Vous êtes sur la page 1sur 50

HYDRODYNAMICS OF THE THREE

PHASE REACTOR
by

Michael Mbrekpadiaha
120510890
Department of Design and Manufacturing

(Mechanical Engineering)

Newcastle University

2013
Table of Contents

Abstract…………………………………………………………………………………………………………………..3

Acknowledgement……………………………………………………………………………………………………4

Nomenclature…………………………………………………………………………………………………………..5-6

List of Figures…………………………………………………………………………………………………………….7-8

1. Introduction…………………………………………………………………………………………………9

1.1 Microreactors – Applications, Advantages and challenges………………………..9-10

1.2 Motivation and objectives………………………………………………………………………….11

2. Literature Review………………………………………………………………………………………….

2.1 Hydrodynamics of a gas –liquid flow in microchannels…………………………….12

2.2 Fundamentals of Multiphase flow…………………………………………………………….13

2.2.1 Dimensionless Numbers…………………………………………………………………..

2.2.1.1 Laminar Flow Condition………………………………………………………

2.2.1.2 Weber Number……………………………………………………………………

2.2.1.3 Capillary Number………………………………………………………………..

2.2.2 Governing Principles………………………………………………………………………

2.2.2.1 Fick’s law of diffusion…………………………………………………………..

2.2.2.2 Henry’s law………………………………………………………………………….

2.2.3 Fluid Properties………………………………………………………………………………

2.2.3.1 Surface tension…………………………………………………………………….

2.2.3.2 Density and Viscosity……………………………………………………………

2.2.4 Parameters of two-phase flow

1
2.2.4.1 Velocities……………………………………………………………………………………

2.2.4.2 Mass, area and volume fraction………………………………………………….

2.2.5 Characteristics of multiphase gas-liquid flows………………………………………...

2.2.5.1 Two-Phase flow regime and Pattern……………………………………………

2.3 Bubble formation and dissolution in microchcannels………………………………………….

2.3. 1 Bubble formation section…………………………………………………………………………

2.3.2 Dissolution of gas bubbles in Liquid……………………………………………………….

3. Two-Phase Flow Regime and Pattern observed

4. Carbondioxide(CO2) Bubble Dissolution

4.1 Experimental Investigation………………………………………………………………………………

4.2 Material and Method………………………………………………………………………………………..

4.3 Result and Discussion………………………………………………………………………………………

4.3.1 Bubble Dissolution…………………………………………………………………………………..

4.3.2 Mass transfer coefficient correlation………………………………………………………

4.3.3 Initial volumetric liquid fraction……………………………………………………………...

4.3.4 The effect of microchannel length…………………………………………………………….

4.3.5 The effect of velocities…………………………………………………………………………….

4.3.6 The effect of flowrate ratios……………………………………………………………………

4.3.7 The effect of continuous liquid phase properties……………………………………

4.3.8 The effect of inlet conditions………………………………………………………………….

5. Conclusions and Recommendations……………………………………………………………………

References…………………………………………………………………………………………………………………….

Appendix

2
ABSTRACT

The present project is aimed at improving the design of three phase microreactors. These
miniaturized devices are widely used in microfluidic industrial applications such as in fast
exothermic reactions involving oxidations, fluorination polymerization and photochemical
reactions. A good number of this application utilizes two-phase gas-liquid flow for their
reactions. The hydrodynamics characteristics of the two phase flow can be utilized to
control various mechanisms such as the mass transfer rates of these microfluidics devices.
Hence the understanding of these flow characteristics will help in the design and process
control of these reactors. In this work, the two-phase gas-liquid flow operating parameters
such as the gas and liquid flowrate, the bubble and superficial velocities, the fluid properties
are experimentally investigated to determine their influence on the rate of gas to liquid
mass transfer which is evidence in the dissolution of the CO2 microbubbles. The
microbubble is generated in a T-Junction geometry mixing section where dispersed gas
phase and continuous liquid comes in contact. For the experiment, CO2 was used as the
test gas while the test liquid consisted of aqueous solutions of triton, sodium dodecyl
sulphate (SDS), polyethylene glycol (PEG) in various concentrations. These solutions were
used to investigate the effect of surface tension and viscosity on the dissolution of carbon
dioxide (CO2) in aqueous liquid phase. From the experimental results, it was observed that
fluid properties and operating conditions play an important role in the two-phase gas-liquid
flow pattern. Having to establish a relationship between the bubble size and the velocity
and the role this relationship plays in determining the flow regimes namely coalescing,
saturating and dissolving regime.

3
ACKNOWLEDGEMENTS

To the glory of God almighty who has brought me this far and to whom I owe my existence
and to the memory of my late father. To the relentless sacrifice of my Mum, the love and
support of my siblings and to everybody who has affected my life.

I express my sincere thanks and appreciation to my supervisor Dr. Ketan Pancholi for his
immense support and for sharing his insight and experience in this field of study as well as
his continuous evaluation to make sure I was at par with the aim of my study and to Dr.
Mohammed Alamiry for the help and support with the experimental work towards the
success of this work

I am also grateful to all the lecturers and MSc students in the Department of Mechanical
Engineering Newcastle University for their support, companionship and for making my time
in this institution worthwhile.

4
NOMENCLATURE

Roman symbols

A Flow cross-sectional area, m2

a Interfacial area, m2

CO2 Carbondioxide

D Difussion coefficient, m2s-1

dh inner diameter, m

J diffusion flux

𝐽G Superficial gas velocity, m/s

𝐽L Superficial liquid velocity, m/s

𝐽M Superficial mixture velocity, m/s

KLa Liquid side volumetric mass transfer coefficient, ms-1

LB Length of bubble, m

Lc Length of capillary

Ls Length of slug, m

PG Gas partial pressure, mbar

QL Liquid volumetric flowrate, m3/s

QG Gas volumetric flowrate, m3/s

Rei Reynolds number

RB Bubble radius, m

RC Capillary radius, m

S Surface area

T Temperature, K

t Time, Secs

Wei Weber Number

5
V Volume

Vo Original volume

Greek symbols

αL Initial volumetric liquid ratio

σ Surface tension, N/m

µ Dynamic viscosity, Pa.s

π Pi constant

∆ Difference

ɛV gas void fraction, rel. units

6
List of Figures

Fig 1. The three types of three-phase reactors (a) the slurry reactor (b) bubble column (c) fluidized
bed……………………………………………………………………………………………………………………………………………………..

Fig. 2.1 General layout of gas-liquid flow patterns……………………………………………………………………………

Fig. 2.2 Sketch of two phase flow patterns in microchannels(Sun 2010). (a) Bubbly flow, (b)
Taylor/Slug flow, (c) Churn flow, (d) Annular flow and (e) the Dispersed flow…………………………………….

Fig. 2.3 Different flow regime in a long serpentine micro channel. (a)Saturating (b)coalescing
(c)dissolving regime………………………………………………………………………………………………………………………….

Fig. 2.4 Microfluidics mixing section (a) T-junction (b) Hydrodynamic focusing (c) Y-junction (d) Flow-
focusing (e) Co-flowing

Fig. 2.5 A schematic illustration of the dissolution and pressure effect on the bubble volume along
the flow direction

Fig. 2.6 Dissolution of CO2 at two individaul pH level of the continous liquid solution.

Fig 2.7 Illustrations of the total component of gas-liquid mass transfer in the segmented
flow modelled by Vandu, Krishna and Baten (2004).

Fig. 3.1 The evolution of bubble sizes along the 2m length, 150μm diameter capillary tube
at different inlet gas pressure, similar liquid flowrate in water triton/deionized water
mixture(a) PG =118mbar, Liquid flow rate, QL =0.006ml/min; (b) 𝑃𝐺 =172mbar, Liquid flow
rate, 𝑄𝐿 =0.006ml/min (c) PG =288mbar, Liquid flow rate, QL =0.006ml/min

Fig. 3.2 The evolution of bubble sizes along the 1m length, 150μm diameter capillary tube
at different inlet gas pressure, similar liquid flowrate in water triton/deionized water
mixture(a) PG =99mbar, Liquid flow rate, QL =0.005ml/min; (b) 𝑃𝐺 =172mbar, Liquid flow
rate, 𝑄𝐿 =0.005ml/min (c) PG =288mbar, Liquid flow rate, QL =0.005ml/min

Fig. 4.1 Schematic diagram of the experimental apparatus.

Fig. 4.3 KaL as a function of the liquid flowrate , QL; comparison of the experimental
calculated results with the theoretical results.

