Fabrication and Characterization of a microfluidic ...

POLITECNICO DI TORINO Master Thesis

Fabrication and Characterization of a microfluidic platform for cells 3D analysis

Author:

Supervisor:

Mohamed Zakarya Rashed

Prof. Matteo Cocuzza

A thesis submitted in fulfillment of the requirements for the degree of Masters of NANOTECHNOLOGIES FOR ICTs in the ChiLab and IRCC Electronics And Telecommunication Engineering Department October 2014 Author signature : ..............................

Supervisor signature: .....................................

Declaration of Authorship I, Mohamed Zakarya Rashed, declare that this thesis titled, ’Fabrication and Characterization of a microfluidic platform for cells 3D analysis’ and the work presented in it are my own. I confirm that:

This work was done wholly or mainly while in candidature for a research degree at this University.

Where any part of this thesis has previously been submitted for a degree or any other qualification at this University or any other institution, this has been clearly stated.

Where I have consulted the published work of others, this is always clearly attributed.

Where I have quoted from the work of others, the source is always given. With the exception of such quotations, this thesis is entirely my own work.

I have acknowledged all main sources of help.

Where the thesis is based on work done by myself jointly with others, I have made clear exactly what was done by others and what I have contributed myself.

Signed:

Date:

i

“NanoTechnology in Medicine is going to have a major impact on the suvival of the human race .”

Bernard Marcus

POLITECNICO DI TORINO

Abstract College of ELectronics, Telecommunication and Physics (ETF) Electronics And Telecommunication Engineering Department Masters of NANOTECHNOLOGIES FOR ICTs Fabrication and Characterization of a microfluidic platform for cells 3D analysis by Mohamed Zakarya Rashed

This work presents a development of a MultiWell Microfluidic Chip for Chemotaxis Analysis on Embryonic Stem Cells using a PDMS (or thermoplastic) basic structural material which is a biocompatible, cheap and transparent to permit in situ visualization as well as allowing easy bonding to glass surface where the culture media can be grown.The fabrication process is developed through a LIGA like process involving SU8, electroplating and hot embossing technology. As a proof-of-concept, a fabrication of a device composed by three inlets and outlets was obtained,that can be observable with confocal microscopy, with an array of pillars inside the central channel in which spheroids of cells could be inserted and immobilized and at whose opposite extremities a specific and controllable gradient could be established.The device has demonstrated a well diffusion along the channel which has been successfully achieved without either any collapse of the pillars of the middle channel nor lateral leakages through the channels themselves thanks to the well bonded PDMS structure. New design, fabrication and processes were realized to obtain a 3D monolithic buried microfluidic platform onto silicon or glass substrates which is considered to be a promising technique compared to micro-machining processes which requires complex design and realizations. A standard processes was obtained to allow a repeatable fabrication onto silicon or glass substrates ...

Acknowledgements “I don’t know half of you half as well as I should like; and I like less than half of you half as well as you deserve.” J.R.R. Tolkien, The Fellowship of the Ring I would like to give special thanks to all people who contributed to my work either by their teachings or by their encouragement especially Dr.Matteo Cocuzza , Dr.Domenico Mombello , Dr.Simone Marasso, Dr.Simone Benetto and Dr.Denis Perrone for giving me a chance to learn from them. As well as my colleague and friend Cosimo for his support and smile during our work together. Thanks to everyone who supported me during this study especially my family members and friends. . .

iv

Contents Declaration of Authorship

i

Abstract

iii

Acknowledgements

iv

Contents

v

List of Figures

viii

List of Tables

x

Abbreviations

xi

Physical Constants

xii

Symbols

xiii

1 Introduction 1.1 Microchip-Based Cell Culturing Assays . . . . . . 1.2 Chemotaxis and sprouting . . . . . . . . . . . . . 1.3 Cell culturing . . . . . . . . . . . . . . . . . . . . 1.4 Physics and Properties of Microfluidics . . . . . 1.5 Microfluidic approach to chemotactic phenomena 1.6 Newton Project . . . . . . . . . . . . . . . . . . . 1.6.1 Newton Device First Version . . . . . . . 1.6.2 Newton Device Second Version . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

. . . . . . . .

2 Background And Related Work 2.1 µ-Slide Chemotaxis 3D TM . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1 Principle of work . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.1.1 3D/2D Chemotaxis Experiments in/without Gel Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1.2 Mechanical shear stress . . . . . . . . . . . . . . . . . . . . . . . 2.2 Hydrogel/poly-dimethylsiloxane hybrid bioreactor facilitating 3D cell culturing by Bart Schurink & Regina Luttge . . . . . . . . . . . . . . . . . v

. . . . . . . .

1 1 2 5 6 8 9 9 11

14 . 14 . 14 . 16 . 17 . 20

Contents 2.2.1 2.2.2

vi A brief about the presented work . . . . . Methods and materials . . . . . . . . . . . 2.2.2.1 Two-step molding process . . . . 2.2.2.2 Transport through the barrier . 2.2.2.3 Characterization of the barrier . 2.2.2.4 Cell culture in the bioreactor .

3 Equipment and materials used in fabrication 3.1 Materials . . . . . . . . . . . . . . . . . . . . . . 3.2 Technological processes . . . . . . . . . . . . . . 3.2.1 Micro-Milling . . . . . . . . . . . . . . . . 3.2.2 Physical Vapor Deposition (PVD) . . . . 3.2.3 Lithographic Processes . . . . . . . . . . . 3.2.4 Wet Etching . . . . . . . . . . . . . . . . 3.2.5 Replica Techniques and Soft Lithography

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

. . . . . . .

. . . . . .

20 21 21 22 23 23

. . . . . . .

26 26 29 29 29 30 31 32

4 Newton Device Fabrication Technological Processes 4.1 Newton device fabrication limits due to using copper to realize the mold . 4.1.1 Adhesion of SU-8 onto silicon and copper . . . . . . . . . . . . . . 4.1.2 Limitations of electroplating process during the fabrication of copper master mold . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2 Fabrication of SU-8 Newton device master mold (Second Technique) . . . 4.2.1 Mask production . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.2 Wafer preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.3 SU-8 Lithography & deposition of Anti-adhesion layer . . . . . . . 4.2.4 UV Lithography & Developing . . . . . . . . . . . . . . . . . . . . 4.2.5 PDMS casting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.6 Glass bonding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2.7 Coring fabrication procedure to realize off-chip fluidic connections 4.3 3D Monolithic buried Newton device fabrication by double deposition of SU-8 Resist (Third Technique) . . . . . . . . . . . . . . . . . . . . . . . . 4.3.1 Technological Processes flow to Fabricate a monolithic buried microfluidic platform over a silicon substrate . . . . . . . . . . . . . 4.3.2 Fabrication Results & problems . . . . . . . . . . . . . . . . . . . 4.3.3 Technological Processes flow to fabricate a monolithic buried microfluidic platform over a glass substrate . . . . . . . . . . . . . . .

34 35 35

5 Newton Device Testing 5.1 Experimental setup for diffusion test . . . . . . . . . . . . . . . . . . . . 5.1.1 BD MatrigelTM Handling and injection . . . . . . . . . . . . . . 5.2 Filling the Central/Lateral Channels with Fluids . . . . . . . . . . . . . 5.2.1 Filling the BD MatrigelTM into the central channel . . . . . . . . 5.2.1.1 Results obtained after filling the central channel of the PDMS device obtained from Newton SU-8 Master Mold 5.2.2 Filling the Lateral channels . . . . . . . . . . . . . . . . . . . . . 5.2.2.1 Results obtained for filling Lateral channels of the PDMS device obtained from Newton SU-8 Master Mold . . .

52 52 53 55 56

. . . .

36 37 37 37 38 39 41 42 43 44 46 47 49

58 . 60 . 61

Contents 5.2.3 5.2.4

vii Results obtained for filling Lateral & central channels of the 3D Monolithic 2-layers SU-8 device fabricated onto Silicon substrate . 63 Results obtained for filling Lateral & central channels of the 3D Monolithic 2-layers SU-8 device fabricated onto Glass substrate . . 65

6 Conclusion And Perspectives

66

A Technological processes Parameters used in experiments 68 A.1 Deposition of a thin film Teflon-like layer onto SU-8 (DRIE parameters) . 68 A.2 Technological process flow for 3D monolithic buried microfluidic platform fabrication onto glass substrate . . . . . . . . . . . . . . . . . . . . . . . . 69

Bibliography

72

List of Figures 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9

Positive and negative chemotaxis . . . . . . . . . . . . . . . . . . . . . . Cellular deformation and orientation . . . . . . . . . . . . . . . . . . . . Endothelial cells becoming Tip cells . . . . . . . . . . . . . . . . . . . . multiwell microfluidic chip . . . . . . . . . . . . . . . . . . . . . . . . . . Pictorial image illustrates the idea of the first version of Newton device Newton device 1st version structure . . . . . . . . . . . . . . . . . . . . Newton device structure second Version . . . . . . . . . . . . . . . . . . Diffusion Simulation inside Newton device . . . . . . . . . . . . . . . . . Pressure Simulation inside Newton device . . . . . . . . . . . . . . . . .

. . . . . . . . .

3 4 4 9 10 10 12 13 13

2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11

µ-Slide Chemotaxis 3D Device . . . . . . . . . . . . . . . . Homogeneous concentration . . . . . . . . . . . . . . . . . . Difference between 3D/2D chemotaxis experiments . . . . . Difference between 3D/2D chemotaxis experiments . . . . . 2D Versus 3D Chemotaxis . . . . . . . . . . . . . . . . . . . Unidirectional Laminar Flow . . . . . . . . . . . . . . . . . Turbulent and the Oscillating Flow . . . . . . . . . . . . . . Hydrogel/poly-dimethylsiloxane hybrid bioreactor structure Bioreactor final structure . . . . . . . . . . . . . . . . . . . Profiles of the diffusion of FITC-BSA . . . . . . . . . . . . Cells cultured in Matrigel in the bioreactor . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

. . . . . . . . . . .

15 15 16 16 17 18 19 22 23 24 25

3.1 3.2 3.3 3.4 3.5 3.6 3.7

PDMS structure . . . . . . . . . . . . . . . . Benchman VMC 4000 Micro-Milling machine E-Beam Evaporation . . . . . . . . . . . . . Photolithography UV lamp . . . . . . . . . . Injection Molding Machine . . . . . . . . . . Hot Embossing Technique . . . . . . . . . . . Soft-lithography process . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

27 29 30 31 33 33 33

4.1 4.2 4.3 4.4 4.5 4.6 4.7 4.8

Adhesion of SU-8 over Si and Cu . . . . . . . . . . . . . . . . . . . . Microscopic image showing Adhesion of SU-8 over Si and Cu . . . . Electroplated Newton Device sample . . . . . . . . . . . . . . . . . Mask for SU-8 lithography . . . . . . . . . . . . . . . . . . . . . . . Final SU-8 mold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Optical microscope measurements in µm ( . . . . . . . . . . . . . . Table showing Mixing Ratios to obtain PDMS with specific hardness Newton PDMS final device . . . . . . . . . . . . . . . . . . . . . . .

. . . . . .

. . . . . . . . .

. . . . . . . .

35 36 37 38 40 40 41 42

viii

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

List of Figures 4.9 4.10 4.11 4.12 4.13 4.14 4.15 4.16 4.17 4.18

5.1 5.2 5.3 5.4 5.5 5.6 5.7 5.8 5.9 5.10 5.11 5.12 5.13 5.14 5.15 5.16 5.17

ix

Microscopic images for PDMS final device . . . . . . . . . . . . . . . . . PDMS newton device bonded to glass . . . . . . . . . . . . . . . . . . . . Master PMMA mold . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3D Buried microfluidic layer fabrication techniques . . . . . . . . . . . . . Transmission Versus the thickness of SU-8 resist for Various UV wavelengths Mask for exposure of newton device (Third technique) . . . . . . . . . . . Microscopic image of 3D Monolithic Newton microfluidic buried platform Limitations of Monolithic buried layer fabrication . . . . . . . . . . . . . Monolithic Newton device after developing . . . . . . . . . . . . . . . . . Microscopic image showing the Monolithic device fabricated on glass substrate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

42 43 43 44 45 47 48 48 49

Twin syringe pump by Harvard ApparatusTM . . . . . . . . . . . . . . . Fluid experiment setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . Experiment Equipment and guidelines . . . . . . . . . . . . . . . . . . . Filling MatrigelTM into the central Channel . . . . . . . . . . . . . . . . MatrigelTM filling inside the central channel . . . . . . . . . . . . . . . . MatrigelTM after gelation process . . . . . . . . . . . . . . . . . . . . . . Formation of Bubbles inside the MatrigelTM . . . . . . . . . . . . . . . . Newton device with a cap PDMS top layer . . . . . . . . . . . . . . . . Air bubble traping techniques . . . . . . . . . . . . . . . . . . . . . . . . First Lateral channels filling . . . . . . . . . . . . . . . . . . . . . . . . . Second Lateral channels filling . . . . . . . . . . . . . . . . . . . . . . . Lateral channels filling with colored Chemoattractant-free liquid . . . . An optical microscopic image showing the meniscus due to the diffusion of the Chemoattractant-Free . . . . . . . . . . . . . . . . . . . . . . . . . Lateral channels filling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3D Monolithic 2-layers SU-8 Newton Device . . . . . . . . . . . . . . . . Lateral channels filling . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3D Monolithic 2-layers SU-8 Newton Device onto glass substrate . . . .

. . . . . . . . . . . .

53 54 56 57 58 58 59 59 60 61 61 62

. . . . .

62 63 64 64 65

50

List of Tables 4.1

Process flow for the Fabrication of Newton device using double SU-8 layers 46

x

Abbreviations ESC

Embryonic Stem Cell

PDMS

Poly Di Methyl Siloxane

PMMA

Poly Methyl MethAcrylate

PMMA

Poly Methyl MethAcrylate

MEAs

Mult Electrode Arrays

LOC

Lab On Chip

LIGA

Lithographie, Galvanoformung Abformung

PBS

Phosphate Buffered Saline

FITC-BSA

Fluorescein Isothiocyanate Conjugated Bovine Serum Albumin

sccm

standard cubic centimeters per minute

xi

Physical Constants Boltzman Constant

KB

=

1.3806488 × 10−23 Kg.m2 .s−2 K −1 (exact)

xii

Symbols Re

Reynolds number

(dimensionless quantity)

Pé

Péclet Number

(dimensionless number)

ρ

Density

(Kg/m3 )

µ

Viscosity

Pa·S (Kg/m.s)

ν

Velocity

(m/s)

D

Diffusion Coefficient

(m2 /s)

D0

Maximum Diffusion Coefficient

(m2 /s)

d

Mean Distance Traveled

(m)

t

time

(s)

EA

Activation Energy for diffusion

(J.M ol−1 )

T

temperature

(K)

R

Gas Constant

(JK −1 mol−1 )

J

Diffusion Flux

(M ol/m2 .s)

N

Concentration

(M ol/m3 )

xiii

For/Dedicated To my family,Dad, Mom, Rosa,Hilal and science. . .

xiv

Chapter 1

Introduction 1.1

Microchip-Based Cell Culturing Assays

Over the past decade, there has been an increase in demand for high-throughput cellbased assays for drug screening and system biology. Microfabrication technologies enable the realization of performing assay on chip with the advantages of high through-put screening, reduced sample volume and mass-fabrication of high-density arrays. Recently, with the advancements in microfabrication and microfluidic technologies, there have been many developments of lab-on-chip devices for cell culturing [1] [2] [3]. Some on-chip system has the ability to culture the cells under the gradient of a single compound [4]. However, most devices only have the ability to screen for only one compound and there has been no cell culture device capable of assaying for the combinatorial effect of several compounds on cells. The ability of the chip to screen for the combinatorial effect of several compounds has great importance because cells receive more than one extracellular signal at once. Cells divide, survive, grow, differentiate, or undergo apoptosis as a result of the integration of the various different signals that they receive [5]. Also,biological phenomenons such as gene regulation or stem cell differentiation control are governed by the combinatorial effects of several gene regulatory proteins or extracellular factors. In addition, combination therapy has been standard in treatment of HIV, and study has also shown that the combinatorial effect of two or more drugs can improve chemotherapy [6].

