Switchable Retroreflector Films for Enhanced Visible and Infrared Conspicuity

Home , Diffuse reflection, Night vision, Retroreflector

Switchable Retroreflector Films for Enhanced Visible and Infrared Conspicuity: Naked eye response switchable retroreflectors with integration into an optical interrogation and response system

A dissertation submitted to the Division of Research and Advanced Studies of the University of Cincinnati

in partial fulfillment of the requirements for the degree of

DOCTOR OF PHILOSOPHY (Ph.D.)

In the Department of Electronic and Computing Systems

of the College of Engineering and Applied Science

2015

Phillip Jordan Schultz

B.S., University of Cincinnati, 2009

Dr. Jason C. Heikenfeld, Committee Chair

Abstract

Retroreflectors are common optical devices used in several commercial conspicuity applications. Since the 1980's, there have been several investigations aimed at switching the retroreflected signal. Most of the developed switching technologies are based on MEMS actuation and quantum mechanics, among others.

These devices were aimed at the free-space optical communications markets, lending them to have high operating speeds. However, they are limited by single devices/small areas, spectral range (short wave infrared), low contrast, physical rigidity, and high fabrication costs. For naked eye conspicuity applications, these limitations are not suitable.

The approach presented in this dissertation does not provide the high switching speeds of the previous MEMS and MQW methods, but provides superior performance in nearly all other metrics of interest to visual conspicuity applications, most notably safety.

These metrics include high contrast of >2000:1 at 635 nm, and >400:1 at 850 nm, large area of 75 cm2 +, visible and infrared operational spectral range from 400 - 1600 nm, input angles of ±38o, optical efficiencies of 25%, low power, thin and flexible construction at <0.6 mm thick, and switching speeds sufficient for rapid visualization (<100 ms).

This dissertation investigates several novel types of switchable retroreflectors, and compares them to a set of desired metrics for naked eye applications. It determines that the use of a polymer dispersed liquid crystal (PDLC) based switchable retroreflector would be the ideal choice for such applications. This dissertation further achieves a complete system level integration and demonstration, including development of an encoded laser interrogation system. It provides optical models for both day and night operation. It also provides day and night field demonstration results using visible (2.5 -

10 mW - 635nm) and infrared (1.5 mW - 850 nm) light sources (with night vision) out to

400 meters to validate the technology’s viability and potential use. Overall, the goal of this dissertation is to introduce and demonstrate an improved switchable retroreflector and electronic system satisfying all of the ideal requirements for enhanced naked-eye optical conspicuity.

iii

Acknowledgements

I would not have been able to complete, much less pursue, this dissertation without the help and support of several individuals throughout my personal and professional life. I would like to thank some of them here.

First, I would like to thank some of those closest to me. I would like to thank my wonderful wife, Carrie, for being there when I needed her on several levels. Though there were ups, and downs, she hung in there and allowed me to finish without worries.

I would like to thank my son, Sullivan, for always being there and putting a smile on my face when I needed it most. Next, I would like to thank my parents for emphasizing the importance of education and doing everything to the best of your ability. I would also like to thank my twin brother, Alex, for always giving me that on-the-spot competition that you cannot find anywhere else.

I would like to thank my advisor and mentor, Dr. Jason Heikenfeld. His support, and seemingly non-human levels of energy, helped push me to work harder to achieve what I wanted on a daily basis. His advice, encouragement, and guidance, allowed me to complete this dissertation knowing I would be proud of the work I had accomplished. I would also like to thank all of my committee members for their encouragement and acceptance throughout my Ph.D. candidacy.

I would like to thank my financial, material, and equipment support from , Xetron, and Reflexite. Without their help, none of this would have been possible.

I would also like to thank everyone from NDL. Their encouragement, friendship, training, and assistance, helped get me through the occasional grind and allowed me to understand what a cohesive work environment can do. I would especially like to thank

Brad Cumby, who took time out of his day to help me collect data for my various field demonstrations. Life would have been a lot more difficult without that. I still owe him.

I would finally like to thank all of the faculty and staff of the University of

Cincinnati. Their assistance, support, and guidance allowed me to pursue my aspirations for the last 9+ years with relative ease. I am proud of my undergraduate and graduate tenure here. I will encourage others to follow the same path.

