Visible light Laue diffraction from woodpile

photonic crystals

Björn Brüser,1 Isabelle Staude,2 Georg von Freymann,2,4

Martin Wegener,2,3 and Ullrich Pietsch1,*
1
University of Siegen, Department of Physics, Walter-Flex-Strasse 3, 57072 Siegen, Germany
2
Institute of Applied Physics and DFG-Center for Functional Nanostructures, Karlsruhe Institute
of Technology, D-76131 Karlsruhe, Germany
3
Institute of Nanotechnology, Karlsruhe Institute of Technology, D-76021 Karlsruhe, Germany
4
Fachbereich Physik und Forschungszentrum OPTIMAS, TU Kaiserslautern, 67663 Kaiserslautern, Germany
*Corresponding author: pietsch@physik.uni‐siegen.de

Received 13 July 2012; accepted 15 August 2012;

posted 30 August 2012 (Doc. ID 172524); published 25 September 2012

Bragg diffraction is often used as a tool to assess the structural quality of two-dimensional and
three-dimensional (3D) photonic crystals. However, direct conclusions from the Laue diagrams to the
underlying crystals structure cannot be drawn, as multiple scattering due to the high index contrast
takes place. Here we systematically study the scattering of visible light by 3D woodpile photonic crystals
with varying internal refractive index contrast Δn, to determine the limits of the single (kinematic) scat-
tering approach. We aim to describe the intensity distribution of diffracting Bragg peaks with analytic
expressions similarly to x-ray scattering at electronic crystals. Measured scattering curves of selected
Bragg reflections are classified in terms of Δn. We find that the kinematic approach describes the shape
and intensity distribution of experimental scattering curves in acceptable accuracy as long as Δn < 0.15.
The transition between single and multiple scattering is observed for Δn ≈ 0.16 − 0.25 before multiple
scattering dominates for larger Δn. The classification of the scattering regimes is confirmed by simula-
tions in terms of numerical solution of Maxwell’s equations. © 2012 Optical Society of America
OCIS codes: 050.1960, 160.5298, 260.1960, 350.4238.

1. Introduction at definite angular positions with respect to the in-

Since the seminal work of Yablonovitch [1] and John
[2], huge efforts have been made to investigate the
interaction of light with photonic crystals (PhCs).
Up to now most research is focused on PhC formation
and their application as waveguides, amplification or
suppression of spontaneous emission, chirality and
others [3–5]. However, three-dimensional (3D) PhCs
could also find application as dispersive optical ele-
ments in light scattering. Similar to x-ray scattering
on crystals, the PhC will lead to displaying distinct
Bragg peaks upon illumination with light appearing
1559-128X/12/286732-06$15.00/0
© 2012 Optical Society of America
cident beam. Under these conditions diffraction an-
gles and peak intensities are simple functions of
the crystallographic structure of the PhC. However,
despite 25 years of PhC research, a quantitative ana-
lysis of Bragg scattering in 3D PhCs as it takes place
for x-rays in crystals is still missing. In most cases
the appearance of Bragg scattering peaks for scatter-
ing of light at one-dimensional (1D) or two-
dimensional (2D) PhCs is explained using the Bragg
condition [6–9]. Only for a 1D PhC has a complete
description of scattering curves been presented by
Bushuev et al. [10] in terms of full electrodynamical
theory that is in some sense similar to the dynamical
theory in x-ray crystallography [11]. Nothing is
known so far to describe the shape and intensity

