Content uploaded by Mario Rothbauer
Author content
All content in this area was uploaded by Mario Rothbauer on Jun 02, 2016
Content may be subject to copyright.
Registered charity number: 207890
Featuring Work from the Ertl Lab in the Cell Chip Group at
Austrian Institute of Technology, Vienna, Austria.
Title: Multi-layered, membrane-integrated microfluidics based
on replica molding of a thiol–ene epoxy thermoset for
organ-on-a-chip applications
Simple, versatile and robust fabrication of multi-layered,
membrane-integrated microfluidics using the biocompatible
thermoset OSTEMER enables the development of customized
tissue-, organ and body-on-a-chip systems capable of
mimicking tissue interfaces and biological barriers.
As featured in:
See Peter Ertl et al. Lab Chip, 2015,
15, 4542.
www.rsc.org/loc
Lab on a Chip
PAPER
Cite this: Lab Chip,2015,15,4542
Received 28th August 2015,
Accepted 22nd October 2015
DOI: 10.1039/c5lc01028d
www.rsc.org/loc
Multi-layered, membrane-integrated microfluidics
based on replica molding of a thiol–ene epoxy
thermoset for organ-on-a-chip applications†
Drago Sticker,‡Mario Rothbauer,‡Sarah Lechner, Marie-Therese Hehenberger and
Peter Ertl*
In this study we have investigated a photosensitive thermoset (OSTEMER 322-40) as a complementary
material to readily fabricate complex multi-layered microdevices for applications in life science. Sim-
ple, versatile and robust fabrication of multifunctional microfluidics is becoming increasingly important for
the development of customized tissue-, organ- and body-on-a-chip systems capable of mimicking tissue
interfaces and biological barriers. In the present work key material properties including optical properties,
vapor permeability, hydrophilicity and biocompatibility are evaluated for cell-based assays using fibroblasts,
endothelial cells and mesenchymal stem cells. The excellent bonding strength of the OSTEMER thermoset
to flexible fluoropolymer (FEP) sheets and polyIJdimethylsiloxane) (PDMS) membranes further allows for the
fabrication of integrated microfluidic components such as membrane-based microdegassers, microvalves
and micropumps. We demonstrate the application of multi-layered, membrane-integrated microdevices
that consist of up to seven layers and three membranes that specially confine and separate vascular cells
from the epithelial barrier and 3D tissue structures.
Introduction
Cell-based in vitro assays are a vital research tool for biomedi-
cal scientists, tissue engineers, biologists, and pharmacolo-
gists. Despite their routine application, current cell-based
assays still do not address the fact that the cellular microenvi-
ronment can gravely influence how cells physiologically
respond and function in vitro.
1,2
It is well known that cell
morphology, protein expression, differentiation, migration,
functionality and cell viability differ distinctly in the presence
of varying media components and surface modifications and
between 2D and 3D environments.
3
Furthermore, close cell-
to-cell contact and proper spatial orientation between cells is
an important issue when using co-cultures and stem cells to
model complex tissue analogues and organoids.
4
To prevent
altered phenotypes and reduced functionality, microfluidic
cell culture systems have been developed to analyse multiple
tissues or organoids of distinct origin under physiologically
relevant ‘in vivo-like’culture conditions.
5
These so called
‘organs-on-a-chip’can potentially be interconnected within a
single microdevice using microchannels and microvalves to
simulate the complex interplay of multiple tissue or organ
types with an intact circulatory system, establishing a ‘body-
on-a-chip’. Such microfluidic cell culture systems provide con-
stant nutrient supply, waste removal, and controlled stimuli
(e.g. drug administration, shear force) in the presence of rele-
vant temperature and gas gradients.
6,7
Another fundamental
aspect in developing organ-on-a-chip systems is concerned
with the integration of multiple organ functions within a sin-
gle microdevice to fully recapitulate organ–organ interactions.
In other words properly scaled geometries need to be
employed for each different organ compartment according to
organ function level and tissue and organ structure as well as
fluid volumes and gas exchange rates.
8–10
It is important to
note that recreating a complex biology requires a
multifunctional design strategy that accounts for cell trap-
ping, culturing, and administration of compounds (e.g. nano-
materials, pharmaceuticals) as well as the application of phys-
iological cues.
11–13
Since physiological cues can range from
defined fluid shear forces to mechanical stretching and bend-
ing to biochemical and electrical stimulation as well as spatial
orientation of cells,
14,15
higher level system architectures are
involved to operate these microfluidic cell culture systems. To
accommodate the added functionality membrane-integrated
microfluidic cell culture systems have been recently intro-
duced to integrate fluid handling using valves and micro-
pumps and to separate cell cultures using integrated
4542 |Lab Chip,2015,15,4542–4554 This journal is © The Royal Society of Chemistry 2015
BioSensor Technologies, AIT Austrian Institute of Technology GmbH,
Muthgasse 11, 1190 Vienna, Austria.
E-mail: peter.ertl@ait.ac.at; Web: www.cellchipgroup.com
†Electronic supplementary information (ESI) available. See DOI: 10.1039/
c5lc01028d
‡Both authors contributed equally to the manuscript.
Lab Chip,2015,15,4542–4554 | 4543This journal is © The Royal Society of Chemistry 2015
membranes.
16,17
Despite the benefits of using membrane-
integrated microfluidic cell culture systems,
18–20
to date auto-
mation, integration, miniaturization and assay parallelisation
of multi-layered, membrane-integrated devices is still in its
infancy.
To foster the development of next generation advanced
in vitro cell culture systems simple and reliable fabrication
methods that allow rapid prototyping of multi-layered,
membrane-integrated microfluidic devices need first to be
established.
21
In general, microdevice fabrication involves a
variety of methods including etching, micromachining,
photolithography and e-beam lithography, hot embossing,
molding, and laser photo-ablation as well as more recently
developed 3D printing techniques.
22,23
While the choice of
fabrication method is guided by multiple factors including
available infrastructure, fabrication throughput and costs per
device, the desired feature resolution, precision and material
is selected by biocompatibility and size requirements of the
biology under investigation. Initially, microfabrication
focused on glass, silicon and hybrid devices since the tech-
niques applied originated from the semi-conductor industry
and material properties were convenient especially for cell
culture and analysis applications.
24
More recently, replica
molding has been introduced as an attractive alternative that
can be applied to many biocompatible materials including
polyIJdimethylsiloxane) (PDMS), hydrogels, thermoset compos-
ites and thermoplastics.
25–30
Replica molding techniques can
also be used for scale-up production and high-throughput
fabrication using hot embossing and injection molding of
various polymers including polymethylmethacrylate (PMMA),
polycarbonate (PC), cyclic olefin copolymer (COC), polysty-
rene (PS), and polyethyleneterephthalate (PET).
30–34
Among
these, replica molding of PDMS, also called soft lithography,
is most frequently used by scientists to rapidly fabricate
membrane-integrated microfluidic cell culture systems at
low costs.
19,35,36
An important aspect concerning the inte-
gration of membranes between two microchannels is
sealing, which can be readily accomplished with PDMS by
pressure (reversible bonding) using van der Waals interac-
tion and by plasma activation or chemical bonding using a
solid-to-solid interfacial poly-addition reaction.