Table of Contents

Abstract ...... i

Acknowledgements ...... iv

Table of Contents ...... vi

List of Figures ...... xi

List of Tables ...... xiv

Chapter 1: Introduction ...... 1

1.1 Introduction to this Chapter ...... 1

1.2 Conspicuity ...... 1

1.3 Retroreflectors: Theory in Brief ...... 2

1.4 Research Aims and Outline ...... 5

1.5 References...... 7

Chapter 2: Background and Prior Art ...... 8

2.1 Introduction ...... 8

2.2 Prior Art ...... 8

2.2.1 Micro-Electro-Mechanical Systems (MEMS) Retroreflectors ...... 9

2.2.2 Multiple Quantum Well (MQW) Retroreflector ...... 13

2.3 Display Technologies ...... 16

2.3.1 Electrowetting ...... 16

2.3.2 Liquid Crystal ...... 19

2.4 Summary ...... 23

2.5 References...... 24

Chapter 3 - Comparison of 5 Types of Switchable Retroreflectors ...... 29

3.1 Introduction ...... 29

3.2 Background ...... 29

3.3 Choosing the Retroreflecting Optical Film ...... 30

3.4 Selection of Electrical Modulation Methods ...... 33

3.4.1 Electrowetting ...... 34

3.4.2 Liquid Crystal ...... 34

3.5 Optical Efficiency Model ...... 35

3.6 Electrowetting Lenslet Scattering ...... 36

3.6.1 Fabrication and Construction ...... 36

3.6.2 Electrical Switching ...... 36

3.6.3 Retroreflection Results ...... 38

3.6.4 Discussion ...... 39

3.7 External Electrowetting Light Valve ...... 40

3.7.1 Fabrication and Construction ...... 40

3.7.2 Electrical Switching ...... 40

3.7.3 Preliminary Results ...... 42

3..7.4 Discussion ...... 43

3.8 Integrated Electrowetting Light Valve ...... 43

vii

3.8.1 Construction and Fabrication ...... 43

3.8.2 Electrical Switching ...... 45

3.8.3 Retroreflection Results ...... 45

3.8.4 Discussion ...... 46

3.9 Conventional Liquid Crystal Light Valve ...... 46

3.9.1 Construction and Fabrication ...... 46

3.9.2 Electrical Switching ...... 46

3.9.3 Retroreflection Results ...... 48

3.9.4 Discussion ...... 49

3.10 Liquid Crystal Scattering ...... 49

3.10.1 Construction and Fabrication ...... 49

3.10.2 Electrical Switching ...... 51

3.10.3 Retroreflection Results ...... 51

3.10.4 Discussion ...... 52

3.11 Discussion, Conclusions, and Demonstration ...... 52

3.12 References...... 55

Chapter 4 – Switchable Retroreflector and Interrogation System ...... 58

4.1 Introduction ...... 58

4.2 Background ...... 58

4.3 System Description ...... 59

4.3.1 Introduction ...... 60

viii

4.3.2 Retroreflector Film Characterization ...... 61

4.3.3 Electronics and Optics Design and Integration ...... 67

4.4 Optical Model for Maximum Distance ...... 73

4.5 Demonstration ...... 78

4.6 Discussion and Conclusions ...... 81

4.7 References...... 82

Chapter 5: Applications and Future Work for Switchable Retroreflector Films and

Visual Identification Systems ...... 84

5.1 Introduction to this Chapter ...... 84

5.2 Retroreflective Tag ...... 84

5.2.1 Tag Specific Applications ...... 84

5.2.2 Tag Specific Future Work ...... 86

5.3 Visual Identification System ...... 92

5.3.1 System Level Applications ...... 92

5.3.2 System Level Future Work ...... 92

5.4 Conclusion ...... 94

5.5 References...... 95

Chapter 6: Summary and Conclusion ...... 96

6.1 Introduction to this Chapter ...... 96

6.2 Summary of System Performance and Limitations ...... 96

6.3 Separation of Work ...... 98

6.4 Conclusion ...... 99

Appendix A – Fabrication Sheets ...... 101

A.1 Integrated Electrowetting Light Valve Retroreflector ...... 101

A.1 Polymer Dispersed Liquid Crystal Retroreflector ...... 103

List of Figures

Figure 1.1. Examples of civilian applications for conspicuity ...... 1

Figure 1.3. Simplified retroreflector basics...... 3

Figure 1.4. Glass bead retroreflector basics...... 3

Figure 1.5. Corner cube retroreflector basics ...... 4

Figure 1.6. Full cube retroreflector basics...... 4

Figure 2.1. Tilting mirror MEMS based switchable retroreflector ...... 10

Figure 2.2. MEMS based retroreflective modulators ...... 10

Figure 2.3. Structure and operation of the MARS based switchable retroreflector...... 11

Figure 2.4. Examples of deformable mirror based MEMS retroreflectors ...... 12

Figure 2.5. Structure of a multiple quantum well...... 13

Figure 2.6. Quantum well conduction and valence bands ...... 14

Figure 2.7. Quantum well absorption spectra for various applied biases...... 15

Figure 2.8. Arrayed Multiple Quantum well retromodulator ...... 16

Figure 2.9. Basic concept of an electrowetting system ...... 17

Figure 2.10. Liquid crystal basics...... 211

Figure 3.1. Corner cube retroreflectors...... 322

Figure 3.2. Electrowetting lenslet retroreflector...... 377

Figure 3.3. Collinear images of the lenslet retroreflectors ...... 377

Figure 3.4. Retroreflection vs. input angle for integrated electrowetting lenslets...... 388

Figure 3.5. External electrowetting light valve retroreflector ...... 411

Figure 3.6. Collinear images of the external electrowetting light valve retroreflector .. 411

Figure 3.7. Retroreflection vs. input angle for the external electrowetting light valve

retroreflector...... 422

Figure 3.8. Integrated electrowetting light valve ...... 444

Figure 3.9. Collinear images of the integrated electrowetting light valve retroreflector 444

Figure 3.10. Liquid crystal light valve retroreflector ...... 477

Figure 3.11. Collinear images of liquid crystal light valve retroreflector...... 477

Figure 3.12. Retroreflection vs. input angle for the TNLCD retroreflector...... 488

Figure 3.13. Polymer dispersed liquid crystal retroreflector ...... 500

Figure 3.14. Collinear images of the polymer dispersed liquid crystal scattering

retroreflector ...... 500

Figure 3.15. Retroreflection vs. input angle for the polymer-dispersed liquid crystal

retroreflector...... 522

Figure 3.16. Field demonstration of a switchable retroreflector using night vision...... 544

Figure 4.1. Block diagram of the interrogator and receiver/retroreflector...... 600

Figure 4.2. Fundamentals of corner cube retroreflectors...... 622

Figure 4.3. Polymer dispersed liquid crystal switchable retrorefelective film ...... 633

xii

Figure 4.4. Operation of the polymer dispersed liquid crystal switchable retroreflector

...... 644

Figure 4.5. Spectrum response characterization setup ...... 655

Figure 4.6. Retroreflection intensity vs. input angle for the PDLC retroreflector...... 655

Figure 4.7. Diffraction intensity at various input angles as a function of percent

retroreflection vs. wavelength...... 666

Figure 4.8. Interrogator and electronics block diagram feeding into its optics train. ... 688

Figure 4.9. Receiver/switchable retroreflector driver electronics block diagram...... 69

Figure 4.10. Custom optical bandpass filter used for daytime demonstrations...... 700

Figure 4.11. Images of the assembled system...... 711

Figure 4.12. System temporal response...... 722

Figure 4.13. Diagram of the optical model of the system...... 733

Figure 4.14. Theoretical plots of retroreflected irradiance vs. distance...... 777

Figure 4.15. Example daytime demonstration of a switchable retroreflector...... 79

Figure 4.16. Example nightime demonstration of a switchable retroreflector...... 800

Figure 5.1. Examples of enhancing conspicuity by flashing ...... 855

Figure 5.2. Potential improvements to the PDLC based switchable retroreflector ...... 877

Figure 5.3. Surface reflections of planar and thermoformed PDLC retroreflector ...... 89

Figure 5.4. Thermoformed surface reflection initial results...... 900

xiii

Figure 5.5. Far field (~10 m) observation of a planar and thermoformed tag...... 911

List of Tables

Table 2.1. Summary table of prior switchable retroreflector devices...... 233

Table 3.1. Base retroreflector technology comparison...... 300

Table 3.2. Comparison of key attributes of the five types of switchable retroreflectors...... 533

Table 4.1. Laser diode limits for receiver detectable spot sizes and associated irradiances with/without a filtered photodiode ...... 711

Table 4.2. Values used for optical model calculations ...... 788

Table 6.1. Specifications and results for visual identification system and its various components ...... 977

xiv

Chapter 1: Introduction

1.1 Introduction to this Chapter

This chapter first discusses the motivation behind this dissertation based on improving visual conspicuity for safety applications. It identifies currently used methods, their limitations, and their need for improvements. It calls for an alternative method and provides a basic theory behind the main technological component used throughout this dissertation, the retroreflector. This chapter then provides the specific dissertation aims along with an outline of this document.