To correlate such a model with in vivo conditions, the culturing of cells in a threedimensional (3D) matrix has been proven to be of vital importance. Besides the difference in cell–cell and cell–extracellular matrix interaction, 3D systems outperform 2D cultures in their response to biochemical and biophysical factors. In general, 3D cultures 1

Chapter 1. Introduction

2

better represent native cellular behavior in comparison to planar cultures. One way to establish 3D cultures is the use of a hydrogel matrix, consisting of cross-linked fibers with high water content that are abundantly found in native animal tissues.Taking these materials in vitro, the challenge for creating appropriate culture conditions arises from a lack of a vascular structure for the transport of oxygen and nutrients into the cell culture. In recent years, microfluidic culture systems, or so-called microbioreactors, have been developed to address this problem and enable the use of continuous flow and small culture volumes. Potentially, the research field of neurology can benefit from the fidelity of 3D culturing on a chip. Of particular interest are chips with planar microelectrode arrays (MEAs), which were developed for the study of in vitro neuronal networks and which are generally accepted tools in this field of research. The basic design of a MEA consists of a glass surface with embedded, densely packed microelectrodes for neuron coupling, which are connected to macropatch electrodes for the signal readout. Previously -and still now-, polydimethylsiloxane (PDMS) has been used for the fabrication of microfluidic chips that can be sealed to a glass surface by spontaneous and reversible adhesion, which provides an appropriate technology platform for cell culturing. However, many of the microfluidic cells-on-a-chip designs exert high shear forces onto the cells and their matrices through the flow of culture medium. Therefore, these systems are limited in their utility for prolonged 3D culturing. A proposed solution to such problem is to integrate a barrier, separating the culture chamber from the culture medium channel, which can avoid this direct shear stress onto the culture. Such a design sustains an appropriate mass transport and prevents the cells from migrating out of the culture into the channel, that kind of solution will be described later in the next chapter while the next pages will be devoted to give an idea about the cell culturing and the chemotaxis techniques and definitions.

1.2

Chemotaxis and sprouting

Chemotaxis (originated from the Greek language word which means (chemo= chemistry and taxis = line-up) is a biological phenomenon whereby single cells or multi-cellular entities, such as bacteria or somatic cells, move according to the gradient of a certain chemical concentration in their environment. This is a fundamental deportment exploited for example by bacteria to reach nutrients or useful substances (chemo-attractants), as well as to escape from dangerous or poisoning species (chemo-repellents) and this mechanism is similar also for eukaryotic cells like steam ones.


3

The first behavior described is called positive chemotaxis (movement toward increasing gradient), whilst the second one negative chemotaxis (movement in the opposite direction) and it is sketched in the figure below:

Figure 1.1: Positive and negative Chemotaxis.

It is a very complex mechanism, since it involves the synchronization of many cellular functions: • Sensing, by a dynamic and polarized distribution of specific transmembrane receptors, of the external gradient of concentration of some species in the environment. • Sending specific signals outside the membrane to change the collective cellular behavior. • The mechanical deformation of the cellular structure as a consequence of sensed species. The strategy to obtain a cellular displacement results from the deformation of the cytoskeleton in response to the gradient detected. As a consequence the new structural shape of the cell leads to a different orientation in the space aiming to a preferential movement afterwards (Figure.1.2). Moreover the sprouting occurs due to the sensing of specific chemical stimuli , results in the formation of endothelial sprouts, which is the base of angiogenesis in tumor masses, which are considered to be the base of the process of vascular genesis in vivo, leading to the tumor progression in organisms. Sprouts are composed by tip cells (which define the sprouting direction by sensing chemical gradients) and stalk cells (following the tip to construct vessels in the defined direction) Figure.1.3.


4

Figure 1.2: Cellular deformation and orientation in response to chemical gradients TM c

Ibidi

Figure 1.3: (a), (b) Competition of endothelial cells to become tip cell. c) Sprouting c

web.mit.edu

These two mechanisms have attracted big interest in the field of oncology, since motility mechanism are strongly involved in cancer progression and metastasis, causing the spreading of the tumor mass. One of the most important and best understood example in this field is the epidermal growth factor/epidermal growth factor response (EGF/EGFR) system, highly involved in cancer development. Due to the release of these factors, the gradients resulting may provide chemotactic signals that direct tumor cells motility towards blood vessels, where they can enter the blood stream and move to other sites in the body inducing metastasis phenomenon. For that reason there are extensive studies on chemotaxis and it is the focus of many researches and therapeutics intervention for the cure and the prevention of cancer, inhibitors may block at least certain aspects of the chemotactic response, and may result in new approaches for the treatment of metastasis, by affecting cell motility.


1.3

5

Cell culturing

The cell migration mechanisms play a fundamental role in physiological adult cells processes as well in cancer metastasis as in embryogenesis. Cells move in a 3D environment where they are subjected to different kinds of stimuli. However, nowadays the ability to understand their formation, function, and related pathology has often depended on twodimensional cell cultures. The exploitation of microfluidic platforms leads to a better mimic of a 3D environment. Moreover, microfluidic devices ensure a high throughput testing due to the possibility of their integration in a small area of substrate, allowing also an easier exploitation of parallelism in analysis. Nevertheless it permits a minimum loss in reagents, solutions and chemicals used to perform experiments. Cell culturing is the process by which living cells are grown in similar controlled conditions outside their natural environment. Traditionally, researchers grow cells in glass or plastic Petri dishes in an attempt to replicate in vivo conditions. Cell growth is promoted by periodically replacing the cell culture nutrient media in which the cells reside. Traditional cell culturing occurs on the flat interior surface of a Petri dish, which can be considered as a 2D culture. Four problems with current, manual 2D cell culturing approaches requires some techniques to be maintained. These are replicating a natural growth environment, eliminating contamination,supporting multiple cultures simultaneously, and maintaining consistency and quality. The first issue, unnatural growing conditions, arises from the fact that cells grown in the traditional 2-D fashion do not resemble those that are observed in vivo. Cells do not achieve in vivo like phenotypes and activity when attached to and propagating on a glass or plastic surface. 3D cell culture systems use growth substrates that more closely resemble those in the body. These bio-mimetic surfaces are known to stimulate the growing cells to form organ-like attachments and behaviors. Suspending cells in the cell culture media is advantageous to facilitate feeding and harvesting the cell cultures. One of probable solution could be growing cells on porous, hydrogel beads containing magnetic particles enables the cells to be maintained in suspension by rapid back and forth agitation of the growth vessel (LevitubeTM ) using a stepper motor equipped device (Bio WigglerTM ). The growth beads may be aspirated and dispensed by a liquid handling robot. Harvesting the growth beads is accomplished by using an externally applied magnetic field. This project focuses on developing subsystems that support these 3D cell culturing techniques. Contamination is an issue in many veins of biological research. In most cases, contamination by foreign bacteria or viruses is easy to be detected due to cell death. There are however some hard to detect protozoan contaminants, single cell organisms such as amoeba, that do not kill cells through infection but cause abnormal behavior in tests [7].


6

Because these protozoa are ubiquitous, the one effective method for eliminating contamination risk is to maintain a closed, sterile environment with minimal human contact. To reduce or even eliminate this issue can be by providing a sterile enclosure in which a robot performs all the cell culturing functions a researcher would typically perform.

Lab grown cell populations, called cell-lines, require constant nourishment and monitoring. Limits on human capabilities prevent most researchers from growing more than four cell lines without introducing significant human errors that results in unsatisfactory results or misidentified cell lines. It is estimated that over 8% of published cell lines are misidentified [8] [9]. This limits the amount of usable and consistent cell cultures [9]. By developing automated subsystems that provide scheduled and consistent cultivation of cell lines. This eliminates variance introduced by human error and allows the system to be scaled up to maintain larger cell line populations due to the consistency of an automated system. Several 2D cell culture automation systems have been previously developed, but the application of automation to 3D cell culture is still in its infancy [10] [11]. Automation improves the consistency of cell cultures while also improving the quality of results and repeatability of experiments [12].

1.4

Physics and Properties of Microfluidics

Dealing with microfluidics when at least one dimension of the device (i.e. one size of a channel) is of the order of some hundreds µms, leading to the manipulation of quantity of liquids in the order of µl and less. In case of our study, the size of the channels is comparable to cell dimensions, which are typically around 10µm (for ES cells between 5µm and 50µm). At this scale some macroscopic behaviors are no more preserved,hence by reducing the scale the, same results are not preserved. This is mainly due to the change of the surface to volume ratio, which increases by scaling down the dimensions. As a consequence, at microscopic scale different phenomena compared to macroscale can be observed. The most relevant and evident ones are: • Predominance of viscous effects over inertia in the Navier-Stokes equation, effect that can be estimated by the Reynolds number: Re =

ρν∆ν µν∆2

∼

ρνι µ

(Where µ = viscosity, ρ = density, ν = velocity).

If Re < 2000, then inertia is damped by viscosity till it become negligible. As a consequence the flux in the channels becomes laminar and, if Re < 1, no turbulence is possible.


7

• Mixing modes change from physical mixing to diffusion at interfaces, since it is impossible to obtain the first one owing to absence of turbulence. For that reason in microfluidics, chaotic motion obtained by magnetic beads , as well as centrifugal microfluidic platforms are necessary to break and enlarge interfaces, in order to allow an efficient mixing. • Surface tension of liquids play a fundamental role at microscopic scale, allowing the exploitation of different phenomena. Such microscopic scale physical phenomena were largely employed in this project in order to avoid the penetration of liquid in certain zones, despite some apertures were created. The project is mainly depending on surface tension force of liquids on the microscopic scale to prevent the passage of the fluids through the lateral channels from the main central channel. It was performed using pillars distanced by specific micrometers to prevent the fluids to enter depending on their surface tensions. • Capillarity effects, as a consequence, also become much more relevant at small scale, leading to effects very different from macro-scale correspondents. Formation of bubbles is a big issue in narrow channels and devices, they can lead to the failure of the working of the device itself, and besides small bubbles can provoke fatal damages in the technological process i.e pillars of the channel can be broken due to the force exerted by the bubbles upon them due to their big effect in micro or nano scale unlike in macroscale effect . Diffusion is a dominant transport process in microfluidics, and in particular plays a fundamental role in our device. Due to diffusion mechanism , particles move from high concentration to low one thanks to the Brownian motion. In 1-D, as it is in practice in the case of study, diffusion can be modeled by the following equation: d2 = 2Dt (where d = mean distance traveled, D = diffusivity coefficient, t = time ). The dependence of the one-dimensional molecular net flux (j) on the concentration (N) is described by the Fick’s Law: j = −D

∂N ∂x

(1.1)

Diffusivity D[cm2 /s] describes how slow or fast an object diffuses and It has a great dependence on Temperature (T): D = D0 exp−

EA KB T

(1.2)


8

An important factor which is the Péclet Number (Pé), which is an indicator of the advection/diffusion ratio.

The advantages of the microfluidic usage in technology and especially in Lab-on-Chip field which are:

• It permits a good portability and lower price due to the reduction of materials for the manufacture and the reagents involved in the analysis as well. • lower energy required to perform specific experiment due to the miniaturization of the devices as well as volumes needed of reagents or stimuli. • Higher throughput due to parallel and Faster processing ( Surface/Volume reduction ) . • Possibility to analyze smaller samples and so no waste of important materials, as well as no need for higher blood volumes for example to perform some tests.

1.5

Microfluidic approach to chemotactic phenomena

The principal function that have to be performed in the study of chemotaxis, as explained previously, is the creation of chemical gradient, which can be constant or time-varying, but it must be well controlled and reproducible. Traditional methods are unable not only to act locally around one or few cells, but also to create a real gradient, since at macro-scale mixing dominates over diffusion, so exploiting complex membranes or dispenser cannot be kept constant and stable in time. Moreover microfluidics allows to perform automatic and dynamic analysis too. As a consequences, as well as the small amount of reagent and cells needed, these are the great advantages of micro-gradient generators in those biological studies. Any microfluidic device designed for performing chemotactic processes can be based on two principles: flow-based or flow-free. The first one is the most powerful, laminar flux can introduce stable gradients which can be dynamically varied. While the second one minimizes the stress on cells and structures and the complexity in the control system, but apart from its simplicity, it gives the possibility of introducing only static gradients and working only with equilibrium states of the system.


1.6

9

Newton Project

Newton Project is the name given to the multi-well microfluidic chip For chemotaxis analysis on embryonic stem cells by The research group operating at the Chilab of the Politecnico di Torino in collaboration with IRCC (Institute for Cancer Research and Treatment) and University of Genova, and with other interdisciplinary partners.

1.6.1

Newton Device First Version

The aim of this project is the fabrication of a microfluidic multi-well platform to analyze chemotaxis mechanisms on ES cells Figure.1.4.These cells, which could be induced to differentiate in about 220 adult cells types, represent a well-known model to be studied for a widespread number of applications.

Figure 1.4: Pictorial image of multi-well microfluidic chip

The main applied idea is a two parallel micro-channels, in which the solutions flow independently around some chambers in which the biological samples can be positioned and remain fixed; the shape must guarantee that the lateral fluidic flow can not enter the central (main) chambers, but only lap the cavities to contact the culture field (to be able to diffuse inside) in which spheroids are immersed. This is the key to produce the desired controlled gradient, allowing also to change it dynamically, without direct contamination and physical contact with the cells under study. The first version of Newton device schematic view is shown in Figure.1.5. The chip contemplates an array of 7 chambers in series. The influence of different flow pressure and the incidence of performed diffusion due to the presence of previous chambers is nowadays matter of study, as it will be explained later.


10

Figure 1.5: Pictorial image illustrates the idea of the first version of Newton device with 7 chambers in series. a) Pictorial image of the central chamber. b) Pictorial image of the central channel.

Figure 1.6: Newton device structure

A PDMS microfluidic platform was designed and fabricated using standard MEMS technology and connected to an external pumping system as described elsewhere [13] [14]. The depth of channel is about 160µm, and it is imposed and tuned during the step of SU-8 spinning. Instead their width is imposed by mask design and it is equal to 800µm. A parallel micro-channels configuration was chosen to bring stimulating or inhibiting liquids to the micro-chambers. A micro array of pillars allows for the interface


11

with the micro-chamber containing the spheroid Figure.1.6. The main problem of this configuration is related to the fact that liquids passing into the channels could enter the central chamber, because those windows were not able to really confine them exploiting surface tension only. Pillars are disposed along an arch and their diameter is 50µm, which is also equal to the distance between themselves. In the center there is a lodging of dimension comparable to a spheroid of cells so as to keep it in position. The aim of these micro-structural pillars is to act as a barrier that maintains the liquid in contact with the micro-chamber and let the reagents diffuse from the mainstream avoiding any mixing with the reagents in the parallel micro channel. The spheroids (size: 500 µm in diameter), containing the ES cells, were maintained in a hydrogel and pipetted in the micro-chamber. As shown in Figure.1.6 b, the latter has confinement walls for the spheroid centering and a fluidic system for the evacuation of the excess of hydrogel. The PDMS device was obtained from a SU-8 master and bonded on a cover slip glass slide Figure.1.6 C. The inlet and outlet of the chip were obtained by using a PMMA counter mold, which is mechanically aligned on the SU-8 pattern.In order to analyze the response, the ES cells were engineered to express a fluorescent protein when a specific subtype was obtained from a given induced stimulus Figure.1.6 d.