Fig. 4.4 the Evolution of normalised bubble lengths as a function of time for CO2 /Triton mixture. αl=
0.01, 0.02,
0.04………………………………………………………………………………………………………………………………………

Fig.4.5 Change in volume per unit volume as a function of the bubble velocity VB. CO2/1%
conc. of Triton in deionized water………………………………………………………………………………………

7
Fig. 4.6 Relationship between the superficial gas velocity and the bubble velocity……………..

Fig. 4.7 The change in bubble volume as a function of the (A) ratio of liquid to gas flowrate
(B) ratio of gas to liquid flowrate……………………………………………………………………………………………

8
Chapter 1

Introduction
1.1Microreactors – Applications, Advantages and challenges
To satisfy the demand to creating innovative processes with reduced pollution and high
energy efficiency as required from chemically related applications in recent years, it is
important for the industries concern to consistently deliver high yield and cost effective
products at various scales to be used by other industries or for their own processes
(Guagwen, Jun and Quan 2008). One of the ways of achieving this is by enhancing the
production process/device used in these applications by reducing the size of the production
units using the concept of microtechnology. The result of which is the microstructure
reactors. The use of microfluidic multiphase microreactors has made tremendous impact in
numerous applications in industries ranging from medical (drug recovery, delivery and
detection and for diagnostic purposes) to chemical industries. In chemical and process
industry, microreactors are used in applications such as the processing of chemicals and fuel,
production of food, paper and even pharmaceuticals substances. Over the years there has
been a consistent effort aimed at “process intensification” which involves the development
of novel apparatus that will bring improvement in processing in addition to reducing the
ratio of device-size to production capacity leading to a cost effective and energy efficient
technology. (N. Shao 2010)

A typical reaction process in gas-liquid-solid microreactors involves the transfer of gas into
the liquid phase and both brought in contact with the solid phase for chemical reaction to
occur. The need to enhance these processes has initiated the current trend of research into
ways of utilizing hydrodynamics to improve the performances of this micoreactors. Hence, a
fundamental understanding of the gas-fluid flow characteristic such as the flow pattern,
pressure drop and mass transfer is very important for an effective engineering design and
efficient operational purpose. Multi phase reactors are classified according to the
positioning of the solid phase in the system. They include; slurry reactors whose solid
particles is evenly distributed over the reactor, with no velocity between liquid and solid.
The system is agitated by turbulence induced gas bubbles.

9
Fig 1. The three types of three-phase reactors (a) the slurry reactor (b) bubble column (c) fluidized
bed.

With the bubble column three-phase reactor the particles tends to settle at the bottom, but are
pushed from the bottom by liquid flow agitated by gas bubbles, hence the solid suspension is
enhanced by the gas-induced liquid motion. The fluidized bed consists of two distinct regions, the
upper solid free region and the lower solid concentrated region. The solid particle is fluidized by the
flow of liquid and gaseous phase from the bottom of the system upward. (Tsutsumi 1987). In
summary microstructured reactors are beneficial to a wide range of applications:
such as reactions involving rapid mixing of reactants(ranging from factions of a second to a few
minutes), fast rate of reagents heat removal, reactions where explosives and highly toxic
components are involved, the formation of uniform foams, dispersions and emulsions that can be
stable for a long time and new pathways to energy generation and management (Shao 2010).

However despite all these benefits, there are some challenges associated with the design structure
of the aforementioned multiphase reactors, one of which is the inefficient contact between the
individual phases involved in the reaction which reduces the interphase energy transportation
capability due to non uniform distribution of the phases and also the non-uniformity of phase
mixture, for example the bubble column may experience a low gas to liquid volume ratio which will
reduce its reaction rate. In order to satisfy the requirement of progressive innovations in chemical
processes, there have ongoing attempt to modify these reactors, such modifications are intended to
give high gas hold up, low energy consumption, high interfacial mass transfer rate, a control of their
flow modes, the acceleration and improvement of fluid or reactant mixture and bonding for efficient
reaction.(Vandu and Shao 2010).

10
1. 2 Motivation and objectives

In recent times there have been a huge amount of industrial applications which depends on
gas-liquid-solid reactions for such processes as hydrogenation, oxidation, isomerisation and
photocatalysis enhanced by using microstructure reactors. The performance of most of
these devices including their yield rate depends on the hydrodynamics characteristics of the
gas-liquid flow in microchannel, the recent approach is the novel fixed bed gas-liquid-solid
microreactor which utilizes microbubbles to enhance the reaction between CO2 and liquid
at the surface of Titanium Oxide catalyst inorder to generate pollution free fuel as the final
product, High surface area to volume ratio is a major characteristics which makes
microbubbles suitable for this application. The efficiency of most interfacial transport
processes such as heat, mass etc. depends on the surface area between the phases. It could
be seen from the representation below that the surface area to volume ratio of a spherical
bubble increases inversely with its radius.
S 4πr2 3
=4 =
V ⁄3πr3 r

Several investigative research invoving experimental, numerical and theoretical study has
been carried out with the aim of understanding the hydrodynamics characteristics, mixing
and mass transfer for gas-liquid flow in microchannels. A summarised review of these
research study is found in section 2. However there is still a limited understanding on the
dynamics of gas phase dissolution into the liquid phase in the microchannel flow. Hence
the main aim of this study is to investigate the mass transfer rate of C02 into the liquid
phase which is evidence by the dissolution of the bubbles(reduction in the bubble volume),
this aim will be achieved through investigating the various flow factors both dependent and
independent (gas and liquid flow rates and velocities, the liquid phase physicochemical
properties of the liquid phase such as the surface tension and viscosity) with respect to their
effect on the dissolution rate of CO2 core microbubbles

11
Chapter 2

Background and Literature Review

2.1 Hydrodynamics of a gas –liquid flow in microchannels


Most of the applications in microreactor utilizes the gas-liquid flow in microchannel for
processes such as gas absorption, desorption, hydrogenation and direct fluorination, there
an indepth understanding of the hydrodynamics and mass transfer charcteristics of this flow
is very important in predicting the reactor performance. The small cross-sectional size of
the microchannel enhances high rate of mass transfer, thereby increasing the performance
of the processes performed on these devices, also it is important to know the major
differences in the characteristics between a two- phase flow in a conventional size channel
and a microchannel. In microchannel flow there is a high emphasis on the surface over
volume forces, they exhibit a laminar flow, hence a low Reynolds number because of the
dominance of viscous forces over inertial forces and finally the effects of wall roughness and,
wettability and flow confinement factors are taken into consideration (Shao 2010). Other
basic characteristics of a two-phase gas-liquid flow include; the flow pattern, the volumetric
gas content, volume fraction/void fraction, pressure drop and mass transfer. It is therefore
important to be able to manipulate the varying parameters that are controllable such as the
gas and liquid velocities, the geometry of the gas-liquid contactor. In addition to all the
advantages of using microchannels it is also important to be familiar with the consequences
which comes as a result of their sizes such as the susceptibilty to clogging and the high
pressure drop, hence it is important to have a good knowledge of the processes and
reactions inside the system, the principle governing these processes and very importantly
the hydrodynamic characteristics of the microchannel flow.

12
2.2 Fundamentals of Multiphase flow
To fully understand the concept of gas-liquid two phase flow there are certain parameters used to
describe the attributes, cahracteistics and general flow behaviour. Hence these parameters will be
discussed under the following sub-headings.

2.2.1 Dimensionless Numbers


Just as with single phase fluid dynamics, dimensionless numbers are often used to describe
the physical process in multiphase flow so as to reduce large number of variables.

Laminar Flow Condition: the Navier stoke equation is used to describe microfluidic
operating within a scale of 1µm to 1m and is represented by the relationship below;
DU
ρDU = - ▽P + ƞ▽2 (2.1)

Where ρ is the fluid density, ƞ is the dynamic viscosity, P is the pressure and U is the velocity.

The equation can also be represented dimensionless;


D′ U′ 1
Dt′
= ▽′ P′ + 𝑅𝑒 ▽′2 U ′ (2.2)

The included Reynolds number, Re helps to identify the type of flow,it defines the ratio of
inertial forces to viscous forces for the multiphase flow in a microchannel, the Reynolds
number is always low signifying a laminar flow.
ui .D
Rei = μ
(2.3)

Where ui is used to represent the fluid phase velocities with respect to the cross sectional
area flown through by the fluid.

Weber Number: used to analyse fluid flow like the multiphase flow which involves an
interface between two different flows. It is used to represent the ratio of inertia to surface
tension forces. (R. Revellin)
ρ.ui 2 . D
Wei = σ
(2.4)

13
Capillary Number: the capillary number relates the viscous force to surface tension as they
exist at the interface of gas and liquid as represented below;
ƞ𝑈
Ca = 𝛾
(2.5 )

Where ƞ is the shear viscosity of the liquid phase, V is the velocity and γ the surface tension.

2.2.2 Governing Principles


Fick’s law of diffusion: defines the concept of mass transfer which is a very important
mechanism in diffusion as it relates to the dissolution of gas in a liquid. Hence Adolf Fick
derived a way of relating the diffusion flux with the concentration gradient as stated below

J = D.∆𝐶⁄∆𝑋 ( 2.6)

Where J is the mass diffusion flux, while D is the diffusivity constant, ∆C⁄∆X indicates the
change in concentration with time.

Henry’s law: while the diffusion law governs the gas diffusion dynamics, henry law limits its
solubility; it relates the direct proportionality of the gas phase to solubility to the gas partial
pressure, as shown below

PG = Hg C (2.7)

Where PG represents the gas pressure, Hg is the Henry’s constant (which depends on the
solute, solvent and the temperature and finally C is the concentration of the solute.