1.2 Conspicuity

Conspicuity, by definition, is the state or quality of being obvious to the eye or mind, attracting attention. Something conspicuous will stand out compared to its surroundings so an observer can be aware of and/or act upon it in some way. Conspicuity is an increasingly important subject when it comes to applications involving safety and information. Figure 1.1 shows examples of civilian applications. Reflective vehicle

Figure 1.1. Examples of civilian applications for conspicuity

1 markings, road markings, road signs, reflective clothing, and flashing lights, are all common methods of making an object or person more conspicuous when a potential hazard may be present. Overall, conspicuity allows an observer to be aware of and act upon a given situation for guidance and/or safety purposes.

There is interest for improved conspicuity in a variety of settings. Most devices are static reflectors (normal or retroreflective), and others are continuously active in all directions. There are no solutions that possess properties of being active only when desired and directional. Switchable and directional devices would provide several advantages. They would be more conspicuous when switching, as the eye responds better to flashing. There is a potential for data transmission to provide specific location information or even distress calls. They would have lower power consumption than those that are continuously running. They would also have increased security, as they are only active when needed. Ideally, they would not be emitting in all directions, like a lamp or beacon, but rather directional, like a retroreflector, which returns incident light back to the source. As is the case, the basis of all devices and systems discussed herein will be on retroreflective technologies.

1.3 Retroreflectors: Theory in Brief

Retroreflectors are a special type of reflector that differ from diffuse reflectors

(reflects in all directions) and specular reflectors (mirror – reflection angle equals incident angle about the surface normal axis) in that they reflect light back to the source in a narrow cone (low divergence). Figure 1.2, below, shows this concept. A majority of the optical power is concentrated in the center ~±0.2o cone of retroreflection and dissipates continually out to ~±2o cone. An observer standing next to the illumination axis will perceive the retroreflected light as 100’s times brighter than a diffusely reflecting surface.

The high brightness, predictable operation, and electrically passive properties lend them

Figure 1.2. Simplified retroreflector basics showing the incident light

retroreflecting back in a low divergence cone [1]. to be useful in many general conspicuity and engineering applications. They are well suited, and ubiquitous, in road signs and markings [2] as drivers are readily in line with

Figure 1.3. Glass bead retroreflector basics showing a) standard structure and

optical path of an incident ray, b) an industry standard glass bead array, and c)

typical application in the garment industry [7].

Figure 1.4. Corner cube retroreflector basics. a) Cross sectional view showing

basic structure and optical path of an incident ray, and b) angular scanning

electron microscope image with optical path.

Figure 1.5. Full cube retroreflector basics. a) Top view showing array and

active area [8] and b) angular sketch showing complex structure [9]. the observation axis due to the illumination from vehicle headlights. Additional applications include surveying [3], range-finding [4], manufacturing line measurement and counting systems [5], and free-space communications [6].

There are three main retroreflective structures all with varying degrees of operational input angles and optical efficiencies. Figures 1.3-1.5 show the bead, corner cube, and full cube structures, respectively. All three are manufactured in lightweight,

4 flexible sheeting form using micro-scaled structures. A more thorough review of the different types of retroreflectors will be presented in a subsequent chapter.

1.4 Research Aims and Outline

Past switchable, retroreflective technologies posed several limitations when applying them to naked eye optical conspicuity applications, as will be reviewed in the following chapter. Chapter 2 will also provide theory on electrowetting and liquid crystal modulation, as these are the basic technologies used for several of the devices investigated in this dissertation.

In Chapter 3, this dissertation compiles a list of ideal requirements for an optical conspicuity system and takes a fresh look at alternate technologies to fulfill them. This leads to The First aim, which is to evaluate which retroreflective and modulator technologies are ideal for multiple applications and choosing an appropriate combination for further development and system level integration. Chapter 3 reports our work on comparing several metrics of the three basic retroreflective structures and choosing the most fitting based on a list of ideal requirements. Chapter 3 then compares performance specifications and measured parameters of the modulator types to make a recommendation for future development and system level integration.

The Second aim is to use the technological decision from aim one and develop a simple and scalable switchable retroreflector tag fabrication process, with an integrated camouflage patterning, and evaluate anti-glare thermoformed patterns. The fabrication process is discussed in brief in Chapter 3, with more details provided in Appendix B. A section in Chapter 5 discusses the thermoforming anti-glare patterns.

The Third aim is to design and test small-sized, optical receiver and switchable retroreflector control electronic circuit to be integrated into an optical interrogation and

5 response system for rapid visual identification applications. The electronic circuit receives an encoded optical signal, processes it, and then drives the switchable retroreflector to retroreflect light back to the source for viewing. Northrop Grumman

Corp. (NGC) Xetron division participates in the completion of this aim, which is discussed in Chapter 4.

The Fourth aim is to design and optics and electronics for a laser interrogator and viewing system. This system has optics for laser beam shaping and driving electronics to encode the laser with user interaction. This is used for interrogating the switchable retroreflector and receiver electronics and viewing the retroreflected signal. Chapter 4 discusses this aim, of which Xetron supported as well.

The Fifth aim is to demonstrate the entire system and compare the results to a theoretical optical model of system performance. The optical model takes into consideration several variables to provide viewable distances under certain conditions.

It also evaluates how retroreflected beam divergence changes with the size of individual retroreflectors and size of the tag. Chapter 4, also, discusses this aim.

Chapter 5 then discusses applications and future work of switchable retroreflectors and the visual identification system developed in this dissertation. Chapter 6 then provides a summary and draws an overall conclusion of the work presented herein.

1.5 References

[1] www.atssa.com, “Retroreflection Basics.” [Online]. Available: http://www.atssa.com/Retroreflectivity/WhatisRetroreflectivity.aspx.

[2] G. V. L. Jr, “Reflectors used in highway signs and warning signals, Parts I, II, III,” Journal of the Optical Society of America ( …, pp. 462–487, 1940.

[3] www.engineersupply.com, “Seco 62 mm Standard Prism Assembly.” [Online]. Available: http://www.engineersupply.com/seco-62-mm-standard-prism-assembly- with-5.5-by-7-inch-target.aspx. [Accessed: 18-Aug-2013].

[4] E. Laskowski, “Range finder wherein distance between target and source is determined by measuring scan time across a retroreflective target,” US Patent 4,788,441, 1988.