6732 APPLIED OPTICS / Vol. 51, No. 28 / 1 October 2012

distribution of light scattering at 3D PhCs. The re-
fractive index of natural materials for x-rays is al-
ways close to unity scattering and can be described
by a sum of single scattering events, known as the
kinematic scattering approach [12] corresponding
to the first order term of Born’s scattering series.
In general, this approximation is not valid for visible
light scattering on PhCs. Multiple scattering takes
place because the refractive index for conventional
optical materials is much larger then unity and a
large scattering contrast appears at material-air in-
terfaces. In a previous paper [13] we quantified the
Bragg scattering measured from a 3D PhC created by
multiple-beam interference in a photopolymer [14]. Fig. 1. (Color online) Scanning electron micrograph of an as-
prepared WP PhC before infiltration of the liquids. Sketch of
At the positions of maximum intensity addressed
the bct (left) and fcc (right) unit cell is shown as inset.
by the maxima of the multiple-beam interference
pattern, the photopolymer becomes polymerized ac-
companied with a small change in the refraction in- chosen as a
0.9 μm, resulting in c
1.27 μm. Al-
dex of about Δn ≈ 10−3 . It has been shown that the ternatively, the structure can be described using a
angular distribution of Bragg scattering peaks can body-centered tetragonal primitive (bct) unit cell
be well described by a sinc function known for kine- with the same c∕a ratio. Here, the rods are aligned
matic scattering. In this paper we will study the parallel to the h100i and h010i directions while the
upper limit of this approach by filling the voids of stacking direction is given by the same vector as
a woodpile WP) PhC with liquids of a known refrac- the fcc unit cell, i.e. h001i.
tive index as proposed by [9]. This allows for gradu-
3. Experimental Setup
ally decreasing and, hence, systematically tuning the
refractive index contrast. The scattering properties The experimental setup for scattering analysis is
are investigated by recording scattering curves shown in Fig. 2. In most cases the blue emission line
through selected Bragg peaks. We will show that (λ
488 nm) of an Ar-Kr laser is used for probing the
the shape and intensity of Bragg peaks can be de- PhC mounted at the center of a diffractometer. A
scribed in terms of the kinematic approach as long white-light fiber laser source is used to probe the dis-
as Δn < 0.15. Our experimental results are verified persive properties of the PhC. A reference beam is
by calculations in terms of numerical solutions of separated from the incident beam for intensity nor-
Maxwell’s equations providing qualitatively similar malization before focusing onto the PhC. Both the in-
conclusion. tensities of the diffracted beam and the reference
beam are recorded by avalanche photodiodes. The
2. Properties of Woodpile Photonic Crystals use of an x-ray diffractometer allows for a very pre-
cise rotation of the PhC with angular steps of about
For our investigation we fabricate WP PhCs [15] on
0.01°. Subsequently, the intensity of a selected Bragg
standard microscope cover slides by direct laser
peak can be scanned as a function of both the inci-
writing (DLW) [16] using the high-resolution photo-
dent and the exit angle. The adjustment of the rota-
resist IP-L from Nanoscribe and a commercial Nano-
tion angles, as well as the subsequent recording of
scribe Photonic Professional DLW system [17].
the corresponding scattering intensities, are con-
Within the spectral range of visible light, the
trolled by a dedicated LabView program.
polymerized IP-L has a refractive index of
roughly n
1.52. In order to allow for measurements
under large incident angles while ensuring that the
probing laser spot is always smaller than the foot-
print of the structure, the WP PhC’s lateral dimen-
sions have been chosen to be relatively large
(100 μm × 100 μm). A scanning electron micrograph
of a typical fabricated WP structure is shown in
Fig. 1. In the stacking direction of the WP 20 vertical
layers corresponding to five repetition units (unit
cells) in total have been realized. The crystallo-
graphic structure of the PhC can be described by rods
stacked along the crystallographic h110i and h11̄0i
directions defining
p a face-centered cubic (fcc) unit cell
with ratio c∕a
2 where c stands for the distance
of the first to the fifth woodpile layer and a is the Fig. 2. (Color online) Sketch of the experimental setup used
in-plane distance between the rods. For our to record scattering intensities as a function of incident and exit
experiments the in-plane lattice parameter has been angle.

