Liquid crystal display as spatial light modulator for diffractive optical elements

and the reconstruction of digital holograms

Günther Wernicke1, Sven Krüger1, Hartmut Gruber1, Nazif Demoli2, Matthias Dürr3 and Stephan Teiwes3
1
Labor für Kohärenzoptik, Humboldt-Universität zu Berlin,
Invalidenstr. 110, 10115 Berlin, Germany
Phone/Fax +49 30 2093 7897
E-mail: [email protected]
2
University of Zagreb,
3
HoloEye GmbH, Dorotheastr. 8, 10318 Berlin, Germany

Liquid crystal displays, originally designed and fabricated for projection systems, are often used as spatial light
modulators in optical correlators or in fringe projection systems.
Operating in a linearized phase mode, they can achieve a performance to be applicable as diffractive elements.
An adapted driver electronics and adapting measurements of the phase modulation behavior can lead to a dynamic
phase modulating system with an almost linear modulation and a maximum phase shift of 2π. The electronic system can
directly address the graphics card signal or can picture various video standards. So, computer generated holograms can
be addressed at video frame rates.
The limiting parameters of the performance are mostly due to physical boundary conditions, such as pixel number and
size, response time, transmission etc.. We can assume that the fast growing micro-structuring technology will serve us
soon with displays of higher resolutions and efficiency.
Beam shaping elements, two-dimensional holograms and the reconstruction of digital holograms will be shown. The
latter opens new possibilities for non-destructive testing devices especially in the field of holographic interferometry.
We will show results of different systems and derive boundary conditions for applications in holographic reconstruction
of coherent masks.

Keywords: liquid crystal display, spatial light modulator, dynamic diffractive elements,
dynamic DOE/ HE

1. INTRODUCTION
Due to the developments in the field of micro-lithography and micro-fabrication diffractive optics is one of the fast
rising optical technologies these days [1]. With only refractive optics, many tasks in the field of laser beam modulation
could only be realized with a high effort or not at all. On the other hand, the high number of laser applications require a
vast number of beam shapes, spot patterns, divergence properties and varying beam profiles. All these can be realized
by diffractive optics.
Diffractive optical elements are already used as pattern projectors in application fields like positioning, material
processing and marking [2].
Here they often compete with scanning technologies or mask technologies, which are used either for pattern
representation or material processing. Laser scanners in connection with shutters are of high resolution and very fast, so
former arguments for DOE, to be able to create spot patterns without any flickering, are vanishing. The temporal
stability of a diffraction pattern remains as a reason, but otherwise, the scanner is a dynamic system ...
Masks have always the impact of the inefficient use of laser energy. Another problem is to place the mask very close
on top of the surface of the item to be engraved. This requirement might be hindering in fabrication processes where
items are moved and processed at very high speeds.
However, diffractive optical elements fabricated with micro-lithographic techniques are, even if there are replication
techniques already, static elements and quite expensive.

Advanced Photonic Sensors and Applications II, Anand K. Asundi, Wolfgang Osten, Vijay K. Varadan,
182 Editors, Proceedings of SPIE Vol. 4596 (2001) © 2001 SPIE · 0277-786X/01/$15.00
A dynamic phase modulating system using a Spatial Light Modulator (SLM) device for addressing the diffractive
structure could open up a wide range of application fields. Beside the already mentioned, also methods of non-
destructive testing based on interferometry, realization of micro-patterns for material processing [3] and applications in
the lighting technologies are possible.
Spatial Light Modulators (SLMs) became very important components in optical systems. Beside shutter and display
applications, also the possibility of phase modulation is more and more subject of current research. SLMs using two-
dimensional arrays of phase-modulating pixels are the basis of many new system proposals in adaptive optics, image
processing and optical switching [4-6]. With increased resolution and performance they could also compete with micro-
lithographic fabricated diffractive elements. So, a wider range of application fields can open up. Here, one can mention
laser beam splitting and beam shaping for projection and material processing application. Various kinds of applications
are possible, from non-destructive testing methods based on interferometry to visual media.

