Beam Steering

 

Light Deflection (Chuck Titus)

OUTLINE:

  1. Introduction
  2. Uses
  3. Beam Steering Technologies
  4. Beam Steering with Liquid Crystals: Tunable Blazed Phase Gratings
  1. Digital Light Deflection
  1. Liquid Crystal Digital Light Deflector (LC-DLD)
    1. What it is and How it Works
    2. Materials
      1. Nematic liquid crystalline deflecting prisms
      2. Smecic A liquid crystalline deflecting prisms
    1. Design
    2. Fabrication
    3. Experimental Results
    4. LC-DLD in Action: a Short Movie Clip
  1. References

 

 

  1. Introduction

    2. Precise and controllable delivery of laser beams or other guided modes to a desired location is an important topic, with telecommunications, military, and other general industrial applications. The most common means of obtaining such delivery is the use of large (i.e. macroscopic) mechanically controlled mirrors. While this technology is mature, it is limited by the mechanical nature of mirror movement. Inertial properties of mechanically driven mirrors limit the speed with which steering direction can be changed. The other well-established beam steering device, the acousto-optic modulator, has a severely limited angular range.

    The potential for new beam steering applications and the inability of existing beam steering technologies to meet those needs have been the driving force behind research into new beam steering technologies.

    Two such efforts make use of liquid crystalline materials: tunable blazed diffraction gratings and liquid crystal digital light deflectors (LC-DLD). Steering direction is controlled via the electro-optic properties of liquid crystals. Accordingly, these devices are free of macroscopic moving parts and performance is relatively free of inertial problems exhibited by large motor-driven mirrors.

    However, efficiency and angular range of both grating and LC-DLD technologies are still an issue. In addition, proposed applications are placing greater demands on steering speed, so device speed has resurfaced as an issue despite the relative freedom of these devices from inertial concerns.

    The objective of our work is to find liquid crystal materials and determine the design approaches which will permit fabrication of an efficient liquid crystal digital light deflector with a wide angular range.

     

  2. Applications

There are many uses for beam steering. Our device may find uses in the following areas (among others):

 

  1. Beam Steering Technologies

Steering of beams or deflection of light can be effected by the following technologies, among others:

 

  1. Beam Steering with Liquid Crystals: Tunable Blazed Phase Gratings

    2. Electrically tunable blazed phase gratings, utilizing liquid crystal materials as the electro-optic filler, have been proposed and fabricated[B][C]. Devices of this sort have also been patented[D]. A uniformly thin layer of liquid crystal is sandwiched between parallel transparent substrates. One of the substrates is patterned with closely spaced electrodes, and the opposite substrate may be coated with an uninterrupted common electrode. Application of (roughly) linearly changing voltage over a small range of the electrodes induces a linear phase profile in that region. This is repeated periodically across the entire thin layer with the intent to form a "sawtooth" periodic phase profile characteristic of a blazed grating.

    With these devices, lack of efficiency and narrow angular range are problems. Most of the inefficiency is the result of the inability of liquid crystals to sharply change orientation at the blaze resets. Such changes in orientation are prohibited by the energy cost of such distortions in liquid crystalline layers. This blaze reset must then occur over some finite distance, resulting in the "flyback" regions as shown in the figure below. Any light propagating through those regions is not steered into the direction preferred by the remaining regions (which may posses the correct phase profile).

     

  2. Digital Light Deflection

An alternate strategy for development of a beam steering device is to build structures from simple two-position beam steering components. Each component consists of two optical elements, a passive birefringent deflector which deflects incident light of two perpendicular linear polarization orientations by different angles, and an optical switch which selects the polarization state to be passed on to the deflector. Such a device is called a Digital Light Deflector (DLD).

The idea of digital light deflection device is not new. Both inorganic crystals and liquid crystals have been utilized in DLD devices. The earliest DLD devices utilized inorganic birefringent materials for the deflecting elements and inorganic electro-optic materials for the polarization-switching elements[E][F][G][H]. These inorganic beam steering devices were very fast but also very large and consumed large amounts of power.