1 October 2012 / Vol. 51, No. 28 / APPLIED OPTICS 6733

4. Scattering Geometry
Figure 3 displays the notations of the diffractometer
angles. Angles ψ and φ are mainly used for PhC
alignment. The scattering curve of a Bragg peak in-
tensity is probed by rotating the PhC around the ω -
axis and recording the scattering beam intensity at
the detector position θ Both angles are connected via
θ
arcsin− sinω  λ∕d − ω;
where λ stands for the wavelength and d for the in-
plane lattice spacing of the considered reflection with
in-plane Miller indices h; k written as
d
a∕h2  k2 1∕2 :
5. Scans in Reciprocal Space
As common for x-ray diffraction experiments, the
scattering curves are recorded in angular space and
transformed into reciprocal space. This is done by
considering the wave vector transfer ⃗q (cf., Fig. 4)
which is composed of a parallel ⃗q∥ and a perpendicu-
lar ⃗q⊥ component. While the parallel component is
always fixed to an in-plane reciprocal lattice vector,
⃗q∥
G∥
2π∕d [see Eq. 2] the perpendicular compo-
nent is probed by rotating both the PhC and the de-
tector by angles ω and θ, respectively. A sketch of the
reciprocal lattice along with the beam vectors is
given in Fig. 4 together with so called Ewald‘s con-
struction which is an appropriate description of
kinematic scattering. Bragg diffraction takes place
whenever a reciprocal lattice point crosses the Ewald
sphere that has a radius kint
2πneff ∕λ within the
PhC. Here, neff is an effective refractive index. Note
that kint differs from kext , outside the crystal causing
refraction of light at the air-PhC interface.
Due to the small size of the PhC, the reciprocal lat-
tice points cannot be described by delta-functions but
have a finite size. Considering the different numbers
of repetition units parallel and perpendicular to the
stacking direction, each reciprocal lattice point is
degenerated into ellipses with the long axes parallel
to ⃗q⊥
h001i implying a broadening of the Bragg
peaks. This functional behavior is similar to the
Fig. 3. (Color online) Schematic of the scattering geometry.
Crystal rotation is possible via three axes ω, ψ, and φ. The detector
can be rotated around the θ-axis.
6734 APPLIED OPTICS / Vol. 51, No. 28 / 1 October 2012
(1)
(2)
Fig. 4. (Color online) 2D Ewald construction. For a thin WP PhC
the reciprocal lattice points are degenerated into ellipses along
h001i. The decomposition of the transfer wave vector q, is shown
⃗ ∥.
as well as the in-plane reciprocal lattice vector G
shape of a slit function. Considering neff and d, ⃗q⊥
and ω are connected by
s















































 

2πneff λ sinω 2
q⊥
1− −
λ dneff neff
  
2πneff sinω
 cos arcsin − : (3)
λ nn eff
From this, the out-of-plane Miller index L is directly
obtained via division by 2π∕c:
L
q⊥ c∕2π: (4)
6. Experimental Results
Δn is tuned by infiltration of commercially available
refractive index (RI) liquids purchased from Cargille
Labs [18]. As a first step, in order to characterize the
PhC 2D scattering pattern of our WP PhC samples,
Laue diffraction images have been recorded under
normal incidence, i.e., the monochromatic laser beam
of λ
488 nm impinges onto the PhC along h001i di-
rection and the diffracted intensity is detected using
a 2D detector system.
As shown in Fig. 5(a), the diffraction pattern shows
distinct spots for weak (kinematic) scattering
(Δn ≈ 0.01) revealing the square plane symmetry of
the PhC. The appearance and arrangement of the
Bragg peaks are defined via the Laue equations
known from x-ray crystallography [19,20]. Based on
bct description, the bright spots (circles) can be in-
dexed by (10L), while the second order (20L) reflec-
tions with much stronger scattering intensity appear
outside the screen. The difference in the scattering
intensity originates from the fact that the 10L spots
are structurally forbidden for infinite crystals but not
entirely forbidden for finite ones. Figure 5(b) shows
the scattering pattern for PhC with maximum
Fig. 5. (Color online) Laue diffraction picture from a WP PhC (a) with Δn ≈ 0.01 (b) and with a maximum contrast of Δn ≈ 0.52. The
incidence direction is parallel to 001. The transmitted beam in the center is blocked by a beam stop. The arrangement of the first order
diffraction peaks [(10L)—circles] displays the square plane symmetry of the h, k plane. Due to the particular scattering geometry the (20L)
reflections will appear outside of the screen.