2.2.3 Fluid Properties


Fluid Properties
Youguang (2010) conducted an experiment to study the formation of bubbles using a T-
shaped microchannel with depth and width 40μm and 120μm respectively, using nitrogen
as the gas phase and glycerol-water mixtures with varying concentration of sodium dodecyl
sulphate (SDS) as a surfactant. He observed that aside from flow parameters such as the
liquid and gas flow rate, the viscosity of the liquid had an effect on the size of the bubble;
there was a decrease in bubble size with increasing viscosity. Surface tension has been
reported in previous studies to have a major influence on the flow characteritics such as lug
formation and slug lengths. While using an experimental approach to study the influence of
various flow and fluid parameters on the bubbles formation mechanism in a flow focusing
micro channel, Dietrich (2013) investigated the effect of surface tension by adding different

14
concentration of surfactant (0.1 and 0.15 wt%) in water, it was reported that the slug length
increased with increasing surface tension and viscosity.

Surface tension: surface tension σ defines the energy per unit area between the liquid
and gas phase. The surface tension influences flow phenomena through the formation of
droplets and curvature of the liquid surface in a liquid of water.
𝐹
σ= (2.8)
𝑙

As a result of this curvature, there will be a pressure difference across the interface. The
difference pressure can be expressed using Young-Laplace equation
1 1
∆𝑃 = σ [𝑅 + ] = σ (𝐾1 + K 2 ) (2.9)
1 𝑅2

∆P is the pressure difference across the curved interface, 𝑅1 and 𝑅2 represents the radii of
curvature and K the surface curvature. With situations involving gas-liquid interface, where
there is gradients in the solute concentration and temperature the surface tension becomes
variable. The surface tension gradient will in turn generate viscous flow in the liquid as a
result of the tangential stress at the interface. This causes the formation of a secondary flow
known as the Marangoni effects which is effective in microchannels due to the dominance
of surface tension forces and can be used to control fluid in these microchannels (N. Shao
2010).

2.2.4 Parameters of two-phase flow


Velocities
There are different types of velocities involved in the two-phase flow, due to the difference
in relative velocity between the gas and the liquid phase.

Gas/Liquid superficial velocity: The individual velocity of each phase in the flow assuming
the flow comprised of a gas and liquid together in a channel.
𝚀𝐺 𝐺
𝐽𝐺 = = 𝜌𝐺x (2.10)
𝐴

𝚀𝐿
𝐽𝐿 = (1-x) (2.11)
𝐴

Hence, the mixture superficial velocity which is also known as the mean velocity in the
liquid slug JM can be represented a

JM = 𝐽𝐺 + 𝐽𝐿 (2.13)

15
Mass, area and volume fraction
The void fraction is an important parameter in the two-phase flow; it is used to define the
ratio of cross-sectional area occupied by either of the phase to the total area of both phase
combine. It is used to determine the average velocities of the gas and liquid.

Cross sectional volume fraction is given by the equations below;


𝐴𝐺
ϵ𝐺 = 𝐴 (2.14)
𝐺 +𝐴𝐿

𝐴𝐿
ϵ𝐿 = 𝐴 (2.15)
𝐺 +𝐴𝐿

Where ϵ is the gas/liquid fraction, 𝐴𝐺 is the area occupied by the gas, 𝐴𝐿 is the area
occupied by the liquid. (Revellin 2006)

2.2.5 Characteristics of multiphase gas-liquid flows


Two-Phase flow regime and Pattern
Most of the studies conduted on gas-liquid two-phase flow emphasis much on the flow
pattern characterization, these flow pattern identified as bubbly, Taylor, churn, slug
annular/annular flow are observed under a wide range of gas and liquid volume flowrates,
superficial velocities and for varying channel geometries such as the circular, rectangular
and triangular section. Investigations are also conducted to understand the effect of of the
initial mixing point of the gas-liquid flow; the T, Y, M-junction, the flow focusing section and
co-flow section and the liquid phase properties; liquid density, viscosity, surface tension and
the moalr mass on the hydrodynamic characteristics and the flow regimes. Understanding
the flow regime plays a major role in designing microfluidics devices involving two phase
flow. For instance, the flow pattern which represents the geometric distribution or topology
of the phase presence in the flow influences mass transfer or dissolution rate. (S. de Loos).
Chemical reaction which involves a gas phase reacting with a liquid phase component is
influenced by the mass transfer rate of the gas component. (Warnier) the two-phase
microchannel flow can be classified into the surface tension dominated, inertia dominated
and the transitional regime and comprises of various flow patterns which are identified in
previous research report as bubbly, slug/taylor, slug-annular, annular and churn flow

16
Fig 2.1 General layout of gas-liquid flow patterns

Bubbly flow. Bubble < 100µm is has attracted a wide range of application which ranges
from biomedical application to chemical processes. This pattern exist at low gas flow rates
as compared to the accompanying high liquid flow rate consist of small bubbles which
according to A. Schumpe (1986), description are uniformly sized with few collisions in
contrast to K. Triplett(1997) description as being distorted (non-spherical) bubbles . Even
with the difference in observation, they both agreed on other characteristics such as the
size (diameter) of the bubble being smaller than the microchannel diameter.

Taylor Flow: Also identified with other names such as segmented flow, slug flow, plug flow
and bubble train flow. It is a common and important type of flow pattern often generated in
microchannels and most often studied by researchers because of its advantage in achieving
other flow characteristics. This flow pattern consist of an arrangement of large size gas
bubbles with length larger than the diameter of the channel size and with varying size of
liquid slugs between them depending on the magnitude of the liquid flow rate. Studies
reveal a recirculation motion that takes place in the liquid plug which enhances the diffusion
of the bubble content.

Churn Flow: the churn flow occurs during the transition from the Taylor to the annular flow
higher proportion of gas which causes the velocity to overcome the surface tension in the
liquid phase causes instability near the trailing end of Taylor bubbles, resulting in the
disruptive movement of the gas into the liquid slug.

Annular Flow Occurs as a result of increasing gas flow rate which causes the taylor bubbles
to merge and the continuous gas core surrounded by a liquid film formed on the tube wall,
sometimes small droplets of liquids can be observed in the liquid because of the waves
induced break-up from the surface of the liquid film.

17
(a) (b) (c) (d)

Fig. 2.2 Sketch of two phase flow patterns in microchannels(Sun 2010). (a) Bubbly flow, (b)
Taylor/Slug flow, (c) Churn flow, (d) Annular flow and (e) the Dispersed flow

Most often times the two phase flow does not maintain a constant flow pattern throughout
the length of the microchannel, the pattern may change with time or position (Cabaud,
2012), these transitional flow pattern are classified into three basic groups; the saturating,
coalescing and the dissolving regime.

(a) (b) (c)

Fig. 2.3 Different flow regime in a long serpentine micro channel. (a)Saturating (b)
coalescing (c) Dissolving regime (Cubaud, Sauzade and Sun 2012)

The dissolving regime is the main regime of interest in this study. The main feature of a
dissolving regime is its transformation from segmented flow to bubbly flow which is
characterised by the decrease in bubble size downstream in the flow direction along the
length of the microchannel. The dissolving regime is also enhanced by the bubbly or slug
flow pattern which are the two most common two-phase flow pattern in microchannel, the
large unit volume gas- liquid interfacial of especially the slug bubble gives it the advantage

18
of enhancing interfacial mass transfer (Sun 2010). Also, the liquid plugs between the
bubbles, encourages recirculation motion which homogenizes dissolved CO2 in the plug,
hence increases the concentration gradient between the CO2 in the bubble and that of
liquid plug therefore increasing the diffusion flux, a phenomenon which is defined by Fick’s
law.

A summary of the literature review of previous studies conducted on two-phase flow in


microchannels including the observed flow patterns and other characterisics are highlighted
in Table 1. Below

Authors Gas-Liquid Experiment setup Flow patterns Results/Observations


A. Elazhary Air-water Impacting T-junction in Bubbly, plug It was assumed that the
a rectangular acrylic churn, annular height of the
block with coss section microchannel could
1.87mm height 20mm have more influence on
width the flow regime
boundaries than its
diameter.
Triplett Air-water Circular d=1.1 and Bubbly, churn, Experimental data for
and 1.45mm and semi- slug, slug- the flow pattern, void
Ghiaasiaan triangular annular and fraction, and the
(1998) microchannel, d=1.09 annular pressure drop had an
and 1.49mm 𝑈𝐺 = overall agreement with
0.02 − 80𝑚/𝑠, 𝑈𝐿 = Suo and Griffith(1964)
0.02-8m/s
Kawahara N2-distilled T-junction mixer Quasi- Liquid properties
and water connected to a circular homogenous influences the bbble
Sadatomi N2-ethanol microchannel, d=250μm and quasi- velocity using the
(2009) seperated equation UG=C0j where
segmented flow C0 the liquid
dependence
distribution an and UG
the bubble velocity
increases with
increasing viscosity and
decreasing surface
tension.
Yue and CO2 –water 667μm diameter Churn, Slug and Good agreement of the
Chen and 0.3M horizontal rectangular Slug-annular experimental data with
(2007) NaHCO3/ microchannel prediction of flow
0.3M Na2CO3 patterns map proposed
C02-1M by Triplett(1999). Liquid
NaOH mass transfer
coefficient increases