We already demonstrated the measurements done on the LC 1004 system from HoloEye[6]. Using the obtained results
concerning the modulation properties and using the design software DigiOpt for the calculation of diffractive structures
we did more effort in the investigation of the SLM system as diffractive element. The obtained results with the
investigated systems are promising for new generations of SLMs to work in dynamic phase-modulating systems.

We will present and discuss the results achieved with our Sony-LCD based SLM LC 2002 system. Here, we focus on
examples for various application fields. The system is a simple Fourier-setup. The input polarization and the analyzing
polarization state are variable. Most images in this paper showing reconstructed beam splitter elements, were recorded
while expanding the diffraction pattern by a microscope lens.

2. THE OPTICAL SYSTEM

2.1. Properties of the system architecture

The phase type diffraction grating, which optical thickness or refractive index changes regularly, and here especially the
multi-phase level to have continuous phase distribution, produces high diffraction efficiency. An exclusive phase
modulation is hard to achieve with nematic Liquid Crystal Displays (LCDs ), mostly the amplitude and phase
modulation is coupled [6-9].
Our new system is shown in Figure 1.

Figure1: Optical set-up with diode laser (right) and LC 2002 (center)

Proc. SPIE Vol. 4596 183

 The laser source is here a laser diode with an optical output of about 10 mW. The beam is expanded with a lens.
The correct input polarization is realized by the orientation of the laser polarisation and the lens forms a collimated
beam. The display is positioned in front of a second lens, which was implemented in order to realize the Fourier-
transform of the phase-modulated signal. The convergent beam is directed onto a microscope lens, which enlarges the
diffraction pattern onto a screen or scattering plate. A final analyzer separates a certain linear polarization state of the
out-coming elliptical polarized beam.
The SLM is the LC 2002 [4], with a pixel size of 32 µm and SVGA resolution. Its properties, especially the phase
behavior for different electronic settings was investigated quite intensively. These experiments will be described later.
For the complex modulation one can realize a phase shift of nearly 2π at 532 nm. Results of these measurements will be
published later. The measurement set-up, which is based on a Mach-Zehnder-interferometer in connection with a fringe
locker, is shown in Figure 2.

Figure 2: Measurement set-up for the determination of the properties of the LC 2002. In the center the Mach-Zehnder
interferometer and the LC 2002

Due to the pixel size the diffraction pattern is small

[10], just suitable for CCD-camera detection. But
for intensity measurements on a larger scale Amplitude- and Phase-Modulation
(P= 140, A= 5)
(Photo-detector) and of course for visualization and
4.5
recording with a high resolution digital camera, the
4
pattern was enlarged.
3.5
Beside the diode laser also a Nd:YAG laser at 532
3
nm, an Argon-Ion laser at 488 nm, and a He-Ne
2.5
laser at 632.8 nm was introduced into the system.
2
All Figures which show CCD-shots of the
1.5
diffraction pattern were taken from the Nd:YAG Signal in relative units
1
set-up. We obtained comparable results with the Phase-shift
0.5
blue line of the Argon-Ion laser, here also a 2π
0
phase shift can be realized. The Diode laser
0 100 200
wavelength application showed, even if the Gray-levels
maximum phase shift did not exceed 1.6π, a quite
good performance. Here, we could also measure
the highest percentage of intensity in the signal Figure 3: Representation of the coupled amplitude and
concerning the whole system. We could achieve phase modulation of the LC1004
25% of the laser input intensity in the signal from a

184 Proc. SPIE Vol. 4596

binary diffraction pattern. It was less, around 10%-15%, for the Nd:YAG and the Argon-Ion laser, obviously due to the
reflection and absorption processes and also caused by scattering.

