To see the tendency
toward more realistic D values, let us look at the effect of diffusion
coefficients. Figure 7a and b represents the mixing efficiency and
the mixing factor, respectively, as a function of the diffusion
coefficient of the species for a zigzag channel featuring s ) 400
ím and for an equivalent straight channel (dashed lines). Figure
7b is directly derived from Figure 7a and illustrates the mixing
gain corresponding to the zigzag configuration. This mixing factor
is defined as the ratio between the mixing efficiencies obtained
with the zigzag and straight microchannels. Because of computational
limitations regarding the values of the mesh size, D values
range from 1 10-6 to 2 10-7 m2 s-1. To have a significant
value of the mixing efficiency for high Reynolds numbers, such
as 533, the minimal value of D studied is fixed to 4 10-7 m2 s-1
(minimum efficiency of 10% for the straight channel). As portrayed
in Figure 7a, the mixing efficiency decreases when the diffusion
coefficient decreases for three different Reynolds numbers. For
Re ) 267, the shapes of the curves in Figure 7a for both
geometries are similar, highlighting the limitations of the zigzag
configuration in this range of diffusion coefficients. When the
Reynolds number increases, the mixing efficiency in the straight
channel is strongly affected by the decrease of the residence time.
For the zigzag geometry, this decrease is greatly compensated