ΔN
0.52, corresponding to the uninfiltrated poly-
mer PhC. It displays additional spots and lines of
strong scattering (multiple and diffuse scattering)
in between the Laue spots.
In the following we restrict our analysis to the in-
tense (20L) reflection. At least one of these reflec-
tions becomes accessible rotating the sample by ω.
In order to study the transition between both cases
shown in Fig. 5, the PhC has been infiltrated by a RI
liquid of defined refraction index. For better compar-
ison, the same PhC has been used for all diffraction
studies. After the measurement with one particular
RI, the liquid used has been removed from the PhC
and replaced by an RI with another RI. Each time,
before refilling the sample, the scattering curve of
the empty PhC has been recorded and compared to
the scattering curve of the virgin sample recorded
prior to the first infiltration. A set of subsequent re-
cordings of the nominally empty PhC are shown in
Fig. 6. The curves clearly demonstrate the appear-
ance of multiple scattering characterized by the fact
that not all the peak positions can be explained by
Laue equations. Similar features show up at same
scattering angles but with different intensities and
sometimes different intensity ratio among the peaks.
These differences are explained by the remaining
parts of liquid after removal and small changes of
the sample position with respect to the incident
beam. The qualitative agreement of all curves justi-
fies the procedure used to modify Δn.
Figure 7 shows selected scattering curves of (20L)
Bragg reflection as function of the refractive index of
the infiltrated liquid. Due to different illumination
conditions, the main scattering features seen for
maximum Δn in this figure (nliq
1.30) differ from
that shown in Fig. 6. The fluctuations of angular po-
sitions of curve maxima can be explained by a
slightly different setting of the sample with respect
to the beam after refilling. However, for low Δn the
curves show one distinct peak at angular positions
predicted by Laue equations. Its FWHM, Δω, is re-
lated to the finite size of the scattering crystal which
can be described by Scherrer’s equation
λ
Δω ≈ (5)
D  cos θ
Considering D
5  c, Δω becomes the order of 5°.
The general approach of kinematic scattering from
PhCs can be found elsewhere [13]. Due to the relation
I ∼ Δn2 [12] an increasing Δnneff
1.52 increases
0
10

1,0
0,9
normalized intensity

0,8 10
-1
st
normalized intensity

1 cycle
0,7
nd
∆n=0.22
0,6 2 cycle
rd
0,5 3 cycle
-2
10
0,4
th ∆n=0.19
0,3 4 cycle
∆n=0.13
0,2 st ∆n=0.05
5 cycle
0,1 -3
∆n=0.02
10
0,0 10 20 30 40 50 60 70
10 20 30 40 50 60 70 80 sample rotation angle ω [°]
sample rotation angle ω [°]
Fig. 7. (Color online) Scans throughout the (20L) reflection of
Fig. 6. (Color online) Set of scattering curves of the (20L) reflec- WP PhC filled with RI liquids. For clarity, the curves are vertically
tion for a WP PhC after removal of the RI liquids. displaced.

1 October 2012 / Vol. 51, No. 28 / APPLIED OPTICS 6735


the reflection intensity I. Whereas multiple scatter-
ing shows up for large Δn, curves with low contrast
can be described by a sinc-like function. Considering
Eqs. (3) and (4), its curve shape can be described by
  2
sinNπL π
I∝ · Δn · cos h  k  L
N sinπL 2 simulation. The variation of R2 as function of nliq
The first term in Eq. (6) is the lattice sum and de-
pends on the number of illuminated unit cells, N,
the second term, is the structure factor describing
the phase sum of elements within a single unit cell.
Comparison with the experimental curves shown in
Fig. 7 are possible using N
6 and h
2 (not
shown). The position of the scattering maximum ap-
pears at L
0, 1 and those of subsidiary maxima on
the size of the PhC.
Figure 8 shows selected experimental scattering
curves and the respective kinematic simulation.
Polymer shrinking effects cause a reduction of the
out-of-plane lattice parameter and, as a consequence,
small shifts of the Bragg peak position have been ta-
ken into account in our simulations. Depending on
the RI liquid used, the best agreement with experi-
mental curves has been found considering c lattice
parameter shrinking by 1–3%. Below Δn
0.05
(i.e. nliq > 1.46) the measured scattering curve can
be well described by Eq. (6). By increasing Δn, both
intensities and position of subsidiary maxima in the
experiment start to deviate from the expectation of
the kinematic approach. The further increase of
Δn results in the appearance of additional intense
peaks which makes identification of a main maxi-
mum impossible.
The agreement of the experiment with the result
of kinematic simulation is characterized by an R2 -
factor defined as
-0,6 -0,4 -0,2 0,0
0,2
0,1
0,0
1,5
relative intensity
P
I exp − I kin 2
R2
1 − P 2 ; (7)
I kin
being a measure for square deviation between mea-
sured and simulated intensity values summed up
over the whole rocking curve. R2
1 stands for
: (6) the complete agreement between experiment and
is shown as a round symbols in Fig. 9.
If we take the correct description of the main max-
imum of the scattering curves as the distinctive fea-
ture, PhC filling with nliq > 1.37 (Δn < 0.15) fulfills
the kinematic approximation. The index contrast of
the PhCs investigated in [13] (Δn ≈ 0.001) is clear be-
low this limit. In detail both angular position of
subsidiary maxima and scattering efficiency increase
continuously from a refractive index difference of
Δn
0.01 to Δn
0.15. Between Δn
0.16
(nliq > 1.36) to Δn
0.22 (nliq > 1.31) the scattering
efficiency stays nearly constant but a splitting of the
main Bragg peak occurs. The intensity of side peaks
increases and additional peaks appear. Finally, for
larger (maximum) contrast, no distinct main maxi-
mum can be identified (not shown), i.e., no Bragg
peak occurs due to the strong influence of multiple
scattering. Other reflections like (10L) show similar
behavior as show for (20L), but the quality of the
curve fit is reduced due to much lower experimental
intensities.
The limit of the kinematic approach has been ver-
ified by calculations in terms of numerical solutions
of Maxwell’s equations using the commercially avail-
able solver DiffractMod [21]. The underlying algo-
rithm is based on a method known as the rigorous
coupled-wave analysis in combination with modal
transmission line theory [20]. Here, we have calcu-
lated the scattering curves for different refractive
index contrasts, Δn, using the full Maxwell code
and compared them with the respective kinemati-
cally calculated scattering curve using Eq. (6). The
0,2 0,4 0,6 agreement between both approaches has been
∆n=0.05
simulation
1,0
∆n=0.13 0,9