[5] Banner Engineering, “PicoDot PD Series.” [Online]. Available: http://www.bannerengineering.com/en-US/products/8/Sensors/38/Laser- Sensors/67/PicoDot-PD-Series/. [Accessed: 18-Aug-2013].

[6] L. Zhou, J. Kahn, and K. Pister, “Corner-cube retroreflectors based on structure- assisted assembly for free-space optical communication,” Microelectromechanical Systems, Journal of, vol. 12, no. 3, pp. 233–242, 2003.

[7] 3M, “Scotchlite Store.” [Online]. Available: http://scotchlitestore.3m.com/. [Accessed: 19-Aug-2013].

[8] J. Lloyd, “A brief history of retroreflective sign face sheet materials,” 2008. [Online]. Available: http://www.rema.org.uk/pdf/history-retroreflective-materials.pdf .

[9] K. Smith, “Retroreflective sheeting including cube corner elements,” US Patent App. 13/430,137, vol. 2, no. 12, 2012.

Chapter 2: Background and Prior Art

2.1 Introduction

This chapter critically reviews MEMS and semiconductors for use in communications and general optical switching applications. This chapter also provides theory on electrowetting and liquid crystal, as these are the two main technologies behind the devices presented in the following chapter, as alternatives for improved switchable retroreflection for naked-eye conspicuity. The main goal of this chapter is to provide pre-requisite background information for the subsequent chapters.

2.2 Prior Art

As mentioned in the previous chapter, retroreflectors have a wide range of applications. Most, however, require only static retroreflection. When it comes to dynamic activation and free-space communications, static operation is obviously a major shortcoming. There have been several attempts at making retroreflectors switchable.

Early attempts included the use of the Stark Effect [1] in molecular gases to modulate light [2] by causing an optical transmission spectral shift. This was placed over a large corner cube [3] to provide one of the first means to modulate a retroreflector. A US

Patent was granted for specific use in a friendly identification system [4] for this work. A similar method included the use of a lanthanum-modified lead zirconate titanate (PLZT) crystal sandwiched between a polarizer/analyzer optical path placed over a retroreflector

[5]. PLZT is a piezoelectric material that, in the presence of a voltage, responds by physical re-orientation of domains, causing a change in the material’s optical birefringence, allowing modulation. This, also, was granted a US Patent for use in a friendly identification system [6]. Another attempt includes the use of a voice coil behind one of the three walls of a corner cube to allow for electrical modulation of the

8 retroreflected light by changing the wall’s position. Yet another attempt was an unpowered version whereby a thin Mylar sheet was stretched onto a frame to have an acoustic resonance of 3 kHz. This replaced one of the sidewalls of a corner cube, which could then be modulated by a change in external pressure differentials [7].

All of these attempts suffered from several disadvantages including low retroreflected optical powers, high interrogating optical powers, use of long wave lasers, low spectral bandwidths, low-pressure operation, high electric field strengths, and large physical sizes. In general, the most important trends were to decrease size, weight, and power consumption. In the present case, a MEMS approach was attempted, offering potential improvements in all desired attributes.

2.2.1 Micro-Electro-Mechanical Systems (MEMS) Retroreflectors

Micro-electro-mechanical systems make use of semiconductor fabrication techniques to create micro-scaled, electrically actuated, mechanical components, as the name suggests. Their small size lends them to be lightweight and have low power operation, further reducing weight as they allow for smaller batteries in remote applications. The first attempt was published in 1995 [8] and is based on a planar fabricated, micro-assembled, corner cube where one of the side wall mirrors can be electrically actuated to tilt ~3 degrees, allowing switching of the retroreflected signal.

Figure 2.1 shows this device. This type of modulator has been investigated, with slight variation, for several different publications [9–12].

An alternative method using MEMS based actuation includes modulation of the retroreflected signal’s path length. This has been accomplished a couple of ways. One uses electro-thermally active interdigitations that create a vertical deflection of a stage with a retroreflective optical element mounted to it [13] (Figure 2.2 (a)). The second

Figure 2.1. Tilting mirror MEMS based switchable retroreflector [10].

Figure 2.2. MEMS based retroreflective modulators using path length

modulation by use of (a) electro-thermally actuated stage with a retroreflective

optical element [13] and (b) an active phase grating [14]. method uses an electro-active reflective phase grating as one of the side wall mirrors

[14], and can be seen in Figure 2.2 (b). The phase grating is composed of an array of ridges to which an electric potential is applied, causing the ridges to deflect downward, effectively creating an increased optical path length [15]. As both of these devices are changing the path lengths, it translates to a change in phase of the retroreflected signal,

10 which requires specialized measurement systems to detect. The human eye is incapable of doing so, forcing these devices to be used strictly for free-space communications, and impractical for a naked-eye optical conspicuity system.

Another MEMS embodiment is a transmissive element that makes use of a deformable membrane that can change the device’s surface optical properties from being a specular reflector to an anti-reflective coating. This is known as a Mechanical

Anti-Reflection Switch (MARS) [16], which was placed over a corner cube retroreflector for a retrocommunication application [17-18]. Figure 2.3 shows its structure and operation. The MARS device makes use of a quarter-wave (λ/4) thick transparent

Figure 2.3. Structure and operation of the MARS based switchable retroreflector

in its (a) specular reflection state, and (b) retroreflection state. suspended membrane over an air gap of a thickness of an odd multiple of a quarter- wave (mλ/4). In this condition, the optical path is an even multiple of λ/4, which forms a high reflectivity mirror, as shown in Figure 2.3 (a). When a voltage is applied, the membrane deflects to the substrate, effectively eliminating the air gap, providing a single layer anti-reflection coating with an odd multiple of λ/4 optical path, making the MARS device transparent, and allowing retroreflection.

Single sidewall mirror deformation is, perhaps, the most successful MEMS based switchable retroreflective device. As corner cubes need all sidewalls properly aligned and planar, deforming one of them will change the optical path preventing three wall reflections. Figure 2.4 shows examples of this.

Figure 2.4. Examples of deformable mirror based MEMS retroreflectors using

deflection in an array of ridges [20]. (c) shows a basic macro-view example [20].

(b-c: © Boston Micromachines Corporation.)

Figure 2.4 (a) shows an earlier attempt using an array of pores to create voids in the substrates where the mirror can be deflected when a voltage is applied [19]. This later progressed to an array of ridges [20–22], similar to the grating of Figure 2.2 (b), but with larger dimensions. This latter form has since progressed to a commercialized product, being sold by Boston Micromachines Inc. [23].