19
with increasing
superficial liquid
velocity.
Volkel Air-ethanol, T-Junction mixer with a Bubbly,
(2009) N2-Water rectangular Slug/taylor,
microchannel 1000μm churn and
width and a height of annular flow
500 μm
Kapoor Air-Ethanol Three dimensional Bubbly flow, Gas slug length varies
and Buwa model of a T-Junction Taylor flow, and directly with the cross-
(2009) 0.8x0.8 mm2 square Slug-annular sectional area of the gas
cross section flow inlet. Also the gas and
microchannel liquid flowrates can
influence the bubble
size/slug length.
Kim and Air- Poly-methyl Four main flow It was observed that the
Evans Deionized methacrylate(PMMA) regime; pressure drop in both
(2011) water microchannels of capillary, microfluidic chip has a
different dimensions segmemnted, linear dependence on
fabricated on two types annular, dry and the capillary number.
of micro-fluidic chip. two transitional;
with injection channel segmented/
of 50μm for the width annular,
and height of the micro- annular/dry fro
milled type chip and 50 the two
μm, 200 μm for the different chip.
width and the height 0f
the hot embossed type
chip
Rajesh and Air-water A T-junction For the The dimesionless length
Buwa +SDS hydrodynamic section air/water- od bubbles was said to
(2013) Air- connected to an end- kerosene decrease with
kerosine(oil) milling fabricated experiment, the increasing liquid phase
rectangular following flow flowrate QL. both in
microchannel with regime water with and without
various configurations, identified as; the surfactant. However
bubble drop, the addition of the
slug drop, slug surfactant showed a
long slug, slug decrease in water-oil
two slugs and interfacial tension
slug long slug. resulting in the
deformation of the
bubble at the
downstream

2.3 Bubble formation and dissolution in microchcannels

20
2.3. 1 Bubble formation section
Microbubble of sizes <100µm are Used in most reactors for applications such as ultrasound
imaging in the biomedical sector, bubbles of this dimension are achieved using such
methods as sonification and shear emulsification. Unfortunately these methods provide
limited control over bubble dimensions and their size distribution. Hence the microfluidic
method is an optimal method that can be used to generate microbubbles whose sizes and
distribution can be controlled by tuning the frequency the bubble breakup using the
continuous liquid flowrate and the inlet gas pressure.

The inlet and outlet geometry of microreactors can create a huge impact on their
performances, the way in which gas bubbles are first introduced is very important in
determining the two-phase gas-liquid characteristics such as the flow pattern mixing rates
and the mass transfer rates. The initial size of bubble can be influenced by the geometry of
the mixing section. Microfluidics uses various geometries such as the focusing mechanism,
the co-flowing and the cross flowing geometry in the production of bubbles. The cross
flowing geometry is used by the Y-junction and T-junction mixing section as shown in fig
below; which consist of a continuous phase inlet, dispersed phase inlet and an outlet. The
cross flowing geometry mixing section has proven to have a good effect on the uniformity
on the generated bubble size due to its breaking mechanism

Fig. 2.4 Microfluidics mixing section (a) T-junction (b) Hydrodynamic focusing (c) Y-junction
(d) Flow-focusing (e) Co-flowing

The T-junction and the Y-junction has the cross-flowing geometry as shown in fig. this
geometry has been widely applied in numerous microfluidics processes, the layout enables
the continuous and dispersed phase to meet at perpendicular or opposite directions to

21
each other , with the Y-junction, both phases meet at ˂1800 . Shao (2011) conducted an
experiment to investigate the effect of inlet conditions on the frequency and size of bubbles
formed during gas-liquid flow in a microchannel using a T-junction, Y-junction and the M-
junction he claimed that amongst the three type of inlet junction observed the T-Junction
and the Y-junction generated smaller size bubbles. By using a numerical method, Kapor used
the volume of liquid (VOF) method to investigate the effect of T-junction on bubble
formation period and slug length; he therefore concluded that the bubble/slug length can
be controlled by altering the T-junction configuration. In Donata’s (2008) numerical and
experimental approach to investigate the influence of the channel geometry on the
hydrodynamics properties and mass transfer rate in microchannels, he stated that the gas
inlet geometry can have a great impact on flow characteristics such as the pressure drop
and slug length.

2.3.2 Dissolution of gas bubbles in Liquid


The capability to generate and control dissolving gas carrying bubbles is a long existing
phenomenon that can be utilized to enhance the performances of various applications. This
process is used in daily living for gas exchange in the respiratory system and in blood
circulation. Also known as the shrinking process of microbubbles, one of the main cause of
bubble dissolution is due to the transfer of the gas molecules from the bubbles into the
sorrouding liquid phase as a result of the pressure difference between the internal pressure
of the bubble and its surrounding pressure. The internal pressure increases as the bubble
radius decreases; this enables the gas inside the bubble to diffuse out from the internal
region of high pressure to the surrounding region of low pressure

Various research approach has been channelled towards the study of bubble size evolution
as it flows through a microchannel. According to the ideal gas law relationship, the
shrinkage of bubble can be related to the either the dissolution of the gas molecules in the
adjacent liquid slug, or can be as a result of pressure drop along the microchannel as
illustrated in Fig.2.5. In summary, the volumetric gas liquid mass transfer is influenced by
the bubble velocity VB, the length of liquid slug between each bubble Lslug, and the diffusivity
of the gas in the liquid phase,D. Also it is important to note that the magnitude of gas
solubility in the liquid at a fixed flow conditions varies in location along the microchannel.

22
Fig. 2.5 A schematic illustration of the dissolution and pressure effect on the bubble volume
along the flow direction.( Abolhasani, Kumacheva and Gunther)

Reports extracted from the soft matter journal (2010), shows the role of pH of the liquid
continous phase in the dissolution of CO2 bubbles through chemical absorption as
illustrated schematically in fig . The basic continous liquid phase enhances the dissolution of
CO2 as illustrated by stronger reduction in size compared to the acidic solution eventhough
both had a similar initial size, hence he claimed that pH component as part of the continous
liquid properties can be used to control the magnitude of this dissolution as described by
Henry’s law (CO2)l = kH Pco2 where (CO2)l is the concentration of molecular dissolved CO2, kH
is the henry law constant, and Pco2 is the applied inlet gas pressure.

Fig. 2.6 Dissolution of CO2 at two independent pH level of the continous liquid solution.

23
2.3.2.1 Mass Transfer in Gas-liquid flow
The intensification of mass transfer processes which occur as a result of short diffusion path
and greater interfacial area is one of the main advantages of using microrecators. (Sobieszuk,
Pohorecki, Cyganski and Grzelka, 2011)This is enhanced in segmented flow by the
recirculation motion within the liquid slug, a phenomena claimed by previous studies to
depend on the dimensionless bubble velocity and the length of the slug. The mass transfer
can be described in a system using the mass transfer coefficient, KLa, which serves as an
easy way to analyse results and develop correlations in experiments (Perez 2011). An
immense amount of research has been carried and models generated to predict the KLa
values of gas-liquid flow in microchanels using flow parameters. In Yue and Yuan (2009)
experimental study to investigate the mass transfer and flow properties of air- water Taylor
flow in two square microchannel of hydraulic diameters 400μm and 200μm respectively,
they obtained an empirical correlation model; which shows a depenedency of the KLa on the
bubble velocity and the liquid slug length
𝐷𝑈𝐵 𝐿𝐵
KLa =(𝐿 )0.5 (𝐿 )0.3 0 (2.14)
𝐵 +𝐿𝑆 𝐵 +𝐿𝑆

With an aim of measuring the real time, accurate KLa of a gas-liquid flow in rectangular
microchannels of dimensions between 40μm x240μm and 100μmx2000μm, with working
fluids consisting of CO2-ethanol, CO2-water and CO2-n-propanol Ji, Ma, Zhu and Wang
(2010), carried out an experimenta study which utilized the pressure variation measured
from the microchannel to calculate the KLa using the pressure-volume-tempearture
correlation substituted in the model generated by Yue (2006)
𝐽 𝑐 −𝑐′
KLa = 𝐿𝐿 In( 𝑐𝑒 −𝑐 2 ) (2.15)
𝐶 𝑒 2

Where C0 and C2 represents the concentration of CO2 and are calculated from the measure
pressure, volume and temperature as follows;
𝑉𝐶𝑂2 𝑃𝑖𝑛 −𝑉(𝑃𝑜𝑎 −𝑃𝑜𝑏 )
C2 = (2.16)
𝑅𝑇𝑉𝑙𝑖𝑞𝑢𝑖𝑑

𝑉𝐶𝑂2 𝑃′𝑖𝑛 −𝑉(𝑃′𝑜𝑎 −𝑃′𝑜𝑏 )


C’2 = (2.17)
𝑅𝑇′𝑉′ 𝑙𝑖𝑞𝑢𝑖𝑑

Where 𝑃𝑖𝑛 is the inlet pressure of absorption processes and 𝑃′𝑖𝑛 is for the blank experiment,
𝑃𝑜𝑎 is the outlet pressure after absorption in the experiments, while 𝑃′𝑜𝑎 is same for the
blank experiment, 𝑃𝑜𝑏 is the outlet pressure befor absorption and 𝑃′𝑜𝑏 is same for the blank
experiment, T is the temperature of the system.

