Lab-on-a-Chip for Studying Growing Pollen Tubes - Plant Cell Morphogenesis: Methods and Protocols - page 237

18. Handle with care. Use a clean, sharp cutter to dice the cured
PDMS replicas before extracting them. The cutter should
touch the silicon wafer as the dicing takes place. If silanization
was properly performed, the PDMS clearly detaches from the
mold as the PDMS is being cut. Removing the replicas from
the mold should be done effortlessly. If PDMS is stuck to the
mold, the most likely reason is a problem with the silanization.
Unfortunately, the mold is almost inevitably lost if the PDMS
is stuck since it is next to impossible to remove the PDMS
without destroying the SU-8 layer. However, should this hap-
pen, a complete wafer cleanup can be performed (piranha bath
and HF cleaning) and the silicon wafer can be recovered to
restart mold fabrication.
19. Punching of inlets and outlets can be easily done with a revolv-
ing punch. The size of the round holes must match the PVC
tubes used.
20. Oxygen plasma bonding is recommended for fast and reliable
results. A matrix test must be carried out to determine the
optimal parameters for bonding glass and PDMS. In our setup
(Harrick Plasma PDC-001), the bonding time for a glass/
PDMS interface is 40 s with a high voltage application.
Another (low-cost) alternative is to spin-coat a thin layer of
PDMS on the glass slide (2,000 rpm for 30 s), semi-cure the
PDMS layer, make the bond, and then completely cure. The
duration of the semi-curing depends on many factors; how-
ever, a good starting point is 3 min at 90 °C. We found that it
is preferable to leave it longer since if the bonding fails, another
thin layer of PDMS can be added on top, whereas if less time
is used then the features on the PDMS replica will be fi lled by
the PDMS gel.
21. Since the fl uid pressures inside the microfl uidic platform are
relatively low, the PVC tubes do not need to be glued to the
PDMS replica. Friction is suffi cient to keep the tubes in place
(given the PDMS is at least 1 mm thick). Inlets and outlets on
the side of the chip are discouraged since this requires more
complex connections.
22. Although fresh Camellia japonica pollen is ideal, this may be
diffi cult to obtain as this species fl owers only once a year for a
few weeks.
23. An easy way to hydrate pollen grains is to wet a piece of paper
towel with hot water and place both pollen and tissue in an
enclosed glass container (Petri dish). Very importantly, avoid
any direct contact between pollen and water to prevent the
pollen grains from absorbing liquid water at this point.
24. Liquid growth medium has already been optimized for
Camellia japonica pollen [ 7 , 12 ]. Usually, we prepare 10 ml of
medium and use 1 ml plastic capsules for testing.