1,0
0,5 kinematic limit
R² factor

0,0 0,8
1,5
∆n=0.16
1,0
0,5 0,7
0,0
1,5 experiment
1,0 ∆n=0.19 0,6
simulation
0,5
0,0 0,5
-0,6 -0,4 -0,2 0,0 0,2 0,4 0,6 1,30 1,35 1,40 1,45 1,50
reciprocal lattice coordinate L refraction index of RI liquid
Fig. 8. (Color online) Kinematic fits to the corrected scattering Fig. 9. (Color online) R2 factor versus refractive index of the
curves of the (20L) reflection for selected refractive-index index-matching fluid in the WP PhC. The range where the kine-
contrasts. matic approximation applies is indicated.

6736 APPLIED OPTICS / Vol. 51, No. 28 / 1 October 2012

characterized by the same R2 factor as used to qua-
lify the experimental data replacing I exp by I max in
Eq. (6) where I max stands for the normalized inten-
sity calculated by Maxwell theory. The results are de-
picted in Fig. 9 by square dots. Up to Δn < 0.12
(nliq > 1.40), the calculated R2 factors stay nearly
constant and close to 1, while for larger Δn the R2
factor decreases. This is in good qualitative agree-
ment with our experimental findings. Quantitatively,
the theoretical values are found to be slightly smaller
than those found in the experiment, which is very
likely caused by fabrication imperfections. Our re-
sults demonstrate that under conditions of a rather
large refraction index contrast, the simple kinematic
approach is indeed able to reproduce the scattering
curves for visible light with some impact for technical
application such as designing PhCs as dispersive
devices.
7. Summary and Conclusions
We have shown that shape and intensity distribution
of scattering curves of visible light on 3D WP PhCs
can be described by the kinematic scattering theory
in a similar manner as used for x-ray diffraction if
Δn < 0.05. For Δn < 0.15, the scattering curve can
be described by a sinc function dominated by a main
maximum of definite width and intensity. Conse-
quently, the limit of applicability of the kinematic ap-
proach appears at much larger Δn than expected
from extrapolating the x-ray case toward visible
light. In x-ray diffraction from small crystals Δn is
in the range of 10−5 − 10−6. For PhCs made by holo-
graphic exposure, we demonstrated that the kine-
matic approach holds for Δn ≈ 0.001 [13]. Here we
show its applicability for systems where Δn is larger
by three orders of magnitude. This finding has impor-
tant consequences for applications of PhCs as optical
elements. Our result offers the possibility to use tai-
lored PhCs as optical elements for light-scattering
experiments. Similar to the traditional Laue experi-
ments known from x-rays, such an optical element,
will be able to select monochromatic spots from a
white incident beam by redirecting different colors
toward different positions in space. Since the scatter-
ing intensity scales with Δn2 in the regime of validity
of kinematic approach, one can gain several orders of
magnitude in scattering efficiency for only slight
variations of Δn. Furthermore, even a complete
separation of neighboring Laue spots can be expected
from theory if the number of stacked unit cells of the
WP structure can be increased to exceed a limit of
10–15 (corresponding to 40–60 layers), thereby pro-
viding additional exciting opportunities for future
applications.
This work was supported by VDI-BMBF in the fra-
mework of NAMIROS. The Karlsruhe team has been
supported by the Deutsche Forschungs-gemeinschaft