2.2 Diffractive Structures

For the measurements we calculated diffraction gratings and beam splitters in order to derive the accuracy of the phase
shift level implementation indirect by measuring efficiencies and homogeneity of the diffraction pattern. In order to
overcome the speckle effect of non-periodic diffractive elements, we replicated the calculated pixel matrix over the
display size. Because of the 4:3 ratio of the display we calculated 160 pixel x 160 pixel matrixes and aligned them in a
4x3 mode. So, for illumination with a circular beam aperture at least 8 of them should be nearly complete illuminated.
For very simple patterns also a 80 x 80 matrix was enough to realize the diffraction pattern.
The iteration procedure of the calculation requires degrees of freedom for the noise. So, in general, especially for
image-signals, we took a 80 x 80 image embedded into a 160 x 160 matrix.
For addressing the eight level and 256 level elements we used the look up tables calculated and to be seen in [6] to
linearize the phase modulation. Here the gray values or the voltages could be adapted to the wavelength and the system
settings. The driver electronics is just able to set the voltages in 256 steps, what was also to be seen during the phase
shift and contrast measurements. So, we can not base on the fact, that really all 256 steps can be resolved electronically
and optically, respectively.

2.3. Optical performance

We started with binary gratings and binary diffractive

elements. Using the complex modulation values from the
display we had to choose two gray scale values with the
same transparency value concerning the amplitude
modulation and a π phase-shift between them. Figure 3
demonstrates the amplitude and phase modulation for a
certain driver and polarizer / analyzer setting. So a binary
diffractive element showing a circle as diffraction pattern
was addressed with a set of gray level pairs. These pairs were
determined either by the same transparency value or by the π
phase shift as criterion.
Figure 4 shows a comparison of the signal and the dc energy
of the diffraction pattern. The achieved ratio is quite
impressive. The matching pair was finally 148–176 for these
settings. Such a ratio of around 20:1 between signal and dc
intensity is sometimes hardly to achieve with micro-
lithographic fabricated diffractive elements. Nevertheless, Figure 4: Comparison of signal and dc
the light efficiency of the whole system is not comparable. energy (in µW) for different pairs of
But for the evaluation of the implementation of the phase
gray levels for the LC 1004
levels by the SLM, the efficiency values in the diffraction
plane are representative.

3. FROM GRATINGS TO DIGITAL HOLOGRAMS

Since we achieved quite impressive results of diffraction patterns, we calculated sample structures for certain
applications. It was not very difficult to realize some geometrical diffraction patterns like arrays of beams, off-axis
arrays, circles and marking patterns. For smaller diffraction angles - bigger structures - also the homogeneity of the
spots increased We do also calculate diffractive structures for micro-lithographic DOE, so, it was quite surprising, that

Proc. SPIE Vol. 4596 185

 the diffraction patterns showed such a good quality, even the matrices were just 160 x160 pixels. Using higher matrices,
like 256 x 256, the image pixels came closer and the speckle effect became more and more obvious.

In Figure 5 an example for a possible projection target is shown. As we mentioned also applications for visual media,
the next sample in Figure 6 show a possible way of dynamic information representation. Logos, signs and grayscale
images are possible to create.
Talking about medial applications, one will think immediately of colored diffraction patterns. We did some experiments
just overlaying the three laser beams in one system. Due to the different diffraction angles the images were of different
scale on the screen. So, we got a first idea, how those colored images could look like, but three systems addressing three
different diffractive structures, according to the RGB-channels of the image, had to be installed.

Figure 5: Diffraction pattern for a calibration target

Figure 6: Diffraction pattern of a DOE of a logo

186 Proc. SPIE Vol. 4596

 4. BEAM SHAPING ELEMENTS AND FRESNEL HOLOGRAMS

Diffractive optical elements are often designed in order to realize several functions, for instance beam focussing and
splitting. So different phase functions are super-positioned. One motivation for the additional implementation of a lens
phase is the dc intensity. While addressing an additional lens phase the diffraction pattern moves out of the focus and
the remaining dc intensity will, depending on the focal length, spread all over the diffraction plane. But here we have
two different functions, which both contain in optical realization a deviance to the designed structure.