While comducting an experimental study to investigate the rate of methane absorption in


water for taylor flow through small tubes, Bercic and pintar (1997) deduced a correlation
from their experimental results which expressed the liquid-side volumetric mass transfer

24
coefficient, KLa to be to vary inversely with the liquid slug length. With subsequent studies
involving mass transfer in gas-liqiud flow, various attempt have been made in correlating KLa
with the hydrodynamic characteristics of the flow. The model developed by Vandu, Krishna
and Baten (2004) took into consideration all the components of mass transfer from the
bubble to its surrounding liquid. It expresses the KLa as the sum of mass transfer(i) from the
bubble cap to the liquid slug and (ii) from the bubble body to the surrounding liquid film.
Henece is the volumetric mass transfer coefficient is stated as;

Fig 2.7 Illustrations of the total component of gas-liquid mass transfer in the segmented
flow modelled by Vandu, Krishna and Baten (2004).

KLa = KL,capacap + KL,filmafilm (2.18)

Where acap is the hemispherical shape of the bubble caps, afilm is the interfacial area of in the
film region and the KL which is formulated according to Hibie penetration theory

𝐷
KL = 2√𝜋𝑡 ( 2.19)

Is modified into

2 𝐷𝐽𝐵 4(𝐿𝐵 −𝑑ℎ ) √2 𝐷𝐽 4


KLa = 𝜋
√(𝐿 −𝑑 ) 𝑑
+ 2 𝜋 √ 𝑑 𝐵 (𝐿 (2.20)
√ 𝐵 ℎ ℎ (𝐿𝐵 +𝐿𝑆 ) ℎ 𝐵 +𝐿𝑆 )

Where D is the diffusivity of the gas in the liquid,

Finally, In Abiev and Lavretsov (2011) analysis of the various hydrodynamics and mass
transfer studies involving segmented gas-liquid in microchannels and from the review
presented they concluded that the KLa varies directly with such parameters as the bubble
velocity, length of liquid film . Table 2 contains a summarised review of other studies
related to gas dissolution and gas-liquid mass transfer.

Author Aim Gas-Liquid Test Section Observation


Cubaud, Sun Study the rate of CO2 CO2 - pyrex glass Cross Bubble dissolution
(2011) dissolution in Liquid, Water, geometry depends on the injection
Studied the dynamics ethanol hydrodynamic gas pressure and fluid
ofCO2 formation and and section, pyrex pair composition
diffusion of CO2 in low methanol glass, square
viscosity liquid cross section
microchannel,

25
d=100μm
Park (2009) Investigate the N2/CO2 - Flow focusing Final dimensions of the
generation and water bubble generator bubble depends on the
controlled dissolution N2/CO2- connected to a initial dimensions which
CO2 bubbles to a hydrogen serpentine can be controlled by gas
desired size of about carbonateAC microchannel pressure, flowrate and pH
10µm by manipulating FV of continuous liquid
the pH of the phase
continuous aqueous
solution
Sun (2010) Experimentally CO2 – water, Lithographically CO2 diffusion
investigation to study ethanol and fabricated cross- determined by gas
the mass transfer during methanol section geometry, pressure and liquid
CO2 absorption Gas and silicon solution. The
microchannel transformation of
segmented flow into
dilute bubble flow was
observed
Cubaud, Studied the relationship CO2 - water Hydrodynamic Three characterisrics
Sunn and between bubble focusing section regime saturating,
Sauzade hydrodynamics and connected to a dissolving and coalescing
(2012) dissolution in confined long serpentine which have their
geometrics channel. dependent on the liquid
volumetric fraction is
established. The
dissolving regime is ideal
for CO2 dissolution in
water.
Abolhasani Investigation of gas C02-ethanol Automated set- Shrinkage of bubbles
and solubilty and mass and up involving a inside the microchannel is
Kumacheva transfer in a microfluidic fluorocarbon computer due to the dissolution of
(2011) platform FC40 controlled syringe gas molecules in the
pump and adjacent liquid plug,
pressure while its expansion is as
regulator for the a result of pressure drop
inlet gas pressure. along the flow direction.
Fries (2008) An experiment to For flow T-Junction
investigate the charaterization. hydrodynamic
influence of channel N2-ethanol, mixer
design and the fluid toluene, water connected to a
properties on gas-liquid or glycerol in rectangular
flow characterization diferent microchannel
and mass transfer rate. concentration. of 100 and
C02- 0.1M of 400μm and
sodium height of 120
hydroxide for and 360 μm
mass transfer.

Sang and A new route for Perfluorocarbon


Wahl producing microbubbles +CO2- 50%
(2013) based on liquid-gas glycerol+water+

26
phase transition using a 9.4x10-3 g/g of
gas that is soluble in the surfactant
water with a an tetradecyltrimet
insoluble gas and a hylammonium
liquid phase of an bromide
acqueous surfactant
solution
Shao, Numerical 5Vol.% CO2/N2
Gavriilidis investigation of mass mixture-0.2M
and Angeli transfer of gas to liquid NaOH or water
(2009) during taylor flow with
and without chemical
reaction,
Aoki, The influences of flow CO2 – 0.1M T-junction On investgating the
Tanigawa parameters such as NaOH mixer relationship between
and Mae flowrate, void fraction connected to the flow parameters slug
(2012) and the channel size on PTFE tubes at size, it was observed
mass transfer rate in a both the inlet that the slug volume is
gas-liquid slug is and outlet of dependent on the inner
experimentally the mixer diameter of the T-
investigated with length Junction and the void
and inner fraction, also the rate of
diameter CO2 obsorbant in liquid
was increased with
increasing liquid void
fraction, α.
Sobieszuk Experimetntal study to CO2/N2 mixture Y-junction The absorption of CO2
and determine the and PFD microreactor increased with
Pilarek solubility of CO2 solution comprising of increasing superficial
(2012) perfluorodecalin(PFD) circular velocity of two-phase
a liquid carrier of microchannel flow as expected. It was
respiratory gases in of inner also concluded that the
bioprocess diameter and gas-liquid microchannel
engineering. length 0.4mm system is suitable for
and 100mm experimental
respectively investigation of CO2
spolubilty in PFD

27
Chapter 3

Two- phase flow regime and pattern observed


The evolution of the bubbles was carefully observed along the serpentine tube from its
point of formation to dissolution fig shows Photographs of visual observation of different
flow pattern together with their various fluid properties and flow parameters. These
regimes from names given in previous studies include; Coalescing, segmented (Taylor) and
the dissolving regime. Also observed during the experiment were the common gas-liquid
flow patterns namely the bubbly, Taylor, churn and the annular.

As a result of the claim from previous literature that the inlet gas pressure and liquid
flowrate will affect the formation size and the dissolution of the bubbles. We carried out an
experiment at a constant liquid flowarate and varying gas pressure to verify this hypothesis

At high inlet gas pressure, a coalescing flow regime characterised by bubbles with diameter
greater than the channel size and with short length of liquid slug between them was
observed. Also the slight increase in the bubble size downstream of the tube, may be
attributed to a combination of decompression and lower solubility condition as often
experienced in low pressure region (T. Cuband 2012), with the reduction of the inlet gas
pressure, a new flow regime with a reduced bubble size and an increased in the slug length
is formed. This flow regime which is known as the saturating flow regime maintains the
same bubble size at both upstream and downstream of the capillary tube. This can occur as
a result of the saturation of the liquid phase with an high concentration of dissolved
carbondioxide hence a reduction in further diffusion of CO2 into the liquid. With further
reduction in gas pressure the dissolving flow regime which is the main regime of interest
was achieved. This regime is characterised by the reduction in bubble size along the length
of the capillary tube an indication of enhanced CO2 diffusion and gas-liquid mass transfer.

28
Flow (a) 400μm

(b)

(c)
Fig. 3.1 The evolution of bubble sizes along the 2m length, 150μm diameter capillary tube
at different inlet gas pressure, similar liquid flowrate in water triton/deionized water

29
mixture(a) PG =118mbar, Liquid flow rate, QL =0.006ml/min; (b) 𝑃𝐺 =172mbar, Liquid flow
rate, 𝑄𝐿 =0.006ml/min (c) PG =288mbar, Liquid flow rate, QL =0.006ml/min

a.) 𝑃𝐺 =99mbar, Liquid flow rate, 𝑄𝐿 =0.005ml/min

b.) 𝑃𝐺 =139mbar, Liquid flow rate, 𝑄𝐿 =0.005ml/min

c.) 𝑃𝐺 =185mbar, Liquid flow rate, 𝑄𝐿 =0.005ml/min, microchannel length, L=1m


Fig. 3.2 The evolution of bubble sizes along the 1m length, 150μm diameter capillary tube
at different inlet gas pressure, similar liquid flowrate in water triton/deionized water
mixture(a) PG =99mbar, Liquid flow rate, QL =0.005ml/min; (b) 𝑃𝐺 =172mbar, Liquid flow
rate, 𝑄𝐿 =0.005ml/min (c) PG =288mbar, Liquid flow rate, QL =0.005ml/min

In summary, a visual analysis of the flow pattern photos shows correlation with T. Cubaud
claims that the bubble sizes increases with increasing gas flow rate while the liquid slug
increases in length with increasing liquid flowrate and that the bubble size is inversely
proportional to the initial volumetric liquid fraction.