Overall, MEMS devices offer relatively high speeds up to approximately 1 megabit per second. However, they suffer from low optical performance (contrast, efficiency, incident angle), high power density, small area (not easily scalable), and low fabrication yield (mostly due to them being difficult to fabricate). Many of them are also

12 limited to a narrow spectral band (~100nm), forcing limited use at specialized wavelengths, mostly in the infrared, for communications applications.

2.2.2 Multiple Quantum Well (MQW) Retroreflector

Multiple Quantum wells are heterostructure semiconductor electro-absorption devices comprised of thin (~10nm) layers of alternating bandgap materials sandwiched between bulk semiconductors. Figure 2.5 shows an example.

Figure 2.5. Structure of a multiple quantum well [24].

These alternating bandgaps create energy wells on both the conduction and valence bands where electrons and holes are confined. As such, this confinement gives way to slightly more separated electron and hole wavefunctions (compared to the bandgap) [25]. This can be explained more thoroughly by Schrodinger’s wave function [26], and can be seen in Figure 2.6, which shows the waves confined in a well. The subsequent layers’ larger bandgaps (surrounding the well) ultimately affect these waves, due to the small thickness of the layers. This energy separation between the electrons and holes (through the bandgap) has an associated zero bias optical absorption related to a range of photonic energy levels. When an external

Figure 2.6. Quantum well conduction and valence bands showing the first few

electron and hole energy levels and their associated probability wave functions

without an applied bias (E = 0) and with an applied bias (E ≠ 0). [25] field is applied perpendicular to the quantum well layers, the energy levels shift accordingly, and the electrons are pulled to one side of the well, and holes to the other. In the case of Figure 2.6, the hole energy levels also raise, which effectively shortens the energetic distance between the electrons and holes. The shortened distance shifts the lower edge of the spectra to allow lower photonic energy levels to be absorbed and create an exciton (electron-hole pair). Figure 2.7 shows an example spectral shift. In gallium arsenide based semiconductors, exciton size is approximately 30 nm, substantially larger than the width of the well. This causes the excitons to be confined, and compressed within the well without immediately tunneling into the bulk material and field ionizing. This gives rise to peaks in the

Figure 2.7. Quantum well absorption spectra for various applied biases. As the

applied voltage is increased, the absorption peak shifts to lower photonic energy

levels [27]. absorption spectra. The confined exciton induced peaks and shift in the spectra is known as the Quantum Confined Stark Effect (QCSE) [28].

Multiple quantum well modulating retroreflectors exploit the QCSE phenomenon by placing arrayed transmissive versions of these devices over top of bead or corner cube retroreflectors to modulate the retroreflected light [24, 29].

Figure 2.8 shows the array. MQW modulating retroreflectors offer high speeds in the

Mbps range [30–32] and have been tested over long distances [33]. This lends them to be useful in long range optical communications systems where large amounts of data need to transmit in a lightweight package. However, they have relatively low contrast, are narrow in spectral bandwidth, have small apertures, have relatively high power over larger areas, and are expensive to fabricate.

Figure 2.8. Arrayed Multiple Quantum well retromodulator [30]

2.3 Display Technologies

In an effort to address the limitations (scaled size, contrast, spectral bandwidth, etc.) of the MEMS and MQW based switchable retroreflector devices, alternative technologies were investigated. These technologies are all based on inherently transmissive display technologies, most notably liquid crystal and electrowetting. It is appropriate to preview the theory for these display technologies to allow for a better understanding of the devices and materials that will be presented in Chapter 3.

2.3.1 Electrowetting

Electrowetting is an electrostatically induced modification of the wetting abilities of a surface. It was first described by Gabriel Lippman in 1875 with his analysis of mercury’s wetting properties under a directly applied potential on a metal substrate [34].

The word ‘electrowetting’ was first used in 1981 with a description of a new display technology [35]. A majority of the early work used liquid metals as the manipulated fluid.

Due to electrochemically induced reactions when applying direct current to water

(electrolysis), a new form of electrowetting was pursued, which employed an insulating

16 material between the conductive fluid and the electrode [36]. Electrowetting-on-dielectric

(EWOD) is the term coined for such a system, which it is the basis for the electrowetting theory and devices discussed here.

Figure 2.9 shows a typical electrowetting-on-dielectric system. Such a system consists of a polar fluid (water) sitting on top of a hydrophobic-dielectric coated electrode, surrounded by an immiscible, non-polar, insulating fluid (oil). Since the surface is hydrophobic (low surface energy), the zero potential equilibrium state takes on the form shown in Figure 2.9 (a), characterized by the polar fluid forming a spherical shape. It can be mathematically described with a balance of a three-phase interfacial surface tension ( γ ) force diagram, defined by Young’s equation:

(2.1)

Figure 2.9. Basic concept of an electrowetting system of a conductive fluid in (a)

its equilibrium state, and (b) its voltage induce electromechanically manipulated

state.

17 where θY is Young’s contact angle, and γdi, γpd, γpi are the surface tensions of the dielectric surface-insulating fluid, dielectric surface-polar fluid, and the polar fluid- insulating fluid interfaces, respectively.

An applied voltage introduces an electromechanical force (FEW) to the system in the form of a capacitive load related to the dielectric layer’s thickness (Figure 2.8 (b)).

The force is defined by:

(2.2)

where ε0 is the permittivity of free space, εr is the dielectric constant, d is the dielectric thickness, and V is the applied voltage. This force causes a reduction in the perceived contact angle of the polar fluid ( θV ), and adjusts Young’s equation to:

(2.3)

Substituting equation (2.2) into (2.3) and rearranging, results in the Young-Lippman electrowetting equation:

(2.4)

When the voltage is taken away, the polar fluid returns back to its equilibrium, spherical shape, creating a form of reversible switching. This voltage induced fluidic movement has been used in several areas of research, including fluidic transfer for Lab-On-Chip

[37–39], and various optical manipulation applications including lenses [40–42] and electronic displays [43-44]. It has become one of the preferred methods to inject fluids or displace other materials on the micro-scale. A more thorough review of electrowetting theory can be found elsewhere [45].

The materials used are typically transmissive, conformal coatings. This is especially true when it comes to arrayed optics and displays, where optical transparency is a crucial parameter with respect to efficiency. Transparent conductive layers are typically indium tin oxide (ITO). Dielectric layers come in the form of ceramics (Al2O3, etc.) or polymers (Parylene) of thicknesses in the range of 100 nm - 1 micron. These dielectrics are normally coated with a low surface energy fluoropolymer (Teflon,

Fluoropel, etc.) providing the hydrophobic surface. Photoresists, such as SU-8, often define the arrayed channels and pixels. Substrates are typically glass, but polymer sheets can be used if the device is to be flexible. Overall, several of the materials and processes used for electrowetting fabrication are similar to liquid crystal display fabrication, meaning mass production retooling would be minimal.