37
Alternative
back-end processing techniques are based on ‘gluing’proce-
dures to integrate membranes (e.g. polycarbonate, polyester,
PDMS) within three-dimensional PDMS and PDMS-hybrid
devices for application as degassers, actuators and cell cul-
ture devices. Despite the apparent advantages of low cost fab-
rication, simple bonding procedures, high gas permeability,
and good optical properties,
38–40
PDMS comes with several
limitations including high compliance resulting in channel
deformation, high water-vapour permeability leading to bub-
ble propagation and osmolality shift, and high permeability
for small hydrophobic molecules as well as the presence of
uncured PDMS oligomers affecting cellular physiology and
function.
41–43
To overcome the limiting material properties of PDMS, we
have investigated an alternative thermoset plastic that can be
readily used to rapidly fabricate microdevices containing inte-
grated actuators such as valves, pumps and degassers as well
as multi-layered, membrane-integrated cell-on-a-chip and
organ-on-a-chip devices. The main strength of the photosensi-
tive thermoset OSTEMER is the versatile and robust bonding
ability combined with well-established soft lithography pro-
cessing methods.
44–46
Although similar to PDMS fabrication
the resulting microdevice consists of a non-elastomeric and
non-gas-permeable material with excellent chemical resis-
tance.
47
For instance, the OSTEMER precursor contains three
structure giving components (e.g. allyl, thiol and epoxy mono-
mers) and two initiators and is polymerized in a two-step pro-
cess. The first curing step is initiated by UV light which trig-
gers the polymerization reaction between allyl–thiol
monomers. Due to its off-stoichiometric ratio of thiol over
allyl groups, an excess of reactive thiol groups, along with
unreacted free epoxy monomers, is present in the bulk mate-
rial and the surfaces, which are needed for follow-on assem-
bly and bonding procedures. At this point, the polymer is still
elastic with a storage modulus of 3.2 MPa and can be
removed from the master and transferred to any substrate.
Next, curing of the remaining thiol and epoxy groups is ther-
mally accelerated, resulting in robust and versatile bond-
ing
45,48
that creates a polymerised stiff material with a stor-
age modulus of 2.3 GPa.
44
Even though the bulk material
price of OSTEMER is twice as much as that of PDMS, cost-
effective fabrication of microdevices can still be accom-
plished by either reducing material consumption in small-
scale fabrication or by increasing production volume using
injection molding techniques.
In the present study, the suitability of the photosensitive
thermoset OSTEMER is investigated for applications in
life science and its biocompatibility, autofluorescence and
bonding properties are compared to those of existing cell
chip materials. In addition to the physical, chemical and
biological characterization of the OSTEMER thermoset, we
have combined replica molding with microinjection molding
techniques as described in Fig. 1 to fabricate valves,
degassers and pumps as well as highly sophisticated cell-
based microdevices composed of up to three membranes and
four vertically aligned microchannels within a single device
architecture. The practical application of the fabricated
multi-layered, membrane-integrated microdevices is demon-
strated using various in vitro models employing two- as well
as three-dimensional on-chip cultures of adipose-derived
mesenchymal stem cells (adMSCs), human umbilical vein
endothelial cells (HUVECs), and placental epithelial chorio-
carcinoma cell lines (ACH-3P and BeWo ATCC).
Experimental
Fabrication of dry film resist and PDMS master molds
The epoxy-based dry film resist (DFR) TMMF S2045 (Tokyo
Ohka Kogyo Co., Ltd) was used for microfabrication of mas-
ter molds for PDMS soft lithography. Prior to lamination sili-
con wafers were sonicated in 2% Hellmanex III solution
Lab on a Chip Paper
4544 |Lab Chip,2015,15,4542–4554 This journal is © The Royal Society of Chemistry 2015
(Hellma Analytics), ddH
2
0 and isopropanol for 10 min at 30
°C. The DFR was laminated under heat to the wafer using a
HeatSeal H425 A3 office laminator (GBC) to achieve 45 μm
and 90 μm high structures. Following lamination, a polymer
film mask (Photo Data Ltd, UK) was applied and the wafer
was exposed to UV light using a mask aligner (EVG 620). For
TMMF crosslinking, a postexposure bake (PEB) was
performed at 90 °C for 5 min. The resist was then developed
in EBR solvent (PGMEA/1-methoxy-2-propyl-acetate, Micro-
Chemicals) under magnetic stirring until non-crosslinked
TMMF was completely removed from the wafer surface (typi-
cally after around 90 s) followed by isopropanol and ddH
2
0
rinsing. Next, the developed and N
2
blow-dried photoresist
structure was hard-baked in an oven at 200 °C for 1 h. To
avoid adhesion of OSTEMER, the DFR/silicon wafers were
spin-coated with 0.5% Teflon AF (60151-100-6, Dupont) in
Fluorinert FC-40 solution (F9755, Sigma Aldrich) at 3000 rpm
for 60 s and baked for 60 s at 125 °C and 5 min at 175 °C
prior to usage.
Fabrication of PDMS master molds commenced with
mixing Sylgard 184 and curing agent (Dow Corning) in 10 : 1
mass ratio and then pouring onto the cleaned DFR master
mold to form a 1 mm thick PDMS layer overnight at 70 °C.
Fabrication of microchannel structures above 100 μm was
accomplished using in-house micromachined acrylonitrile
butadiene styrene (ABS) polymer plates featuring 350 μm high
structures. In all cases, the 1 mm thick PDMS replicas were
gently delaminated from the underlying master molds and
carefully cleaned using detergent, isopropanol and ddH
2
0,
and dried at 70 °C for 30–60 min in an oven. Prior to microin-
jection molding of OSTEMER, a PDMS microfluidic mold was
fabricated and after delamination, two injection ports, each
with a diameter of 3 mm, were punched into both ends of the
microfluidic channel (Fig. 1).
OSTEMER preparation and replica molding
For all devices OSTEMER 322-40 (Mercene Labs AB, Sweden)
was used. The two components were weighed into a glass
Fig. 1 Schematic illustration of the versatile OSTEMER microfabrication process based on either DFR or PDMS master molds. For fabrication of
freestanding microstructures on porous membranes, microinjection molding was used using microfluidic PDMS molds. While the first UV polymerization
step results in a rubbery PDMS-like product, the final thermal curing step initiates bonding and generates hard polymeric microdevices.
Lab on a ChipPaper
Lab Chip,2015,15,4542–4554 | 4545This journal is © The Royal Society of Chemistry 2015
container under a fume hood according to the ratio specified
by the manufacturer. For homogenization of the viscous com-
ponents the container was vortexed for approximately 3 min
and thereby emerging bubbles were removed by degassing
for approximately 20 min. UV curing was performed with 365
nm wavelength Hg-tubes in a crosslinker equipped with an
integrated energy irradiation sensor (Bio-link BLX
Crosslinker, Vilber Lourmat).