So, we measured the quality of the beam splitter concerning homogeneity and appearance of a symmetrical image for
comparison with the single beam splitters. The sample structure was again the 1 to 8 beam splitter, which is to be seen
in Figure 7 . Here, just slightly within the noise the dc intensity could be guessed. Measurements using different lens
functions and scaling by using the digital zoom lead to the same efficiency and uniformity values as without any lens
phase. The homogeneity variations were from 8-15% and the signal to dc ratio varied from 5:1 to 7.5:1, but not
significant to the change of the lens phase.

Figure 7: Binary 1 to8 beam splitter superimposed with a Fresnel-Lens structure

In Figure 8 is shown the diffracting structure of an axicon on the laptop screen and the result of the diffraction: a ring
structure.

Figure 8: Example of an axicon structure

Proc. SPIE Vol. 4596 187

 While we are combining elements, we are turning from Fourier-holograms to Fresnel-holograms, which describe all
holograms realizing even three-dimensional objects in a certain distance from the hologram. In order to estimate the
resolution properties for the reconstruction of digital holograms calculated Fresnel-holograms can be used. Figure 9
shows the reconstruction of a Fresnel-hologram calculated at the BIAS (Institute of Applied Beam Technology
Bremen). Beside the resolution also the expected or desired reconstruction quality can be visualized.
The reconstruction of digitally recorded holograms was in general made digitally, using the computer [11]. A dynamic
phase modulating system opens here new possibilities for non-destructive testing devices based on interferometry. This
new approach is really the biggest challenge for
SLMs, but the obtained results are more than
just a feasibility study [12]. The pixel matrix
from the CCD-camera, which pixels are in the
range of 6-8 µm, can just resolve interferometric
fringe patterns representing a diffraction angle
of about 1 degree. The gray scale matrix is
comparable to the interference pattern stored in
holographic films. A common technique to
increase the diffraction efficiency of the
amplitude hologramm is a bleaching process,
when for instance the silver grains in the
emulsion transform to a complex structure
leading to a change of refractive index and
thickness. This is the transition from an
amplitude to a phase hologram. Theoretical
considerations give evidence that diffraction
efficiency of phase holograms can reach 37%.
F
Figure 9: Resolution measurements on a calculated Fresnel-
Hologram

In Figure 10 the reconstruction of a digital hologram in the phase mode of the LC display is shown. The image is quite
full of noise and speckles reduce the resolution drastically, but by moving a screen or a scattering plate along the optical
axis the three-dimensional impression was to be seen. Experimental problems consist here obviously again in the large
pixel size and the limited pixel number. Geometrical exact reconstruction is hard to achieve, just with varying
reconstruction wavelengths and a pixel size in the range of the recording device an efficient system can be build up.
As with the combined diffractive elements also here it is possible to superimpose, simple digitally to digital

Figure 10: Reconstruction of Digital Holograms: Single Hologram (left), digital superposition
of two digital holograms (right)

188 Proc. SPIE Vol. 4596

holograms, gray level matrixes , respectively. The result is again the realization of both optical functions, what is the
same that happens with double exposure techniques in interferometry, fringes are visible. The addition of two digital
phase masks leads in reconstruction to interferences, as to be seen in Figure 10.

4. APPLICATIONS AND BOUNDARIES

Some critical remarks are necessary.

The SLM system based on an off-the-
shelf LCD realizes dynamic
diffractive elements. But for technical
applications it is only for conditional
use, not only due to the displays
parameters, also the driver
possibilities and the liquid crystal
material itself leads to restrictions.
Furthermore, the strong intensity
dependence of the LCD should be
mentioned, which was just roughly
investigated [13].
So we have to assume that the present
systems suffers from its own
restricted properties. We hope to get
better with the LC 2002, with a pixel
size of 32 µm and an SVGA
resolution (Figure 10).