Based on the materials used, electrowetting can offer low power operation at high speeds (for naked eye applications), operate in a wide spectral range, provide efficient, high contrast optical switching, and can be relatively simple to scale and fabricate. These qualities lend electrowetting devices to be highly useful in large area optical switching applications. This dissertation analyzes and discusses three electrowetting based switchable retroreflectors in Chapter 3.

2.3.2 Liquid Crystal

Liquid crystals are materials that exhibit properties of both liquids (flow) and solid crystals (near lattice-ordered molecular orientation) that are typically rod-shaped organic materials. The molecules are capable of being re-oriented with an applied electric field due to their dielectric anisotropy. Such materials exhibit exploitable switchable optical properties including thermochromically induced diffraction [46-47] and birefringence.

These have been successfully taken advantage of in several industries, but most notably, in the electronic display industry.

Prior to 1888, liquid crystal materials were only concluded to have ‘interesting’ color shift effects at cold temperatures. In 1888, Friedrich Reinitzer, while studying cholesteryl benzoate (a cholesteric liquid crystal), discovered several other important properties of liquid crystals including two reversible melting points, reflection of circularly polarized light, and rotation of polarized light [48]. From here through the mid 1900s, liquid crystals were seen more as a scientific curiosity. It was not until 1962 when

George Heilmeier et al., of RCA laboratories, began research using liquid crystals in an electronic display to replace cathode ray tubes [49]. Early devices proved impractical because of the necessity for high temperatures to operate. In 1966, scientists at RCA laboratories discovered that mixtures of different nematic phase liquid crystals resulted in suitable operating temperature ranges for commercial products [50]. This led to seven-segment display prototypes that were subsequently integrated into early digital clock and cockpit display prototypes [51]. RCA predominantly used dynamic scattering effects [52], a reflective display that did not rely on dyes or polarizers, which suffered from image washout when viewed under bright light. To alleviate this, Wolfgang Helfrich

(RCA) suggested an alternate approach but was denied. He subsequently left RCA and went to work for Hoffman-La Roche. Here, Helfrich and Martin Schadt, with the use of polarizers, developed the twisted nematic liquid crystal display [53], which has become the dominant approach for the mobile, PC, and television display market.

As mentioned previously, typical liquid crystal molecules are rod-shaped and exhibit optically birefringent properties. This can be seen in Figure 2.10 (a) which shows the direction of the ordinary ( nO ) and extraordinary ( nE ) refractive indices of a liquid crystal molecule. A majority of liquid crystals/mixtures have a higher extraordinary refractive index, making them positively birefringent ( Δn= nE - nO ) [58]. The differences in refractive indices cause incident light to transmit differently based on its polarization

Figure 2.10. Liquid crystal basics. (a) Optical anisotropy of a liquid crystal

molecule showing the ordinary refractive index ( nO ) and the extraordinary

refractive index ( nE ). The orientation director typically describes the bulk

molecular orientation. (b-e) shows the different phases and associated directors

of bulk liquid crystal materials due a change in temperature. state and incident direction. By controlling the orientation of the liquid crystal (described by its director), one can control how incident light transmits through the material.

Orientational control is typically performed by substrate structures and electrostatic actuation.

Most of the liquid crystals known today were discovered in the early to mid

1900s. There are several different types and phases of liquid crystals found naturally and in the laboratory. Listing every known type is beyond the scope of this dissertation, which one can review elsewhere [54–57]. A vast majority of the displays available, and the basis of the devices discussed here, are thermotropic liquid crystals found in the nematic phase.

Thermotropic liquid crystal phases are temperature induced changes in long range molecular ordering (position of orientation) of the liquid crystal molecules, as can be seen in Figures 2.10 (b-e). At high temperatures the molecules tend to have an isotropically random ordering, making them more fluid-like, but no ability to align the molecules with an external field for optical manipulation. At lower temperatures the molecules exhibit long axis molecular alignment in layers, making them more crystalline.

These are known as smectic phases. The nematic phase is the most common phase and is characterized by the molecules having long axis molecular ordering while maintaining the ability for the molecules to flow. This phase is the most useful as an applied field can re-orient and align the molecules, allowing the desired changes in optical properties. These thermally induced phases are important when it comes to synthesizing and choosing a liquid crystal mixture for a specific application. Mixtures can be customized to provide nematic operation in ambient temperature ranges where the final device will be utilized.

Liquid crystal devices offer low power operation at high speeds (for naked eye applications), and can operate in a wide spectral range. The general switching method used determines the optical switching properties. Several devices use polarizers, leading to high transmissive losses, while others can operate without them. The material’s optical anisotropy causes angular effects that can lower transmissive efficiency when viewing off axis. Regardless, liquid crystal devices, especially displays, are readily available in large areas, lending them to be an obvious choice when a transmissive, optically switchable device is required. Chapter 3 analyzes and discusses two liquid crystal based switchable retroreflectors.

2.4 Summary

This chapter has presented several different devices that have been devised and tested to perform switching of retroreflected light, as summarized in Table 2.1. It is obvious that the desire for such a device is immense. The early approaches using

MEMS and semiconductor physics have proven to be useful for high speed communications, but have the disadvantage of being difficult to fabricate for operation over a large area, along with limited spectral range, and high power densities. The approach this dissertation has decided to pursue for switchable retroreflector devices is on the basis of using display technologies. As such, this chapter has presented background information on electrowetting and liquid crystal technologies. This background information should allow the reader to better understand the devices presented in the subsequent chapters.

Table 2.1. Summary table of prior switchable retroreflector devices.

Electrical Switching Device Method Speed Size

Gaseous Layer Polarization < 1 MHz < 1 cm2 PLZT Deformation - < 1 cm2 Acoustic Resonance Deformation 3 kHz Single Device MEMS Tilt Deformation 1 MHz Single Device Stage Deflection Deformation <10 kHz Single Device Phase Grating Deformation 100 kHz Single Device Mechanical Anti-Reflection (MARS) Deformation 1 MHz Single Device Sidewall Pores/Ridges Deformation 100's kHz Single Device Multiple Quantum Well Polarization 80 MHz < 1 cm2

2.5 References

[1] J. Stark, “The effect of an electric field upon spectral lines,” Sitzungsber. Akad. Wiss. Berlin, 1913.

