Organ on a Chip (OoC) systems are of high interest through its use for medicine testing in a small time scale without the need for animal testing. Membranes used in OoCs form a base to grow cells on and need to be suitable for cell-attachment and porous. Fouling is the 'Achilles heel' in membrane performance. Research shows that as a result of fouling the viability of the skin cells grown in the chip decreased to zero after 3 weeks. If research on organs is to extend and research on cell- or tissue growth will include longer time spans, the influence of membrane fouling with conditions similar to the OoC is an important factor to understand. In this research a microfluidic flow cell is produced and used to explore fouling within the OoC. Static and dynamic fouling experiments are executed on membranes having pore sizes ranging from 0.4 to 5 μm. Scanning electron microscope (SEM) images indicate that standard blocking and cake layer formation are dominating fouling mechanisms. Membranes with 1 μm pore size are the most susceptible to standard blocking. Further a decrease in pore area of 0%, 11% and 20% and a decrease in uncovered amount of pores of 27%, 34% and 80% for Glycine, BSA and λ-DNA respectively are measured after one week of fouling. Cake layer formation is seen after fouling for a shorter duration for BSA (one day) than for Glycine (one hour) and a higher concentration of BSA particles is permitted through all tested membranes than Glycine, therefore the conclusion is drawn that BSA causes less fouling than Glycine.

Implanted intra-peritoneal insulin infusion (‘i4’) devices administer insulin into the peritoneum of diabetic patients with extreme blood glucose swings, that are highly difficult to control. The flushing procedures of i4 devices to clear obstructions from its conduits caused by insulin degradation are not always fully effective. The goal of this research was to investigate how the flushing effectiveness can be increased by adapting conduit geometries and adapting the flush flow rate. It was hypothesized that flushing of conduits with complex geometries (featuring a.o. widened sections, or parallel flow paths) would be less effective than simple geometries, and that a higher flush flow rate would increase flushing effectiveness. An ex vivo laboratory experimental setup was created by producing polymethyl methacrylate (PMMA) test conduits of differing geometries that model the in vivo implanted intra-peritoneal insulin infusion (i4) device’s conduits, but are transparent to allow optical inspection. These test conduits were filled with acidified insulin, and precipitation thereof was artificially induced by heat incubation of the filled conduits, as a model for obstructed i4-device conduits. Experimental flushing procedures using sodium hydroxide (NaOH) as a flush liquid were performed on these test conduits at 2 different flush flow rates: 1.5mL/min, and at 20mL/min. After the flushing, the effectiveness of the procedure to clear the conduit of precipitates was evaluated by microscopically analysing the quantity of residual precipitates, and measuring Total Organic Carbon (TOC) of a liquid sample taken from the post-flushing conduit. All conduits with more complex geometries than straight tubes showed lower effectiveness of cleaning by flushing. The high flow rate showed an averaged 19% higher effectiveness of the flushing. In the sections of the conduits featuring widening, flushing often resulted in a narrow path being effectively cleaned, but significant regions of the conduit showed presence of residual precipitates. This research provides insight into the negative influence of complex geometries on effectiveness of flushing, and the positive influence of increased flush flow rate to improve effectiveness. Experimental repetition in this research is low, and the experimental methods were prone to much influence of material or measurement disturbances. This research merits elaboration and improvement of materials and methods to more precisely investigate the proposed relations. Outcomes suggest that future i4-devices should be designed to avoid complex conduit geometries as much as possible, and to be flushed with significantly higher flush flow rates than currently clinically applied.