Replica molding of OSTEMER microfluidics involved casting
of the polymer mix using either DFR or PDMS master molds as
indicated in Fig. 1 (Casting). A conventional PET transparency
film (3M) was used between PDMS (250 μm) or glass (1.1 mm)
spacers to adjust the total height of the microfluidic layers. Fol-
lowing UV exposure at a dose of 300–700 mJ cm
−2
,the
OSTEMER device was gently delaminated from the master
mold, cleaned of unreacted monomers in a stream of ethanol,
dry boosted and cut into shape. As needed, access holes were
drilled using a conventional bench top drill press followed by
extensive cleaning with ethanol. For membrane-free micro-
devices, the OSTEMER replicas were aligned, clamped and
fixed in place and baked at 110 °C overnight to ensure that
all layers were bonded properly to each other. In case of
mono-membrane microdevices a porous Whatman®track-
etched polycarbonate membrane (GE Lifescience) was layered
in between replica molded OSTEMER microfluidic parts prior
to bonding. Access holes of 1 mm diameter were punched at
the inlet points using a biopsy puncher.
For microinjection molding (Fig. 1), porous Whatman®
track-etched polycarbonate membranes with pore sizes of 0.4
μm and 3 μm (GE Lifescience) were first laminated on 250
μm thick PDMS sheets (HT-6240, Silex, UK). Next, a 200 μl
drop of OSTEMER was transferred to one inlet and gravimet-
rically injected into the microchannel. Following UV exposure
at a dose of 300 mJ cm
−2
, the PDMS layer was gently
delaminated from the porous membrane, trimmed to shape
and aligned as an intermediate layer in between replica
molded OSTEMER layers. Assembly and bonding of the mul-
tiple layers was performed as described above.
Autofluorescence measurements
From a 24-well plate (NUNC930186, VWR) the bottom polysty-
rene (PS) was removed and 1 mm thick samples of OSTEMER
322-40, COC (Topas 8007, microfluidic ChipShop GmbH) and
glass (631-1550, VWR) were inserted, while the bottom PS
was analyzed too. Optical scans were performed with a Tecan
Infinite®M1000a plate reader averaging 10 flashes with 40
μs integration time each.
Fabrication of microdegassers
The integrated degasser consisted of a microfluidic channel
separated by a PDMS membrane from the pneumatic
layer. Following fabrication of the OSTEMER fluidic layer
using DFR master molds a silanized PDMS sheet was
layered between the UV-cured OSTEMER microchannels (2 h
at 70 °C), clamped together and covalent linkage was
accomplished after baking at 100 °C for 1 h. Silanization of
the activated PDMS (corona treatment) was carried out
using a 5% solution of 3-mercaptopropyltrimethoxysilane
(MPTS) (AB111219, abcr GmbH, Germany) in absolute etha-
nol for 1 h at room temperature (RT). Silanization using 5%
(3-aminopropyl)triethoxysilane (APTES) (A3648, Sigma
Aldrich) for evaluation of bonding strength was performed
in the same manner. A Tesla coil (BD-20 V, ETP, USA) was
used for the corona discharge.
Fabrication and characterisation of micropumps
90 μm high pneumatic and fluidic layers were replica
molded using a DFR master following UV curing and dril-
ling, and the layers were sandwiched between a 50 μm thick
fluoroplastic film (200C-20, Teflon, DuPont). Access holes
for the fluid in- and outlets were punched into the film
using a biopsy puncher and the aligned fluidic–pneumatic
film assembly was fixed using an aluminum manifold. The
manifold was used for pneumatic control, exhibiting a nega-
tive pressure of −75 kPa during 72 h of curing at RT to pre-
vent bonding of the film onto the fluidic gate structure (see
Fig. 5a). Pneumatic actuation was controlled via solenoid
valves (S070M-6 DC-32, SMC, Austria) using Lab-View soft-
ware (National Instruments, Austria). Flow velocities were
determined using a flow meter (SLI-1000, Sensirion,
Switzerland).
Cell culture
Adipose tissue-derived human mesenchymal stem cells
(adMSCs) were kindly provided by Prof. K. Kasper (BOKU
Vienna). Cells were cultivated using alphaMEM (12000-063,
Gibco) medium supplemented with 10% FCS (A15-101, PAA)
and 0.5% gentamicin (PAA, P11-004). Cell culture mainte-
nance commenced following 80% confluence, where cells
were enzymatically detached using Accutase (A6964, Sigma
Aldrich) and split in a 1 : 3 ratio. Induction of stem cells was
performed using differentiation medium (130-091-678,
StemMACS™, Miltenyi Biotec) supplemented with 1 ng ml
−1
of IL1-β(I9401, Sigma Aldrich). Confirmation of stem cell dif-
ferentiation was performed after fixation of cells with 4%
formaldehyde (F8775, Sigma Aldrich) for 30 min and subse-
quent rinsing with ddH
2
O, and staining of osteocytes using
50 mM Alizarin Red S (0348.2, Roth) dissolved in ddH
2
O and
pH adjusted to 4.3 with HCl.
The employed vascular model, kindly provided by W.
Holnthoner (LBI, Vienna), consisted of adMSCs and
lentivirally transduced GFP-HUVECs purchased from Olaf
Pharmaceuticals (Worcester, USA). adMSCs were cultivated in
untreated 75 cm
2
cell culture flasks using endothelial growth
medium (EGM-2, Lonza, Switzerland) and split upon
reaching 80% confluence. In turn, HUVECs were cultivated in
gelatin-coated 25 cm
2
cell culture flasks and maintained in
EGM-2. The coating was carried out using 1% gelatin
solution (9000-70-8, Sigma Aldrich) in DPBS and was incu-
bated for 30 min at 37 °C. NIH-3T3 fibroblasts (ATCC 59,
Lab on a Chip Paper
4546 |Lab Chip,2015,15,4542–4554 This journal is © The Royal Society of Chemistry 2015
DMSZ, Germany) were cultivated in uncoated 75 cm
2
cell cul-
ture flasks (VWR, Austria) using Dulbecco's modified Eagle
medium (DMEM) containing high glucose culture medium
with stable glutamine (E15-101, PAA) supplemented with
10% fetal calf serum (FCS) (A15-101, PAA) and 0.5% gentami-
cin (P11-004, PAA). BeWo ATCC epithelial choriocarcinoma
cells, kindly provided by Tina Bürki (EMPA, Switzerland),
were maintained in RPMI 1640 medium supplemented with
10% FCS. ACH-3P epithelial choriocarcinoma cells, kindly
provided by Prof. Berthold Huppertz (Medical University Graz,
Austria), were maintained in DMEM/Ham's F12 medium
mixed in a ratio of 1 : 1 and supplemented with 10% FCS. All
cells were cultivated in a humidified atmosphere at 37 °C and
5% CO
2
atmosphere. For viability/cytotoxicity/nuclei staining 1
μM Calcein AM, 1 μM EthD-1 (L-3224, Thermo Fisher) and 10
μM Hoechst 33342 (40047, Biotium) in full medium were
applied for 30 min. Prior to EthD-1 staining NIH-3T3 cells
were treated with 1% Triton X for 10 min.
