Chemistry Reference
In-Depth Information
For coloring the aqueous phase injected in the lower side channel, we used 2mM
DPhyPC doped with 2 molar % of 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine-
N
-(carboxy fluorescein) in the upper picture, and 1,2-dipalmitoyl-sn-glycero-3-
phosphoethanolamine-
N
-(lissamine rhodamine B sulfonyl) (both dyes from Avanti
Polar Lipids) in the lower picture. The dye lipid was sonicated in the aqueous phase
to create liposomes. We see that the droplets form a periodic structure with an
'elementary cell' withstanding transport along the channel. The relative positions
of the droplets within the arrangement are fixed, as is obvious due to the droplet
color.
Most of the results discussed here were obtained with channel dimensions on the
order of 100 microns or more, for the sake of simplicity of generating the devices and
optical transparency. However, we repeated some of the experiments with channels
etched in silicon which were much narrower (down to 20
m wide). We found
that the smaller the droplets and channel dimensions were, the more stable were the
membranes. A straightforward explanation is that the Laplace pressure, which scales
as the inverse of the droplet radius, sets the energy scale for any disturbance leading
to substantial rearrangements of deformations of membranes, and thus potentially
to coalescence. Extrapolating this general trend, further down-scaling is expected to
yield emulsion structures which are even much more stable. It should be noted that
the formation of stable, mono-disperse emulsions with droplet diameters of 50nm
or less is routinely possible by suitable methods [
19
].
µ
2.3.1 Formation and Stability of Membranes
Spontaneous membrane formation is well known to occur if two surfactant-laden
oil/ water interfaces are brought in intimate contact on their lipophilic sides [
21
].
Recently, this has also been demonstrated with aqueous drops placed on a substrate
next to each other in an oil background phase [
22
,
23
]. The stabilization of the
oil-water interface is limited by the diffusion of the amphiphiles to the interface.
This means that the droplets have to be “pre-stabilized” in the oil phase to obtain
sufficient surface coverage before they are brought together to prevent coalescence of
the droplets and a destruction of a lipid bilayer. The waiting times for this are on the
order of 10's of minutes and the operations are done manually. We found that when
there is a flow of the lipids around a stationary aqueous droplet, the surface coverage
of the oil-water interface occurs rapidly, thus stabilizing the droplets (Fig.
2.4
). It is
to be expected that a similar process will occur if droplets are formed in flow within
a microfluidic channel, such that the droplets are stable against coalescence almost
immediately after formation. Subsequentlywhen a suitable gel emulsion, thus formed
in a microfluidic channel, is at sufficiently low continuous phase volume fraction, we
expect spontaneous membrane formation. That this is indeed the case is demonstrated
in Fig.
2.5
. Here we have produced a zig-zag structure of mono-disperse aqueous
droplets by step emulsification [
13
].