The structuring of polymers on the micro and nanoscale is performed largely by lithographic techniques, which require complex steps and suffer from geometrical and scaling limits. Techniques such as Nanoimprint Lithography and Hot Embossing are attractive alternatives as they can produce high quality nanoscale structures with a high throughput. Colloidal building blocks offer a way to self-assemble the masters required for these replication steps. However, the difficulty with self-assembly lies in the creation of stable structures with little defect that are well controlled. This work sets out to provide a comprehensive overview of the aspects of the fabrication of colloidal crystals, the most influential parameters and how they can be qualified. Offering a basis for future research on microscale replication by colloidal crystals and resulting in development of a method for the fabrication of colloidal crystal, which can be used as a master to replicate microscale patterns through Hot Embossing.

Kidney stone disease is an increasing worldwide issue. There is a current lack of understanding of the exact mechanisms involved, partially due to the need for experimental instrumentation able to mimic the microenvironmental conditions present in vivo. As crystal nucleation often initiates at liquid-solid interfaces, the interface morphology plays a significant role in the rate of nucleation. Within the nephron, the functional unit of the kidney, four distinct segments can be distinguished that contain different surface morphologies. Particularly, the cells lining these segments contain protrusions in the shape of nanopillars that vary in length, diameter and spacing. Exploiting the opportunities provided by organ-on-chip technology, we designed and manufactured a microfluidic device proposed to increase our understanding of the exact mechanisms that drive kidney stone crystallization. We used two-photon polymerization to fabricate surfaces that contain these morphologies in materials possessing a Young’s modulus matching the value of the biological structures. After optimizing design and process parameters like laser power and scanning speed, we manufactured nanopillars with diameters in the range of 150-250 nm and aspect ratios up to 100 with which we can mimic the protrusions in the nephron. Pillar arrays were printed on a glass slide, and assembled with a polydimethylsiloxane (PDMS) microfluidic component (channels 150 um wide). We used high-resolution 3D printing to create master molds used for PDMS soft-lithography. After precise alignment, the two components were brought together and clamped mechanically with a 3D printed holder to ensure a watertight seal between the glass and PDMS. This microfluidic device was used to study crystallization of the most common type of kidney stone, calcium oxalate monohydrate (COM). Because of the chip dimensions combined with a surface morphology this device closely resembles a human nephron.

Bone is a tissue with many important functions, such as structural support, organ protection and mineral homeostasis. Additionally, after trauma, it can naturally regenerate into fresh, fully functional bone tissue. However, several factors can significantly impede the healing process and cause the formation of non-union tissue. Costs for hospitalization and treatments that prevent such non-unions can become a significant personal and socio-economic burden. To develop more efficient treatments that can tackle this, a thorough understanding of the bone fracture healing process is needed. It is currently known that bone heals via a multistage sequential process of inflammation, soft-callus formation, hard-callus formation and then bone remodelling. Moreover, during this process there is a need for restoration of a vascular network for nutrient delivery and waste removal. Whereas the regulation of vascular network during later stages is understood, knowledge of the mechanisms and driving forces of vascularization during earlier phases in bone healing is currently lacking. Nevertheless, during this process, cell migration towards the site of fracture is of great importance. This process is affected through a range of extracellular signals, which are dependent on the properties of the extracellular matrix. It has been shown that the structural properties and the composition of the fracture tissue are two of the key regulators in guiding cellular migration. This study has integrated the use of a novel electrospinning technique to mimic the effects of the early fracture composition on (endothelial) cell migration. Fibrinogen and gelatin fibers, which are reflective of the early fracture environment, have been spun using an acidic solvent system. Characterization of the fibers showed that the electrospinning process caused minor changes in the protein structure of fibrinogen, and none in that of gelatin. Furthermore, endothelial cells were cultured with a cell free zone on top of fibrinogen and gelatin patterned substrates to mimic cell migration during fracture revascularization. Cellular migration increased with higher fiber density, and gelatin fibers generally showed higher rates of cell migration. Furthermore, gelatin fibers showed higher amounts of cell polarization than fibrinogen fibers, likely driven by the higher stiffness of gelatin fibers. Moreover, the addition of fibers showed to reflect different modes of in vivo vascularization of cell migration on the substrates. To further analyze the effects of fracture composition on endothelial cell migration, fiber membranes were electrospun on a transwell-like device, and cell migration from a fibrin gel was assessed. The physical constraints posed by the fiber membranes were shown to impede cellular migration from the fibrin gel, and no significant differences were found between cell migration on the membranes. Overall, it became evident that the structural properties and the composition of the fibrous micro-environment regulate cellular migration, and can in turn affect the bone regeneration as a whole. Ultimately, further optimization and characterization of the used models are needed, such that they can more closely reflect the fracture healing environment.

