF
Copyright Undertaking
As a bona fide Library user, I declare that:
  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.


Quick Search


More Information