Chapter 1

General Introduction to Microfluidics

What is Microfluidics...!!
Where does it come from…!!
What for?

Lab on a Chip !!!

Microfluidics MEMS Lab on a Chip
(Micro Electro
Mechanical Systems)

Miniaturisation
Automation

Advantages of
Microfluidics
Integration

Fast processing
Low operational
cost

Many Advantages of Miniaturisation

• Size: down to fL, single cell and molecule manipulations
• Superior flow control
• Protection against contamination and evaporation
• Kinetics easy to study
2
Advantages Flow is intrinsically laminar in microfluidics

• Reduced volume of the studied

sample,
• Reduced analysis time,
• Reduced costs,
• Optimization of signal/noise ratio
• Matching between the size of the
microfluidic chips and the size of
living cells, …

µFluidic Mixer

3
Generation of solution and surface gradients using microfluidic systems

(a) Scheme of the PDMS microfluidic gradient generator used in this work.
(b) Schematic design of a representative gradient-generating microfluidic network. Solutions containing
different chemicals were introduced from the top inlets and allowed to flow through the network. The fluid
streams were repeatedly combined, mixed, and split to yield distinct mixtures with distinct compositions in
each of the branch channel. When all the branches were recombined, a concentration gradient was established
across the outlet channel. The width of this channel and the total number of the branches determined the width
of the gradient and the resolution of the steps making up the gradient, respectively.
(c) Equivalent electronic circuit model of the pyramidal microfluidic network.

Whitesides et al. , Langmuir 16, 2000, 8311– 8316.

6
An old science !

7
What is new ?
New micro fabrication techniques

8
9
Perspectives of Microfluidics
- Food industry
- Chemistry
- Biotechnology
- Oil industry
- Drug discovery

Scales for microfluidics

10
Physics at the microscale

(Sandia labs)
 PDMS

100μm

w=100 μm glass

What happens as things become smaller ? A smaller body has a much greater
surface area to volume compared to a larger body.

Consider a cube of length L and a smaller cube with a length L’=L/2

L L/2
• Volume V=L3 of the cube decreases by a factor 1/8, that is V’ = L3/8
• surface area S =6L2 decreases by a factor 1/4 to S’ = 6L2/4.
• surface / volume increases by a factor 2 (from 6/L to 12/L).
• surface / volume is inversely proportional to the size of the system.
12
Physical quantities appearing in the scaling law depend on the scale.
As the size of the system scales down, the surface to volume ratio & surface forces vs. bulk forces
increase.

Variation of some physical quantities with

the size of the system L.

Centrifugal force fc=mw2R=mv2/R

13
Centrifugal force becomes very weak in microsystems
Example: Electrostatic force

Electrostatic forces are dominant over gravitational forces and inertia (which vary as l3).
z
dz
Electrostatic energy stored by a parallel-plate capacitor is written:
d E=V/d
1 ε SV 2
We = CV =
2

2 2d
a displacement normal to the plane of the capacitor requires a force whose intensity is:

𝑑𝑊! 𝜀𝑆𝑉 "

𝐹= =
𝑑𝑧 2𝑑"
"
𝜀𝑙" 𝑉
𝐹 ∼ ∼ 𝜀𝐸 " 𝑙"
2 𝑑
• Not difficult to exert rapid accelerations in microsystems with the aid of electric
fields.
• Microsystems could be used as accelerometers as their weak inertia allows for
the detection of sudden impacts (such as those involving cars) and the detection
of airbags.
14
Main dimensionless Numbers in microfluidics

Reynolds number Weber number

Péclet Number

Diffusion time TD
Pe == =
Convection time TC

Capillary Number 𝑃𝑒 =
𝐷𝑖𝑓𝑓𝑢𝑠𝑖𝑜𝑛 𝑡𝑖𝑚𝑒
=
𝑇#
𝐶𝑜𝑛𝑣𝑒𝑐𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒 𝑇$

15
Hydrodynamics: Navier-Stokes equations simplify !

Conservation of mass density (continuity equation)

Conservation of momentum

! "! !
µ∇ u − ∇ p = 0
2
Flow is perfectly laminar

( l is a characteristic distance of the system).

Re<< 1 in microfluidic flows

Typically, Re ≈ 10-2 to 10-6

For microflows, flow is ‘always’ laminar

(because of the low Re)

16
Be careful when using scaling laws !!!
l l
Convection time tC scales as l0: tC ∼ ∼ ∼ l0
u l(/T )
l2 2
Diffusion time tD scales as l2: tD ∼ ∼ l
D
• This result (if correct) would mean that in microsystems, convection should
be much slower than diffusion (tC >> tD) !!!
• Does it mean that trying to mix a sample by agitating it would be useless, as
diffusion would be the dominant process ?
• This reasoning is not true. Basically, one should take account (also) of the
coefficient in the considered scaling law. In the present case, one should
take account of the value of the velocity (u) and the diffusion coefficient (D)
to determine the Péclet number to deduce the right result.