With increasing prevalence of neurodegenerative diseases the need for an effective treatment increases. To date pharmacological treatment is the golden standard. However, the medicines used have many adverse effects and, only work for a short time period. The best solution would be to cure the disease and thus stop the degradation of neurons and replace the already lost neurons by healthy ones. With the development of regenerative medicine, strategies to replace the lost neurons are emerging. One of the approaches includes the use of neuronal progenitor cells to differentiate and grow into functional neurons. To achieve this, a suitable in vitro environment is needed that includes an optimal use of culture medium and substrate. Laminin has been shown to be an effective substrate for neuronal cell culture purposes as it is part of the native extracellular matrix in the nervous system. However, laminin represents a family of 20 isoforms as known so far and there is still a lot unknown about the functions of each isoform on neuronal cultures. In this study the effects of eight available human recombinant laminin isoforms (i.e., -111, -121, -211, -221, -411, -421, -511 and -521) on cell attachment, differentiation, neurite outgrowth and connectivity of cerebellar progenitor cells have been assessed by means of immunofluorescence techniques. A general laminin substrate stemming from mice sarcoma (L2020) was used as a control. The results show that Laminin- 211, -221 and -411 do not possess any neurite outgrowth and differentiation abilities and thus were not included in the remaining experiments. Between the remaining substrates used no difference was found on cell attachment and neuronal differentiation. Laminin-511 and -521 increased neurite outgrowth the most and laminin-421 the least. Laminin-111 took third place. Laminin-111, -511, -521 and L2020 were included in the connectivity experiments. Laminin-521 stimulated synaptogenesis to the largest extend followed by laminin-511 and then -111. Overall, laminin-511 and -521 showed the best results in this study and performed better than a general laminin substrate that studies mostly use.

Microfluidics has changed the way we think about laboratory experimental design. With the use of lab-on-chip devices, we are capable of carrying out experiments that include several functions on a single chip. Despite a lot of progress in the miniaturization of these devices, the peripheral systems which control the microfluidic functionalities like valves and pumps, remain bulky, making the microfluidic devices less portable. Conducting polymers are potential candidates that show promise to solve this issue and make the microfluidic devices portable. They are a class of polymer materials with intrinsic conductive properties and the ability to be formed and deformed at a low applied electric potential. This thesis presents the different aspects of the conceptualisation and creation of the actuator for an all polymer conducting polymer microfluidic valve. The proposed system uses the widely studied multi-layer bending beam actuator, commonly used as an artificial muscle or soft actuator. The actuation of the all polymer system is the result of the different expansion rates during an electrochemical reaction. The system consist of three polymer layers. The first layer is the flexible base layer that also functions as a fluid barrier in the microfluidic system. The second layer is the electrode material that is used as an electrode in both the fabrication and actuation of the expanding layer. The final layer is the electrochemically polymerized layer of expanding conducting polymer. The all polymer actuator can eventually form the basis for new opportunities regarding the miniaturisation and commercialisation of Lab on Chip devices