1.6.2

Newton Device Second Version

Another version of Newton device has been designed and realized to fulfill different requirements of chemotactic analysis. The target is always the study of cell chemotaxis, but the main difference is that in this case, instead of working with a single spheroid fixed into a chamber, a high number of cells can be spread everywhere in the culture field, which is distributed in a greater volume along a channel. It gives the possibility to work on a higher number of cells, interacting also between each other, and having the possibility of moving in a larger volume. A controllable gradient has to be created in the culture region, and the movement of cells in their 3D culture environment can be observed through microscopy. The design is simpler compared to the first version one. It consists of a central channel of width 500µm with an inlet and outlet,used as culture field, divided by 75µm-pillars from two lateral channels in which solutions can flow. The structural design can be shown in Figure.1.7.with the detailed dimensions of the realized device. To be sure about the idea of using the surface tension force of fluids to design pillars with specific dimensions and distanced between each others with specific distance to prevent the stimuli flow from leaking into the central channel where the cells are cultured . As it


12

Figure 1.7: a ) Optical microscopic image of SU-8 master of Newton-2. b) Zoomed in optical microscopic image of the pillars between channels (dimensions expressed in µm

is so difficult and complex to simulate the MatrigelTM material, so the dimensions and distances were obtained empirically by performing different tests with different pillars dimensions and distances and by allowing the passage of the lateral fluidic flow, its possible to study the perfect pillar dimensions which can prevent the passage of the fluids into the central channel. While the diffusion occurs in the space between pillars, with the same mechanism of the Newton 7-chambers chip version, but through a longer path. A simulation have been done to study the effect of passing a fluid flow in the lateral channels onto the central one, and so one can study if the design of the pillars with 75µm diameter and distanced between each others with same value, would cause a passage of the lateral fluid flow or not. A schematic view of the diffusion mechanism simulated using R which was performed by one of my colleague(Ribet Federico). COMSOL Multiphysics

The 2D simulation has been done considering these two important aspects: • It took into account only half of the module as long as it is symmetric • It simulates the worst case condition when water is entering a channel full of air only,since the entrance among pillars is of course easier when there is only atmospheric pressure than when a MatrigelTM is present inside because the MatrigelTM is more viscous than the air .


13

As shown in Figure.1.8, Figure.1.9, even with a high flow rate, the surface tension of water is strong enough to prevent the entrance of the flow in the chamber, despite the presence of air only ,in the real case with collagen the counter-pressure would be indeed higher, facilitating further the restrain of pillars, also an evident slowdown in correspondence to the narrow passage created by pillars, this results also in a temporary increase in pressure which accumulate until the front passes this region and can vent in the next part of the channel, strongly increasing the flow velocity.

Figure 1.8: Passage of fluid around pillars without invading the chamber simulated R . a) Imposed initial interface in the model. After a using COMSOL Multiphysics duration of: b) 0.14 s . c) 0.22 s.

Figure 1.9: Pressure along the channel during the fluid flow along the pillars a) During the passage . b) After the passage. Note: The increase of pressure revealed is from 100 Pa to 200 Pa during the passage, so it doubles until the fluidic flow is able to be vented and recover a significant lower value again.

Chapter 2

Background And Related Work This section will be dedicated to discuss the previous related & similar work to the project, especially those which were proved to be experimentally successful and some of them are already in market . One of the first and most sold product in market is µ-Slide Chemotaxis 2D & 3D

TM ,

made by IbidiTM . The need to mimic the 3D behavior of the human body and in vivo silmulations,more focus will be given to the 3D one.

2.1

µ-Slide Chemotaxis 3D

TM

The µ-Slide Chemotaxis 3DTM was developed to investigate the chemotactical behavior of fast or slow migrating, non-adherent cells being embedded in a 3D gel matrix, and exposed to chemical gradients. It is possible to observe the migration in linear and stable concentration profiles for over 48 hours. As gradients are rapidly established, fast responses (occurring in less than 30 min) can also be measured. The µ-Slide Chemotaxis 3DTM is for use with all adherent cells that will be observed in 2D or in in vivo-like 3D gel conditions, such as endothelial cells, fibroblasts, cancer cells, and others and nonadherent cells, such as T-cells, dendritic cells, neutrophils, monocytes, granulocytes, lymphocytes, and others. The design of the device can be show in Figure.2.1.

2.1.1

Principle of work

Two large-volume reservoirs are connected by a small gap. If the large reservoirs contain different chemoattractant concentrations (indicated in this document by the red and blue colors), then there is a linear concentration gradient inside this gap. Cells that

14

Chapter 2. Background And Related Work

15

TM c Figure 2.1: µ-Slide Chemotaxis 3D device structural view Ibidi

are placed into this gap (=observation area) are exposed to linear concentration gradients. In steady state, there is a homogeneous concentration inside the large reservoirs as shown in Figure.2.2.

Figure 2.2: Pictorial image showing the homogeneous concentration inside the TM c gap,Note the bright brown color is the gel Matrix doing the chemotaxis Ibidi

Chapter 2. Background And Related Work 2.1.1.1

16

3D/2D Chemotaxis Experiments in/without Gel Matrices

When conducting 3D experiments, the observation area is filled with cells that are surrounded by a gel. This gel hinders the convective flow of liquid, so that both large reservoirs can be filled with neutral medium and chemoattractant. Typical aqueous gels (like collagen gels) are not thought to hinder diffusion. When using stiff hydrogels with pore sizes, in the range of the diffusing molecules, this approach is invalid and no chemical gradient can be established inside the gel. It is also possible to conduct 2D experiments without the use of gel. For this method, cells need to be slightly adherent to the surface of the observation area. referring to the color code shown in Figure.2.1.the difference between the two experiments can been in Figure.2.3 ,2.4.

Figure 2.3: Pictorial image showing difference between a) 3D chemotaxis experiment TM c in Gel matrices b) 2D chemotaxis experiment without gel Ibidi

Figure 2.4: Pictorial image showing difference between a) 3D chemotaxis experiment TM c in Gel matrices b) 2D chemotaxis experiment without gel Ibidi

It is commonly accepted that cells in a culture behave very differently when they are attached to a flat 2D surface, compared to being inside a 3D gel matrix. In most cases, the 3D environment mimics the in vivo situation much better. Gradients for chemotaxis can easily be built up in water-based gels (like Collagen I gels), because the gel structure does not hinder diffusion. To give a clear idea about the difference between 2D and 3D


17

Chemotaxis experiments this Figure.2.5.shows the microscopic image taken for HT-1080 cancer cells on a 2D & 3D surfaces which shows the big difference in the obtained microscopic images, ad how the 3D environment can reveal more information to the cells in culture. This shows that the results obtained by 2D culturing can’t be trusted due to the lack in more details in the information which can be obtained by studying them.

Figure 2.5: An Image of HT-1080 cancer cell taken over a) 2D surface b) 3D surface TM c

Ibidi

2.1.2

Mechanical shear stress

One Important parameter which can have a great impact on the in-vivo adherent cultured cells is the mechanical shear stress exerted upon them from the walls or the channel. While inside the human body that mechanical shear stress can be exerted by the biofluidic systems such as blood or lymphatic vessels and nephrons.This mechanical stimulus has a great impact on the physiological behavior and adhesion properties of cells.In some cases it can induce undesired movement of the cell which can get false results related to chemotactic experiments. It was noted that: • a small channel generates high shear stress values • a large channel generates low shear stress values. There are different types of shear stress that can be investigated inside the cultural channel,depending on each one of these types, there must be a desired fluidic pump system which can be used to mimic these forces or even neglect them depending on the type of the conducted experiment.


18

• Shear Stress. It is the mechanical force induced by the friction of liquid against the distal cell membrane. Cells are able to countervail deformations caused by shear stress by rearranging their cytoskeleton. Other shear stress dependent effects include changes in metabolism, gene expression, and differentiation. Physiological shear stress values vary from 0.5 to 120 dyne/cm2 depending on the vessel type (e.g., artery or vein) and the size of the organism (e.g., mouse, rat, or human). • Unidirectional Laminar Flow. It is encountered in most small healthy biological vessels, such as small arteries and veins. Certain cells, such as endothelial cells and kidney epithelial cells, are in-vivo constantly exposed to flow. Experimentally, this is achieved by perfusing medium through low walled channels, and by keeping the flow constant over time for both direction and velocity. As seen from Figure.2.6.homogeneous laminar shear stress covers the whole channel area of the slide, except for a thin band on either side of the channel walls and near the reservoirs. The width of these bands is proportional to the channel’s height. While Pulsatile Laminar Flow is encountered in large arterial vessels due to the fluctuations caused by the heartbeat. Experimentally, this type of flow can be mimicked by employing unidirectional flow with a periodically changing flow rate while keeping the flow direction constant.

TM c Figure 2.6: Unidirectional Laminar Flow Ibidi

• Turbulent and the Oscillating Flow. Near surfaces is characterized by changes in flow rate and direction. Direction and velocity change over time, thus the flow profile is not constant. In vivo, turbulences are rare and can only be found during pathophysiological processes. In many cases, turbulences are falsely considered to appear around plaque that has developed at the largest branching of the Arteria Carotis. While Oscillating Flow as seen in Figure.2.7.is accepted as a means of simulating turbulences when using flow chambers. Although the flow is laminar, there is no main direction due to the fact that the direction of the flow is changed


19

at regular intervals (often this interval is every 0.5 s). Other than during valve switching, the flow rate is kept constant.

TM c Figure 2.7: Turbulent and the Oscillating Flow Ibidi

• Non-Uniform Laminar Shear Stress. This occurs in vivo at branching sites and at other regions of disturbed blood flow in vessels. Experimentally, nonuniform laminar shear stress can be achieved by spatially varying flow rates. It can be used for investigating cells at different shear stresses using a single sample. In more complex experiments, it can be used for studying cells and cellular signaling when the cells are exposed to areas of strongly varying shear stress. The µ-Slide yshaped(product of IbidiTM ) was designed to help conduct studies of non-uniform shear stress. In the branched region of the slide, the prevalent shear stress is approximately half of that found in the single straight channel regions.

After noticing the impact of the mechanical shear stress upon the cell inside the human body,so its so important to try to mimic these forces upon the cells so as it can be fully studied in-vivo. There are a lot of different commercial fluidic pump systems which try to mimic these forces upon the cells cultured in-vivo to have results which really can be trusted. While also in same time its necessary to avoid the shear forces upon the cells which is exerted by the PDMS walls during the cultural cell flow, that is because the forces exerted from PDMS walls can affect the cell is behavior not as same as the one in the human body veins.That can be achieved by making a barrier between the culture dish and the channel walls itself. That solution can be discussed fully below.


2.2

20

Hydrogel/poly-dimethylsiloxane hybrid bioreactor facilitating 3D cell culturing by Bart Schurink & Regina Luttge

2.2.1

A brief about the presented work

The authors present a hydrogel/poly-dimethylsiloxane (PDMS) hybrid bioreactor. The bioreactor enables a low shear stress 3D culture by integrating a hydrogel as a barrier into a PDMS casing. The use of PDMS allows the reversible adhesion of the device to a commercially available microelectrode array. A two-step molding process facilitates this relatively simple, cost effective, and leakage-free add-on microculture system. Agarose (2%) is used as hydrogel barrier material and mass transport is evaluated by fluorescein isothiocyanate-albumin fluorescence under static conditions which yields a diffusion coefficient of average value of 2.2 × 107 cm2 s− 1 across the barrier. Previously, PDMS has been used for the fabrication of microfluidic chips that can be sealed to a glass surface by spontaneous and reversible adhesion,which provides an appropriate technology platform for cell culturing. However, many of the microfluidic cells-on-a-chip designs exert high shear forces onto the cells and their matrices through the flow of culture medium. Therefore, these systems are limited in their utility for prolonged 3D culturing[15] An integrated barrier, separating the culture chamber from the culture medium channel, can avoid this direct shear stress onto the culture. Such a design sustains an appropriate mass transport and prevents the cells from migrating out of the culture into the channel. The bioreactor consists of a round culture chamber which is surrounded by a ring shaped channel separated by a hydrogel barrier. This design maximizes the surface coverage between the barrier and the culture to ensure a uniform inflow into the cell culture. By means of a two-step molding process using a polished mold surface with interchangeable parts that ensures a leak-tight connection between the barrier, the PDMS casing and the MEA as well as a central alignment of the barrier, the flow channel and the culture within the PDMS casting. Recently, microfluidic systems with an integrated hydrogel as a barrier for cell culture on a chip have been described. In these systems, cells are cultured inside of a straight channel layout (e.g., to create a stimuli gradient) in which a hydrogel barrier ensures a controlled inflow of nutrients and oxygen [16] [17].


2.2.2

21

Methods and materials

Due to the fabrication by soft-lithography, that utilizes a photoresist mold of SU-8, the channels are restricted to a uniform height. This molding principle is limited in its maximal height to realize a vertically orientated 3D culture atop of the prearranged electrodes of a commercially available MEA. Although very thick and advanced multilayer SU-8 photoresist processing is described in literature [18], it is cumbersome to realize a mold structure by SU-8 with aligned parts at different heights. To cope with this limitation of an SU-8 mold, it is introduced a fabrication process for a microfluidic bioreactor that complies to the dimensions of the MEA and provides a simple 3D addon microculture system. The fabrication of this microfluidic bioreactor is based on soft-lithography from a precision engineered metal mold.

2.2.2.1

Two-step molding process

A two-step molding process is performed using an assembled mold (Figure.2.8). Figure.2.8 a. shows all the interchangeable parts of the mold. The mold is based on a surface polished brass plate of 40mm (part A) and fabricated by computer numerical control milling. The parts D and F are also made of brass, whereas the other interchangeable parts B, C, E, and G are made of stainless steel. The inner and outer diameters of parts E and G are 3 and 5 mm, and their heights are 4 and 3mm, respectively. Parts B and C are each 10mm long and their diameters are 1 and 1.5 mm, respectively. The outer diameter of the brass part D is 3mm to conform with the inner diameter of part E. The diameter of the two holes in both parts D and F (diameter 2.5 mm) corresponds with the diameter of the two stainless steel parts B. Additionally, a borosilicate glass tube (part H) is put on top of the mold base plate (part A) to form the circumference of the PDMS casing. The tube has an inner diameter of 15 mm. For the molding of the PDMS casing, Sylgard 184 elastomer and curing agent (10:1 wt.%) are mixed and poured onto the mold configuration as shown in Figure.2.8b. The PDMS mixture is degassed and cured at 80◦ C for 30 min in a convection oven. Afterward, the PDMS casing is removed from the mold, cleaned with 70% ethanol (Sigma), and thoroughly rinsed with ultra pure water (Millipore). In the second step of the molding process, the PDMS casing is placed onto the mold configuration shown in ( Figure.2.8c.)to form the hydrogel barrier. Agarose (Sigma) is used for the barrier material. Agarose powder is dissolved in phosphate buffered saline (PBS) (Sigma) at 2% and heated to 80

◦

C. The barrier is formed by inserting the agarose in the PMDS

casing at the inlet resulting from part C. After gelation of the agarose, the PDMS with


22

Figure 2.8: a) (Color online) Overview of the metal mold and its parts. b) For the molding of the PDMS casing, parts B, C, D, and E are assembled into the mold baseplate (part A).The borosilicate glass tube (part H) will be placed on the base plate during molding. c) For the molding of the hydrogel barrier, parts B, G, and F are assembled into the mold base plate (part A). [19]

the integrated hydrogel barrier is removed from the mold. The resulting bioreactor is adhered to a glass or MEA surface Figure.2.9.

2.2.2.2

Transport through the barrier

The bioreactor is fabricated as described above and adhered to a microscope glass slide (76mm × 225mm × 4mm × 1mm). The channel and culture chamber are filled with PBS to avoid that the hydrogel dries out prior to the measurement. Before use, the PBS is removed strictly from the channel. A (1 mg/ml) fluorescein isothiocyanate conjugated bovine serum albumin (FITC-BSA) solution in PBS is injected into the channel. Diffusion of the labeled protein into the agarose barrier is observed by a fluorescence microscope (Leica DM IL LED) over distance at 25◦ C. The fluorescent images for quantifying the diffusion through the barrier were analyzed using IMAGEJ software for intensity profiling at three different locations in the bioreactor. The obtained data were fitted to


23

Figure 2.9: (Color online) Bioreactor can be adhered to the commercially available MEA simply by the adhesive properties of the PDMS casing, resulting in a leak-tight connection between the PDMS casing and the MEA, while the culture chamber is aligned to the electrode array. Colored dye solutions are injected into the bioreactor to indicate the channel and the culture chamber. [19]

a Gaussian function for dimensional diffusion with ORIGIN 7.5 software to simulate the diffusion profiles for further analyzes.

