BioMEMS Magnetic Bead trapping

Microfluidic device for paramagnetic bead trapping
Date: Winter Semester 2014-2015

Team K10-2014: Papanagiotou Dimitris,Tzannetos Giwrgos,Tsiantas Fanis
Students, Dept of Mechanical Eng, NTUA

Short Abstract:
We developed a microfluidic device to trap beads for biological experiments. We also designed some chambers to test our system and performed experimental work to verify its efficiency.


We would like to thank George Kanakaris, Nikos Kavalopoulos and Angeliki Minia for their valuable guidence in the multidisciplinery field of BioEngineering.


We use bead trapping devices to desolate minute particles of interested fluids. One of the most effective means of doing so, is by using a magnetic field to attract and guide paramagnetic beads (act as common magnets when inside a magnetic field) that exist in the fluid.By activating and de-activating the magnetic interaction we achieve our aim.

Why such a Bio-MEMES?

Beads possess paramagnetic properties. Using electro-magnets or conventional ones, we can easily and economically trap the beads by placing the magnet under the chip's chamber.

The basic prerequisites of our design were:
- Economic construction
- Easily manufactured
- Easy to repair
- Low energy requirements
- Abiluity to mix the beads in the fluid so as to achieve a better flow

Design and implementation

A. Existing solutions

The most advanced existing solution found in after research is shown below.
sepmag.pngExisting system for bead trapping.

It is comprised of a cylindrical magnet. By placing a bottle containing the beads and a liquid inside the whole of the magnet we can achieve a trapping round the edge of the bottle. This system is expensive because the used magnets have a high cost. Moreover, it cannot mix the beads in the liquid.

B. Considered solutions


1ο σχέδιο.jpg
Solution_1: Linear actuator in conjuction with servo motor.

The above solution was the first one considered. The linear actuator is used as a means to approach the chamber. The servo-motor materializes the mixing of the beads by rotating itself. The whole system is controlled by a micro-contoller. The estimated cost was 150$. This solution, regardless its advantages offers many drawbacks as well:
- Containes many parts (difficult assembly)
- The magnet attracts beads as it leaves the chamber (influence on trapping efficiency)
- Difficult implementation (construction of base, electro-mechanincal assembly etc.)


In order to solve the second from the aforementioned drawbacks (the most important) we considered a vertical linear actuator that will not affect the orientation of the magnetic field as it approaches or leaves the chip's chamber. However, such a solution is very difficult to implement and requires a complex structure to support the actuator and the servo. It also has high energy requirements, thus, making the whole project unprofitable.


Trying to find an econimical way to implement a vertical trapping we came up with an idea to use gears and a linear screw actuator. In this way the implementation cost drops imidiately as the system costs about 80$. However the same does not apply to the energy requirements. These remain about the same as before so the overall improvemnt of the system is medium.

C. Our system

C1. Structure

Bio-MEMS using arduino and servo-motor:
View_1 of our system View_2 of our system

We decided to use rotation both for the trapping as well as the mixing of the beads. Although the drawback of the influence of the magnetic's field orientation during the approach of the magnet is not totally eliminated, experiments (see below) showed that this influence is negligible.

We can see that the system is made of four basic components: An energy source, a microcontroller, two buttons for the two functions of the system (trapping & mixing) and the servo motor which leads the cylinder. This cylinder contains four strong magnets as shown in the picture below.

For the construction se used the laser printer of NTUA's BioLab.


Cylinder of the servo motor with the positions of the magnets.

The above magnetic orientation was tested experimentally and yielded very satisfactory results (see below). Testing further formations constitutes a matter of future research.

C2. Function

The external power source of 9V ensures low energy requirements. Also, notice that the components are simple, easy to find and purchase and very economical (arduino is the basic kit). The wooden base ensures strenght and lightness. The plexiglass structure is cheap and efficient. The servo motor consumes small amounts of energy and produces enough torque to rotate the magnets. Finally the two buttons make it easy to controll the system which is adjustable, meaning, we can change some of its parameters in the microcontroller to achieve higher efficiency (as in the mixing for instance).

Below we can see a video demostrating the function:

Function of our system

It is evident that the system can mix as well as trap the beads. Using electronic devices, such as amplifiers, we cound increase the current reaching the servo thus changing the response of the cylinder and making it faster or slower. This was not studied in this project and it is a matter of further research.