To control fluid transport, microfluidic systems often make use of pressure driven flow and pneumatically actuated valves. However these require bulky external instrumentation. An integrated fluid control mechanism would make microfluidic systems more portable, closed and automated. The aim of this project was to create a pH-responsive membrane for integrated fluid control in microfluidic systems. pH-responsiveness will allow the membrane to interact directly with analytes in a microfluidic system. First pH-responsive materials and fluid control mechanisms to create a pH-responsive membrane were reviewed. Cross-linked polymer hydrogels with pH-dependent volumetric swelling were identified as the most suitable material for the membrane. These materials respond to changes in pH by large volumetric transitions, which creates potential for control over a wide range of flow rates. Furthermore they have an inherent interaction with water and are permeable to small molecules such as H+ ions. The material is synthesized from a liquid precursor that can be cured through photo-initiated free radical polymerization. We tested a system for measuring the pH response of the material that consisted of a hydrogel disk that was vertically constrained inside a fluidic channel. Due to this constraint, expansion only occurred in the lateral direction and could be measured using an optical microscope. Furthermore, we developed a method for manufacturing macro-porous hydrogel membranes that gives control over pore size, shape and position. A photo-lithography approach was used to pattern the membranes and thereby create pores, using a photo-mask that was manufactured on an office printer in a very fast and low-cost process. The resulting membranes had a thickness of 140-190 um and pore diameters of 100-400 um. The pore size was measured for environmental pH of 1.6 and 7.1, within this range the pores doubled in diameter. Furthermore the pH-responsive deformation ratio of the pores increased significantly with increasing curing time and decreasing pore diameter. The results suggest that there is a difference in material properties around the pores that develops due to a local difference in received exposure dose during curing. The fluidic properties and pH-response of the membrane can be adjusted to suit a specific application by changing the design, the curing time or the chemical composition of the membrane. The constrained disk system can then be used to measure and compare the pH-response of different potential materials.

The integration of luminescent nanoparticles into transparent polymer matrices opens the way to affordable, scalable and efficient luminescent solar concentrators. A key challenge in the fabrication is to prevent agglomeration of the nanoparticles as this will drastically reduce the performance due to scattering effects. The incompatibility of inorganic nanoparticles with organic media typically leads to irreversible agglomeration upon direct mixing. In this study, inorganic luminescent Y3Al5O12:Ce3+ nanoparticles are transferred from a polar to a nonpolar solvent without any noticeable agglomeration using amphiphilic statistical copolymers as stabilizing agents. The process parameters that determine the success of stabilization are studied both theoretically and experimentally and a general procedure is proposed to find the optimal conditions for the phase transfer process. The importance of compatibility between the amphiphilic copolymers and the polymer matrix was demonstrated by integrating the stabilized nanoparticles into various polymer matrices. Fully transparent polymer nanocomposite thin films were prepared without any sign of agglomeration. The simple, universal and scalable method presented in this report allows for the fabrication of nanocomposite luminescent solar concentrators using a wide variety of luminescent materials and polymers.

Nanoparticles have unique properties that are sought after for the development and improvement of applications. Prerequisite to achieve these applications, methods are required to put the nanoparticles in different patterns and arrangement on a target substrate. In the course of developing an aerosol-based nanoparticle printing system, this work explores the results of different aerosol deposition configurations. The work is based on a spark ablation process that generates an aerosol of argon and copper nanoparticle agglomerates. The deposition configuration deposits the aerosol on a target substrate. Three configurations are explored, one which deposits at subsonic velocities, and two which deposit at sonic velocities. The aim is to find the most favourable deposition conditions for the direct writing of nanoparticle patterns. It is found that a sonic deposition configuration with a low pressure ratio at small substrate distances has the smallest deposit diameter. The configuration allows for the patterning of narrow lines, that consist of two deposition regions: a micro-aggregate region and a nanoparticle region. It was found that the density of these regions is mainly influenced by the process variables of the spark ablation process, and can thus be used to gain a high consistency and well-defined edges.