Preparation of cell-laden hydrogels
Hydrogel-based 3D co-cultures were prepared using fibrin gel
(Baxter, Austria) by warming fibrinogen to RT and diluting 4
Uml
−1
thrombin in calcium chloride (CaCl
2
) to 1.6 U ml
−1
.
For the co-culture model cells, fibrinogen and thrombin (0.2
Uml
−1
) were mixed with fibrin gels (25 mg ml
−1
final concen-
tration) containing 500 000 adMSCs or HUVECs and inserted
into the microchambers. adMSCs were prior stained with red
CM-Dil cell tracker dye (C-7001, Life Technologies) by adding
3mlof8μM dye solution to a 75 cm
2
cell culture flask and
incubating for 10 min at 4 °C and 30 min at 37 °C. Following
injection of the cell-laden hydrogels into the microfluidic
Fig. 2 Biocompatibility evaluation of OSTEMER. a) Viability assay (MTT) of epithelial cells (H441), mesenchymal stem cells (adMSCs) and fibroblasts
(NIH-3T3) cultured on gelatin-coated OSTEMER discs and in polystyrene well plates (n= 4). b) Adhesion progress of adMSCs on plain OSTEMER
and standard polystyrene well plates 4 h and 17 h post-seeding. c) Osteogenic differentiation of adMSCs within OSTEMER/TiO
2
-glass microfluidic
channels. Matrix mineralization was confirmed by oil red staining (orange) 11 days post-induction. d) Cell morphology is visualized by F-actin
staining (green) of NIH-3T3 cells and adMSCs (red) 48 h post-seeding on glass (left), gelatin-coated OSTEMER (center) and plain OSTEMER (right).
For adMSCs nuclei were stained with DAPI (blue).
Lab on a ChipPaper
Lab Chip,2015,15,4542–4554 | 4547This journal is © The Royal Society of Chemistry 2015
channels the hydrogel was allowed to polymerize at 37 °C for
15 min, while injection ports were sealed off using a qPCR
foil. For the membrane-free co-culture model 90 μm high
microchannels were cast, UV-cured, peeled off from the
mold, drilled with access holes and extensively rinsed with
ethanol. The elastic thermoset was then applied onto a COC
plate and fibrin containing adMSCs were inserted through
the access holes. After 10 min of incubation at 37 °C the ther-
moset was gently peeled off from the COC substrate and the
top perfusion layer was aligned, filled with DPBS and bonded
at 37 °C overnight.
Cell viability
Biocompatibility studies involved the assessment of the meta-
bolic activity of different cell lines cultured on OSTEMER
substrates residing at the bottom of the chamber slides
(178599, Lab-Tek) and cell culture substrates, while
OSTEMER was baked for 64 h at 50 °C. All chambers were
extensively rinsed with ethanol and coated with a 2% gelatin
solution in DPBS for 30 min at 37 °C. Following washing with
DPBS, NIH-3T3, adMSCs and H441 cells were seeded to 50%
confluence in quadruplicates and incubated over a period of
48 h. The supernatant was discarded, and the cells were
washed and incubated for 4.3 h at 37 °C/5% CO
2
using 140
μl per well of 0.5 mg ml
−1
thiazolyl blue tetrazolium bromide
(MTT) (M5655, Sigma Aldrich) solution dissolved in cell cul-
ture medium. To solubilize the formed crystals 130 μl of 10%
sodium dodecyl sulfate (SDS) (A0676, Applichem, Germany)
supplemented with 0.1% of 10 M hydrochloric acid (HCl)
(H1758, Sigma Aldrich) was added to each well and incubated
overnight. Aliquots were transferred to a 96-well plate and
Fig. 3 OSTEMER material characterisation. a) Autofluorescence frequency scan of polymerized OSTEMER at a resolution of 20 nm and 10 nm for
excitation and emission, respectively. The intensity values were normalized to glass. b) Comparison of the autofluorescence intensity of OSTEMER
to frequently used cell culture materials. Selected excitation/emission frequency pairs represent absorption/emission maxima of standard
fluorophores used in cell-based assays (TRITC, FITC and DAPI). c) LIVE/DEAD staining as well as nuclei staining (Hoechst) of NIH-3T3 cells on
OSTEMER 24 h post-seeding. Pictures were obtained using a standard fluorescence microscope with TRITC, FITC and DAPI filter blocks.
Fig. 4 a) Qualitative comparison of water vapour permeability of 45
μm high OSTEMER and PDMS microchannels 4 h after injection of 1.4
μl of water and incubation at 40 °C. b) Representative image of the
hydrophilic polymer surface at the air/water interface of full polymeric
OSTEMER microchannels.
Lab on a Chip Paper
4548 |Lab Chip,2015,15,4542–4554 This journal is © The Royal Society of Chemistry 2015
the optical density was analyzed at 570 nm. The arithmetic
mean from cell-free wells (blanks, n= 4) was subtracted from
the value from each cell-containing sample (n= 4) and STD
was normalized accordingly.
Immunohistochemistry
Cells were fixed in 4% glutaraldehyde (340855, Sigma
Aldrich) and permeabilized in 0.2% Triton X-100 solution
(X100, Sigma Aldrich) for 20 min and 5 min, respectively. To
reduce unspecific antibody binding a blocking solution of
5% w/w human serum albumin (HSA) (A1653, Sigma Aldrich)
in DPBS was added for 30 min at RT. Next Alexa Fluor 488
phalloidin (A12379, Life Technologies) or TRITC-conjugated
phalloidin (90228, Millipore) solution in 1% HSA/DPBS at a
concentration of 6.6 μM was incubated for 1 h at RT and
0.4 μgml
−1
DAPI (90229, Millipore) in DPBS was added for 5
min. E-cadherin staining of ACH-3P cells was performed by
fixation using 2% paraformaldehyde (158127, Sigma Aldrich)
for 10 min at RT, permeabilization with 0.2% Triton X-100
for 10 min at RT and blocking in 5% goat serum (G9023,
Sigma Aldrich) for 1 h. Finally anti-CD324 Alexa488 (14-3249,
eBioscience) diluted 1 : 50 in 0.5% HSA/DPBS was incubated
for 3 h at 4 °C. Samples were embedded in Vectashield®
mounting medium (Thermo Scientific) prior to analysis.
Between all steps samples were washed three times in DPBS.
Fluorescence microscopy
Fluorescence images were taken using a TE2000-S inverted
fluorescence microscope (Nikon) equipped with a DS-Qi1Mc
digital camera. All conventional fluorescence images were
processed using NIS-Elements software (Nikon). In turn,
CLSM imaging was performed using a Leica TCS SP5 II sys-
tem (Leica). Images were recorded with a 63×oil immersion,
a20×water immersion and a 10×objective using the manu-
facturer's LAS AF imaging software.