2.2.2.3

Characterization of the barrier

The barrier is integrated to separate the fluidic channel from the cell culture while maintaining mass transport into the culture. To test this hypothesis, agarose hydrogel was used, which is a commercially available sol–gel, relative inexpensive, and biocompatible. Further, it has been shown elsewhere that it prevents adhesion of cells and migration out of the culture at a mass volume ratio of 2%. [20] Within a static culture setup, mass transport through the barrier is observed by diffusion of FITC-BSA Figure.2.10. The simulated diffusion profiles are found to be in agreement with the experimental results (R = 0.97±0.02). The profiles show a decrease in FITC-BSA initial concentration at the channel side and an increase in FITC-BSA concentration at the interface of the culture chamber, as a result of the static setup. This result shows a diffusion coefficient of an average value of 2.2 × 10−7 cm2 s−1 , which is in agreement with values reported elsewhere [21].

2.2.2.4

Cell culture in the bioreactor

The add-on bioreactor was used to culture neuronal cells in Matrigel. To validate the function of the barrier within the bioreactor for 3D cell culturing, the setup was compared


24

Figure 2.10: (Color online) Profiles of the diffusion of FITC-BSA (1 mg/ml, 66 kDa) for the experimental and simulated data. The data are collected from the microscope intensity images according to the pathway following the channel toward the culture chamber, thereby crossing the agarose barrier (this pathway is indicated by the arrow in the schematic insert). The finite FITC-BSA concentration in the channel results in a decrease of initial concentration as the fluorescent protein diffuses from the channel toward the barrier and culture chamber. [19]

with a specifically designed control experiment. This experiment allowed us to evaluate cell viability dependent on the degree of supply of culture medium. The diffusion of culture media in the bioreactor is limited by the barrier. However, in the control culture medium is added directly on top of the culture. Cells in both the bioreactor and the control were distributed in a 3D fashion within the Matrigel. For the bioreactor, an increased cell density is observed in the peripheral region compared to the center of the culture (Figure.2.11). Then after counting the viable and dead cells for the total population, peripheral and center regions. Even with the integrated barrier in the bioreactor, an enhanced cell viability is observed compared to the control. For the bioreactor, the fraction of viable cells in the peripheral region is higher compared to the center. For the control culture, these fractions are of the same order of magnitude in both regions. Thanks to the integrated hydrogel barrier in the bioreactor, mass transport into a 3D culture chamber can be realized without applying shear forces to the cells. The hybrid bioreactor design has the potential for batch processing by means of the two-step molding process, which could be used to enable high throughput screening. To accommodate specific biological experiments, the barrier material in the bioreactor can easily be varied. For example, different types or concentrations of hydrogel can be used to alter mass transport or to add functional properties. The observation of diffusion of FITC-BSA in the agarose hydrogel provided us with an initial indication of culture medium inflow as shown before in Figure.2.10.


25

Figure 2.11: (Color online) Cells cultured in Matrigel in the bioreactor [(a) and (b), respectively, 10 × and 20× magnification], a higher amount of cells is seen in the peripheral region compared to the center. The control [(c) and (d), respectively, 10× and 20× magnification] shows a comparable amount of cells in both regions. Cells in the bioreactor and the control were cultured for 6 days and afterward stained by Calcein AM for viable cells (green) and ethidium homodimer for dead cells (red). [19]

The size of FITC-BSA (66 kDa) is significantly larger than most nutrient molecules in the culture medium. Also, the stokes-radius of the BSA is around 3.5 nm, while the pore size in the 2% agarose is around 80 nm. Therefore, diffusion for these nutrients is expected not to be restricted by the barrier and to have higher diffusion rates than FITCBSA, rendering our design fit for 3D cell culture. Nevertheless, culturing is a process at a much larger time scale than the diffusion process (rather weeks than minutes) with regular intervals for medium refreshment. Therefore, the time required toward a diffusion equilibrium in the bioreactor is ought to be shorter in comparison to a total nutrient depletion in the culture chamber. In other words, starving of cells in the culture chamber is not likely due to the integration of a hydrogel barrier.To obtain a more homogeneous distribution, the barrier thickness can be reduced to decrease the residence time of a nutrient molecule in the barrier. Also, by a continuous and infinite flow of culture medium, the concentration of nutrients in the bioreactor could reach an increased homogeneous distribution of nutrients in all zones, compared to the static setup which was tested and aforementioned.

Chapter 3

Equipment and materials used in fabrication 3.1

Materials

Different materials are used in microfluidic devices. Historically silicon (Si) substrates were used, because of the huge disposal of microelectronic techniques available for the micromachining. Despite this factor, Si presents several problems for those applications: • It is very expensive and its processing are expensive as well • It is not transparent. • It is not biocompatible. For these reasons, it was replaced by glass and polymers. Glass is inert, optically transparent, insulating and it is not crystalline, giving the possibility to have rounded profiles for channels according to the process; however it is difficult to be processed to obtain complex 3D structures. Microfluidic research is nowadays mainly based on polymeric materials, which open new interesting possibilities. Thanks to the huge variety and possibility to be tuned, polymers allows to exploit very different properties. They are typically low cost, chemically inert, insulating and biocompatible. They allows an easier and faster prototyping and their surface properties can be easily modified, but specially they are easily be processed and permit mass replication techniques, as afterwards described. Polymers are macromolecules composed by the repetition of a fundamental units (monomer), which are interconnected to form chains, thanks mainly to the covalent 26

Chapter 3. (Laboratory) experimental setup

27

bonds. In our project we are mainly interested in cross-linked polymers, in which the inter-chain bonds determine the mechanical behavior. In particular PMMA and PDMS were exploited.

• PMMA (PolyMethylMethacrylate) is a transparent thermoplastic synthetic polymer and it is often used as an alternative to glass. It is a strong and lightweight material, with a 92% transparency in the visible range. Moreover it has a good biocompatibility. • PDMS (PolyDiMethylSiloxane) is a silicone elastomer, which looks as a viscous liquid. This organic polymer is optically transparent, inert, biocompatible, hy-

c Figure 3.1: PDMS Structure Wikipedia.org

drophobic and it is particularly known for its flow properties. When the oligomer dimethylsiloxane is mixed with a cross-linking agent (proportion 10:1), such as methyl hydrosiloxane containing a platinum-based catalyst, it is able to create a solid structure thanks to the cross-linking derived to the curing treatment at high temperature.

MicroChem Primer 80/20 is based on a combination of 20% HMDS and 80% PM Acetate. HMDS is most well know chemical pretreatments for increasing photoresist adhesion to Oxides,Nitrides,PolySilicon,Glass,Quartz and other difficult surfaces. PM acetate acts as an active pre-wetting agent. Advantages: • Improved wet etch performance. • Comptatible with most of positive photoresists. • Improved critical dimension control. • Combination pre-wet and adhesion promoter.

Chapter 3. (Laboratory) experimental setup BD Matrigel

28

TM

Basement membranes are thin extracellular matrices underlying cells in vivo. BD MatrigelTM Basement Membrane Matrix is a solubilized basement membrane preparation extracted from the Engelbreth-Holm-Swarm (EHS) mouse sarcoma, a tumor rich in extracellular matrix proteins. Its major component is laminin, followed by collagen IV, heparan sulfate proteoglycans, entactin/nidogen [22],[23]. BD Matrigel Basement Membrane Matrix also contains TGF-beta, epidermal growth factor, insulin-like growth factor, fibroblast growth factor, tissue plasminogen activator[24]and other growth factors which occur naturally in the EHS tumor. BD MatrigelTM Basement Membrane Matrix is effective for the attachment and differentiation of both normal and transformed anchorage dependent epithelioid and other cell types. These include neurons, hepatocytes [25], Sertoli cells [26],[27] , chick lens, and vascular endothelial cells. BD Matrigel Basement Membrane Matrix will influence geneexpression in adult rat hepatocytes,vascular endothelial cells, as well as three dimensional culture in mouse[28],[29] and human mammary epithelial cells. It is the basis for several types of tumor cell invasion assays, will support in-vivo peripheral nerve regeneration,[30] and provides the substrate necessary for the study of angiogenesis both in vitro and in vivo.[30] BD MatrigelTM Basement Membrane Matrix also supports in- vivo propagation of human tumors in immunosupressed mice. BD Matrigel Basement Membrane Matrix can be used for the transplantation of unsorted mammary cells, as well as sorted epithelial subpopulation embedded in BD Matrigel. This matrix has also been used as a cancer stem cell model and shown to enhance tumor growth rates in vivo [31]. The forumulation as stated by the manufacturers: Dulbecco’s Modified Eagle’s Medium with 50 µg/mL gentamycin. BD MatrigelTM Basement Membrane Matrix is compatible with all culture media. This mixture resembles the complex extracellular environment found in many tissues[18]. MatrigelTM is more dense with respect to collagen and at room temperature crystallizes rapidly, forming a porous net exploited as substrate for cell culture. Finally an important material exploited in the device fabrication is SU-8 IBMTM . It is an epoxy-based resin which can be used as a negative photoresist in a normal UV photolithography process. Its best properties are: good sensitivity, high contrast with possibility to obtain good aspect ratio (15:1), good thermal resistance and high transparency in UV range. Being very viscous it allows to obtain also thick films in the order of hundreds of micrometers. Moreover is it quite cheap and in particular it permits to perform replica technique simpler than LIGA. Some drawbacks are the difficulty of the process to be set up and the extreme difficulty to be removed from substrates, in particular after the treatment


3.2

29

Technological processes

A brief overview on the techniques exploited in our technological process will be presented. As mentioned before, the origin of the production techniques is in the semiconductor industry, i.e. those were designed for inorganic materials. Due to the growing exploitation of organic and polymeric materials, those have evolved in order to satisfy the new requirements. In particular polymers are very suitable for replica and printing techniques.

3.2.1

Micro-Milling

This technique is based on a conventional cutter, in which a small tip, that can be around 300µm in diameter minimum, mechanically removes material from the substrate to be modeled by engraving and drilling its surface. It is a very flexible technique since it permits to obtain very fast and simple prototyping, but of course is not suitable (or at least not convenient) for a mass production, due to the slowness of the serial process. Moreover it cannot guarantee a high precision ( ∼ 10µm) and indeed the minimum resolution is limited by the tip dimension: a 300µm tip allow to realize relief parts of minimum 100µm and excavated parts of the order of the diameter. However it is ideal to produce some relatively big parts on PMMA, such as inlet and outlet plugs.As well as the fluidic holder which is used during the testing part of the device.It is noted also that an annealing process (thermal treatment) can limit stress in the critical zones of the modeled substrate.

Figure 3.2:

3.2.2

Benchman VMC 4000 Micro-Milling machine, by Light Machines CorporationTM

Physical Vapor Deposition (PVD)

To deposit a thin film of a certain material onto a substrate this machine can be used for that purpose. The Work principle of this machine is the accumulation (condensation) of a material vapor onto a cooler substrate and it is used mostly for metal layers or alloys. There are two main types of PVD, i.e. Thermal Evaporation and Sputtering. In


30

this project, a technique from the first category is used: the Electron Beam Evaporation (E-beam) (Figure.3.3). A wire is heated up to the electrons emission regime and those are accelerated and deviated by a magnetic field in order to hit the material in a crucible (called liner), proper for each material to avoid contamination, which is cooled continuously.

c Figure 3.3: E-Beam Evaporation Principle intechopen.com

E-Beam PVD allows high deposition rate and usually, through rotating crucibles, permits the deposition of different materials, without opening the chamber hence avoiding oxidation, as it is required in the case of adhesion layers. Typically many substrates can be exposed simultaneously being placed on the same sphere on rotating planetary wafer holder, so that all are exposed almost to the same conditions (in fact, lower distance is compensated by lower angle).

3.2.3

Lithographic Processes

Lithography is big set of techniques able to select areas to work on (i.e. deposit, implant, etch) by transferring the desired pattern from a mask to a proper layer on the substrate underneath. Typically photo-lithography is performed using UV light, produced by a mercury lamp (λ ∼ 400nm), impinging on a polymer layer (photoresist) which changes its properties according to the light intensity and power impinged onto it. The selection of the desired pattern is performed by a mask made of quartz on which the dark layer (obscuring UV light) is composed of chromium. It can be selective or transparent depending on the polarity of the resist:


31

Figure 3.4: Photo-lithography UV lamp and mask aligner

• Positive photoresist: UV photons break polymer chains making them soluble in a proper developer, so that the pattern replicates the mask itself • Negative photoresist: UV photons induce cross-linking of polymer chains, becoming etch resistant, so that the negative of the mask is transferred on the substrate. Other important properties of photoresist are: • Contrast (slope that can be obtained). • Resolution (minimum dimension of features). • Sensitivity (energy required to obtain transformation). • Adhesion (ability to adhere to the substrate, may require adhesion promoters). The exposure tool can be based on three different principles: • Contact mode, in which the mask is kept in contact with the substrate during the exposure, it is fast and cheap but not recommended sometimes due to the contamination it can cause to the mask. • Non-contact mode, using the same tool but with a certain gap. It avoids the contamination of the mask, but increases the shadow effect, i.e. decreases in resolution. • Projection mode, which is the most used for mass production. It allows demagnification and gain in resolution, however it has a very high cost, being very complex (precise optical tools and stepper motors required).

3.2.4

Wet Etching

Etching is the process in which unwanted geometries are removed, whilst the other ones are typically protected by a photoresist or a hard mask; in the case of photoresist


32

removal, the areas are selected in the previous lithography step, by dissolving or crosslinking the polymer chains. Wet etching is based on acid or alkaline bath, i.e. the immersion of the sample in chemical solution able to attach and remove the selected regions. It is quick, cheap and more common with respect to dry etching, but it is also less precise in term of resolution and aspect ratio. In our case (working only on exposed resists). The most important characteristics of any etching process are:

• Selectivity, ability to etch mainly the desired target with respect to other ones. • Etch rate, speed of the etching. (depending on both reaction speed and diffusion). • Surface Roughness, flatness of the surfaces obtained.

3.2.5

Replica Techniques and Soft Lithography

Some techniques which are used to fabricate molds to generate and replicate micropatterns and micro-structures using an elastomer, instead of using photo-lithography technique with a rigid mask on every single device fabrication are much more used in patch process due to their fast processing and fabrication properties which can be promising. The most common process is LIGA standing for lithography, plating and molding. The first step is the classical photo-lithographic exposure of a photoresist, typically a X-Ray or deep-UV resist with high viscosity (corresponding to higher thickness) to obtain the desired structures, after being developed, PMMA is widely used for this purpose. Then electroplating is performed by the immersion inside an electrolytic bath in which the wafer acts as cathode onto which cations are deposited. In this way mold inserts are built. And the result is the so called (master mold) for the replica.

Finally

there is the molding step, that allows to replicate the pattern produced onto the master, among these types of molding that can be used, two possibility are: • Injection Molding technique : the polymer is injected at high pressure into the mold cavity kept under low pressure. • Hot Embossing technique : in which the tool and the thermoplastic material (which can be PMMA for example) is firmly pressed for several seconds, after being heated up to the glass transition temperature(Tg) of the polymer. It is particularly suitable for large scale production. In some cases, i.e. when the dimensions of the device are big enough and the resolution required is not under the micro-metrical scale, an alternative and much cheaper


33

c Figure 3.5: Injection Molding Machine maximintegrated.com

c Figure 3.6: Hot Embossing Technique rsc.org

c Figure 3.7: Soft-lithography process istpace.org

technique can be adopted and it is calledSoft Lithography (or Casting in-situ or MicroMolding) as shown in Figure.3.7. The silicon master is patterned using SU-8 to fabricate the final mold. SU-8 as aforementioned is a negative UV photoresist developed by IBMTM , which can be processed in a standard clean-room environment with typical photo-lithography process, allowing to build thick and high aspect-ratio structures with polymers [32]. After the construction of the mold, low-viscosity PDMS (composed by pre-polymer plus curing agent) can be poured in this mold (casting), in which it follows the SU-8 surface wells and channels by capillarity action and, after being put in an oven to cross-link and solidify, the polymeric replica can be simply peeled-off. Despite the lower resolution and the lower aspect ratio achievable, LIGA technique can be very useful in a microfluidic device process, since the dimensions and the characteristics required are often satisfactory, furthermore it allows to produce samples without the use of very expensive facilities (X-Ray synchrotron), but only using standard and simple machines normally present in every clean-room.