The Digital Light Deflector (DLD) can be used to deflect a laser beam without using moving parts. The DLD consists of cascaded pairs of deflecting and switching elements. In each stage, a birefringent element deflects the incident beam by two possible angles. The accompanying electro-optic switch selects the incident linear polarization state, thus deciding the output deflection for that stage. By cascading together N stages with successively increasing deflection angles, 2-to-the-power-N output angles are possible. Many earlier DLD devices utilized inorganic electro-optic materials and required high power. The only prior low-power DLD, using nematic liquid crystal deflectors, suffered from significant scattering losses and was not optimized for wide-angle steering.

 

  1. LC-DLD, Liquid Crystal Digital Light Deflector

The idea of using low molecular mass nematic liquid crystalline materials in such devices was published more than eight years ago[I].

Liquid crystalline materials are an attractive medium for polarization switches because power consumption is reduced (relative to inorganic switches). In addition, the deflection angle of a liquid crystal deflector is easily customized by choosing the appropriate wedge angle and/or material (i.e. choosing appropriate birefringence). This first Liquid Crystal Digital Light Deflector (LC-DLD) employed a nematic filler for the passive deflecting prisms and separate liquid crystal polarization rotators. Scattering losses were noted and there was some speculation about potential remedies (including the use of smectic A liquid crystals), but no work was conducted in that area. The device was fabricated for small angle use only, avoiding the steering errors discussed later in this report.

Based on prior reports, conclusions pertaining to the current state of the art can be drawn: An accurate, wide-angle, low-loss LC-DLD has not been reported prior to this project. Those are the goals of our research. Losses can be reduced by appropriate choice of material. Accuracy can be increased through optical design. Angular range is increased through both material and design improvements. Material and design considerations are discussed below, after a brief review of the operating principle of this type of beam steering device.

The published work of McRuer, et.al.[I] suggests possible solutions to the material problem in general terms (suggesting the use of smectic A liquid crystalline materials to fill the prisms) but does not say which smectic A liquid crystals are suitable for this application. Nor does that prior work attempt to build such a device. The angular range problem is related to the first, but is never mentioned. The accuracy range problem is not anticipated in the prior work on liquid crystal light deflection devices.

Over the course of the last year, our investigations led to an embodiment of the liquid crystal light deflection device which solves those problems.

Solutions to the problems were presented at a conference[J].

 

6.A. What it is and How it Works

The prismatic version of the digital light deflector consists of cascaded binary light switches (shown in the figure below). Each switch consists of a passive birefringent deflecting element, and a polarization rotation switch. Linearly polarized light is directed into the polarization rotation switch. That switch either propagates the incident light unaltered, or rotates the plane of polarization by 90 degrees. The resulting light is then directed into the passive birefringent deflecting prism. The prism consists of a birefringent medium (e.g. liquid crystal) and a means of supporting such (glass substrates). The optic axis of the birefringent medium is oriented parallel to the vertex of the prism. When light of polarization parallel to the optic axis is received by the prism, it path is deflected. Light of the orthogonal polarization is deflected by a lesser amount. The angular separation between these two deflections is the characteristic angular deflection of that prism.

Stacking together N such binary light deflecting switches (following figure illustrates for a two-stage device) can create a device which is capable of deflecting light by any of 2N angles.

 

6.B. Materials

The basic two-position LC-DLD component consists of two elements, each with a separate function. The polarization selection switch serves only to choose the orientation of linearly polarized light to be passed on to the deflecting prism. Ideally, the polarization selecting switches should be capable of switching quickly between two states whose transmitted linearly polarized light orientations are perpendicular.

The deflecting wedge then serves only to refract ordinary and extraordinary modes passing through it by different angles. It should do so without inducing significant losses. There are three types of losses. First are losses resulting from scattering induced by the liquid crystalline passive deflecting wedges. The second type of loss results from Fresnel reflections at refractive index mismatches. The third type of loss is leakage into an unselected steering direction, caused by imperfect polarization rotation or misalignment of a deflecting wedge with respect to its predecessor. Losses of the second and third type are important, but are not the primary subject of this study.

We worked with two types of liquid crystal in an attempt to find low-loss materials for deflecting wedges: low molecular mass nematics and smectic A phase materials.

Why not use inorganic crystals, machined to the desired prismatic shapes? The machining process is more time-intensive than fabrication of liquid crystal wedges. It is relatively simple to design and fabricate liquid crystal with any desired prism angle.

 

6.B.1. Nematic liquid crystalline deflecting prisms

In general, a high birefringence is desirable in order to make an effective passive deflecting element for wider steering angles. But the trade-off is that scattering in a liquid crystal is proportional to the square of the birefringence. Scattering is caused by fluctuations in the orientations of liquid crystal molecules. Such fluctuations are quite noticeable in nematic-type liquid crystals, which possess only orientational order.