Diffusion time TD
Péclet Number Pe == =
Convection time TC
17
 • Diffusion and convection

» Dt D

convection
W=100 µm
diffusion
L ~ ut
L » Dt
D D
L = ut
C C

Diffusion time Convection time

𝐿"# 𝑤 " 100𝜇𝑚 "
𝑡# ≈ ≈ ≈ ≈ 250 𝑠
Péclet number 𝐷 𝐷 40𝜇 𝑚" ⁄𝑠
𝑡# ∼ 4 min

Convection time:
Example: For a small molecule (protein) tC = LC u @ t =tD
D= 40 µm2/s ; u = 1000 µm/s
-> Pe = 2500 ⇒ LC = ut D ≈ (1000 µ m / s ) (250s) ≈ 25cm
This large Pe number suggest diffusion Conclusion: Mixing will occur above 25 cm
phenomena are acting much more slowly (along the channel) and 4 min of flow.
than convective transport phenomena
(mainly because of the large velocity)

• For Pe >> 1 convection dominates (diffusion slow: td large and/or tc small)

• For Pe << 1 diffusion dominates (diffusion is fast: td small and/or tc large). 18
Example of applications based on the laminarity of microflows: H-Filter
Sorting particles versus size : H-filter microfluidic device

Diffusion distance
» Dt
convection
L ~ ut
D

diffusion 𝛅
W=100 µm

δ ∼ Dt
2

𝐿"# (diffusion) Einstein-Stokes equation

Péclet 𝑇# 𝑇# =
𝑃𝑒 = 𝐷 𝑘% 𝑇
Number 𝑇$ 𝐿$
𝑇$ = (convection) 𝐷=
𝑣 6𝜋𝜂𝑅

For two particles with two different diffusion coefficients and sizes :
D1/R1 and D2 /R2, since the convection time doesn’t depend on the size of the particle:

Pe1 D2 R1
= =
Pe2 D1 R2 19
ANR Project (El Abed et al. 2021): sorting products of water splitting photolysis (H2 and O2) in
microfluidics channels.
• Passive sorting :
Laminar flow (Re < 1) & convective flow (Pe >> 1)
Trajectory of particles follows lines of flow

ΔPR 2 ⎛ ⎛ r ⎞ ⎞
2

vz (r) = ⎜⎜1− ⎜ ⎟ ⎟⎟ Sorting of particles versus size in

4η L ⎝ ⎝ R ⎠ ⎠ microchannels

21
Active Sorting: apply an external force
Principle of active sorting
« force » field

Input Microfluidic channel output

for sorting

Type of force / involved quantity

Electrophoresis à Charge of surface

Acoustophoresis à Compressibility

Dielectrophoresis à Polarisability

Magnetophoresis à magnetic permeability

22
Magnetic sorting

Without magnetic field

With magnetic field

23
• Dielectrophoresis :
- Sorting depends on the size and polarisability of the particle (or cell)

Claussius Mosotti coefficient

Ø 0> : Positive DEP à Mvt towarsd

maxima of electric field.
Ø < 0 : Negative DEP à Mvt towards
minima of electric field.

24
Extraction of « CTCs » using microfluidic device
CTCs: Circulating tumor cells

Nature Protocols 9, 694–710 (2014) 25

• Acoustophoresis :
• sorting depends on size, density and compressibility of
particles (cells)
à Standing acoustic wave in the microfluidic channel.

max nodes max

Flow direction

Laurell, T.; Petersson, F. and Nilsson,

Section : 350 μm x 125 μm A. Chem. Soc. Rev., 2006, 36, 492-506
Flow rate: 0.3 mL/min

26
 MIXING
Different Strategies for mixing reactants (quickly)
Microfluidic systems are suitable for studying diffusion,
but not efficient for fast mixing of reactants !

Strooke et al. , Science 2002, "Chaotic Mixer for Microchannels".

Knight et al. Phys. Rev. Lett. 80, 3866 (1998).

“ Hydrodynamic Focusing on a Silicon Chip: Mixing Nanoliters in Microseconds”

27
 Introduction to Droplet Microfluidics

Thorsen et al.,
PRL, 86, 4163, 2001

T-junction Flow focusing

28
Droplets, microemulsions and interfaces

Monodisperse micro-emulsions

Solids and soft microspheres

29
Highly monodisperse emulsions = each microdroplet as a independent microreactor

50 μm

Fast mixing (ms range)

Microfluidics 2019 A. El Abed 30

Experimental setup

31
Microfluidics 2019 A. El Abed 32
Microdroplets Spectroscopy

Emission Spectra of 0.1 M fluorescein solution obtained with microdroplets spectroscopy

33
Droplet formation physics governed by the balance between
interfacial forces and viscous forces:

Parallel flow

Capillary number (Ca)

Ca >> 1

Jetting

Ca~1

Dripping

Ca<< 1

34
 Fabrication process of a microfluidic device

Photolithography

Spin-coating Exposure Development

Si wafer subsrtate
UV insolation

Si wafer subsrtate Si wafer subsrtate SEM image of a microfluidic design

Si wafer subsrtate

• Speed
• Rotation time
• Heat treatments

35
Design of a microfluidic system

PDMS molding

36
Design of a microfluidic system
Chip Assembling by Plasma Oxidation

Glass Slide Inlet/outlet punched,

PDMS structure

Channel sealing by
Plasma bonding

Plasma Chamber

Final Device

37
38
High-throughput realtime dilution analysis

39
Flexibility of microdroplets manipulation

Merge Slice

Pico-injection M. Belloul et al. , Appl. Phys. Lett. 103, 033112 (2013).

40
Flexibility of microdroplets manipulation at high
throughput Optical detection, analysis and high throughput sorting

Encapsulation of living organisms (cells,

Microdroplets spliting bacteria, …) Gradient microdroplets
& sorting generator

41
Active control of merging and sorting droplets

42