Chapter 4

Newton Device Fabrication Technological Processes As aforementioned before, the aim of this work was to develop a mold from copper that can be used to replicate microfluidic PDMS LOC replica either by Injection molding or hot imposing or even Soft Lithography techniques. So the technological processes can be done as following, first the mold which should be processed by electroplating to form the Cu mold have to be prepared , secondly, performing the replication processes to replicate the microfluidic PDMS devices. Then Finally the testing part,which will be dedicated to either check the fluidic diffusion mechanisms as well as testing at the end the microfluidic Lab-On-Chip if it is valid for the requested field of study of culturing the stem cells or not. The core of the device, i.e. the channels with the chambers and the inlets, is made of R 184 Silicone Elastomer). The process exploited to create PDMS PDMS (SYLGARD

devices is soft lithography. It expects to build a mold, on which the silicone elastomer (mixed with the cross-linking agent) has to be cast; then it has to be heated up to allow the polymerization of the PDMS, operation called curing, to define the geometries. Once obtained this result, the PDMS has to be detached from the mold and to be bonded on a glass, closing the channels and the structures patterned by SU-8 mold, and from which the behavior under study will be observed with microscope. The Project has taken three different trajectories regarding the way of realization and processes as well as the materials used. Firstly as stated before the need for a patch production which can be realized by having a copper mold was the aim, but due to the realization problems originated from the physical properties of both copper and SU8 which was discovered that SU-8 has not a good adhesion over copper as it has for Silicon which forced us to study the adhesion differences between silicon/copper and 34

Chapter 4. Newton Device Fabrication Technological Processes

35

SU-8, which drove the project to the second trajectory of realizing the Newton device . Secondly the project took another way which was to design the Newton device by fabrication of a SU-8 mold onto silicon substrate and obtaining the replica by casting PDMS into the mold and peeling it off after being processes, even if the results of that fabrication and design was satisfactory, some problems aroused during the testing phase -as will be discussed later-leaded to a third way of fabrication process. Thirdly due to the problems related to fluidic filling or handling inside the channels , there were a great need to design a 3D monolithic SU-8 microfluidic device using 3D techniques specially (Lamination, double layer exposure using different UV dosages) and it will be discussed at the end.

4.1

Newton device fabrication limits due to using copper to realize the mold

4.1.1

Adhesion of SU-8 onto silicon and copper

An experiment has been done to mark the differences of adhesion of the SU-8 onto both silicon/copper . A mask was designed to create an inter-mixed shaped design of silicon substrate and copper seed layer (upon the same silicon substrate). Then SU-8 will be spun upon the substrate by the same process which will be mentioned in the next chapter . As shown in Figure.4.1, lithography was performed in such a way in order that each pillar can be realized onto silicon and copper so as the difference of adhesivity can be observed easily and simultenously between the two materials. The device will be inspected by the optical microscope to discover the behvaior of SU-8 in terms of adhesion property , which was observed to be worse on copper than onto silicon as shown in Figure.4.2.

Figure 4.1: a) Newton device mold used to study difference of adhesion of SU-8 resist upon both Si & Cu substrates b) Microscopic image showing the Pillars over both Si & cu.


36

Figure 4.2: Microscopic image clarifying the difference of week adhesivity of SU-8 on Cu on the left , while very good adhesion on Si on the right side.

That results has leaded to a new design which done on the project to allow growth of the SU-8 onto silicon spots when there is a great need to a good adhesion, especially when designing the pillars which must be well attached onto the substrate so as they will not collapse. Not Only the problem related to the adhesion of SU-8 over the copper arose but also the problems related to the electroplating process , which wasn’t successful as will be mentioned below.

4.1.2

Limitations of electroplating process during the fabrication of copper master mold

As the aim of the project was to obtain a master mold made from copper in order to allow patch processing and creating replica by hot imposing replica technique. An electroplating process for copper over the silicon substrate has been done by depositing a copper seed layer onto the silicon substrate , which is used to allow the electric contact for the master substrate to grow specific dimensions during electroplating process . But unfortunately this process didn’t succeed till now and there are new modifications into the design and the processes so as to fix that problem. The problem is that it is always coming up with many wrong final structures or collapsing of pillars as can be seen in Figure.4.6.while also the negative obtained mold didn’t succeed as well due to the dimensions of holes obtained were not as desired , as well as the surface of the structure is not flat, the roughness is worse too. It was advised to continue the fabrication process without using a copper mold neither by doing an electroplating process ( for a seed layer of copper over a silicon substrate)


37

Figure 4.3: Optical microscope measurements for different samples in µm showing the dimensions obtained from electroplated device and how much far they are from the design dimensions, besides the collapse of pillars.

4.2

Fabrication of SU-8 Newton device master mold (Second Technique)

As aforementioned, the copper mold fabrication was not successful enough to carry on doing it. Fabrication of the Newton device over a silicon substrate with an anti-adhesion layer (Teflon-Like) to allow the peeling-off of the PDMS replica after the PDMS casting was done . In the following sections, the process flow will be stated. Starting from the mask production till the final device.

4.2.1

Mask production

R & CleWin R ) , the first step Once the CAD design is done (by using Rhinoceros

is to create a lithographic mask to produce the mold itself. The mask was performed to pattern the SU-8 photoresist on which channels and small features have to be formed, so the mask pattern has to be impressed by photo-lithography which can be achieved by a laser writer system (MICROTECH Laser Writer LW405ATM ), which is able to print patterns on a proper photoresist, which can be consequently transferred on the chromium present on the quartz mask . The result can be seen in Figure.4.4.

4.2.2

Wafer preparation

The substrate used for the master is a 4” single-crystal < 100 > p-type single side polished silicon wafer.

Four chips of dimensions 3.5cm × 3.5cm are obtained by cutting

the obtained wafer then they have to be rinsed with Acetone and 2-PROPANOL, and dried with N2 air. If the samples are not clean enough , so it will be cleaned in a boiling


38

Figure 4.4: Mask for SU-8 lithography (left) and details of the mask geometry for Newton V2 (right) which is desired to obtain holes (not pillars).

Piranha solution (H2 SO4 : H2 O2 = 3 : 1 ) for 5 min. In case of a substrate of seed layer of Copper that can be used for the electroplating process later, so a deposition of copper using an electron beam physical vapor deposition must be done.

4.2.3

SU-8 Lithography & deposition of Anti-adhesion layer

First of all the wafer has to be dehydrated on a hotplate at 110◦ C for 2 min. After that some drops of MCC Primer 80/20( MicroChem Corp., MA, USA) have to be dropped on either the Silicon/Copper substrate and after waiting 10 s it can be spun at speed of (4000 rpm for 30 s with an acceleration of 1000 rpm/s2 ). After that it will be baked on the hotplate at 110◦ C for 2 min. After that a layer of SU-8 2150 (MicroChem Corp., MA, USA) is spun over the substrate in two phases so as to achieve a thickness of 350µm : • First phase for 5 s at low rotation speed (500rpm) with acceleration of 100 rpm/s2 so as to distribute the resist over the whole surface. • The second phase, longer and with higher rpm (30 s at 2000rpm and acceleration of 100 rpm/s2 ) is designed to obtain the desired and uniform thickness, which in our case is 350µm. At this point a soft-bake is necessary to reduce the remaining concentration of solvent in the photoresist which its presence can reduce the adhesion, contributes to mask dark erosion and is responsible of contamination as well as sticking.Its baked on the Hotplate: • First for 8 min at temperature of 65◦ C • Then while the chips are still on the hotplate,ramp the temperature to reach 95◦ C then they are kept for 75 min.


39

• They will be cooled down to room temperature for 30 min at least on the hotplate. Its strictly advised to check the planarity of all the working surfaces and equipment used in the process, since the resist has a relatively good reflow capability also helped by the long bake and relaxation steps required, so as to avoid final non-uniform thicknesses of the mold. As the master mold will be fabricated using SU-8 and hence this mold will be used to create new replica, so the problems due to the adhesion of PDMS and the SU-8 must be minimized to allow an easy peeling off process of the PDMS casted into the mold. This is achieved by plasma polymerization of C4 F8 passivation layer deposited onto the SU-8 layer which is necessary to allow for ease of demolding PDMS [33]. Fluorine containing plasmas are known to decrease surface energy and increase the hydrophobic behavior of surfaces [34]. The process parameters and details can be found in Appendix A.1.

4.2.4

UV Lithography & Developing

UV exposure process is performed on the substrate is put into the mask aligner (NEUTRONIX QUINTEL NXQ-4006). The mask is placed just centered above (since there are not further lithographic steps later) and afterwards they are approached and brought in contact. The UV exposure is performed using an energy dose of 11 mW/cm2 .s for (36.3 s for silicon substrate ) and (72.6 s for copper ). Then its followed by a postexposure bake on the hotplate as follows: • All the chips must be put on the hotplate. • While the chips are still on the hotplate, temperature will be increased to reach 65◦ C with a plateau of 5 min. • Increasing the temperature till it reach 95◦ C, then they are kept there for 20 min • Cooling down to room temperature while the samples are kept on the hotplate and Its switched off. After cooling, the master is ready for the development in the specific SU-8 developer. The chip must be in a vertical position and it is kept inside the developer and taken out every 5 min and checked its status, sometimes it is possible to spray Ethanol on the chip until its discovered that there are no more white spots,which means that the resist is not well developed yet. Shaking continuously the solution in the beaker, to allow the total detachment of the undeveloped resist. Finally it is rinsed with 2-PROPANOL and dried very softly with nitrogen so as not to break the pillars, or it can be left on the


40

Figure 4.5: a) Final Newton device V2 after developing b) after bonding to the Aluminum pad to allow casting of PDMS.

room air to be drought. The result (Figure.4.5) can be checked and analyzed with an optical microscope, to verify the correct geometries and dimensions of the patterns the sample.Figure.1.7b) can show the dimensions obtained which was for Newton(second version). It is almost very close to the mask dimension which was supposed to have pillars with height of 350µm with a diameter of 75µm and the width of 500µm.Other results can show different samples conducted by different experiments and it can show the variance of dispersion of the dimensions from the desired one(Figure.4.6). Subsequently the thickness of the SU-8 structures can be measured with a Profilometer, it should be almost constant over all the wafer surface.

Figure 4.6: Optical microscope measurements in µm


4.2.5

41

PDMS casting

First of all, the SU-8 mold will be attached to an aluminum substrate which can be considered as a holder to allow easy handling as well as casting or peeling off the PDMS into or from the mold easily. That can be done by heating up a piece of wax on a hotplate and then drop some droplets onto the aluminum substrate and then stick the SU-8 mold to the surface.And then its kept to cool down. R 184 SILICONE ELASTo create the PDMS to be casted, first of all SYLGARD

TOMER and curing agent have to be mixed in ratio 10:1 and afterward the mixture has to be degassed in vacuum to eliminate bubbles that process can take from 15∼20 min. Bubbles are one of the most critical issues in this kind of device, in fact bubbles can compromise the correct behavior of microfluidics if they stay trapped in critical places, such as in correspondence of channels or features. The ratio of which its possible to mix both the silicon pre-elastomer and the curing agent to obtain specific elasticity or hardness depending on the mold which will be casted into, can be shown in Table 4.7

Figure 4.7: Table showing Mixing Ratios to obtain PDMS with specific hardness

Special care must be given to that ratio of mixing because it will result to either a very soft PDMS replica device which will be damaged while the demolding process from the mold, or a very hard replica which can also be damaged while demolding . The ratio been used was (10:1.5). Then it is baked inside the convection oven for almost (25∼30 min) at 90◦ C for cross-linking. After that the mold containing the PDMS device are both put for 5 min inside a beaker filled with Ethanol to promote and facilitate peeling off the PDMS replica from the mold. The final device can be show in Figure.4.8. The device was inspected by the optical microscope to verify the dimensions, as see from Figure.4.9.that the length of the pillars was very good compared to the desired dimension.


42

Figure 4.8: Newton PDMS final device a) showing the device with pillars .b) Top view of the final device

Figure 4.9: Microscopic images for PDMS final device

4.2.6

Glass bonding

After demolding the PDMS device, holes will be created to provide space to the tubes which be used to inject the fluids.To perform that, pierces must be made to the inlets and outlets of the channels then device will be bonded to the glass substrate (MENZELTM ). The object obtained contains all the micro-metric features, but at this ¨ GLASER step the channels are still opened. The simplest solution, which is also perfect to allow to observe the internal of channels and specially chambers (this is in fact the objective of the device), is to bond the device with a thin glass substrate. To perform this bonding a little quantity of PDMS is spin over the glass surface to obtain a very thin (10-12 µm) and uniform layer (5 s at 500rpm, plus 60 s at 4000rpm with an acceleration of 1000rpm/s2 ). Subsequently it is put on a hotplate at 75◦ C for 5 min to semi-reticulate. Then the chip is placed (on the opened channel side) over the glass and finally it will be baked inside the convection oven at 90◦ C for 10 min to complete the curing. The bonding obtained is irreversible. The final device after being bonded to the glass can be seen in Figure.4.10 b).


43

Figure 4.10: The PDMS device.a) After being bonded to glass.b) With a Cap Layer

4.2.7

Coring fabrication procedure to realize off-chip fluidic connections

Inserts have been designed to allow a tight connection of the tubes used to inject the liquids or the cells to be cultured inside the channel as well as to generate the fluxes inside the channels. They are produced by PDMS as well, using molds formed from PMMA carved through micro-milling. The mold design is shown in Figure.4.11a.showing the design used to create the holes for tubes entry. The design allow to the tubes connecting the lateral channels to be inserted laterally. In this way the space required is decreased and the live observation at the microscope is possible. After the casting of the PDMS, it is cured at 80∼ 90◦ C for about 30 min. The whole device can be kept in ethanol for about 5 min to allow the easy peeling off of the PDMS chip later. Afterwards the obtained inserts must be bonded on the top of the chips, in correspondence of inlets and outlets. This binding is done again using PDMS, in this case it is used as a glue and then its placed again in the convection oven to reticulate completely. The obtained device is finally complete and it is ready to be tested(Figure 4.10 b).

Figure 4.11: Master PMMA mold used to cast the inserting tubes for the Newton device a) CAD design. b) After milling. c) PDMS casted device


4.3

44

3D Monolithic buried Newton device fabrication by double deposition of SU-8 Resist (Third Technique)

There have been many micromachining approaches to fabricate buried micro channels, such as wafer bonding after bulk micro machining, sacrificial etching by surface micromachining , and through mold plating , etc. Besides, LIGA or micro stereolithography can also realize micro channel structures [35]. However, those processes either are complex, expansive[36], require high temperature/electrical field, or need special metal mask to protect from second exposure for forming channel spaces .As shown in Figure.4.12.the R film to form fabrication processes are either employing laminated SU-8 and Riston channel spaces, or it can be performed by utilizing proton beam to partially expose SU-8 to form buried channels [37] . As its shown the different ways to fabricate a buried layer of SU-8 in Figure.4.12 e,f. The optimal way to obtain a buried microfluidic platform using the double UV exposure for two different layers of SU-8 is to use an anti-reflection film to prevent the bottom exposure of the lower layer(buried layer) because the problem related to SU-8 as its a negative photoresist, is that any unwanted UV exposure will cross-link the SU-8 and so it will not be able to be developed to the desired design.

