The early reports on high-efficiency
DWs did not have much follow-up. In
some instances, they could have demon-
strated high efficiency if the wavelength
of the test laser beams satisfied the
half-wave condition. In other cases, the
ability of LCs to maintain as dramatic a
deformation as the cycloidal pattern of
orientation in DWs was overestimated.
Even if both the materials and the
technology were ripe for it, apparently
the range of perceived applications did
not warrant the investment of time and
resources to further those developments
at that time.
Geometrically, axial waveplates are obtained by wrapping a cycloidal waveplate around
an axis like a folding fan (a). The optical axis orientation a of LC (or a LC polymer) in an
axial waveplate of “polar wavevector” q w rotates by 90° for each 90°/q w change of the
azimuthal angle w. Those are revealed as 4q w dark tails between crossed polarizers.
The patterns of arrows illustrating distribution of optical axis orientation in waveplates
with q w = 1 (b) and q w = 2 (c) are superimposed on actual photos.
(a)
(b)
(c)
a
w
The increase in the size of the doughnut in the far-field patterns—the polar analog of the
diffraction angle for a cycloidal waveplate—with increasing q w is demonstrated in (a-c)
for axial waveplates (AWs) characterized with q w = 4, 16 and 64, correspondingly, using
an input laser beam of 325-nm wavelength and homogeneous intensity profile. The bottom row shows corresponding images of AWs taken between crossed polarizers with
white light illumination. AWs can be stacked or organized in arrays for controlling light
beams. The photos in (d) show the diffraction pattern generated in the far field by an
array of AWs of q w = 4 for an incident laser beam of homogeneous profile at 532 nm
and its image between crossed polarizers.
(a) (c) (b) (d)
incorporated an analysis of the mechanical stability of cycloidal orientation
patterns of LCs: The thickness of the LC
layer must be smaller than the spatial
period of the orientation pattern by a
factor determined by elastic constants of
the LC—roughly, L < 0.5L. Thus, the
floodgates were opened, and LC and LC
polymer DWs with 100 percent efficiency have since been demonstrated using
many different materials.
Fabrication of LC and LC polymer
DWs is a multistep process, and each step
must be optimized to obtain high-quality
DWs. The holographic photoalignment
process is part of the problem due to the
long exposure times involved. The recently developed technology for printing
DWs from a master DW paves the way
for their large-scale production with high
quality and large areas, avoiding all complexity, cost and the stability problems
of holographic setups. The printing technique makes use of the rotating polarization pattern obtained at the output of the
master DW from a linearly or circularly
polarized input beam. The wavevector of
the printed DW is doubled when one uses
a linearly polarized input beam.
DWs optimized for a given wavelength range are produced by properly
choosing the optical anisotropy and the
thickness of the LC layer. In the case of
LC polymers, this is easily done by varying the spin-coating regimes or stacking polymer layers. There is no need
for multiple photoalignment layers,
since each LC polymer layer enforces
its alignment pattern on the new layer.
To vary the angle between orientation
patterns of subsequent layers in order to
further widen the diffraction spectrum,
researchers introduce a twist with the
aid of an intermediate cholesteric LC
polymer film. The twist can be obtained
in each layer with the aid of chiral dopants as well.
Diffractive waveplates
with axial symmetry
The cycloidal orientation pattern can be
wrapped around the axis normal to the
waveplate plane (z-axis) like a folding fan
44 | OPN Optics & Photonics News
www.osa-opn.org
