In the last decade, Additive Manufacturing (AM) has evolved from a specialized niche application to
a widely used standard tool that has become indispensable in many areas of research and industry.
Due to new technologies and materials enabling the fabrication of high quality products, AM is not
only relevant for Rapid Prototyping, but also in the fabrication of products for end users. Especially
when a high degree of customization or geometrical complexity is involved, AM methods can be
considered as viable options. In medicine and biochemical engineering, AM is typically employed
for applications like the fabrication of dental implants and mouthguards or for customized lab
equipment, microfluidic devices and even chromatography columns. The combination of biological
materials and living cells with AM methods has resulted in the establishment of bioprinting as a
separate field with new opportunities and challenges. Bioprinting methods allow the fabrication of
soft, water-based materials suitable for the physical entrapment of enzymes. This allows biocatalytic
reactors to be directly printed using enzyme-loaded inks.
The present thesis aims at extending the toolbox for the fabrication of biocatalytically active
... mehr
materials with a focus on extrusion-based bioprinting. Novel inks are established in combination
with specifically adapted printing setups to achieve enhanced printability. To assess the performance
of different materials regarding the resulting biocatalytic activity, microplate-based activity assays
are established for two different enzymes. The inks and hydrogels are characterized using a range
of additional analytical methods like rheology, mechanical testing or scanning electron microscopy.
To determine the permeability of hydrogels for substrate molecules, a microfluidics-based method
for the estimation of diffusion coefficients in hydrogels is developed. As a general contribution to
the improvement of process monitoring and control in extrusion-based bioprinting, a PID-based
pressure control is established to generate a constant and reproducible ink flow.
In an initial study, a novel material system for the printing of enzymatically active structures
was established by employing high internal phase emulsions (HIPEs) as inks. HIPEs are emulsions
that contain at least 74 % (v/v) of internal phase which corresponds to the densest possible packing
of droplets before deformation occurs. Polymerizable oily monomers were used as the external
phase of the HIPEs and poly(ethylene glycol) diacrylate and acrylic acid were added to the aqueous
internal phase. Oil- and water-soluble photoinitiators were added to the respective phases, as well.
Polymerizing these inks resulted in the formation of an open-porous polymer scaffold filled with
an interconnected hydrogel. This approach enables the fabrication of composite materials with
hydrogel-like properties like the diffusibility for substrate and product molecules, while exhibiting
higher mechanical stability due to the supportive effect of the polymer scaffold. Also, the formulation of inks as emulsion results in rheological properties favorable for extrusion-based printing. To allow
the production of small HIPE batches, a customized setup was established based on a 3D-printed
helical stirrer blade that allowed the production of HIPEs in 50 mL Falcon tubes. The small scale
production allowed adding the enzyme β-galactosidase to the aqueous phase by minimizing material
waste. Rheological measurements with a range of different HIPE compositions showed that HIPEs
with a high amount of surfactant in the external phase and with a high volume fraction of inner
phase displayed higher yield stress values which correlate with printability. In general, the produced
HIPEs displayed excellent rheological properties. Electron scanning microscopy showed that both
the external and internal phase of the HIPEs could be polymerized. A cure-on-dispense setup with
four UV LEDs was developed to allow the polymerization of the inks during extrusion which reduced
ink spreading and improved printing quality further. Hollow cylinders of enzyme-containing inks
were printed to perform activity assays in 48-well microplates. The results showed that the HIPEs
were more biocatalytically active when they contained high amounts of monomer in the aqueous
phase and a high aqueous phase volume fraction. The presence of at least 7 % (w/w) monomer in
the aqueous phase caused a more than fivefold increase in measured activity compared to HIPEs
without monomer in the aqueous phase. The diameter of the printer nozzle could be identified
as another important parameter influencing the resulting specific activity. This observation could
be attributed to the mass transfer limitations caused by the matrix material which makes high
surface-area-to-volume ratios favorable.
To cover a wider range of ink types, the second and third study of the thesis aimed at investigating
inks based on agarose and agar. These inks demonstrated very different material properties compared
to HIPEs, both in a fluid state as inks and in a solidified state as hydrogels. To adapt to the
requirements of these inks, the hardware of the printing setup and the employed analytical techniques
were again specifically adapted to the investigated inks. Diffusibility for substrate and product
molecules was identified as one of the most important properties of the materials regarding their
suitability for the immobilization of enzymes. Thus, a microfluidics-based method to estimate the
diffusion coefficient of an analyte within transparent hydrogels was established. A microfluidic
chip with three inlets and a y-junction was employed to create an interface between the hydrogel
to be investigated and an analyte solution. For that purpose, liquid ink was injected at elevated
temperature into the chip from one side, until it reached the y-junction. After the gelation of
the ink, the analyte solution was injected through one of the other inlets and the diffusion of the
analyte through the hydrogel was monitored using a UV area imaging system. Diffusion coefficients
could be estimated by fitting the obtained analyte concentration profiles along the microfluidic
channel with an analytical solution of Fick’s second law of diffusion. As a case study, the diffusion
coefficient of lysozyme was compared in a range of hydrogels made from different concentrations of
unmodified agarose and low-melt hydroxyethyl agarose. It was found that the diffusion coefficient of
5(6)-carboxyfluorescein was slightly higher in unmodified agarose hydrogels compared to low-melt
agarose hydrogels. This aligns well with the theoretical prediction that the polymer networks of
low-melt agarose hydrogels exhibit smaller pore sizes. The same trend was found for polymer
concentration where higher concentrations were associated with lower diffusion coefficients and
smaller pores.