30
Chapter 4

𝑪𝑶𝟐 Bubble dissolution


CO2 Is a strongly diffusive gas which can be used to achieve small and soluble bubbles.
Although many researches have been conducted on bubble formation and distribution, a
few works has been done towards the diffusive phenomena in microchannel. T. Cubaud
performed an experimental study which was aimed at investigating the evolution of carbon
dioxide bubbles dissolving in water along a serpentine microchannel, he observed three
operating regime which he identified as the saturating, coalescing and dissolving flow
regime. Citing the dissolving regime which is characterised by a transition from elongated
segmented pattern to a bubbly pattern as the best regime for enhanced gas bubble
dissolution he assumed that the initial shrinkage rate of the elongated bubbles depends on
the gas inlet pressure, a fact which is correlated by R. Sun in his study of the dissolution of
carbondioxide into liquid. He stated that the rate of dissolution of large (elongated) bubbles
depended on the fluid pair composition and the gas injection pressure. Other observation
included the effect of the length of microchannel to the dissolution rate, he claimed that the
lengthy microchannel will cause a pressure drop from high pressure to low pressure region
along the length of the channel hence reduction solubility and a slowing down of dissolution
dynamics. He also observed that bubbles increases in size at the low pressure region down
the channels length due to decompression and lower solubility. Our study will be focused on
investigating the relationship between the rate of dissolution with the flow and fluid
parameters such as the flow rates, the inlet gas pressure, the pH and the viscosity of the
liquid.

EXPERIMENTAL SET-UP

Fig. 4.1 Schematic diagram of the experimental apparatus.

31
Material and method
The investigative experiment to study CO2 bubble dissolution was carried out using an
experimental set-up which consists of a hydrodynamic mixing section with a T-junction
geometry and a serpentine output channel as shown in fig channel maniaturization
produces a high surface area to volume ratio hence, enhancing the interaction between gas
and liquid which allows the formation of slug flow pattern an ideal flow pattern for
increased mass transfer (Aoki and Mae 2011). The T-junction was used because it enhances
the opportunity of controlling the initial bubble dimension and size distribution (Park 2010).
Aoki (2011) claims that maniaturised microchannel with T-Junctions can be used to form
stable gas-liquid slug flow . All the experiment was carried out using CO2 gas and
combination of deionized water with various concentrations of other liquids. Here the two
individual phases flow through orthogonal (perpendicular) microchannels. Liquid at variable
flow rate is injected at one side of the pyrex glass block by a syringe pump through a circular
rubber michrochannel with diameter, d = 150μm and CO2 gas at variable pressure is
supplied to the setup from a compressed tank. The liquid solution contains water with
various concentrations of surfactant such as sodium dodecyl sulphate (SDS), polyethylene
glycol (PEG) and Triton. The experiment was performed at room temperature. An inverted
microscope was used to observe the flow behaviour and the recorded images transferred to
an image processing application computer software connected to the microscope for high
speed imaging. A typical procedure of the experiment involved controlled supplying of the
gas inlet pressure with a pressure control knob from the gas tank, before supplying the
liquid from the syringe pump at flowrates ranging from 0.1 to 1.1ml/hr and the live video of
the flow is monitored, video recordings and images of the flow were capture and saved via a
real-time computer recording application software connected to the microscope.

The experiment was performed in both static and dynamic condition, under the static
condition the bubbles where collected immediately after their formation and kept for
specified period of time to observe their dissolution while the dynamic condition involved
observing the bubbles behaviour as it flows through the michrochannel.

Solvent Viscosity pH Molecular Surface


Weight tension
Polyethylene 90cP at 5% aqueous 43.50
0
glycol (PEG) 25 𝐶 solution (5-7)
Triton 240cPs at 6-8 6.25 33
0
25 𝐶
Table 3 Physical Properties of the liquid used in the experiment
for each set of flow and fluid parameters such as the gas pressure P G, the liquid phase
flowrate QL and liquid viscousity µl, the following measurements were taken from recorded
images;

32
(i) The length of the bubble both upstream and down stream of the microchannel.
(ii) The evolution of the bubble at specific time intervals
(iii) The distance travelled by a reference bubble in successive frames
secondly, the measured measurants and other inlet conditions where used to calculate parameters
such as the the gas and liquid superficial velocity(JG and JL respectively) are calculated by dividing
their individual volume flowrate with the cross-sectional they occupy.

Result and Discussion


Bubble dissolution
The Eulerian approach which describes the properties of the flow at particular position in
space will be used to compare the changes in behaviour of bubbles at the inlet and outlet
side of the microchannel in relation with the varying fluid properties and flow parameters.
For all experiments, measurements consisting of the bubble length at both upstream and
downstream of the microchannel and the change in bubble position with time was taken
using ImageJ,. The bubble length was then converted from pixel into micrometer and used
to calculate the its volume at both upstream and downstream of the micro channel and
finally the change in bubble volume deduced using the soft matter(2010) journal model
which expressed the magnitudeof bubble dissolution as the ratio of change in upstream
and downstream bubble volume to the original bubble volume and this expression will be
used to calculate the magnitude of CO2 bubble dissolution in this study

∆𝑉0⁄
𝑉 (4.1)

Where

∆𝑉0(t) = 𝑉𝑢 − 𝑉𝑑 (t) (4.2)

Where Vu and Vd represents the upstream and downstream bubble respectively.

Mass transfer coefficient correlation


The mass transfer is a very important characteristics which is used in predicting the
performance of gas-liquid flow in microchannels, the volumetric mass transfer coefficient,
KLa can be used to characterise the mass transfer efficiency. Therefore it is very important
to understand the KLa as it relates to other parameters of the two phase flow. Many
publications has shown an indepth study in the experimental and numerical investigation of
mass transfer in the gas-liquid capillary flow. The liquid-side mass transfer coefficients for
this study was calculated from experimental results using the Higbie penetration’s theory

33
2 𝐷𝐽𝐵 4(𝐿𝐵 −𝑑ℎ ) √2 𝐷𝐽 4
KLa = √ + 2 𝜋 √ 𝑑 𝐵 (𝐿 (4.3)
√𝜋 (𝐿𝐵 −𝑑ℎ ) 𝑑ℎ (𝐿𝐵 +𝐿𝑆 ) ℎ 𝐵 +𝐿𝑆 )

Where JB is the velocity of the taylor bubble, D is the liquid phase diffusivity, LB and LS are
the bubble and slug length.

Previous studies have indicated QL and QG as the two main factors affecting mass the
volumetric mass transfer. The volumetric mass transfer coefficient is calculated from the
experimental results and the influences of the flow parameters such as gas flowrate and
superficial gas velocity on KLa is investigated It can be seen from fig.4.2 that the value of
KLa increases with the increasing superficial gas velocity and also that KLa increases with
increasing gas flowrate QG. increasing QG will result in a decreasing liquid slug length, which
will help intensify the the circulation/mixing within the liquid phase of the gas-liquid flow
therefore increasing the KLa and increasing the QL intensifies the circulation mechanism
within the liquid slug giving an homogenous liquid phase which increases KL. Also noted was
the increase in KLa with increasing gas superficial velocity. Increasing the gas flowrate,
increases the number of gas bubbles thereby increasing the interfacial area for mass
transfer.

1.60E-02

1.40E-02

1.20E-02

1.00E-02

8.00E-03

6.00E-03 QL=0.01ML/min

4.00E-03

2.00E-03

0.00E+00
2000 4000 6000 8000 10000

Fig.4.2 The volumetric mass transfer coefficient as a function of the superficial gas velocity.

34
1.05E-02

9.50E-03

8.50E-03

𝑲𝑳𝒂 7.50E-03

6.50E-03 0.03

5.50E-03

4.50E-03

3.50E-03
0.005 0.007 0.009 0.011 0.013 0.015 0.017 0.019
𝑸𝑳

(A)

1.20E-02

1.00E-02

8.00E-03

𝑲𝑳𝒂
6.00E-03
QL = 0.02ML/min
4.00E-03

2.00E-03

0.00E+00
0.58 0.63 0.68 0.73 0.78
𝑸𝑮

(B)

Fig. 4.3 Effect of the (A) Liquid flowrate at a constant liquid volumetic ratio (B) Gas flow rate
at a constant liquid flowrate

35
1.20E-02
Expt result using Higbie
penetration’s theory
1.00E-02
Yue and Yaun (2009)

8.00E-03
Bercic and Pintar (1997)
𝑲𝑳𝒂 6.00E-03

4.00E-03

2.00E-03

0.00E+00
0.58 0.63 0.68 0.73 0.78

𝑸𝑳

Fig. 4.4 KaL as a function of the liquid flowrate , QL; comparison of the experimental
calculated results with the theoretical results.