Scanning electron microscopy
Sectioned LOC devices were mounted at a 45°angle on alu-
minum stubs (Christine Gröpl Elektronenmikroskopie) and
coated with a 5 nm layer of gold using an EM SCD005 sputter
Fig. 5 Establishment of OSTEMER microvalves, -pumps and -degassers for on-chip fluid handling. a) Representative image of an OSTEMER chip
containing a microdegasser in-line with a micropump (three microvalves) and exploded schematics for pneumatic and fluidic layers. The fluidic
channel is highlighted in blue, while the pneumatic control lines for the micropump are depicted in orange. b) Flow velocity of OSTEMER micro-
pumps during a single pumping cycle applying a pressure of ±70 kPa and a delay of 800 ms between each actuation step. The inlay shows a valve
actuation and fluid displacement (orange) sequence with closed (crossed circle) and open microvalves (blank circle). c) Net flow velocity for
increasing actuation pressure at constant negative pressure of 70 kPa. Values are expressed as integrated flow velocity divided by time.
Lab on a ChipPaper
Lab Chip,2015,15,4542–4554 | 4549This journal is © The Royal Society of Chemistry 2015
coater (Leica). SEM imaging was performed using an Inspect
S50 scanning electron microscope (FEI) at 15 kV, spot size of
3 and high vacuum mode.
Results and discussion
Cell culture performance of OSTEMER
A crucial aspect of any cell culture maintenance and cell-
based assay is evaluating the influence of surface properties
on cell viability and function. Consequently, to assess the
biocompatibility of OSTEMER as a suitable cell culture mate-
rial, the viability, proliferation, cell adhesion characteristics
and morphology alterations of epithelial (H441), fibroblast
(NIH-3T3) and mesenchymal stem cells (adMSCs) are investi-
gated in subsequent experiments. Results of a comparative
analysis are shown in Fig. 2 highlighting the similarities
between standard cell culture substrates and OSTEMER ther-
moset. For instance, in addition to identical proliferation
rates adMSCs (visual inspection, data not shown), cell viabil-
ities shown in Fig. 2a using three different cell cultures of
H441, adMSCs and NIH-3T3 revealed similar metabolic activi-
ties between gelatin-coated OSTEMER substrates and glass
slides. Interestingly, higher viability is observed in the pres-
ence of adMSCs, which can be explained by a higher meta-
bolic activity on OSTEMER. To further analyse the effect of
OSTEMER on stem cell cultivation, adMSC adhesion kinetics
using time-lapse microscopy was used to compare dynamic
cell–substrate interactions. A series of images taken over a
period of 17 h revealed similar adhesion dynamics between
untreated OSTEMER and polystyrene substrates (Fig. 2b).
Additionally, the ability of adMSCs to undergo osteogenic dif-
ferentiation within an OSTEMER-based microfluidic device
has been investigated. Fig. 2c shows an image of Alizarin red
stained adMSCs (at day 11) indicating on-chip matrix miner-
alization and formation of osteoblasts. These results are simi-
lar to differentiation rates obtained using standard polysty-
rene well-plates (data not shown).
49
Since surface properties
can directly influence anchorage dependent cell cultures,
additional morphological analyses are conducted in subse-
quent experiments. Therefore cytoskeletal structures of
adMSCs and NIH-3T3 grown on untreated OSTEMER, gelatin-
coated OSTEMER and gelatin-coated glass are investigated by
F-actin staining. Fluorescence images shown in Fig. 2d dis-
play appropriate actin filament development and adhesion
characteristic for both cell types using OSTEMER culture sub-
strates. As an example visualized stress fibres of NIH-3T3 are
disorganized in a crisscross manner, while those of adMSCs
are robust and elongated irrespective of the surface material,
indicating similar cell–substrate interactions.
Material properties of OSTEMER for cell culture analysis
An important aspect of plastic materials for microfluidic cell
cultures is concerned with autofluorescence, hydrophilicity
and water vapour permeability, because these material prop-
erties are known to influence the quality of analysis, wettabil-
ity and bubble formation. An initial fluorescence scan
probing relevant frequency regions of fluorophores used in
cell biology (e.g. excitation wavelength 300 to 640 nm and
emission from 300 to 800 nm) is shown in Fig. 3a and reveals
a broad photoluminescence emission at wavelengths below
400 nm, which is characteristic for most plastic materials.
50
This result is in good agreement with recently published
data, which shows UV light absorption of OSTEMER below
380 nm.
44
However, a direct comparison to common mate-
rials used for microdevices including glass, COC and polysty-
rene showed luminescence emissions which are 3, 1.8 and 4
times higher compared to polystyrene at certain excitation/
emission maxima pairs of most prominent fluorophores used
in cell culture (TRITC 550/570 nm, FITC 500/530 nm and
DAPI 360/450 nm), respectively. Despite the inherent elevated
autofluorescence of OSTEMER, fluorescence images (Fig. 3c)
taken 24 h post-seeding of NIH-3T3 fibroblasts using the
well-known LIVE/DEAD (calcein AM and EthD-1) and nucleus
stain (Hoechst) resulted in excellent picture quality when
using a conventional fluorescence microscope.
Although often neglected in recent literature, bubble for-
mation within microchannels poses a serious problem for
most microfluidic cell cultures, since the air–liquid interface
of a moving microbubble can effectively remove and injure
cells. Since bubbles are preferentially formed at the surface
of gas-transparent materials, the water vapour permeability
of OSTEMER was investigated in subsequent experiments. To
assess the water vapour permeability, two 45 μm high
OSTEMER microchannels were filled with water, sealed and
allowed to rest for 4 h at 40 °C. While no visible water evapo-
ration took place using OSTEMER microchannels, for PDMS
devices with the same geometry (Fig. 4a) more than half of
the water volume evaporated in the same time period with
37.9% (left PDMS channel) and 29.4% (right PDMS channel)
remaining volume. Consequently, the low water vapour per-
meability of OSTEMER (0.15 g mm
−1
per 24 h per m
2
)
51
allows for bubble-free cell culture handling over long periods
of time, which can be challenging for PDMS based micro-
devices. Another crucial parameter in long-term cell cultures
and organ-on-a-chip applications is the applied oxygen level.
Since thiol–ene polymers are reported to exhibit low oxygen
permeability that is similar to other frequently used poly-
meric materials,
52
adequate oxygen tensions need to be
maintained by adjusting medium flow using active liquid
handling protocols. In addition to high biocompatibility,
acceptable autofluorescence and low vapour permeability, the
ability to readily fill polymeric microchannels with aqueous
solutions is also an important criterion for microfluidic cell
cultures. It is important to note that a high energy surface of
an untreated polymer is beneficial for the generation of opti-
mum cell culture conditions, since hydrophilic surfaces are
known to promote direct cell adhesion (see also Fig. 2d).
Fig. 4b shows an image of a water–air interface taken after
pressure-free filling of a microchannel, thus demonstrating
the strong capillary forces and inherent low interfacial ten-
sion of OSTEMER surfaces (see also the ESI; Video S1†).
Although material characterisation did reveal increased
Lab on a Chip Paper
4550 |Lab Chip,2015,15, 4542–4554 This journal is © The Royal Society of Chemistry 2015
background fluorescence, the observed high degree of wetta-
bility within microfluidic channels and apparent low water
vapour permeability during storage and operation make
OSTEMER an attractive material for microfluidic cell
cultures.