Figure 4.12: Different techniques used to fabricate a buried SU-8 channel. a) SU-8 plus filling material to fill the spaces. b) Metal layer used to protect from exposure for R forming channel space. c) Laminated SU-8 and Riston film to form channel spaces. d) Proton beam is used to partially expose SU-8 to form the buried channel. e) Normal UV-lithographic exposure method. f) UV exposure with an anti-reflection coating.[38]


45

In this project,the aforementioned techniques were not used for forming the buried channels. The proton beam technique is quite complex one, while the normal UVLithographic process without anti-reflection coatings is quite challenging as well. As can be seen from (Figure.4.13) that the exposure of the second layer must carried with a higher wavelength but it must be still be within the limits of SU-8 cross linking exposure wavelength range. The technique that was adopted was to fabricate Newton device by using the default process , but in this time,before developing the SU-8 2150 (buried layer), It will be coated with another layer of SU-8 3005 (MicroChem Corp., MA, USA), which the relationship between thickness versus spinning speed was calibrated , and it was found to be almost 4µm for 4000 rpm spin speed. The SU-8 3005 layer will be exposed by a lower energy dosage than the default obtained value from calibration test. It is noted that a calibration test had to be done in order to just match the data sheet values with our equipments’ performance and parameters so as to prevent the exposure of the buried layer as well. In that way, the advantage due to the differences between the energy dosage requirement will be taken for the buried layer SU-8 2150 of a thickness of almost 350µm which will be exposed for 136 s with an energy dosage of 2.94 mW/cm2 .s (Manual mask aligner), and the upper layer of SU-8 3005 which needs only to be exposed for 22 s to achieve a cover layer of 4µm . Many experiments have been conducted to achieve a buried microfluidic layer with the desired dimensions. The results will be discussed in the following sections.

Figure 4.13: Transmission Versus the thickness of SU-8 resist for Various UV wavelengths [38]


4.3.1

46

Technological Processes flow to Fabricate a monolithic buried microfluidic platform over a silicon substrate

First experiment was conducted as a start up for the project to calibrate the process parameters. It was adopted the following technological process flow to obtain the buried microfluidic platform : Starting from the same process flow for SU-8 2150 in (section 4.2.3), but developing process will not be performed for the SU-8 2150 because the fabrication of the second layer will be directly carried, as shown in Table.4.1. The masks used for the exposure of the first and second layer can be seen in Figure.4.14. Process sequence SU-8 photoresist Type Spinning speed step N◦ 1 Spinning speed step N◦ 2 Soft Bake Cooling Exposure time (@ 2.94 mW/cm2 .s UV Energy dosage) Post exposure Bake (PEB) Cooling

Developing (Only After 2nd process)

First process SU-8 2150 (∼ 350µm) 5” @ 500 rpm , acceleration 100 rpm/s2 30” @ 2000 rpm , acceleration 100 rpm/s2 8’ @ 65◦ C then for 1h.15’ @ 95◦ C 30 min (minimum) 136 s

Second process SU-8 3005 (∼ 4µm) 5” @ 500 rpm , acceleration 100 rpm/s2 30” @ 4000 rpm , acceleration 300 rpm/s2 3’ @ 95◦ C

5’ @ 65◦ C then for 20’ @ 95◦ C Must be kept on the hotplate to cool down to Tamb no developing

1’.30” @ 65◦ C then for 2’ @ 95◦ C 1 min at least

1 min at least 22 s

30∼180 min

Table 4.1: Process flow for the Fabrication of Newton device using double SU-8 layers. Note : (’) stands for minutes , (”) for seconds

Many experiments were conducted to study the device performance and design, the only parameter which was modified in each experiment is the second exposure time which had a range from 52 s (high energy dosage) to 22 s (low energy dosage), as well as allowing the pulsatile exposing, allowing exposure for 25 s with an On/Off cycle of 5 s interval. As shown in Figure.4.14 b.that the second mask is just to mask the inlet and outlet holes to develop them later and allow the SU-8 developer to reach beneath the SU-8 3005(top layer) to develop the buried channels. The developing process may last for ∼ 2 h sometimes, and it is advised to agitate strongly to allow the SU-8 developer to enter inside the inlet/outlet holes and reach the channels to develop them.


47

Figure 4.14: a) Mask used to expose SU-8 2150 layer is the negative of that mask. b) Mask used to expose the SU-8 3005 layer to create holes to allow developing the buried layer

4.3.2

Fabrication Results & problems

The best devices obtained were those exposed only for 22 s , 25 s and 25 s (pulsatile) respectively . The results can be seen in Figure.4.15.which shows a good results in terms of pillar’s dimensions and stabilizing as well as they showed a great performance during the fluidic test as will be shown in test section. The device dimensions were very promising, also as it can be seen that the view of the images was so clear even if the buried channel was covered by a layer of SU-8 3005. Problems occurred during the fabrication were firstly due to the clogging of the etched SU-8 material inside the channel (Figure.4.16) which partially blocked the channels and didn’t allow any passage of liquids during the fluidic test. To overcome such type of problems, It is strongly recommended to perform a good agitation during the developing process as well as rinsing the device inside 2-Propanol for about 10 min while agitating as well,so as to clean the developed SU-8 material. But care must be taken during both the developing process and rinsing inside the 2-Propanol because collapsing of the top layer can occur. Regarding the developing process, devices should be checked from time to time because an over developing can cause etching of the cross-linked top layer which will collapse later and break the pillars as can be seen from Figure.4.16 which in this case, the whole sample is ruined. The same problem can happen if an over rinsing inside 2-Propanol lasts for a long time hence, the devices must be checked every 2 min while being agitated inside the 2-propanol.


48

Figure 4.15: Microscopic image of Monolithic Newton device (buried layer). a) The device after developing, as can be seen white and black spots shows the difference in thickness and development which shows dark sports because the device is buried under a cover layer. b) Device channels,inlets and pillars. c) Clear view of the buried pillars in good condition. d) Pillars dimensions showing a good match between the expected value (75µm).

Figure 4.16: Optical microscopic images taken to show the collapsing of the top layer, as well as clogging of the developed SU-8 buried layer inside the channels

Second problem arose was the uncontrolled range of thickness obtained after the developing process as can be seen in Figure.4.17b.Instead of having a top layer thickness of ∼ 4µm we obtained ∼ 17µm. That can be explained as the following:


49

• During the exposure of the top layer, not only the ∼ 4µm layer was exposed but also the layer beneath(SU-8 2150) will be partially exposed as well, and that will cause an integration between the two layers, which made that increase of the thickness. • Second reason could be due to the SU-8 2150 calibrated values in the data sheet stated that for 4000 rpm spinning speed to obtain ∼ 4µm, was supposed to be only over the silicon substrate, but in our case, it was spun over another SU-8 layer, and so the behavior can’t be expected.

Figure 4.17: Monolithic Newton device after developing, a) The device after developing, noting the white spots. b) Microscopic image showing the thickness of the cross sectioned device’s layers.

The Testing phase will be carried out and will be discussed in Section 5.2.3.

4.3.3

Technological Processes flow to fabricate a monolithic buried microfluidic platform over a glass substrate

After verifying the ability to fabricate a 3D 2-layers SU-8 monolithic device onto a TM ¨ silicon substrate, is to fabricate the device upon a glass substrate(MENZEL-GLASER Microscope Cover Slips). The Reason to obtain a device upon glass substrate is to make the channel and whole structure applicable to cell culturing studies by obtaining a transparent substrate for an inverted microscope. That is because unlike the compound microscope, the inverted one needs the object under investigation to be transparent to obtain magnified image.


50

The fabrication process flow as can be seen in Appendix A.2, is similar to that on silicon substrate but the problem is that the SU-8 has a poor adhesion onto glass even with the MCC primer 80/20 adhesion prompter layer, The other problem is the shear stress during fabrication process can cause collapsing of the pillars, but fortunately due to the other SU-8 top cover layer which works also as a support layer as well, provide enough strength to stabilize the high aspect ratio structures. Due to the different thermal expansion coefficient differences between glass and SU-8, sometimes after both soft bake and post exposure one, there were some cracks on the substrate which can lead at the end to a detachment of the SU-8 layer from substrate, and this will ruin the whole device. And so its very recommended to leave the device always on the hotplate after both soft bake and PEB to cool down to room temperature. Its strictly recommended during the lithographic processes to attach the glass substrate onto a piece of silicon wafer by using de-Ionized water, so as to obtain a similar exposure process environment, to obtain better device performance similar to the one obtained onto silicon substrate. That is because the UV light will pass through the glass substrate and scattered back by the chalk, which will cause undesired exposure of different SU-8 spots, which will cause undesired cross-linking of these spots and so it will not be possible to develop them. While if the lithographic process is carried onto piece of silicon wafer that will reflect back the UV light through the same areas which were not masked. Hence better device can be obtained with good dimension resolutions.

Figure 4.18: Microscopic image showing the Monolithic device fabricated on glass substrate, a) The device after developing,NOTE: white spots showing good development. b) The dimensions of the pillars which was almost close to the desired one.

As shown in figure 4.18 the device has succeeded after many experiments which have been conducted. The white spots which shows a good development of the channels and that was achieved after almost 1 hour of developing process with strong agitation. The device inside the beaker containing the developer can be put into ultrasonic cleaner to allow the withdraw of the residuals of uncross-linked SU-8 from the developed buried channels. Testing will be discussed in Section 5.2.4.


51

Standard process tips and methods to obtain well performed monolithic device onto glass substrate

1. The data sheet suggests that it is advised to reduce stress by using a two step bake, 65◦ C then 95◦ C. Although this may help a little bit, it seems bit wrong according to the results obtained because, the glass transition of uncrosslinked SU-8 is 60◦ C so there certainly is no reason to do this on the soft bake. Secondly, it is strongly advised to reduce the cooling speed of the material on both the soft bake and the post exposure bake. At the beginning of both of these bakes the entire volume of SU-8 is essentially uncrosslinked and will go well above the Tg before any significant crosslinking occurs. 2. The data sheet shows an enormous range of exposure which attribute to the very short softbake that they suggest. Unlike most other resists, this bake is only removing solvent from the SU-8, there is no crosslinking happening at this point. It is impossible to over bake the SU-8 by making these bakes longer. On the other hand, if the resist is under baked, there will some of the solvent leaved in the SU-8. This will result in a lower Tg and the solvent acts as a photo-inhibitor which will cause the time of exposure to vary in an unpredictable manner. To eliminate the whole problem, it is advised to bake the SU-8 device two or three times longer than the datasheet says to, then let it cool to 40◦ C or lower on the hotplate. 3. It’s a good idea to cover the SU-8 device while it bakes. On the softbake, using a glass cover like a crystallization dish will help prevent the solvent from baking out too fast and causing bubbles. This is especially true for thicker layers > 50µm. 4. It is not advised to expose SU-8 with a fluence of more than 12mJ/sq.cm/s. Higher fluence may cause the SU-8 temperature to rise above the Tg during exposure resulting in the mask sticking. This can result in poor mask performance later because the of SU-8 residue and since it’s exposed SU-8, it will be very hard to clean off if it will not be done right away. 5. After developing some small stress cracks can be seen in the SU-8. The material shrinks as it crosslinks and a small amount of cracking, at high stress points like sharp corners, is a good indicator that the expose is in the right range. If there is excessive cracking throughout the structure, then it is over exposed. Both over exposure and under exposure will result in adhesion failure. As well as the ramp of the temperature must be slow (3◦ C /min). 6. Hard bake can be done at 100◦ C as the optical properties of SU-8 are better if it does not exceed 150◦ C.

Chapter 5

Newton Device Testing Standardized method to obtain a good device onto a glass substrate The testing part will be split into two different trajectories, First test is supposed to verify that the correct operating of the device specifically related to the diffusion mechanism inside the central channel as well as the lateral ones. Besides the pillars ability to withstand the flow of the fluids from the lateral channels into the central one allowing a homogeneous flow over the whole area. Second testing part is related to the insertion of the cells to be cultured inside the device, and studying their chemotactic process as well as their viability inside the device for specific amount of days or as desired according to the scope of the studies.

5.1

Experimental setup for diffusion test

First of all, the experimental setup used to conduct the testing part will be explained. It is required to use an automatic pumping system to guarantee a very precise, controllable and repeatable flow, those pumps are connected to a twin syringe pump (Pump 33 by Harvard ApparatusTM ) which are able to pump the syringes independently, both infusing or refilling, with a large flow rate range and an accuracy of 0.35% and it is accompanied by a PC lab view by Texas InstrumentsTM , thanks to which is possible to interface PC and laboratory machines to be controlled as shown in Figure.5.1. As well as the automatic pumps, also syringes used for manual liquid pumping (Hamilton) with a volume equal to 250µl have been used (internal diameter 2.3mm). To visualize the fluid flow inside the device, colored fluids are exploited to highlight the visual perception and, in the case of presence of MatrigelTM in the central channel, to see the diffusion within. 52


53

Figure 5.1: a) Twin syringe pump by Harvard ApparatusTM . b) Screen shot of Lab-View software and the parameters which can be controlled .

After preparing the setup devices, the samples are put on the Peltier cooler device stage which is fixed onto a support designed using (Objet30TM , a 3D rapid prototype). The device setup can be explained as following, first step can be shown on Figure.5.2. The Peltier device with two fans, was used to cool down the sample and keep the temperature at 8◦ C and also control the elevation of the temperature according to the given voltage and current parameters which can be controlled by changing the I/V parameters. The experimental setup for testing purpose to keep the temperature at 8◦ C was performed by fixing the current at 1.1 A and the voltage to 10 V , and to increase the temperature , wires polarity are reversed and the current value is readjusted to 0.61 A then the temperature will be increasing till 37◦ C which is the human body temperature which must be fixed to mimic the human body temperature as well as to allow the polymerization of the BD MatrigelTM which will be discussed below

5.1.1

BD MatrigelTM Handling and injection

Recalling back what was mentioned before regarding MatrigelTM and it is properities which will lead to mention the proper way of handling and injection of the MatrigelTM inside the main central channel. First of all the storage of the MatrigelTM must be stable when stored at -20◦ C. Freeze thaws should be minimized by aliquotting into one time use aliquots. Store aliquots in the -20◦ C freezer until they are ready for use and also it must never be stored in frost-free freezer, as it must be kept forzen. some Precautions must be taken when dealing with BD MatrigelTM Membrane Matrix [39], so as to obtain good results and good polymerization of the MatrigelTM inside the central channel:


54

Figure 5.2: Experiment setup showing the Peltier cooling device,which is fixed over a stage made by 3D rapid prototyper, and they are all fixed under the optical microscope to be inspected as well as showing the manual syringe used in fluid injection.

• It is extremely important that BD MatrigelTM Basement Membrane Matrix and all culture-ware or media coming in contact with BD MatrigelTM Basement Membrane Matrix should be prechilled/ ice-cold since BD MatrigelTM Basement Membrane Matrix will start to gel above 10◦ C. Keep BD MatrigelTM Matrix on ice at all times even while operating it. • Reconstitution and use: Color variations may occur in frozen or thawed vials of BD MatrigelTM Basement Membrane Matrix, ranging from straw yellow to dark red due to the interaction of carbon dioxide with the bicarbonate buffer and phenol red. Variation in color is normal, does not affect the MatrigelTM efficacy, that will disappear upon equilibration with 5% CO2 . • Thaw BD MatrigelTM Basement Membrane Matrix by submerging the vial in ice in a 4◦ C refrigerator, in the back, overnight. Once BD MatrigelTM Basement Membrane Matrix is thawed, swirl vial to ensure that material is evenly dispersed. Keep BD MatrigelTM Matrix on ice at all times. Handle with sterile technique. Place thawed vial of BD MatrigelTM Basement Membrane Matrix in sterile area, spray top of vial with 70% ethanol and air dry. • BD MatrigelTM Basement Membrane Matrix may be gently pipetted using a precooled pipet to ensure homogeneity. Aliquot BD MatrigelTM Basement Membrane Matrix to tubes, switching tips whenever BD MatrigelTM Basement Membrane Matrix is clogging the tip and/or causing the pipet to measure inaccurately. Gelled


55

BD MatrigelTM Basement Membrane Matrix may be reliquified if placed at 4◦ C in ice for 24-48 hours. • BD MatrigelTM Basement Membrane Matrix may be used as a thin gel layer (0.5 mm), with cells plated on top. Cells may also be cultured inside the BD MatrigelTM Basement Membrane Matrix, using a 1 mm layer. Extensive dilution will result in a thin, non-gelled protein layer. This may be useful for cell attachment, but may not be as effective in differentiation studies

BD MatrigelTM Basement Membrane Matrix may be used in several ways, in our case, we need a Thick Gel Method which allows the growth cells within a three dimensional matrix, the steps used to use the MatrigelTM for Thick Gel method: • Thaw BD MatrigelTM Basement Membrane Matrix as recommended. Using cooled pipets, mix the BD MatrigelTM Basement Membrane Matrix to homogeneity. • Keep culture plates on ice. Add cells to BD MatrigelTM Basement Membrane Matrix and suspend using cooled pipets. Add 150-200 µl/cm2 of growth surface. • Place plates at 37◦ C for 30 min. Culture medium may now be added. Cells may also be cultured on top of this gel.