Figure 11: The LC 2002 system within an optical correlator

Also here, the addressing of each pixel is more advanced. However, the currently available displays show much smaller
pixel sizes and higher resolutions. The next years will bring us displays and devices, where high resolution diffractive
optical elements can be addressed.

5. CONCLUSIONS
Diffractive Optics turned from the laboratories into many optical systems and replaces refractive optics or leads to new
optical solutions. The introduction of a dynamic phase-modulating system is just a question of time. Actually we do not
know what kind of SLM device will take the lead in dynamic phase modulation. There are quite a few physical effects
and related devices. The LC technology is one of the most developed, but as we saw also restricted to compete with
micro-lithographic fabricated DOE.

REFERENCES
[1] S. Sinzinger, J. Jahns, "Microoptics", Wiley-Vch, 1999.
[2] M. Ferstl, R. Steingrüber, W. Fürst, S. Krüger and S. Teiwes, „High-efficiency computer-generated holograms,“ in
Micromachine technology for diffractive and holographic optics, Proc. SPIE 3879, S.H. Lee, J.A. Cox (Eds.) 147-156
(1999).
[3] T.Haist, E.-U. Wagemann, H.Tiziani, " Pulsed-Laser ablation using dynamic computer-generated holograms
written into a liquid crystal display", Appl Opt. 38, 1999, 428-430
[4]LC 1004 Plug & Display Developers Kit, technical data sheet, Holoeye GmbH, Worldwide Web:
http://www.holoeye.de

Proc. SPIE Vol. 4596 189

 [5] S. E. Broomfiled, M. A. Neil, E. G. S. Paige, R.H. Scarbrough, and G.G. Yang, " Programmable diffractive optics
with an FLC-SLM in a binary phase-only mode," in Holographic International, Proc. SPIE 1732, Y. Denisyuk, F.
Wyrowski (Eds.) 287-296 (1992)
[6] S. Krüger, J. Kamps, G. Wernicke, H. Gruber, N. Demoli, M. Dürr, S. Teiwes," Spatial light modulator system for
the application as a dynamic diffractive element and in optical image processing," Diffractive/Holographic
Technologies and Spatial Light Modulators VII, San Jose, USA 2000, I. Cindrich, S. H. Lee, R. L. Sutherland, (Eds.),
SPIE-Proc. 3951, 179-187
[7] D. A. Gregory, J. A. Loudin, J. C. Kirsch, E. C. Tam, F. T. S. Tu, ``Use of the hybrid modulating properties of
liquid crystal television,'' Appl. Opt. 30, 1374-1378, (1991).
[8] K. Ohkubo, J. Ohtsubo, ``Evaluation of LCTV as spatial light modulator,'' Opt. Comm. 102, 116-124, (1993).
[9] C. Soutar, K. Lu, ``Determination of the physical properties of an arbitary twisted-nematic liquid crystal cell,'' Opt.
Eng. 33, 2704-2712, (1994).
[10] S. Krüger, S. Teiwes, H. Gruber, M. Senoner, G. Wernicke," Digital forward-error-correction in opto-electronic
image analysis - A case study." Optical Pattern Recognition X, Orlando 1999, D. P. Casasent, T.- H. Chao (Eds.) SPIE-
Proc. 3715, 313-320
[11] C. Wagner, S. Seebacher, W. Osten, W. Jüptner, "Digital Recording and numerical reconstruction of lensless
Fourier holograms in optical metrology", Appl. Opt. 38, 1999, 4812-4820
[12] G. Wernicke, S. Krueger, H. Gruber, "New Challenges for Spatial Light Modulator Systems", Holomet 2000,
Strahltechnik 14, W. Osten and W, Jüptner (Eds.), 2000, 27-28
[13] S. Krüger, G. Wernicke, H. Gruber, N. Demoli, M. Dürr, S. Teiwes : Liquid crystal display as dynamical
diffractive element. Projection Displays VII, San Jose 2001, Ming H. Wu (Ed.), SPIE-Proc. 4294, 84-91

190 Proc. SPIE Vol. 4596