(DFG) and the State of Baden-Württemberg through
project A1.4 of the DFG-Center for Functional
Nanostructures.
References
1. E. Yablonovitch, “Inhibited spontaneous emission in
solid-state physics and electronics,” Phys. Rev. Lett. 58,
2059–2062 (1987).
2. S. John, “Strong localization of photons in certain disordered
dielectric superlattices,” Phys. Rev. Lett. 58, 2486–2489
(1987).
3. G. von Freymann, A. Ledermann, M. Thiel, I. Staude, S. Essig,
K. Busch, and M. Wegener, “Three-dimensional nanostruc-
tures for photonics,” Adv. Funct. Mater. 20, 1038–1052 (2010).
4. K. Busch, G. von Freymann, S. Linden, S. Mingaleev, L.
Tkeshelashvili, and M. Wegener, “Periodic nanostructures
for photonics,” Phys. Rep. 444, 101–202 (2007).
5. L. A. Dorado, R. A. Depine, D. Schinca, G. Lozano, and H.
Miguez, “Experimental and theoretical analysis of the inten-
sity of beams diffracted by three-dimensional photonic crys-
tals,” Phys Rev B 78, 075102 (2008).
6. P. L. Gourly, M. E. Warren, G. A. Vawter, T. M. Brennan, and
B. E. Hammons, “Optical Bloch waves in a semiconductor
photonic lattice,” Appl. Phys. Lett. 60, 2714–2716 (1992).
7. S. Richel, C. Kallinger, U. Lemmer, J. Feldmann, A. Gombert,
V. Wittwer, and U. Scherf, “A nearly diffraction limited surface
emitting conjugated polymer laser utilizing a two-
dimensional photonic band structure,” Appl. Phys. Lett. 77,
2310–2312 (2000).
8. C. M. Anderson and K. P. Giapis, “Symmetry reduction in
group 4mm photonic crystals,” Phys. Rev. B 56, 7313–7320
(1997).
9. J. Manzanares-Martinez, P. Castro-Garay, and E. Urrutia-
Babuelos, “Analytical determination of the stop band tuning
of photonic crystals infilltrated with liquid crystals,” Adv. Stu-
dies Theor. Phys. 15, 551–557 (2011).
10. V. A. Bushuev, B. I. Mantsyzov, and A. A. Skorynin, “Diffrac-
tion-induced laser pulse splitting in a linear photonic crystal,”
Phys. Rev. A 79, 053811 (2009).
11. A. Authier, Dynamical Theory of X-ray Diffraction (Oxford,
2001).
12. J. Als-Nielsen and D. McMorrow, Elements of Modern X-ray
Physics (Wiley, 2001), pp. 11, 318.
13. B. Brüser and U. Pietsch, “Kinematic and dynamic light scat-
tering from 3D photonic crystals,” Proc. SPIE 7487, 748709
(2009).
14. S. Orlic, Ch. Müller, and A. Schlösser, “All-optical fabrication
of three-dimensional photonic crystals in photopolymers by
multiplex-exposure holographic recording,” Appl. Phys. Lett.
99, 131105 (2011).
15. M. Deubel, G. Von Freymann, M. Wegener, S. Pereira, K.
Busch, and C. M. Soukoulis, “Direct laser writing of three-
dimensional photonic-crystal templates for telecommunica-
tions,” Nat. Mater. 3, 444–447 (2004).
16. Nanoscribe Homepage, retrieved 05/11, 2011, http://www
.nanoscribe.de/.
17. H. S. Sozuer and J. P. Dowling, “Photonic band calculations for
woodpile structures,” J. Mod. Opt. 41, 231–239 (1994).
18. “Cargille Homepage,” retrieved 05/13, 2011, http://www
.cargille.com/.
19. C. Giacovazzo, Fundamentals of Crystallography, Interna-
tional Union of Crystallography Texts on Crystallography
(Oxford, 2002), pp. 14, 825.
20. B. Brüser, Coherent Light Scattering from Photonic
Crystals and Phase Lattices, urn:nbn:de:hbz:467-5513,
http://dokumentix.ub.uni‑siegen.de/opus/volltexte/2011/551/.
21. Rsoft Homepage, retrieved 02/05, 2012, http://www.rsoft‑
design.com/products.php?sub=Component+Design&itm=
DiffractMOD.
1 October 2012 / Vol. 51, No. 28 / APPLIED OPTICS 6737