Figure 3 shows the amount of light transmitted through a high birefringence nematic liquid crystal (Merck 18349, Dn=0.27) as a function of the thickness of material traversed. Transmitted intensity for this material is very dependent on thickness. By the time a beam has passed through 1.4mm of this substance, 70% of the incident intensity is lost to scattering. A similar test of the low material (Merck ZLI-2806, Dn=0.045) employed in our prototype produced only 7% reduction in transmitted intensity over the same range of thickness.

Larger beam deflections require passive deflecting prisms with steeper angles. Filling those prisms with a nematic liquid crystal produces unacceptably scattering losses, and those losses increase with increasing steering angle (i.e. increasing path length of light beam through nematic material). For this reason, another material must be used.

 

6.B.2. Smectic A phase deflecting wedges

Additional order (beyond the orientational order found in nematics), such as the one-dimensional positional order of smectic liquid crystals, restrains some fluctuation modes and reduces scattering. Smectic A phase liquid crystals are much less prone to the thermally-induced director fluctuations which cause light scattering in nematic materials[K].

Fabrication of smectic liquid crystal deflectors posed a unique challenge. Optical elements, constructed from mechanically rubbed polyimide-coated surfaces, are filled with the smectic liquid crystal material. The desired orientation of molecules is obtained by heating liquid crystal into its nematic phase. The element is then cooled to its smectic A phase. It is during the cooling process that many of these defects

Figure 4. Interdigitated smectic Ad layer structure of 8CB. Layer thickness decreases with decreasing temperature

 

are likely to occur. Most of the available room-temperature smectic A liquid crystals (e.g. 8CB, shown above) are polar molecules. These materials produce a partially interdigitated bilayer structure, the smectic-Ad phase. The degree of interdigitation, and thus the layer thickness, varies with temperature, as shown in figure 4. Because the layers are incompressible[L], layer shrinkage results in a tilting of the layers so as to occupy the same fixed volume. This tilting of layers may promote the formation of focal-conic defects in the smectic-Ad phase[M][N], shown in the photo below.

 

The search for a useful smectic A material centered on those materials which possess a room temperature monolayer smectic-A1 or non-interdigitated bilayer smectic-A2 phase. One such material is a blend of the n=6,7 homologues of dialkyazoxybenzene (DnAOB), shown in the following figure. These molecules possess only a small lateral dipole moment and should produce a monolayer structure.

Figure 5. Monolayer smectic A structure exhibited by blend of D6AOB and D7AOB. Layer thickness does not vary with temperature.

A weight-ratio blend of 46% D6AOB to 54% D7AOB exhibits an ordinary refractive index of 1.52 and an extraordinary refractive index of 1.72. This blend possesses a smectic A phase at room temperature. Prisms filled with this blend do appear to form significantly fewer defects. This blend does have a small drawback. Because of the azoxy bridge group, it absorbs the shorter visible wavelengths. This blend can be used as long as the absorption does not present a problem.

A comparison of the transmissive properties of this blend with that of the nematic discussed earlier is shown in the following plot:

 

 

6.C. Optical Design

 

The stacked prismatic architecture is not free of limitations. One of those limitations is that, as more and more stages with increasingly large characteristic deflections are stacked together, it becomes more difficult to obtain a constant angular separation of adjacent steering positions throughout the device’s entire angular range. This problem becomes more pronounced at the outer reaches of the angular range.

For LCDLD devices with small angular range, 2 degrees or less, this problem is not significant.

A design-based approach to significant reduction of this problem is discussed in our SPIE conference paper [J].

 

6.D. Fabrication

Fabrication of a four-stage demonstration device is described below:

 

  1. The angular separation between adjacent deflection positions was chosen to be twice the far field divergence angle of a common laboratory helium-neon laser (2 x 0.001007 radians).

  2. Choice of materials. In the first stage, the prism was to be filled with a low birefringence nematic liquid crystal, Merck ZLI-2806 which possesses a birefringence of Dn=0.04. The remaining three prisms were filled with a 50:50 (by weight) mixture of the n=6 and n=7 homologues of DnAOB (Di-n-alkyl-azoxybenzene) which possesses a birefringence of Dn=0.2.