[2] V. Early, A. Landman, and H. Marantz, “Light modulation by means of the Stark effect in molecular gases - Application to CO2 lasers,” Dec. 1969.

[3] M. B. Klein and R. H. Sipman, “Large aperture stark modulated retroreflector at 10.8 μm,” Journal of Applied Physics, vol. 51, no. 12, p. 6101, Dec. 1980.

[4] R. Buser, R. Rohde, and N. Nomiyama, “Laser retroreflector system for identification of friend or foe,” US Patent 4,361,911, 1982.

[5] A. Khalafalla, “Birefringence Mode Of Operation Of 9/65/35 Plzt Ceramic,” … , seminar-in-depth, 1973.

[6] R. Carbonneau, J. Dubois, and M. Gravel, “Optical identification friend-or-foe,” US Patent 5,274,379, 1993.

[7] G. J. Olson, H. W. Mocker, N. a Demma, and J. B. Ross, “Coherent CO(2) laser communication system with modulable retroreflectors.,” Applied optics, vol. 34, no. 12, pp. 2033–44, Apr. 1995.

[8] D. Gunawan, L. Lin, and K. Pister, “Micromachined corner cube reflectors as a communication link,” Sensors and Actuators A: Physical, vol. 7, pp. 5–8, 1995.

[9] P. B. Chu and N. R. Lo, “Optical communication using micro corner cube reflectors,” Micro Electro Mechanical …, pp. 350–355, 1997.

[10] V. Hsu, J. Kahn, and K. Pister, “MEMS corner cube retroreflectors for free-space optical communications,” University of California …, pp. 1–53, 1999.

[11] L. Zhou, J. M. Kahn, and K. S. J. Pister, “Corner-cube retroreflectors based on structure-assisted assembly for free-space optical communication,” Microelectromechanical Systems, Journal of, vol. 12, no. 3, pp. 233–242, 2003.

[12] Y. K. Hong, R. R. a Syms, K. S. J. Pister, and L. X. Zhou, “Design, fabrication and test of self-assembled optical corner cube reflectors,” Journal of Micromechanics and Microengineering, vol. 15, no. 3, pp. 663–672, Mar. 2005.

[13] C. Jenkins, G. Brown, W. Johnstone, and D. Uttamchandani, “Microelectromechanical systems actuated small-scale retroreflectors for free space optical communications,” Journal of Optics A: Pure and Applied Optics, vol. 8, no. 7, pp. S384–S390, Jul. 2006.

[14] D. Pedersen and O. Solgaard, “Free space communication link using a grating light modulator,” Sensors and Actuators A: Physical, vol. 83, no. 1–3, pp. 6–10, May 2000.

[15] O. Solgaard, F. S. Sandejas, and D. M. Bloom, “Deformable grating optical modulator.,” Optics letters, vol. 17, no. 9, pp. 688–90, May 1992.

[16] K. W. Goossen, J. a. Walker, and S. C. Arney, “Silicon modulator based on mechanically-active anti-reflection layer with 1 mbit/sec capability for fiber-in-the- loop applications,” IEEE Photonics Technology Letters, vol. 6, no. 9, pp. 1119– 1121, Sep. 1994.

[17] C. Luo and K. Goossen, “Optical microelectromechanical system array for free- space retrocommunication,” Photonics Technology Letters, IEEE, vol. 16, no. 9, pp. 2045–2047, 2004.

[18] C. Luo and K. Goossen, “Free-space optical link by microelectromechanical system array and corner cube reflector,” Photonics Technology Letters, IEEE, vol. 17, no. 6, pp. 1316–1318, 2005.

[19] T. Chan and J. Ford, “Retroreflecting optical modulator using an MEMS deformable micromirror array,” Journal of lightwave technology, 2006.

[20] J. B. Stewart, D. Freedman, S. Cornelissen, M. N. Horenstein, P. Woskov, and J. Tang, “Low power MEMS modulating retroreflectors for optical communication,” Library, vol. 7595, pp. 759505–759505–13, 2010.

[21] T. Bifano, L. Schatzberg, J. Stewart, and S. Cornelissen, “MEMS modulated retroreflectors for secure optical communication,” Proc. of IMECE, pp. 1–5, 2008.

[22] L. Ziph-Schatzberg, T. Bifano, S. Cornelissen, J. Stewart, and Z. Bleier, “Secure optical communication system utilizing deformable MEMS mirrors,” vol. 7209, p. 72090C–72090C–15, Feb. 2009.

[23] Boston Micromachines Corporation, “Modulating Retroreflector (MRR).” [Online]. Available: http://www.bostonmicromachines.com/modulating_retroreflector.htm. [Accessed: 07-Sep-2013].

[24] G. C. Gilbreath, W. S. Rabinovich, T. J. Meehan, M. J. Vilcheck, R. Mahon, R. Burris, M. Ferraro, I. Sokolsky, J. a. Vasquez, C. S. Bovais, K. Cochrell, K. C. Goins, R. Barbehenn, D. S. Katzer, K. Ikossi-Anastasiou, and M. J. Montes, “Large-aperture multiple quantum well modulating retroreflector for free-space optical data transfer on unmanned aerial vehicles,” Optical Engineering, vol. 40, no. 7, p. 1348, 2001.

[25] D. Miller, D. Chemla, and S. Schmitt-Rink, “Relation between electroabsorption in bulk semiconductors and in quantum wells: The quantum-confined Franz-Keldysh effect,” Physical Review B, vol. 33, no. 10, 1986.

[26] E. Schrödinger, “An undulatory theory of the mechanics of atoms and molecules,” Physical Review, 1926.

[27] D. Miller, D. Chemla, and T. Damen, “Electric field dependence of optical absorption near the band gap of quantum-well structures,” Physical Review B, vol. 32, no. 2, 1985.

[28] D. Miller, D. Chemla, and T. Damen, “Band-edge electroabsorption in quantum well structures: The quantum-confined Stark effect,” Physical Review Letters, vol. 53, no. 22, pp. 2173–2176, 1984.

[29] G. C. Gilbreath, W. S. Rabinovich, T. J. Meehan, M. J. Vilcheck, M. Stell, R. Mahon, P. G. Goetz, E. Oh, J. a. Vasquez, K. Cochrell, R. L. Lucke, and S. Mozersky, “Progress in development of multiple-quantum-well retromodulators for free-space data links,” Optical Engineering, vol. 42, no. 6, p. 1611, 2003.