D. Chip design

D1. Existing chips (bibliography)

There is much research regarding the design of efficient micro-structures with the ability to trap beads in an efficient way. We mostly considered the study of J. Nilsson, M. Evander, B. Hammarström, T. Laurell (“Review of cell and particle trapping in microfluidic systems”) for our designs. Some of the chips found from those researchers as well as from further investigation can be seen below:
Existing chips according to “Review of cell and particle trapping in microfluidic systems”, J. Nilsson, M. Evander, B. Hammarström, T. Laurell

D2. Our designs

We created some structures that are similar to the above. In overall we tested fifteen different designs, however, we will present only those that yielded usefull results (four). We decided upon these four designs after some experiments without using our system in which we investigated the flow inside the chambers experimentally to conclude about the necessary formations.

Plenty of different chip features exist but in our case zig-zag formations do not apply to our purpose because our flow is delayed. So we concluded that the below formations are preponderant for the purpose of bead trapping. In the following pictures we can see our designs as well as the constucted chips. For the construction we used the laser printer availabel in NTUA's Bio-Lab and plexi glass with a width 2mm.

Geometry of chip_1:

CAD design Constructed in NTUA-Bio-Lab's Laser printer

Geometry of chip_2
CAD design Constructed in NTUA-Bio-Lab's Laser printer

Geometry of chip_3:

CAD design Constructed in NTUA-Bio-Lab's Laser printer

Geometry of chip_4:
CAD design Constructed in NTUA-Bio-Lab's Laser printer

E. Experiment

We made the experiment to verify that our system is capable of trapping the beads of a fluid. With the help of the biologists we achieved to overcome the issues that arose. The difficulty of the experiment arises form the fact that the fluid's volume is sume μL so it is hard to manipulate it. Furhtermore, the beads have a diameter of nm so we have to use sophisticated methods to measure them (neubaure plate, etc)

E1. Experimental proceedure

For the experiments we used a solution with 500 beads/μL. In every experiment we injected using a properly altered syringe 10 μl of the above solution. This proceduce was followed for each chip. Firstly for the relaese state and then for the trapping state.
We tested many states. These are:
1. Without magnetic field (to verify the efficiency of the manufactured chips)
2. With magnetic field and:
- Feature (chamber) side up
- Feature (chamber) side down.
Before the injection of the solution we used a mechanical device for the mixing of our fluid, so as to spread the beads and make in homogenous. This device is shown below:
5.Αναδευτήρι.JPGDevice that makes the fluid homogenous

Below we can see some snapshots of the experimental procedure that was followed (review of experimental procedure).
Step_1: Forming an injection device to puss the fluid in the chamber. We can also see one chip and the magnets of the system

Step_2: Creating a structure to collect the results after the trapping is carried out.


Step_3: Collectiong our results, adding buffer and finally calculating the trapping that was achieved (neubauer plate).

E2. Results

After determining which of the four chambers was the most efficient in terms of filling with liquid (using the experiment without a magnetic field) we decided to carry the experiments with the magnetic field only to this one. This decision was made because the experimental process is a complicated one and we would like to see the results that correspond to the best performace of our system.
The most efficient chip was the third one of these shown above (eliptical formation). The results are summarized in the table below where it is evedent that the most efficient chip in terms of filling is the elliptical one:
Our most efficient chip.
We carried four experiments in the elliptical chip:
- Two without a magnetic field (feature side up/down)
- Two with a magnetic field (feature side up/down)

The results are summarized in the table that follows:
Summary of our results (four experiments)

F. Conclusions

We arrived in the following important conclusions:

F1. Our system is capable of achieving significant trapping.
F2. Feature down formation yields better trapping results (because of the distance chamber-magnets that is being reduced)
F3. The trapping efficiency is decent (see above table). With further research a TE of 90% might be possible (with feature down).

G. Future work

We considered some matters that could enhance our system and could form the basis for future work and research. These are:

G1. Different orientation of magnets on the plastic cylinder affects the magnetic field that attracts the beads. As a result, a future study that measures the influence of the formation of magnets would be helpful for the optimization of the system.
G2. Furthermore, optimizing the geometry can lead to higher trapping efficiency. One improvement would be to reduce the distance between the magnets and the chamber. This will lead to stronger magnetic forces and better trapping.
G3. Study of the influence of mixing parametres (the amount of degrees the cylinder rotates when mixing). By changing the microcontroller we can determine how many degrees the rotation of the cylinder will last and conclude about the optimum value in terms of mixing (we used a mixing parameter of +/- 10deg in our system).
G4. Experimental work using different chambers. Further testing of our system using other chambers will lead to more trustworhty results regarding trapping efficincy. It may also reveal other flaws or produce ideas about enhancing our system.

In the following picture we can see a summary of these suggestions.
Suggestions for improvement

H. Study of magnetism and bio-Mems (general) / Presentation

When designing and constructing our system we made a study of magnetism and bio-mems in general. This can be found on the following attachment (greek).

Magnetism and bio-Mems

You can also find our presentation (04.02.2015) below.


Our system