Integrated microfluidic components including valves, pumps
and degassers
Although crucial for cell stimulation and maintenance, active
flow control and precise sample delivery of nanoliter volumes
for microfluidic cell culture systems are still not widely
applied. Since microfluidic cell culture systems often require
extended operating periods (e.g. days to weeks), robust and
reliable on-chip fluid handling systems are needed. A critical
parameter for long-term operation of membrane-integrated
microactuators is bonding strength between flexible poly-
meric membranes and the microfluidic substrate. To evaluate
the feasibility of OSTEMER for fabrication of microactuators,
its bonding strength to different flexible membranes was
tested using a 1 mm thick OSTEMER slide with Ø 1 mm
holes covered by FEP and PDMS membranes. Table 1 lists
applied pressures at which polymeric sheets started to delam-
inate from the OSTEMER surface in the presence of varying
surface modifications and operation temperatures. While
corona-treated and reversibly attached PDMS membranes
delaminated already at 20 kPa at RT, covalently bonded mem-
branes withstood pressures up to 200 kPa. In turn, using the
chemically inert, low gas and water vapour permeability FEP
sheet no delamination was observed up to the maximal
applied pressure of 400 kPa. The observed high bonding
strength between OSTEMER and FEP sheet is more than five
times higher than needed to actuate pneumatic valves and
therefore was chosen for fabrication of robust micropumps.
Following the initial assessment of bonding strength, on-
chip micropumps and degassers were fabricated and subse-
quently tested for cell culture applications. The valve design
used to fabricate micropumps consisted of a fluidic layer
containing an integrated gate,
53
which is separated from the
pneumatic layer using a PDMS membrane as shown in
Fig. 5a. Using a four-step actuation procedure (see inset of
Fig. 5b), a series of three valves can be operated as a pump,
where fluid displacement volumes of 96 nl per pump stroke
are defined by the geometry of the middle valve. In addition
to actuation pressures, the step times between valves can also
be controlled to adjust net flow velocities, peak shear forces
and peristaltic profiles. Results shown in Fig. 5c show a
linear increase of net flow velocity in the presence of increas-
ing actuation pressures resulting in physiologically relevant
flow velocities. As indicated in the above section, an impor-
tant aspect of microfluidic cell cultures is the robustness and
durability of the fluid handling systems as well as the elimi-
nation of microbubbles. To prevent any moving bubbles from
reaching the cell culture chamber, an in-line degasser was
integrated downstream of the micropump as depicted in
Fig. 5a. The application of a relative pressure of −60 kPa
across the serpentine shaped microchannel using the in-line
degasser removed air with a velocity of about 180 nl min
−1
.
In a final attempt to evaluate robustness and durability, the
OSTEMER-based micropump was continuously operated at
RT over 28 days with a total transferred water volume of 0.5
litres. Results of the study demonstrated continuous opera-
tion without loss of functionality and bubble formation at
elevated temperatures. In other words, reliable and robust
pneumatically controlled fluid handling systems for auto-
mated microfluidic cell culture handling can be simply fabri-
cated using the OSTEMER thermoset.
Multi-layered cell-on-a-chip and membrane-integrated organ-
on-a-chip systems
The practical applications of OSTEMER-based cell chips are
demonstrated in subsequent experiments using microfluidic
architectures of increasing complexity ranging from
membrane-free single and two-layered devices to single, dou-
ble and triple membrane-integrated, multi-layered micro-
fluidics. Initially membrane-free, single- and double-channel
microdevices are fabricated to validate OSTEMER-based micro-
fluidics for cell-based assays. Fig. 6a shows a schematic and
SEM image of the two-layer microfluidics containing an oval
shaped cultivation chamber to assess the differentiation capac-
ity of stem cells using fluorescence microscopy. Results of
F-actin immunostaining revealed healthy stem cell culture
after a period of 48 h using untreated OSTEMER surfaces (see
also Fig. 2c). Next, a stacked hydrogel-laden, two-channel
microdevice was fabricated to demonstrate 3D cell culture
maintenance, which is crucial for a large range of cell culture
applications. A well-known 3D cell culture model used to study
endothelial sprouting, neo-vascularization and tube formation
is based on fibrin–hydrogel co-culture.
54
Fig. 6b (left) shows a
schematic and SEM image of the device that is filled from the
bottom with a stem cell (adMSC) containing hydrogel and sub-
sequently seeded with endothelial cells to form a layer on top
of the hydrogel. As shown in the side and top view images
(Fig. 6b, right), this configuration is ideally suited to mimic
the vascular tissue interface, since it delivers fluid mechanical
shear forces to endothelial cells, while a 3D tissue-like micro-
environment is provided for the stem cell culture. This cell
chip architecture is possible due to the reactive surface groups
of OSTEMER, which allow efficient bonding at 37 °C, thus
enabling the initial fabrication of a cell–hydrogel filled bottom
layer that is subsequently sealed to a microfluidic top layer
overnight within a cell culture incubator.
Table 1 Bonding strength of PDMS and FEP membranes to OSTEMER.
Pressure applied up to 400 kPa
PDMS FEP
Surface modification MTS APTES w/o w/o
Delamination pressure @ RT [kPa] 200 200 20 *
Delamination pressure @ 37 °C [kPa] 200 200 <10 *
*No delamination observed.
Lab on a ChipPaper
Lab Chip,2015,15,4542–4554 | 4551This journal is © The Royal Society of Chemistry 2015
In addition to membrane-free microfluidic tissue
models, three organ-on-a-chip model systems containing
integrated membranes separating apical and basal culture
compartments and representing tissue biointerfaces with
independently addressable fluid streams are demonstrated
in follow-on experiments. The multi-layered, membrane-
integrated microdevices are fabricated using replica mold-
ing in combination with microinjection molding that fea-
ture up to seven independent layers and up to three inte-
grated porous polycarbonate membranes (see also Fig. 1).
Fig. 7 shows schematic drawings, SEM pictures and fluores-
cent images of (a) single-, (b) double- and (c) triple-
membrane-based microdevices to mimic a simplified
human placental cell barrier, as well as two vascular–tissue–
vascular interfaces, respectively. Fig. 7a shows a single-
membrane device (left) containing HUVECs as well as cho-
riocarcinoma placenta cells as a representative model for
vascular structure and human placenta of the first trimes-
ter, respectively. Following 48 h of incubation both endothe-
lial cells (Fig. 7a, right, HUVEC) and epithelial cells (ACH-
3P) revealed proper cytoskeletal development, cadherin for-
mation and complete intercellular boundaries using immu-
nofluorescent F-actin and E-cadherin stains, respectively.