A major challenge facing the study of bone regeneration today is the inability to mimic the tissue microenvironment and the factors comprising it. Organ-on-chip based microfluidic systems today aim at doing so by dynamically culturing living cells in them. These are preferred owing to their small feature sizes, laminar flows, high throughput, small reagent quantities, and reduced dependency on animal studies. Substrate topography even at the submicron and nanoscale is known to induce osteogenesis by infleuncing cells at ligand/receptor levels but poses a major challenge in terms of fabrication and upscaling. The two photon polymerisation (2PP) process allows producing features in a spatially and dimensionally controlled fashion. The method is typically known for fabricating micro-sized structures but remains largely unexplored for producing features in the submicron/nano scale. This study aims at creating reproducible submicron pillar based topographies using 2PP and integrating them into a microfluidic device for studying their effect on bone regeneration. Uniform submicron patterns were produced using optimal process parameters and writing strategy. Further, the patterns were expanded to areas up to 2 mm $ imes$ 1.7 mm and their uniformity was assessed. To demonstrate multiscale fabrication, these were also integrated onto a 2PP fabricated micro-scaffold in a single step process. The topography enhanced the surface hydrophilicity and could withstand flow rates of up to 8 ml/min. The human mesenchymal stem cells (hMSCs) dynamically cultured in these integrated devices showed a healthy morphology on the pattern with no visible signs of cytotoxicity even after a period of 5 days. This study marks a step towards printing controlled submicron topographies using 2PP that can be upscaled (both spatially and dimensionally), incorporated into biocompatible microfluidic chips, and thus help in achieving the optimal microenvironment for bone regeneration studies in the future.

Administration of drugs can help in the recovery of the brain for example after a stroke. There are many available delivery routes such as intravenous injection or local injection in the brain. Whereas, the former encounters problems with systemic side effects, the latter requires multiple invasive brain surgery. Local drug administration could be achieved with a drug delivery implant to reduce the amount of invasive procedures and to include useful features such as electrodes for monitoring neural activity. In this study an advancement is made towards spatial control of permeability in a membrane. Porous PDMS membranes were fabricated by replica molding in a 3D-printed mold. The smallest pore diameter was 100 μm and the thinnest membrane thickness was 160 μm. The permeability of such a membrane is tested by a dye which must go through the membrane in order to observe staining in 0.6% agarose gel to simulate diffusion of brain tissue. No staining was detected using pristine PDMS membranes. However, a fast flow from hydrophobic to hydrophilic side was observed in a PDMS membrane recently treated with plasma on a single side. Furthermore, slow diffusion of the dye was present in a less hydrophilic, dextran-modified membrane that received plasma treatment on one side as well. Thus, the permeability was controlled by the wettability of the membrane surface without modification of the entire inner pore. The same experiments were performed on a mouse brain to be compared to the results on agarose gel. The use of agar as a brain phantom material is useful for further investigation of a local drug delivery device, but a few improvements are required to better correspond to the brain.

A manufacturing technology is proposed to manufacture polymeric membrane scaffolds for culturing of cells, tissues, and organoids with integrated sensor capabilities and fluidic functionalities. A multilayer substrate is imprinted with a soft mold to manufacture polymeric membrane scaffolds for culturing of cells, tissues, and organoids with integrated sensor capabilities and fluidic functionalities. Aside from exploiting the well characterized, simple and low cost techniques that can be employed in polymer manufacturing, the use of solely polymeric materials increases the opportunities for functional integration, while at the same time allowing for the translation of complex cleanroom fabrication processes into mold-based replication techniques. With this developed manufacturing technology, it is demonstrated that it is possible to manufacture membrane scaffolds with integrated porous electrode up to 1mm by 1mm in effective surface area, with a thickness of 10 µm, with control over the pore diameter (as small as 400 nm), the porosity, and the location of the pores. The morphological characterization is done by scanning electron microscopy. The integrated porous electrode’ performance is characterized electrically by a 4-point probe and electrochemically by cyclic voltammetry and electrochemical impedance spectroscopy. Showing the possibility to use the conductive layer to give cells electrical stimuli and for the use as a biosensor.