5.2

Filling the Central/Lateral Channels with Fluids

Some equipments and materials should be used to fill the device’s channel to obtain the proper behavior for the device as well and the cultured cells.

Apart from the materials

which were aforementioned before, others equipment will be needed such as :

• Time lapse video equipment: CCD camera, video camera, and acquisition software. • Correct 10 - 200 µl pipet tips must be used (others do not work ( Figure.5.3 a,b). • Slant cosmetics tweezers, for convenient plug handling ( Figure.5.3 C). • Optional: Motorized stage and auto-focus (x,y,z), to observe all 3 chambers, in parallel.


56

Figure 5.3: Experiment Equipment a) Proper way of filling using the pipet Tips. b) TM c Beveled Pipet Tip. c) Slant tweezer for convenient plug handling. Ibidi

5.2.1

Filling the BD MatrigelTM into the central channel

These following steps must be followed to obtain good filling of the MatrigelTM inside the main chamber. First and as aforementioned above, After handling the MatrigelTM in the proper way, the Newton device bonded upon the glass sample will be brought and put onto the peltier cooler stage under the optical microscope. Cooling the sample down to (7∼8◦ C) . It is advised that, The day before seeding the cells and conducting the experiment, it is necessary to place all cell media, the µ-Slide, and plugs into the incubator for gas equilibration. The medium should be put into a slightly opened vial. This will prevent the medium inside the slide, and the slide itself, from allowing air bubbles to form during the incubation time. As well as The Beveled pipet tips must be kept on ice for a while because it must be cold when it come in contact with the MatrigelTM . The experiment can be conducted as following : 1. As shown in Figure.5.4. Close Filling Ports C, D, E, and F with the plugs. Handle the plugs with the appropriate blunt tweezers. 2. Use a 20 µl pipet and apply 10 µl of BD MatrigelTM mixture to the top of Filling Port A, leaving space between the tip and the port. It is not recommend injecting the gel directly. (Figure.5.4 a). 3. Immediately afterwards, use the same pipet settings (10 µl) and aspirate air from the opposite Filling Port B. Press the pipet tip directly into the port.

The


57

MatrigelTM from Filling Port A will be flushed inside, filling the entire channel homogeneously. Aspirate until the MatrigelTM reaches the pipet tip as shown in Figure.5.4 b. 4. Leave both Filling Ports A and B filled with gel Level out the liquid heights in both filling ports, as shown in Figure.5.4 c. 5. Gently remove all plugs from Filling Ports C, D, E, and F. Close Filling Ports A and B with plugs. Or they can be leaved with the liquid heights in both ports exposed to air without being covered. 6. Incubate the slide inside a sterile and humid atmosphere to minimize evaporation until the gel is formed. To make sure evaporation is low, use a sterile 10 cm Petri dish with an extra wet tissue around the slide. Or it can be left upon the Peltier Stage and increase gradually the temperature from 8◦ C till it reaches 37◦ C, for 15∼30 min to allow the gelation of the MatrigelTM . 7. Control the surface morphology with a microscope during and after gelation. Checking the amount of MatrigelTM in the channel and Filling Ports A and B to control the evaporation.( Sometimes it advised to close the Port A and B with droplets of water which will be needed later to allow the gelation of the MatrigelTM , that’s because while the gelation process, the MatrigelTM will be dehydrated hence it will aspirate the water from the inlet’s and outlet’s water drops so as it can obtain the proper amount of water again.

Figure 5.4: Filling procedure a) Filling MatrigelTM into the central Channel b) Aspiration of the opposite filling port to allow the homogeneously flow of MatrigelTM . c) TM c Leveling out the liquid heights in both filling ports. Ibidi

Chapter 5. Newton Device Fabrication Technological Processes 5.2.1.1

58

Results obtained after filling the central channel of the PDMS device obtained from Newton SU-8 Master Mold

After doing aforementioned steps , good filling of the central channel was achieved as can be seen from Figure.5.5. Graphite powder were added into the central channel while filling it with the MatrigelTM to visualize the diffusion mechanism and the flow of the MatrigelTM during the incubation time and decide if it is fully gelatinized or not. As can be seen from Figure.5.6. the incubated MatrigelTM was successfully obtained as well as the pillars are stable and were not damaged at all, neither during the filling process nor during the incubation time.

Figure 5.5: An optical microscopic image showing the filling of the red colored MatrigelTM inside the central channel and it looks very good besides there was no collapse of pillars.

Figure 5.6: An optical microscopic image showing the MatrigelTM after incubation a) Graphite powder was observed static inside the MatrigelTM which proves that it is already gelatinized. b) Detailed focused image showing the meniscus shape of the MatrigelTM near the Pillars that shows the gelatinized Gel indicating no pressure differences between the central channel and lateral one

One of the common problems which is always encountered during the filling phase was the bubble formation in different zones and in different samples as can be shown in Figure.5.7. Its clearly visible a bubble in the MatrigelTM .


Figure 5.7:

59

An optical microscopic image showing bubbles formed inside the MatrigelTM inside the Central channel

The solution to this problem can be either to follow the correct steps that were mentioned before or to use the Newton device which a cap layer that was bonded upon it to allow the perfect airtight structure where there is no bubbles could be formed, as can be shown in Figure.5.8. In fact when MatrigelTM reticulates it becomes a crystalline gel, but it is easily shifted or removed if pressure strongly increases or if it is crossed by a high liquid flow. The presence of these caps over the inlets of the central channel will contribute to keep in position the MatrigelTM , as well as preventing the bubbles formation thanks to the airtight structure.

Figure 5.8: Newton device with a cap PDMS top layer to allow airtight structure to prevent air bubble formation.

Moreover another problem related to bubbles regarding those present in the interconnections. Its occurrence is highly enhanced by the temperature ramp from 20◦ C of the syringes environment to 37◦ C inside the confocal microscope box. Solution to solve this


60

problem can be the exploitation of appropriate anti-bubbles filters, for example MillexR PVDF membrane having an average mesh of HV sterile filter units, with Durapore

0.45µm, were available in IRCC as well as OmnifitTM Bubble Trap can be employed. (Figure.5.9).

Figure 5.9:

5.2.2

Air bubble traping equipments a,b) Millex-HV sterile filter unit, c

millipore.com. c) OmnifitTM Bubble Trap.

Filling the Lateral channels

Both large reservoirs are firstly filled with neutral solution, then the chemoattractant is applied without directly reaching the observation area. As a result, there is a short delay in gradient formation and sensitive cells cannot be initially saturated. The lateral channels will be filled with colorful liquids to allow the study and visualization of the diffusion mechanism through the central channel. As performed before while filling the central channel, the same mechanism can be adopted in the filling of both the lateral channels. It can be also explained step-by-step as following: 1. Gently close Filling Ports C and D with plugs (chemoattractant side) as can be shown in Figure.5.10. 2. Fill the first lateral channel by injecting chemoattractant-free medium through Filling Port E. Use 10 µl and the recommended pipet tips explained before. Keep in mind that Filling Ports E and F must be completely filled, but not overfilled. 3. Transfer the two plugs from the Filling Ports C and D to the Filling Ports E and F. This will close the chemoattractant-free side (Figure.5.10 a). 4. Fill the empty lateral channel by injecting chemoattractant-free medium through Filling Port C. Use 10 µl and the recommended pipet tips (Figure.5.10 b).


61

5. Now the chamber is completely filled with chemoattractant-free medium and cells will only grow inside the gel in the observation area. 6. Use a 20 µl pipet (e.g. Gilson P-20) and apply 10 µl Chemoattractant to the top of Filling Port C, as shown. Do not inject directly (Figure 5.11 a). 7. Use the same pipet settings and aspirate 10 µl air from the opposite Filling Port D. Press the pipet tip directly into Filling Port D (Figure 5.11 b). The Chemoattractant on top of Filling Port C will be flushed inside and fill the lateral channel (Figure 5.11 c).

Figure 5.10: First Lateral channels filling with Chemoattractant-free liquid steps TM c

Ibidi

Figure 5.11: Second Lateral channels filling with Chemoattractant-free liquid steps TM c

Ibidi

5.2.2.1

Results obtained for filling Lateral channels of the PDMS device obtained from Newton SU-8 Master Mold

After the filling of the central channel with MatrigelTM , colored solutions have been pumped, through the twin syringe pumps or it can be done manually using the other type of syringe which allows good control of pumping the liquids or aspiration, into the two lateral channels and the diffusion results can be see in Figure.5.12.As can be seen, the well filling of the lateral channels, and also the diffusion mechanism was quite well as expected. In Figure.5.13. Observing the meniscus shape that there was no pressure differences between the lateral channel and central channel and hence, the shape is


62

almost flat at the contact edge between the fluid in lateral channel and MatrigelTM . Its also visible that the diffusion flow was homogeneous longitudinally through the pillars that can be considered as a good result because this is the desired action which should occur when the cultured cells are inside so as they can have equal attractant stimulus over the whole channel to provide better environment to study their Chemotaxis reaction successfully.

Figure 5.12: Lateral channels filling with colored Chemoattractant-free liquid ( The black dots are graphite powder ). a) Lateral channel filling. b) Newton device after filling the central and lateral channels. Note: The liquid level-out the inlets and outlets to prevent entries of air bubbles

Figure 5.13: An optical microscopic image showing the meniscus due to the diffusion of the Chemoattractant-Free colored fluid through the central channel a) The meniscus showing that the diffusion is going well due to no pressure differences between the fluids in the different channels. b) showing the device after few minutes while the lateral fluid is diffusing homogeneously through the whole area (pillars).

Problems which were faced rather than the bubble formation were the collapse of the pillar, as well as the MatrigelTM passing through the central channel towards the lateral channels as can be shown in Figure.5.14, which shows either the filling mechanism wasn’t correctly followed and there was some missing steps not well done, or it can prove that the device sample was not well bonded on the glass and it can be true if the pillars did


63

Figure 5.14: An optical microscopic image showing the breakthrough of the MatrigelTM towards the lateral channels in two different samples (remarked in yellow circles).

not collapse while the filling process but only the liquid passed under pillars to reach the lateral channel. That problem also could originate due to the high pressure obtained during the filling mechanism, because it can cause compressing of the air inside the channel and this can break the pillars in the first place,hence it will allow the leakage of MatrigelTM later. So it is strongly advised to give a special care while filling the channels, very low pumping pressure must be given as possible then it can be increased manually. The same goes for the aspiration processes as well.

5.2.3

Results obtained for filling Lateral & central channels of the 3D Monolithic 2-layers SU-8 device fabricated onto Silicon substrate

Recalling the device fabrication processes mentioned in section:4.3, After investigation both the device structure and dimensions under the optical microscope, the next step was the microfluidic injection test and to verify the performance of the device, it is noted that this device will undergo different kind of modifications before declaring that it can be used directly for the cell culturing purpose. The filling mechanism can be done as explained before in this chapter or it also can be done quickly by using a small injection pipet which is connected to a tube with the same diameter of the inlet and outlet holes. The MatrigelTM can be used as well to fill the central channel by following the same filling process aforementioned in this section, but it is much easier and quicker to just fill all the channels with the same liquid color for quick analysis. As can be seen from Figure.5.15.which shows the device before being filled and after being filled by the colored liquid to signify the ability of the device to be used for culturing the cells. As can be seen that the pillars are still stable and even visible through the


64

SU-8 3005 top layer. The problem which sometimes occurs during the filling process the channels regarding the passage of the liquid from the inlet to the outlet is that the fluid can not fully pass through, either by capillarity mechanism nor by the aspiration process . That can be due to the blockage of the channel by the SU-8 residuals from the developing process which as a consequence deprive the liquids from passing through them even with a strong suction from the outlet. It is also recommended to wait for a while after dropping some droplets of the liquid on the inlets because the surface of the channel must get wet before the flow of the liquid , after this the passage of the fluid will be easier.

Figure 5.15: An optical microscopic image showing the filling of the device(Fabricated onto silicon substrate). a) The device before the filling process. b) The device after being filled through the central and lateral channels.

Figure 5.16: An optical microscopic image showing close-up view for the diffusion process from the Lateral channel to the central one

As shown in Figure.5.16 b.the diffusion mechanism looks to be good as well as it is homogeneous through the pillars. There will be different tests done to study the cultured cells viability and the ability of the device to perform as desired.


5.2.4

65

Results obtained for filling Lateral & central channels of the 3D Monolithic 2-layers SU-8 device fabricated onto Glass substrate

Filling mechanism and fluid test are similar to the ones mentioned before, there is no difference in the filling mechanism or problems related except those ones observed while filling, such as the collapsing of some pillars during filling, as well as leakage of the colored fluid. The leaked liquid was trapped between the Top layer and buried one, which arose from the weak adhesion of the SU-8 onto glass. That detachment as mentioned previously, was observed after the soft and post exposure bakes, especially after the exposure phase cause the SU-8 molecules start to react. And that could be avoided by cooling down the device over the hotplate to the room temperature and never perform any rapid or abrupt changes in temperature during the device fabrication. As shown in Figure.5.17. which shows a good results obtained for the monolithic 3D buried microfluidic platform fabricated onto a glass substrate, with a second step spun over SU-8 3005 top layer thickness ∼ 4µm and it was exposed for 22 s with an energy dosage of 2.94 mW/cm2 .s (Manual mask aligner) using the second mask to open the inlets & outlets. Different optimized devices were obtained which may satisfy the trend towards dense and complex fluidic systems which require a straightforward and versatile technology for the reliable fabrication of highly integrated micro fluidic chips.

Figure 5.17: An optical microscopic image showing the filling of the device. a) The device before the filling process (Fabricated onto glass substrate). b) The device after being filled through the central and lateral channels.

Chapter 6

Conclusion And Perspectives The work presented in this thesis is a development of a multiWell Microfluidic Chip for Chemo-Taxis Analysis on Embryonic Stem Cells using a PDMS (or thermoplastic) basic structural material which is a biocompatible, cheap and transparent to permit in situ visualization and studying of cells. The fabrication process is developed in three different ways, either through a LIGA like process involving SU-8,electroplating and hot embossing technology, which was found to be difficult to obtain due to the lack of some technological devices and some technological problems. The second fabrication process has been done by fabricating a SU-8 master mold onto a silicon substrate into which a casting process of PDMS material is done to obtain multiple replicas after peeling off the PDMS. As a proof-of-concept, A fabrication of a device of three inlets and outlets was obtained,that can be observable with confocal microscopy, with an array of pillars inside the central channel in which spheroids of cells could be inserted (in case of single cell culturing) or multiple cells can be cultured, studied and immobilized, While on the opposite extremities a specific and controllable gradient could be established. The device has demonstrated a well diffusion tested behavior along the channel has been successfully achieved without either any collapse of the pillars of the middle channel nor lateral leakages through the channels themselves thanks to the well bonded PDMS structure. While the third fabrication process was to obtain a monolithic 3D 2-layered SU-8 buried microfluidic platform onto a silicon substrate. The device fabrication processes and design is similar to the other two previous processes but the difference was just on the second process for making a top cover layer of SU-8 3005. That device can be better for future applications cause it is much more easy to use and showed good fluidic diffusion test and performance. The only problem was the process needs a lot of time and care as well as presence , which make it difficult for patch fabrications. A modification of the last process has been done to allow the fabrication of the buried microfluidic device onto glass substrate to allow a better visibility and transparency of the cultured cells from top 66


67

and bottom views. The fabrication process was almost the same with just little differ of hand skills and auxiliary materials used to treat the glass as silicon substrate and try to adapt the whole processes done onto silicon, to the one done on glass. Different tests have been done to prove the ability to provide good fluidic handling inside the channels, as well as homogeneous diffusion from the lateral channels to the central one. After verifying the desired device performance, the device has been used to chemotaxis processes, Embryonic Stem cells was cultured in IRCC laboratory. The device shows a point of interest for further researches to ease the patch fabrication of the device, as well as using a copper mold to allow fabrication of many replicas by using hot imposing techniques. Further future studies can be dedicated to standardize the process flow and fabrication mechanisms. Due to different results could be always obtained depending on the experiment conducted. Also the design can be modified to allow better packaging of the whole device in one single chip of a 3-layers of SU-8 photoresist packed on each others to allow a fabrication of a monolithic device and compact one, the calibration of the parameters and processes can be inherited from the performed 3D 2-layer monolithic device onto glass substrate which can be a good start point for that kind of projects. Thesis written by — Mohamed Zakarya Rashed Contact

Appendix A

Technological processes Parameters used in experiments A.1

Deposition of a thin film Teflon-like layer onto SU-8 (DRIE parameters)

The deposition was performed using a Deep Reactive Ion Etching (DRIE) ( STS 320PC Reactive Ion Etching system is a single-chamber reactor) , an Inductive Coupled Plasma (ICP ) Machine . The equipment had two independent 13.56 MHz radio frequency (RF) power source. The coil around the etching chamber was used to create plasma, while the platen coil was connected to the wafer electrode to control the RF bias potential of the wafer with respect to the plasma But instead of using the etching process, we will only deposit a few nano thick film of a Teflon-Like material C4 F8 . Backside helium(He) pressurization was used to provide sufficient heat transfer between the wafers to the electrode to maintain a constant wafer temperature. When the wafer was placed in the machine, it was clamped by a set of alumina fingers to the electrode. Passivation(deposition) duration time of 30 s was set. The gas used for deposition was C4 F8 with a flow rate of 50 sccm, while the He flow rate is 10 sccm. The pressure in the chamber was maintained at 20 mTorr , the pressure is controlled separately by an APC valve between the chamber and the turbo-pump . The platen power used was 10 W and the coil power (ICP) was 1500 W. The surface of the substrate was kept constant at 20

◦

C.