Expression of E-cadherin reveals the formation of proper
cell-to-cell junctions, thus indicating the presence of a tight
epithelial cell barrier. In our second organ-on-a-chip model a
two-membrane device (Fig. 7b) was fabricated for on-chip
co-cultures of cellularized hydrogels to mimic physiological
endothelial tissue barriers. As a proof of principle, adMSCs
mixed within a fibrin hydrogel were injected between two
porous polycarbonate membranes (top: 0.4 μm, bottom: 3
μm) and cells are visualized using CM-Dil cell tracker dye
(Fig. 7b, right). Additionally, HUVEC cells are introduced
into the top microchannel and propagated over a period of
48 h forming a dense confluent endothelial barrier as indi-
cated by the F-actin stain. This model could be used to
study molecule transport across the barrier by introducing
chemical compounds to channel 1 and analysing media
from channel 3. Our final and most sophisticated multi-lay-
ered, membrane-integrated organ-on-a-chip system is shown
in Fig. 7c where three integrated membranes are used to
separate the epithelial cell barrier from endothelial cells
and stem cell cultures. SEM and fluorescent images clearly
show the three different biological compartments
representing the placental cell barrier separated by a vascu-
lar barrier from a 3D tissue analogue.
Fig. 6 Proof-of-principle application of membrane-free OSTEMER microdevices for in vitro cell-based models. a) Schematic representation of
channel geometry (left) and SEM images of the actual device (center) of OSTEMER microchannels for 2D cell culture. Fluorescent images (right) of
adMSCs grown on plain OSTEMER microchannels stained for F-actin (cytoskeleton; red) and DAPI (blue). b) Schematic representation (left) of a
stacked two-channel membrane-free OSTEMER co-culture device with SEM images of the actual overlapping circle-shaped channel intersections
(center). Fluorescent images (right) of adMSC hydrogel culture (channel 2) stained with CM-Dil cell tracker dye (red) and GFP-HUVECs (green)
grown on top of the hydrogel (channel 1).
Lab on a Chip Paper
4552 |Lab Chip,2015,15,4542–4554 This journal is © The Royal Society of Chemistry 2015
Conclusions
The increasing demand for tissue analogues in pharmacologi-
cal drug screening creates the need for in vivo-like miniatur-
ized, automated and robust tissue and organ models. Due to
the lack of simple and reliable fabrication methods the devel-
opment of customized, complex co-culture in vitro systems is
still in its infancy. Here we report the application of a photo-
sensitive thermoset polymer as a promising tool for the
versatile and simple fabrication of complex biomicrofluidic
devices. For the first time, key material properties of
OSTEMER are evaluated for cell culture applications including
biocompatibility using relevant cell types including fibroblasts,
epithelial cells and mesenchymal stem cells, as well as vapour
permeability and surface hydrophilicity. Although increased
autofluorescence of OSTEMER in comparison to polystyrene
was found in the visible region, excellent fluorescence images
can be achieved using common cell culture fluorophores.
Fig. 7 Proof-of-principle application of multi-layered, membrane-integrated OSTEMER microdevices for in vitro tissue models. a) Schematic (left)
and SEM picture (right) of a single-membrane microdevice used as either an epithelial placental (ACH-3P) or a vascular endothelial (HUVEC) barrier
model. The cytoskeleton of HUVECs was stained for F-actin using phalloidin (green). ACH-3P cells were stained for E-cadherin (green) and nuclei
(blue). b) Schematic (left) and SEM picture (right) of a double-membrane, three-channel device with adMSCs (CM-Dil cell tracker dye) suspended
in fibrin hydrogel and HUVECs separated by a porous membrane representing the proof of concept of an endothelial tissue barrier model. c) Sche-
matic (left) and SEM picture (right) of a triple-membrane, four-channel microdevice applied as an extended placental barrier model system with
co-cultures of three different cell types separated by membranes (dotted lines). Image of a vertical cross section (lower image) of the multi-
layered microdevice filled with adMSCs (CM-Dil cell tracker dye; channel 3) and GFP-HUVEC (green; channel 2) 3D cultures, as well as BeWo ATCC
cells (F-actin, red) cultured on the porous membrane of channel 1.
Lab on a ChipPaper
Lab Chip,2015,15,4542–4554 | 4553This journal is © The Royal Society of Chemistry 2015
Since the excellent bonding strength of OSTEMER to flexi-
ble membranes such as FEP and PDMS, durable and robust
pneumatically actuated membrane-based microdegassers,
microvalves and micropumps can also be fabricated. The
integration of miniaturized fluid handling systems for cell
culture applications is expected to aid automation and stan-
dardization of cell-based microsystems. Practical application
of the investigated thermoset is demonstrated by fabricating
five multi-layered microdevices of increasing complexity
containing up to four independently addressable micro-
channels that can be used to mimic complex tissue interfaces
and organ models. The straightforward fabrication of com-
plex device architectures in addition to the inherent hydro-
philic surface, low water vapour permeability and excellent
bonding properties make OSTEMER an ideal tool for micro-
fabrication of cell-based assays and could therefore bridge
the gap between proof of concept and actual industrial
applications.
Acknowledgements
We thank C. F. Carlborg and F. Saharil (KTH, Micro and
Nanosystems, Sweden) for the first introduction to OSTEMER
microfabrication. We also thank C. Steiniger (AIT), M.
Debreczeny (BOKU, Austria), R. Byrne (MUV, Austria) and I.
Olmos Calvo (AIT) for their valuable technical assistance. The
authors acknowledge M. Purtscher (UAS Technikum Wien,
Austria), W. Holnthoner (LBI, Austria) and P. Slezak (LBI,
Austria) for providing adMSCs, HUVECs and their expertise in
cell-laden hydrogels. We thank A. Weinhäusel for access to
the infrastructure of the Health and Environment Department
(AIT). We thank V. Charwat (BOKU, Wien) and C. Kasper
(BOKU, Wien) for providing adMSCs, T. Bürki (EMPA, Switzer-
land) for BeWo ATCC cells and B. Huppertz (MUG, Austria)
for ACH-3P cells.
References
1 A. K. Capulli, K. Tian, N. Mehandru, A. Bukhta, S. F.
Choudhury, M. Suchyta and K. K. Parker, Lab Chip, 2014, 14,
3181–3186.
2 H. E. Abaci and M. L. Shuler, Integr. Biol., 2015, 7, 383–391.
3 I. Rosser, M. Schlager, M. Purtscher, F. Jenner and P. Ertl,
J. O. C., 2015.
4 N.K.Paschos,W.E.Brown,R.Eswaramoorthy,J.C.Huand
K. A. Athanasiou, J. Tissue Eng. Regener. Med., 2015, 9, 488–503.
5 S. N. Bhatia and D. E. Ingber, Nat. Biotechnol., 2014, 32,
760–772.
6 P. Ertl, D. Sticker, V. Charwat, C. Kasper and G.
Lepperdinger, Trends Biotechnol., 2014, 32, 245–253.
7 S. Giulitti, E. Magrofuoco, L. Prevedello and N. Elvassore,
Lab Chip, 2013, 13, 4430–4441.
8 J.P.Wikswo,E.L.Curtis,Z.E.Eagleton,B.C.Evans,A.Kole,
L. H. Hofmeister and W. J. Matloff, Lab Chip, 2013, 13,3496–3511.
9 C. Moraes, J. M. Labuz, B. M. Leung, M. Inoue, T.-H. Chun
and S. Takayama, Integr. Biol., 2013, 5, 1149–1161.