The calculated KLa from experimental result is in poor agreement with estimates from Bercic
and pintar(1997) presumably because of the differences in operating parameters such as
the channel diameter and the liquid phase capillary number (Ji, Ma, Fu, Zhu and Wang 2010)

Initial volumetric liquid fraction and Dissolution


With the combination of various flow parameters of the gas and liquid phase, different
flow patterns and regimes with a certain degree of dependency on the initial volumetric
where observed in the course of this study

For a low αL , bubbles with large length were observed at the upstream of the microchannel,
with a very small reduction or increased bubble size downstream which indicates a weak
diffusive regime also known as the , at a saturating flow regime, though the bubbles may
shrink in size at the upstream but there is no significant reduction in the size downstream. A
strong diffusive condition occurs at a high αL where the flow pattern changes from a
segmented slug flow to a bubbly flow indicating the characteristics of a dissolving regime
which is the ideal regime to enhance CO2 in liquid. The αL effect has been observed in both
the static and dynamic flow conditions.

We also compared the evolution of the normalised bubble length in the micro channel
using different initial liquid volumetric fraction at specific time intervals in milliseconds, in
order to investigate the magnitude of change in bubble size with respect to the α L. It can
observed that with the increasing αL, the initial and final size of the bubble decreases
thereby increasing the magnitude of bubble dissolution.

36
3.5

2.5

2
d/w

0.01
1.5 0.02
0.04
1

0.5

0
0.00E+00 2.00E-03 4.00E-03 6.00E-03 8.00E-03 1.00E-02 1.20E-02

t(s)

Fig 4.5 the Evolution of normalised bubble lengths as a function of time for CO2 /Triton
mixture. αl = 0.01, 0.02, 0.04

The curves are arranged parallel to one another, this shows that even though the initial
volumetric liquid fraction influences the initial and final size of the bubble but it has no
significance effect on the shrinkage dynamics of the bubble.

Effect of microchannnel length


A comparism analysis in flow behaviour of gas-liquid flow between two capillary tubes of
different length dimension, shows a significance difference in their flow pattern as seen in
the recorded images. visual observation of the flow pattern imagaes shows a significance
change in bubble size in the 2m long capillary compared to the 1m length tube. Also
according to Fig. 3.6 where the gas to liquid flowrate ratio is used as a function of bubble
dissolution, there is a greater magnitude of bubble dissolution in the 2m long capillary. This
correlates T Cubaud 2011 claims that lengthy channels enhances the gas dissolution rate

37
due to an high pressure build up along the length of the microchannel.
4.50E-10

4.00E-10

3.50E-10

3.00E-10

2.50E-10
2m
2.00E-10
1m
1.50E-10

1.00E-10

5.00E-11

0.00E+00
7 12 17 22 27
-5.00E-11

Fig. 4.6

Effect of Velocities

Using the Eulerian approach, The bubble velocity was calculated using the expression ;

∆𝐷
VB = 𝑡

Where ∆D is the change in distance travelled by a reference bubble downstream of the


microchannel in each successive frame of the captured images and t is the reciprocal of the
camera shutter time. The shutter time according to Young and Gebrands (1997) is given by;

𝑁𝑜. 𝑜𝑓 𝑓𝑟𝑎𝑚𝑒𝑠
S= 𝑠𝑒𝑐𝑜𝑛𝑑

∆D was measured by extracting the rear and front spatial coordinates using numerical
methods (imageJ) and velocity for the front and rear of the bubble is calculated and the
average of both is taken as the bubble velocity. Sun (2010) claims that the velocity at the
rear of the bubble is higher than that of the front while the bubble shrinks downstream,
hence to obtain an accurate bubble velocity it is very important to calculate the average of
both bubble front and back velocity VF and VR . It could be observed from figure that the
bubble dissolution represented as the change in volume over the original volume ∆𝑉0⁄𝑉
increases with increasing bubble velocity VB. this is because as VB increases the frequency of
the recirculation motion which occurs in the liquid slug between the bubble increases
improving the homogenous mixing of CO2 in the slug hence the diffusion rate is enhanced

38
(Shao and Gavriilidis 2010). Also figure shows a relationship between bubble velocity and
the superficial gas velocity

0.3

0.25

0.2

0.15
1% Triton QL = 0.02ml/min
0.1

0.05

0
5.00E-03 5.50E-03 6.00E-03 6.50E-03 7.00E-03 7.50E-03 8.00E-03 8.50E-03

Fig. 4.7 Change in volume per unit volume as a function of the bubble velocity V B.
CO2/Triton

1.05E-02

9.50E-03

8.50E-03

7.50E-03 CO2/Triton,
QL=0.02ml/min

6.50E-03

5.50E-03

4.50E-03
0.45 0.5 0.55 0.6 0.65

Fig. 4.8 Relationship between the superficial gas velocity and the bubble velocity, CO2-triton

39
Effect of Flowrate ratio
In previous study on microfluidics bubble formation and dynamics, the flowratio as often
been used as a control parameter. Cubaud claims that QL⁄QG can be used to control the size
of microbubbles, it was also deducted from analysing the result in observing bubble
dissolution in both CO2 /Peg and CO2 /Triton the magnitude of dissolution increases with
increasing liquid to gas flowrate ratio. This is presumably due to the fact that QL⁄QG can be
used to generate small size bubbles with a high internal pressure which enhances its rapid
dissolution. A phenomenon that can be best explained using Young-Laplace Law
2𝛾
∆𝑃 = 𝑅

Where ∆𝑃 indicates a change in pressure at the bubble interface, 𝛾 is the surface tension
and R is the radius. It is clearly seen from the equation that as the bubble size deceases,
their internal pressure increases towards infinity and a sharp pressure gradient is generated
across the bubble interface .

4.00E+08

3.50E+08

3.00E+08

2.50E+08

2.00E+08

PEG
1.50E+08
Triton
1.00E+08

5.00E+07

0.00E+00
0.009 0.011 0.013 0.015 0.017 0.019 0.021
-5.00E+07

(A)

40
0.7
PEG
0.6 Triton

0.5

0.4
DV/V
0.3

0.2

0.1

0
20 25 30 Q /Q 35 40 45
G L

(B)

Fig. 4.9 The change in bubble volume as a function of the (A) ratio of liquid to gas flowrate;
αL <0.02, (B) ratio of gas to liquid flowrate; QL = 0.02mL/min, Lc =2m

The Effect of continuous liquid phase properties


Experiment was also conducted with 1% concentration of polyethylene glycol (PEG) and
triton each in deionized water to investigate the effect of surface tension, density and
viscosity in the dissolution of bubbles. Reports from previous study involving microfluidics
flow has highlighted the effect of certain fluid properties such as the viscosity, molar mass
and surface tension on the formation, distribution and other characteristics of gas-liquid
flow. According to the report from soft matter journal reproduced by Park (2010) on the
experimental study to investigate the effect of the acidity or basicity of the liquid phase on
the formation and dissolution of CO2 bubbles in an acqeous liquid phase, It was observed
that bubbles in the basic solution underwent greater dissolution than that in the acidic
solution; which was attributed to an increased mass transfer of CO2 into the alkaline solution
this claim explains the difference in dissolution observed in the Triton/mixure which has a
high pH value compared to PEG/water mixture. The great magnitude of bubble dissolution
observed with the triton liquid phase is due to the high mass transfer of CO2 into the
solution.

41
Effect of inlet conditions
Dependent flow parameters such as the initial liquid and gas flowrate have been reported in
previous studies to affect the initial and final bubble size. The effect of gas and liquid
flowrate on the dissolution of bubbles and other flow characteristics is investigated using
Triton/water mixture between 0.01 and 0.02ML/min liquid flowrate, it can be seen from
figure that even at higher inlet gas pressure with presumably large initial bubble sizes and
poor dissolution, the rate of bubble dissolution increases with increased liquid flowrate.

0.8
0.7 QL = 0.01ml/min
0.6 QL = 0.02ml/min
0.5
0.4
0.3
0.2
0.1
0
80 130 180 230 280

it can also be observed that decreasing the inlet gas pressure with respect to the liquid flow
rate increases the bubble dissolution. This can be related to the initial bubble size, from
most of our review it is reported that at a constant liquid flow rate, the size of the bubble
dereases with decreasing pressure and in theory, the smaller the bubble size the greater its
internal pressure and according to Henry’s law, the solubility of gas inceases with pressure.

Experimental study of bubble dissolution under a static condition


With the same experimental setup used for studying the dynamic condition of the gas-
liquid flow along the microchannel, with the static condition, bubbles was collected
immediately after formation at the combination of varying flow parameters and kept for a
specified range of time to observe their change in behaviour or characteristics especially
the change. Somes photographs of the monodisperesed bubbles taken immediately after
formation and after specific time intervals are shown below.