68

Appendix A.2 Technological Process flow for 3D monolithic buried microfluidic platform fabrication onto glass substrate

A.2

69

Technological process flow for 3D monolithic buried microfluidic platform fabrication onto glass substrate

Process SU8-2150 / SU-8 3005 over the glass substrate.

TM Microscope Cover ¨ As mentioned before, the glass substrates (MENZEL-GLASER

Slips) are more complex to perform lithographic process onto, compared to silicon ones, so during the exposure process two techniques will be performed , we will recall them later so it is better to label them: 1. Plan A , a piece of silicon wafer will be put under the glass substrate during the lithographic process to reflect back the UV light. 2. Plan B , a pack of glass samples or a plastic transparent box can be put onto the chalk to handle the glass substrate and allow the passage of the UV light without being reflected back to the glass substrate. Preparation of the samples and ensuring the planarity of the work benches & stages 1. Ensure that all the working benches and stages are plan to ensure the planar surface of the SU-8 photoresist layer at the end. TM Microscope Cover Slips) to ¨ 2. Prepare the glass substrates (MENZEL-GLASER

be used. 3. Clean the glass substrates in a boiling Piranha solution (H2 SO4 : H2 O2 = 3 : 1) for 10 min. 4. Put all the samples to be dehydrated on the hotplate @ 110◦ C per 2 min. Spun of MCC Primer 80/20 (over the glass substrate ) 1. Drop the MCC Primer 80/20 using a syringe over the glass substrates. 2. Spin the substrate in the spinner with spin speed: 4000 rpm for duration t= 30 s and acceleration of 1000 rpm/s2 . 3. Bake the substrates on the hotplate for 2 min @ 110◦ C.


70

( SU-8 2150 buried layer (MicroChem Corp., MA, USA) For ∼ 350µm thick layer) 1. Prepare the SU-8 2150 bottle and lean it down till the material can be seen, it is very viscous and so it may take some time, but never use the syringe to drop the SU-8 2150 over the glass substrates. 2. Spin the glass substrate with the following parameters. • First phase , spin speed= 500 rpm , for duration t= 5 s and acceleration = 100 rpm/s2 . • Second phase , spin speed = 2000 rpm, for duration t= 30 s and acceleration = 100 rpm/s2 . 3. Soft Bake • Put the substrates @ T = 120◦ C per ∼ 1h30 min at least. • Ramp down of temperature to T = 60◦ C and keep it for 5 min. — • Ramp down of temperature to T = 50◦ C and keep it for 30 min. 4. Ramp down the temperature to 20◦ C then switch off the hotplate and leave the samples to cool down to room temperature onto the hotplate. 5. Exposure with the first mask Fig.4.14 a, will be performed with an energy dosage of 2.94 mW/cm2 .s (Manual Mask Aligner)with a duration of : • 180 s for glass substrates exposed using Plan A. • 220 s for glass substrates exposed using Plan B. 6. Post Exposure Bake • Starting the hotplate from room temperature and ramp it to T= 60◦ C then leave it there for 30 min. • Switch off the hotplate and let the glass substrates to cool down to room temperature. 7. After cooling down the substrates, no developing should be carried on !


71

( SU-8 3005 (Top Layer ) (MicroChem Corp., MA, USA) For ∼ 4µm thick layer) 1. Prepare the SU-8 3005 bottle and use a syringe to obtain an amount of 1ml and drop them slowly onto the substrate without formation of any bubbles. 2. Spin the glass substrate with the following parameters. • First phase , spin speed= 500 rpm , for duration t= 5 s and acceleration = 100 rpm/s2 . • Second phase , spin speed = 4000 rpm, for duration t= 30 s and acceleration = 300 rpm/s2 . 3. Soft Bake • Put the substrates @ T = 95◦ C per ∼ 20 min at least on hotplate . Or it is recommended to use a convection oven @ T = 95◦ C per ∼ 3 min because we care only about the baking of the top layer. Note:If the convection oven is not in the clean room, then it its strictly required to keep the samples inside a small black box so as not to expose them with any UV light from surroundings . 4. Ramp down the temperature to 20◦ C then switch off the hotplate(if used ) and leave the samples to cool down to room temperature onto the hotplate. 5. Exposure with the second mask Fig.4.14 b, will be performed with an energy dosage of 2.94 mW/cm2 .s (Manual Mask Aligner) with a duration of : • If the glass substrates were exposed firstly using either Plan A, or Plan b, then they must be exposed by the same Plan in the second exposure process. • 22 s for glass substrates exposed using Plan A. • 30 s for glass substrates exposed using Plan B. 6. Post Exposure Bake • If first soft bake was performed on a hotplate then, the samples must be put on the hotplate @ T=65◦ C for 1 min, ramp the temperature till T = 95◦ C and keep it for 3 min. Cool down onto the hotplate to room temperature. • If first soft bake was performed inside a convection oven then it must be put inside the convection oven @ T = 95◦ C per 3 min.Cooling down to room temperature. 7. Developing: developing inside the dedicated beakers filled with SU-8 developer , with strong agitation using the ultrasonic cleaner for 30 ∼ 70 min.

Bibliography [1] Ales Prokop, Zdenka Prokop, David Schaffer, Eugene Kozlov, John Wikswo, David Cliffel, and Franz Baudenbacher. Nanoliterbioreactor: long-term mammalian cell culture at nanofabricated scale. Biomedical microdevices, 6(4):325–339, 2004. [2] Paul J Hung, Philip J Lee, Poorya Sabounchi, Robert Lin, and Luke P Lee. Continuous perfusion microfluidic cell culture array for high-throughput cell-based assays. Biotechnology and bioengineering, 89(1):1–8, 2005. [3] Michael C Liu, Dean Ho, and Yu-Chong Tai. A monolithically fabricated combinatorial mixer for microchip-based high-throughput cell culturing assays. pages 993–996, 2007. [4] Philip J Lee, Paul J Hung, Vivek M Rao, and Luke P Lee. Nanoliter scale microbioreactor array for quantitative cell biology. Biotechnology and bioengineering, 94 (1):5–14, 2006. [5] Michael C Liu, Dean Ho, and Yu-Chong Tai. A monolithically fabricated combinatorial mixer for microchip-based high-throughput cell culturing assays. pages 993–996, 2007. [6] Hans-Guido Wendel, Elisa de Stanchina, Jordan S Fridman, Abba Malina, Sagarika Ray, Scott Kogan, Carlos Cordon-Cardo, Jerry Pelletier, and Scott W Lowe. Survival signalling by akt and eif4e in oncogenesis and cancer therapy. Nature, 428 (6980):332–337, 2004. [7] Alois Knoll, Torsten Scherer, Iris Poggendorf, Dirk L¨ utkemeyer, and J¨ urgen Lehmann. Flexible automation of cell culture and tissue engineering tasks. Biotechnology progress, 20(6):1825–1835, 2004. [8] Rita Barallon, Steven R Bauer, John Butler, Amanda Capes-Davis, Wilhelm G Dirks, Eugene Elmore, Manohar Furtado, Margaret C Kline, Arihiro Kohara, Georgyi V Los, et al. Recommendation of short tandem repeat profiling for authenticating human cell lines, stem cells, and tissues. In Vitro Cellular & Developmental Biology-Animal, 46(9):727–732, 2010. 72

Bibliography

73

[9] Kempner M and Felder R. (2010, september 14) a review of cell culture automation. journal of american association for medical automation, (volume 7) issue 2, 56-62. [10] Wang P et al Kleinhans R, Brischwein M. Sensor-based cell and tissue screening for personalized cancer chemotherapy. med biol eng comput. 2012 feb;50(2):117-26. doi: 10.1007/s11517-011-0855-7. [11] Poggendorf I et al Knoll A, Scherer T. Flexible automation of cell culture and tissue engineering tasks. biotechnol prog. 2004 nov-dec;20(6):1825-35. [12] Felder R and Gasperoni C. (2010, september 14) automating the clinical laboratory. [online]. available: http://www.medicalautomation.org/2010/09/automatingthe-clinical-laboratory/. [13] Simone Luigi Marasso, Eros Giuri, Giancarlo Canavese, Riccardo Castagna, Marzia Quaglio, Ivan Ferrante, Denis Perrone, and Matteo Cocuzza. A multilevel lab on chip platform for dna analysis. Biomedical microdevices, 13(1):19–27, 2011. [14] Marzia Quaglio, Giancarlo Canavese, E Giuri, Simone Luigi Marasso, Denis Perrone, M Cocuzza, and CF Pirri. Evaluation of different pdms interconnection solutions for silicon, pyrex and coc microfluidic chips. Journal of Micromechanics and Microengineering, 18(5):055012, 2008. [15] Nirveek Bhattacharjee, Nianzhen Li, Thomas M Keenan, and Albert Folch. A neuron-benign microfluidic gradient generator for studying the response of mammalian neurons towards axon guidance factors. Integrative Biology, 2(11-12):669– 679, 2010. [16] Yoojin Shin, Sewoon Han, Jessie S Jeon, Kyoko Yamamoto, Ioannis K Zervantonakis, Ryo Sudo, Roger D Kamm, and Seok Chung. Microfluidic assay for simultaneous culture of multiple cell types on surfaces or within hydrogels. Nature protocols, 7(7):1247–1259, 2012. [17] B. mosadegh et al., langmuir 23, 10910 (2007). . [18] Regina Luttge, Erwin JW Berenschot, Meint J de Boer, Dominique M Altpeter, Elwin X Vrouwe, Albert van den Berg, and Miko Elwenspoek. Integrated lithographic molding for microneedle-based devices. Microelectromechanical Systems, Journal of, 16(4):872–884, 2007. [19] Bart Schurink and Regina Luttge. Hydrogel/poly-dimethylsiloxane hybrid bioreactor facilitating 3d cell culturing. Journal of Vacuum Science & Technology B, 31 (6):06F903, 2013.

Bibliography

74

[20] AP Balgude, X Yu, A Szymanski, and RV Bellamkonda. Agarose gel stiffness determines rate of drg neurite extension in 3d cultures. Biomaterials, 22(10):1077– 1084, 2001. [21] A. pluen, p. a. netti, r. k. jain, and d. a. berk, biophys. j. 77, 542 (1999). . [22] Hynda K Kleinman, Mary L McGarvey, Lance A Liotta, Pamela Gehron Robey, Karl Tryggvason, and George R Martin. Isolation and characterization of type iv procollagen, laminin, and heparan sulfate proteoglycan from the ehs sarcoma. Biochemistry, 21(24):6188–6193, 1982. [23] A Albini, Y Iwamoto, HK Kleinman, GR Martin, SA Aaronson, JM Kozlowski, and RN McEwan. A rapid in vitro assay for quantitating the invasive potential of tumor cells. Cancer research, 47(12):3239–3245, 1987. [24] L Vićovac and JD Aplin. Epithelial-mesenchymal transition during trophoblast differentiation. Cells Tissues Organs, 156(3):202–216, 1996. [25] Yi-an Bi, Diana Kazolias, and David B Duignan. Use of cryopreserved human hepatocytes in sandwich culture to measure hepatobiliary transport. Drug metabolism and disposition, 34(9):1658–1665, 2006. [26] K Gassei, J Ehmcke, MA Wood, WH Walker, and S Schlatt. Immature rat seminiferous tubules reconstructed in vitro express markers of sertoli cell maturation after xenografting into nude mouse hosts. Molecular human reproduction, 16(2):97–110, 2010. [27] Xiaozhong Yu, Jaspreet S Sidhu, S Hong, and Elaine M Faustman. Essential role of extracellular matrix (ecm) overlay in establishing the functional integrity of primary neonatal rat sertoli cell/gonocyte co-cultures: An improved in vitro model for assessment of male reproductive toxicity. Toxicological Sciences, 84(2):378–393, 2005. [28] Charles H Streuli, Nina Bailey, and Mina J Bissell. Control of mammary epithelial differentiation: basement membrane induces tissue-specific gene expression in the absence of cell-cell interaction and morphological polarity. The Journal of cell biology, 115(5):1383–1395, 1991. [29] Ren Xu, Celeste M Nelson, John L Muschler, Mandana Veiseh, Barbara K Vonderhaar, and Mina J Bissell. Sustained activation of stat5 is essential for chromatin remodeling and maintenance of mammary-specific function. The Journal of cell biology, 184(1):57–66, 2009.

Bibliography

75

[30] Tatiana Lopatina, Natalia Kalinina, Maxim Karagyaur, Dmitry Stambolsky, Kseniya Rubina, Alexander Revischin, Galina Pavlova, Yelena Parfyonova, and Vsevolod Tkachuk. Adipose-derived stem cells stimulate regeneration of peripheral nerves: Bdnf secreted by these cells promotes nerve healing and axon growth de novo. PloS one, 6(3):e17899, 2011. [31] Elsa Quintana, Mark Shackleton, Michael S Sabel, Douglas R Fullen, Timothy M Johnson, and Sean J Morrison.

Efficient tumour formation by single human

melanoma cells. Nature, 456(7222):593–598, 2008. [32] Gina S Fiorini and Daniel T Chiu. Disposable microfluidic devices: fabrication, function, and application. BioTechniques, 38(3):429–446, 2005. [33] LP Yeo, SL Poh, Yee Cheong Lam, and Mary Bee-Eng Chan-Park. Plasma polymerization of c[subscript4] f[subscript8] thin film on high aspect ratio silicon molds. 2005. [34] Chi-Ming Chan, Lu-Tao Weng, and Yiu-Ting R Lau. Polymer surface structures determined using tof-sims. Reviews in Analytical Chemistry, 33(1):11–30, 2014. [35] F.e.h.tay, j.a.v.kan, f.watt, and w.0.choong, j.micromech. microeng. 11, pp.2732,200. . [36] http://aveclafaux.freeservers.com/su-8.html. . [37] Francis EH Tay, JA Van Kan, F Watt, and WO Choong. A novel micro-machining method for the fabrication of thick-film su-8 embedded micro-channels. Journal of Micromechanics and Microengineering, 11(1):27, 2001. [38] M.o.heuschkel, l.guerin, b.buisson, d.bertrand, p.renaud, sens. actuators b 48, pp. . [39] https://www.bd.com/resource.aspx?idx=17841. .