  3. The optical design principles described above were utilized to design a four stage (four binary deflection switches) cascaded device.

  4. The prisms were fabricated with the design-specified angles. Each consisted of two glass substrates, coated with DuPont polyimide PI-2555. The polyimide surfaces were rubbed mechanically to induce orientation of the liquid crystal filler. The two substrates were adhered together in a wedge-configuration (utilizing apparatus shown below), with the rub directions of the two substrates antiparallel (not perpendicular!) to each other but parallel to the wedge vertex. These wedge cells were then filled with the prescribed liquid crystal filler.

  5. Twisted-nematic (TN) cells were fabricated for use as the polarization rotators.

  6. Each prism was affixed to a TN cell, not abutting the TN cell as shown in Figure 1, but at the design-specified angle (crudely depicted in the figure below).

  7. These four stages were then affixed to each other at complementary angles, the result being that all TN cell substrates were parallel to each other. In addition, the optic axes of the four birefringent prisms were aligned parallel to each other. The stages were affixed in increasing order of their characteristic deflection angles. All gaps were filled with Norland 65, used for index matching to reduce Fresnel reflections at glass surfaces.

  8. As with some of the prior art, an additional crosstalk filtering stage was added, consisting of a fifth TN cell and a dichroic linear polarizing film[I][O].

 

6.E. Experimental Results

A randomly polarized Helium-Neon laser (632.8 nm) was directed at the prismatic light deflector’s entry surface (at normal incidence). A dichroic linear polarization film was interposed between the laser and the entry surface, oriented with its "transmission axis" parallel to the optic axes of the prisms. With each possible permutation of the four stages’ TN cells being turned on or off (i.e. rotating or not rotating the light incident upon them), the transmitted beam was deflected in one dimension. Angular separation of each deflection position was roughly constant throughout the angular range of the device, although this has not yet been measured.

The transmitted intensity at each of the 16 possible deflection positions was measured and normalized to the intensity of the beam after passing through the initial dichroic linear polarizing film. These results are shown in the following plot.

6.F. Short Movie Clip

The following link starts a short movie clip, which shows the performance of a 4-bit (16 position) LCDLD fabricated in our laboratory:

[Start Movie]


Please Note: This movie clip is approximately 37MB, so it is not recommended if you have a modem connection.
  1. References
  1. D. M. Bloom, Proc. Of SPIE 3013, (1998).

  2. R. Matic, Proceedings of the SPIE, V. 2120, pp. 194-205, (1994).

  3. P. McManamon, et. al. Proceedings of the IEEE, V. 84, pp. 268-297, (1996).

  4. T. A. Dorschner and D. P. Resler, U.S. patent 4964701, (Oct. 23, 1990).

  5. W. Kulcke, et. al., IBM Journal of. Research and Development, V. 8, pp. 64-67, (1964).
  6. R. A. Soref and D. H. McMahon, Applied Optics, V. 5,pp. 425-434, (1965).

  7. T. J. Nelson, "Digital Light Deflection", Bell Syst. Tech. J. 43, pp. 821-845, 1964.

  8. U. J. Schmidt, "A High Speed Digital Light Beam Deflector", Phys. Lett. 12, pp. 205-206, 1964.

  9. R. McRuer, L. R. McAdams, and J. W. Goodman, "Ferroelectric liquid-crystal digital scanner", Opt Lett. 15, pp. 1415-1417, 1990.

  10. C. M. Titus, P. J. Bos, Photonics West ’99, San Jose, CA, January 29, 1999, to be published in SPIE Proceedings V. 3633, (1999).

  11. P. G. de Gennes, J. Prost; The Physics of Liquid Crystals; 2nd edition, p. 371; Oxford Univ. Press, New York, 1993.

  12. S. Chandrasekhar; Liquid Crystals; 2nd edition, p. 302; Cambridge Univ. Press, New York, 1992.

  13. Y. Ouichi, Y. Takanishi, H. Takezoe, and A. Fukuda, Jpn J Appl Phys 28, pp. L487-489, 1989.

  14. Y. Ouichi, Y. Takanishi, H. Takezoe, and A. Fukuda, Jpn. J. Appl. Phys. Part I 28, pp. 2547-51, 1989.

  15. G. Hepner, IEEE Journal of Quantum electronics, QE-8, pp. 169-173, (1972).