42
Investigating bubbles dissolution in a static condition

Bubbles where collected and photographs taken and bubbles is left for a certain amount of
time and the change in bubble size is observed.

PEG surfactant,

𝑃𝐺 =93mbar, 𝑄𝐿 =0.8ml/min 𝑃𝐺 =93mbar, 𝑄𝐿 =0.8ml/min after 25minutes

𝑃𝐺 =116mbar, 𝑄𝐿 =0.8ml/min, conc. = 0.5 (𝑏) after 40mins

(a) 𝑃𝐺 =153mbar, 𝑄𝐿 =0.8ml/min conc. = 0.5 (b) after 40mins

43
Conclusions
Investigations on the the evolution of CO2 core bubbles was performed in a circular capillary
tube of 150μm diameter, with the aim of determing the influence of various flow
parameters and fluid properties such as the gas and liquid flowrates, bubble, superficial gas
and liquid velocities and liquid phase properties on the dissolution rate of the bubbles.
We show that the initial and final bubble size is determined by the liquid volumetric fraction,
we could also relate the diffusion to Fickian’s diffusion theory. With the analysis of the
experimental results, it is accurate to assume that at liquid phase pH>5 where the CO2
molecules reacts with the liquid phase molecules, the flow is dominated by a pH
dependent dissolution.

Bubble in the dynamic condition exhibits a fast diffusive mode under the fickians diffusion
theory, while those in the static condition had a time dependent diffusion.

44
References
Abiev, R. and Latretsov, I. (2011). Hydrodynamics and mass exchange in gas−liquid slug flow
in microchannels; Russian journal of general chemistry, vol. 82, pp. 2088-2099

Abolhasani, M. Kumacheva, E. and Günther, A. (2011). Model-Predictive Strategy for Exploration of


Carbon Dioxide Dissolution and Mass Transfer

Ammari, A. Experimental Investigation of Two-phase Flow in Microchannels “Co-current


Absorption of Ammonia in Water to Design an Innovative Bubble Plate Absorber

Bercic, G. and Pintar, A. (1997) The roles of gas bubbles and liquid slug length on mass transport in
the Taylor flow through capillaries.

Biardia, G. and Baldi, G. Three-phase catalytic reactors

Cubaud, T., Sauzade, M. and Sun, R. (2012) CO2 dissolution in water using long serpentine
microchannels; journal of biomicrofluidics, vol 6,

Donata, M. (2008). Multiphase flow in microchannels: Hydrodynamics and implementation in


process engineering.

Drenckhan, W. (2009). A Microfluidic Approach to Chemically Driven assembly of Colloidal Particles


at Gas–Liquid Interfaces

Guangwen, C., Jun, Y. and Quan, Y. (2008) Gas-Liquid microreaction Technology: recent
developments and future challenges; Chinese journal of chemical engineering, Vol. 16, pp. 663-669.

Elazhary, M. and Soliman, H. (2012) Experimental investigation of the two-phase flow regimes and
pressure drop in horizontal mini-size rectangular test section; archives of thermodynamics, vol. 33,
pp.47-65

Ji, X., Ma, Y., Fu, T., Zhu, C. and Wang, D. (2010) Experimental investigation of the Liquid volumetric
mass transfer Coefficient for upward gas-liquid two-phase flow in rectangular Microchannels;
Brazilian journal of Chemical Engineering, vol 27, pp. 573-582.

Loos, S., van der, J. and Tiggelaar, R. Gas–liquid dynamics at low Reynolds numbers in pillared
rectangular micro channels

Kapoor, T. and Buwa, V. Bubble formation dynamics in t-junction micro-channels

kawahara, A. sadatomi M. and shimokawa, S. Bubble and liquid slug length for two-phase flows in a
rectangular microchannel with t-junction type gas-liquid mixer

Kawahara, A., Sadatomi, M., Nei, K. and Matsuo, H. (2009). Experimental study on bubble velocity,
void fraction and pressure drop for gas–liquid two-phase flow in a circular micro channel;
International journal of heat and fluid flow, vol. 30, pp. 831-841

Kim, N., Evans, E. and Park, D. (2011) Gas–liquid two-phase flows in rectangular polymer micro-
channels.

45
Rebrov, E. (2009). Two-Phase Flow Regimes in Microchannels: Theoritical foundations of chemical
engineering, vol. 44, p. 355-367

Perez, A. (2011). Modelling and experiments of microchannels incorporating microengineered


structures.

Rajesh, V. and Buwa, V. (2013). Experimental characterization of gas–liquid–liquid flows in T-


junction microchannels.
Schumpe, A. and Grund, G. (1986). The gas disengagement technique for studying gas holdup
structure in bubble column reactors. Canadian Journal of Chemical Engineering, 64:891–896.

Shao, N. Gavriilidis, A. and Angeli, P. (2010) Mass transfer during Taylor flow in microchannels with
and without chemical reaction.

Shao, N. (2010). Gas-Liquid Two-phase Flow and Reaction in microstructured reactors. PhD
Thesis.Department of Chemical Engineering University College London

Shao, N., Gavriilidis, A. and Angeli, P. (2011) Effect of Inlet Conditions on Taylor Bubble Length in
microchannels; Heat transfer engineering, 32:13-14, pp. 1117-1125.

Sobieszuk, P., Pohorecki, R., Cyganski, P. and Grzelka, J. (2011) Determination of the interfacial area
and mass transfer coefficients in the Taylor gas–liquid flow in a microchannel The role of gas bubbles
and liquid slug.

Sundaresan, S. Modeling the Hydrodynamics of Multiphase Flow Reactors: Current Status and
Challenges Dept. of Chemical Engineering, Princeton University, Princeton, NJ 08544.

Sun, R. and Cubaud,T. (2011). Dissolution of carbon dioxide bubbles and microfluidic multiphase
flows; Lab chip, vol 11, pp 2924-2928.

Sun, R. (2010). Carbon dioxide multiphase flow in microfluidic devices.

Tao, X. Hydrodynamic characteristics of gas/liquid/fiber three-phase flows based on objective and


minimally-intrusive pressure fluctuation measurements. PhD Thesis.

Tsutsumi, A., Kim, Y. and Togawa, S. (1987) Classification of three-phase Reactors Dept. Of Chemical
engineering, University of Tokyo.

Triplett, K., and Ghiaasiaan, S. (1999) Gas-liquid two-phase flow in microchannels Part I: two-phase
flow patterns; International journal of multiphase flow, vol. 25, pp. 377-394.

Revellin, R. (2006) Experimental two-phase fluid flow in microchannels

Warnier, M. Taylor flow hydrodynamics in gas-liquid-solid micro reactors

Vandu, C., Liu, H. and Krishna, R. (2005) Mass transfer from taylor bubbles rising in single capillaries

youguang, M., Chunying, Z. and Xiyan, J. (2010)Formation mechcanism and the size prediction of
bubble in opposite-flowing T-shaped microchannel

46
Yue, J., Luo, L., Gonthier, Y., Chen, G. and Yuan, Q. (2009) An experimental study of air-water Taylor
flow and mass transfer inside square microchannels; Journal of chemical engineering, vol. 64, pp.
3607-3708.

Yue, J. Chena, G. and Yuana, Q. (2007). Hydrodynamics and Mass Transfer Characteristics in Gas–
Liquid Flowthrough a Rectangular Micro channel

47
Appendix

Experimental datas
Fig 4.5 Evolution of normalised bubble length as a function of time
d/w αl = 0.01 αl = 0.02 αl = 0.04
1.00E-03 2.9 1.7 1.53
3.00E-03 2.8 1.65 1.5
6.00E-03 2.8 1.65 1.5
8.00E-03 2.7 1.6 1.4
1.00E-02 2.7 1.53 1.2

Fig 4.2- 4.4 Liquid side volumetric mass transfer coefficient and operating parameters

𝑲𝑳𝒂 𝑷𝑮 𝑽𝑩 𝑱𝑮

4.21E-03 87 2300 230000


7.50E-03 158 3600 400000
8.12E-03 172 4800 460000
0.014 288 8300 730000

𝑲𝑳𝒂 𝑸𝑳

4.21E-03 0.006
7.50E-03 0.012
9.76E-03 0.017

48
QL KLa
0.59 3.73E-03
0.61 6.80E-03
0.68 9.76E-03
0.77 1.00E-02

Figure 4.7 Change in volume as a function of bubble velocity

∆𝑽⁄ 𝑽𝑩
𝑽

0.05 5.00E-03
0.18 6.00E-03
0.25 8.00E-03
0.27 8.30E-03

Bubble velocity and dissolution

0.04 7.00E-03
0.07 5.00E-03
0.39 3.60E-03
0.66 1.20E-03

Figure 4.8 Bubble dissolution (ft) as a function of the (A) ratio of liquid to gas flowrate

𝑸𝑳
⁄𝑸
𝑮 PEG Triton
0.009 4.00E+06 -1.00E+07
0.01 7.00E+06 2.00E+07
0.015 1.40E+07 4.40E+07
0.02 2.70E+07 3.50E+08

49

Vous aimerez peut-être aussi