In a third study, inks based on low-melt agarose and agar were investigated as a less complex
alternative to HIPE-based inks. The thermogelling behavior of agarose- and agar-based inks required
a different printing setup than the photopolymerizable HIPEs. A heatable nozzle consisting of a
3D-printed metal body, a temperature sensor and a heating filament was implemented in the printer
setup to ensure that the inks could be extruded in a liquid state at a defined temperature. The setup also drastically reduced nozzle clogging. The inks were extruded onto a cooled substrate to accelerate
the gelation process and reduce ink spreading. While the customized setup enhanced printability
significantly compared to previous studies with agarose-based inks, it was still drastically inferior to
the printability of HIPEs, both regarding strand thickness and achievable complexity like overhangs.
Only basic grid structures without overhangs could be printed. A polymer concentration of at least
4.5 % (w/w) was found to be beneficial for printing and grid structures of 2 cm in height could be
fabricated. Using rheological methods, the inks were analyzed for their flow properties and their
melting and gelling behavior. The low-melt agarose showed drastically reduced gelling and melting
temperatures compared to the agar inks. The solidified hydrogels were subjected to mechanical
testing. A set of analytical methods established in the previous studies was reapplied to evaluate
the agarose- and agar-based hydrogels with regards to their application in enzyme immobilization.
For that purpose, the thermostable enzyme esterase 2 from Alicyclobacillus acidocaldarius was
added to the liquid inks before printing. Microplate-based activity assays were used to analyze the
enzymatic activity and leaching behavior of printed hydrogel samples, while the microfluidics-based
method was employed to determine the diffusibility of the hydrogels for 5(6)-carboxyfluorescein, the
product of the reaction catalyzed by esterase 2 in the employed activity assay. It was found that the
agar-based hydrogels showed higher diffusibility and activity, but also increased enzyme leaching.
The tendency for enzyme leaching not only demonstrated the lacking suitability of agar-based
hydrogels for the employment in perfused reactors, it also explains the seemingly positive results
in the performed activity assays, as leached enzyme is not exposed to the same mass transfer
limitations as immobilized enzyme resulting in enhanced observed activity. Due to their low enzyme
leaching and acceptable printability, agarose inks with a concentration of at least 4.5 % (w/w) were
recommended as suitable inks for the application in biocatalytic reactors.
Independent of ink type, the previous studies revealed a general lack of reproducibility in
pneumatically driven bioprinting caused by unsteady and poorly reproducible ink flow rates. Besides
ink viscosity and extrusion pressure, additional factors like cartridge fill level, partial nozzle clogging
and ink inhomogeneities were suspected to influence the extrusion flow rate and cause imperfections
in the resulting prints. Batch-to-batch variations and temperature fluctuations influencing the ink
viscosity posed additional challenges. In the study investigating agarose- and agar-based inks, every
printed sample was weighed before being used for activity assays and discarded if it did not comply
with the set target weight within a specified margin of error. When necessary, the extrusion pressure
was manually adapted to ensure the comparability of the printed samples. As a consequence, a
study was initiated to establish in-line process monitoring of the ink flow rate as an essential process
parameter and to develop an automated and reproducible method to generate a constant target
flow rate by continuously adapting the extrusion pressure based on real-time flow rate data. To
obtain the required data in an in-line measurement, a liquid flow meter was integrated into the
setup of a pneumatically driven bioprinter using a 3D-printed mount. A Python-based software
tool was developed to communicate with the flow sensor and to process the data input. A PID
loop was implemented and fed with the real-time flow rate data. Based on the data input, the
software continuously adapted the extrusion pressure of the printer. Three different case studies
were performed to assess the performance of the PID control setup:
a) Continuous dispensing: Several runs of continuous dispensing demonstrated the automatic
pressure adjustment to consistently meet a specified target flow rate independently of the user.
Compared to the constant pressure setting, the adaptive pressure control proved effective in
compensating for environmental or system-related influences like nozzle clogging. b) Adaptation to
ink inhomogeneities: A more realistic use case was investigated by printing hollow cylinders from a cartridge filled with layers of two differently concentrated poloxamer 407 inks to simulate ink
inhomogeneities. The adaptive pressure control was able to generate a constant flow rate by adapting
the pressure appropriately during the printing process. As a result, relatively consistent cylinders
could be printed, whereas the constant pressure setting resulted in cylinders with strongly deviating
wall thicknesses. c) Process transfer to other nozzle types: To demonstrate the simple process
transferability between different experimental setups, test prints were carried out with three different
nozzle types with the same orifice diameter. The adaptive pressure control was able to generate
the same constant flow rate with all three nozzle types within an adjustment phase of 30 to 60 s.
The resulting cylinders were of consistent quality, independent of the nozzle. Prints with constant
pressure setting suffered from a lack or abundance of extruded ink, if not performed with a pressure
specifically determined for the corresponding nozzle type. The performance of the PID-regulated
adaptive pressure control demonstrated that it can contribute to making extrusion-based bioprinting
more reliable and reduce the need for extensive parameter screenings in process development.
Overall, the present work provides a toolbox for the printing of biocatalytically active materials.
Novel inks with individually adapted printing methods and analytical techniques are presented. The
application of emulsion-based inks demonstrates the wide range of materials that can be applied
in combination with enzymes despite not being suitable matrices in tissue engineering. Material
screenings can be accelerated by employing microplate-based activity assays and the adaptation of
the printing setup specifically for each type of ink shows the need for fine-tuning between ink and
printing method. The more universal approach to improve reproducibility in bioprinting using a
PID-based pressure control could be valuable for applications outside the scope of biocatalysis.