10 N. Ucciferri, T. Sbrana and A. Ahluwalia, Front. Bioeng.
Biotechnol., 2014, 2, 74.
11 H. Andersson and A. van den Berg, Sens. Actuators, B,
2003, 92, 315–325.
12 B. Ungerbock, V. Charwat, P. Ertl and T. Mayr, Lab Chip,
2013, 13, 1593–1601.
13 M. Rothbauer, D. Wartmann, V. Charwat and P. Ertl,
Biotechnol. Adv., 2015, 33, 948–961.
14 S. Sheikh, G. E. Rainger, Z. Gale, M. Rahman and G. B.
Nash, Blood, 2003, 102, 2828–2834.
15 K. Ookawa, M. Sato and N. Ohshima, Biorheology, 1993, 30,
131–140.
16 K. Hosokawa and R. Maeda, J. Micromech. Microeng.,
2000, 10, 415.
17 M. A. Unger, H. P. Chou, T. Thorsen, A. Scherer and S. R.
Quake, Science, 2000, 288, 113–116.
18 S. Giulitti, E. Magrofuoco, L. Prevedello and N. Elvassore,
Lab Chip, 2013, 13, 4430–4441.
19 N. Ferrell, R. R. Desai, A. J. Fleischman, S. Roy, H. D. Humes
and W. H. Fissell, Biotechnol. Bioeng., 2010, 107, 707–716.
20 E. Vereshchagina, D. Mc Glade, M. Glynn and J. Ducrée,
Biomicrofluidics, 2013, 7, 034101.
21 E. K. Sackmann, A. L. Fulton and D. J. Beebe, Nature,
2014, 507, 181–189.
22 G. S. Fiorini and D. T. Chiu, BioTechniques, 2005, 38,
429–446.
23 P. F. O'Neill, A. Ben Azouz, M. Vázquez, J. Liu, S. Marczak, Z.
Slouka, H. C. Chang, D. Diamond and D. Brabazon,
Biomicrofluidics, 2014, 8.
24 E. Berthier, E. W. K. Young and D. Beebe, Lab Chip,
2012, 12, 1224–1237.
25 G. S. Fiorini, G. D. Jeffries, D. S. Lim, C. L. Kuyper and D. T.
Chiu, Lab Chip, 2003, 3, 158–163.
26 G. S. Fiorini, R. M. Lorenz, J. S. Kuo and D. T. Chiu, Anal.
Chem., 2004, 76, 4697–4704.
27 D. C. Duffy, J. C. McDonald, O. J. A. Schueller and G. M.
Whitesides, Anal. Chem., 1998, 70, 4974–4984.
28 C. D. Rudd, C. M. Duffy, M. S. Johnson and P. J. Blanchard,
Broader Meaning to Thermosets, 1998, p. 177.
29 A. P. Golden and J. Tien, Lab Chip, 2007, 7, 720–725.
30 R. Novak, N. Ranu and R. A. Mathies, Lab Chip, 2013, 13,
1468–1471.
31 U. M. Attia, S. Marson and J. R. Alcock, Microfluid.
Nanofluid., 2009, 7,1–28.
32 D. A. Mair, E. Geiger, A. P. Pisano, J. M. J. Frechet and F.
Svec, Lab Chip, 2006, 6, 1346–1354.
33 H. Becker and L. E. Locascio, Talanta, 2002, 56, 267–287.
34 K. N. Ren, J. H. Zhou and H. K. Wu, Acc. Chem. Res.,
2013, 46, 2396–2406.
35 Y. N. Xia and G. M. Whitesides, Annu. Rev. Mater. Sci.,
1998, 28, 153–184.
36 D. Huh, B. D. Matthews, A. Mammoto, M. Montoya-Zavala,
H. Y. Hsin and D. E. Ingber, Science, 2010, 328, 1662–1668.
37 Z. Y. Zhang, P. Zhao and G. Z. Xiao, Polymer, 2009, 50,5358–5361.
38 E. Berthier, E. W. K. Young and D. Beebe, Lab Chip,
2012, 12, 1224–1237.
Lab on a Chip Paper
4554 |Lab Chip,2015,15,4542–4554 This journal is © The Royal Society of Chemistry 2015
39 J. C. McDonald and G. M. Whitesides, Acc. Chem. Res.,
2002, 35, 491–499.
40 J. M. Karlsson, T. Haraldsson, C. F. Carlborg, J. Hansson, A.
Russom and W. Van Der Wijngaart, J. Micromech. Microeng.,
2012, 22.
41 M. W. Toepke and D. J. Beebe, Lab Chip, 2006, 6, 1484–1486.
42 E. Berthier, J. Warrick, H. Yu and D. J. Beebe, Lab Chip,
2008, 8, 852–859.
43 K. J. Regehr, M. Domenech, J. T. Koepsel, K. C. Carver, S. J.
Ellison-Zelski, W. L. Murphy, L. A. Schuler, E. T. Alarid and
D. J. Beebe, Lab Chip, 2009, 9, 2132–2139.
44 N. Sandström, R. Z. Shafagh, A. Vastesson, C. F. Carlborg,
W. V. D. Wijngaart and T. Haraldsson, J. Micromech.
Microeng., 2015, 25, 075002.
45 F. Saharil, K. B. Gylfason, Y. Liu, T. Haraldsson, P. Bettotti,
N. Kumar and W. van der Wijngaart, 2012.
46 C. F. Carlborg, T. Haraldsson, K. Oberg, M. Malkoch and W.
van der Wijngaart, Lab Chip, 2011, 11, 3136–3147.
47 C. F. Carlborg, A. Vastesson, Y. Liu, W. van der Wijngaart,
M. Johansson and T. Haraldsson, J. Polym. Sci., Part A:
Polym. Chem., 2014, 52, 2604–2615.
48 F. Forsberg, F. Saharil, T. Haraldsson, N. Roxhed, G.
Stemme, W. V. D. Wijngaart and F. Niklaus, J. Micromech.
Microeng., 2013, 23, 085019.
49 K. Sonomoto, K. Yamaoka, K. Oshita, S. Fukuyo, X. Zhang,
K. Nakano, Y. Okada and Y. Tanaka, Arthritis Rheum.,
2012, 64, 3355–3363.
50 A. Piruska, I. Nikcevic, S. H. Lee, C. Ahn, W. R. Heineman,
P. A. Limbach and C. J. Seliskar, Lab Chip,2005,5,1348–1354.
51 M. L. AB, OSTEMERs, http://www.ostemers.com, 2015.
52 L. Kwisnek, S. Nazarenko and C. E. Hoyle, Macromolecules,
2009, 42, 7031–7041.
53 R. Novak, Y. Zeng, J. Shuga, G. Venugopalan, D. A. Fletcher,
M. T. Smith and R. A. Mathies, Angew. Chem., Int. Ed.,
2011, 50, 390–395.
54 S. Rohringer, P. Hofbauer, K. Schneider, A.-M. Husa, G.
Feichtinger, A. Peterbauer-Scherb, H. Redl and W.
Holnthoner, Angiogenesis, 2014, 17, 921–933.
Lab on a ChipPaper