[30] W. S. Rabinovich, P. G. Goetz, R. Mahon, L. Swingen, J. Murphy, M. Ferraro, H. R. Burris, C. I. Moore, M. Suite, G. C. Gilbreath, S. Binari, and D. Klotzkin, “45- Mbit/s cat’s-eye modulating retroreflectors,” Optical Engineering, vol. 46, no. 10, p. 104001, 2007.

[31] W. S. Rabinovich, “Free-space optical communications link at 1550 nm using multiple-quantum-well modulating retroreflectors in a marine environment,” Optical Engineering, vol. 44, no. 5, p. 056001, May 2005.

[32] J. Ohgren, F. Kullander, L. Sjöqvist, K. Wang, Q. Wang, S. Junique, S. Almqvist, and B. Noharet, “A high-speed modulated retro-reflector communication link with a transmissive modulator in a cat’s eye optics arrangement,” Proceedings of SPIE, vol. 6736, pp. 673619–673619–12, 2007.

[33] M. Plett, W. S. Rabinovich, R. Mahon, M. S. Ferraro, P. G. Goetz, C. I. Moore, and W. Freeman, “Free-space optical communication link across 16 kilometers over the Chesapeake Bay to a modulated retroreflector array,” Optical Engineering, vol. 47, no. 4, p. 045001, 2008.

[34] G. Lippmann, “Relations entre les phénomènes électriques et capillaires,” 1875.

[35] G. Beni and S. Hackwood, “Electro-wetting displays,” Applied Physics Letters, 1981.

[36] B. Berge, “Electrocapillarité et mouillage de films isolants par l’eau,” Comptes rendus de l’Académie des sciences. Série 2, …, 1993.

[37] M. G. Pollack, a D. Shenderov, and R. B. Fair, “Electrowetting-based actuation of droplets for integrated microfluidics.,” Lab on a chip, vol. 2, no. 2, pp. 96–101, May 2002.

[38] P. Paik, V. K. Pamula, M. G. Pollack, and R. B. Fair, “Electrowetting-based droplet mixers for microfluidic systems.,” Lab on a chip, vol. 3, no. 1, pp. 28–33, Feb. 2003.

[39] P. Paik, V. K. Pamula, and R. B. Fair, “Rapid droplet mixers for digital microfluidic systems.,” Lab on a chip, vol. 3, no. 4, pp. 253–9, Nov. 2003.

[40] S. Kuiper and B. H. W. Hendriks, “Variable-focus liquid lens for miniature cameras,” Applied Physics Letters, vol. 85, no. 7, p. 1128, 2004.

[41] M. MAILLARD, “OPTICAL ELECTROWETTING DEVICE,” WO Patent …, 2007.

[42] Parrot SA, “Varioptic Liquid Lens.” [Online]. Available: http://www.varioptic.com/.

[43] R. A. Hayes and B. J. Feenstra, “Video-speed electronic paper based on electrowetting,” Nature, vol. 425, no. 6956, pp. 383–385, 2003.

[44] J. Heikenfeld and A. J. Steckl, “High-transmission electrowetting light valves,” Applied Physics Letters, vol. 86, no. 15, p. 151121, 2005.

[45] F. Mugele and J.-C. Baret, “Electrowetting: from basics to applications,” Journal of Physics: Condensed Matter, vol. 17, no. 28, pp. R705–R774, Jul. 2005.

[46] R. Parker, “Transient Surface Temperature Response of Liquid Crystal Films,” Molecular Crystals and Liquid Crystals, vol. 20, no. 2, pp. 99–106, Mar. 1973.

[47] P. T. Ireland and T. V Jones, “The response time of a surface thermometer employing encapsulated thermochromic liquid crystals,” Journal of Physics E: Scientific Instruments, vol. 20, no. 10. IOP Publishing, p. 1195, 01-Oct-1987.

[48] F. Reinitzer, “Beitr�ge zur Kenntniss des Cholesterins,” Monatshefte f�r Chemie, vol. 9, no. 1, pp. 421–441, Dec. 1888.

[49] J. Castellano and B. Gold, “Liquid Gold: The story of liquid crystal displays and the creation of an industry,” Hackensack, NJ: World Scientific, 2005.

[50] J. Goldmacher, “Electro-optical compositions and devices,” US Patent 3,540,796, 1970.

[51] B. Gross, “How RCA Lost the LCD,” Spectrum, IEEE, no. NovEmbEr, pp. 46–52, 2012.

[52] G. H. Heilmeier, “Dynamic Scattering in Nematic Liquid Crystals,” Applied Physics Letters, vol. 13, no. 1, p. 46, Jul. 1968.

[53] M. Schadt, “Voltage-Dependent Optical Activity of a Twisted Nematic Liquid Crystal,” Applied Physics Letters, vol. 18, no. 4, p. 127, 1971.

[54] P. J. Collings and M. Hird, Introduction to Liquid Crystals: Chemistry and Physics (Google eBook). CRC Press, 1997, p. 324.

[55] P. J. Collings, Liquid Crystals: Nature’s Delicate Phase of Matter. Princeton University Press, 2002, p. 204.

[56] P. G. De Gennes and J. Prost, The Physics of Liquid Crystals, vol. 46, no. 4. Oxford University Press, 1995, p. 597.

[57] Google Inc. and LMGTFY, “Liquid Crystal Search.” [Online]. Available: http://lmgtfy.com/?q=liquid+crystals.

[58] J. Li, C.-H. Wen, S. Gauza, R. Lu, and S.-T. Wu, “Refractive Indices of Liquid Crystals for Display Applications,” Journal of Display Technology, vol. 1, no. 1, pp. 51–61, Sep. 2005.

Chapter 3 - Comparison of 5 Types of Switchable Retroreflectors

3.1 Introduction

The previous chapters concluded that there was a need to improve upon current methods of conspicuity, and highlights how the disadvantages of prior art technologies are not conducive for cost effective broad scale implementation. They also provided theoretical background information on retroreflectors and scalable modulator technologies based on electrowetting and liquid crystal. This will prove to be useful for this chapter as it goes into further detail about specific devices based on these technologies. Chapter 3 provides a list of device requirements necessary for naked-eye conspicuity applications. Chapter 3 then reviews and chooses which specific base retroreflective structure would be most appropriate to integrate with the various modulation methods to be explored. The chapter then considers designs, fabrication methods, and analyzes five different modulating technologies, coupled with the chosen retroreflector, and determines which one will best fit the system-level requirements for naked-eye optical conspicuity. A demonstration is also performed to validate long-range and nighttime capabilities using infrared (night-vision).

3.2 Background

Retroreflectors are commonplace in conspicuity and range finding applications as they efficiently and predictably reflect light back to the source. By modulating (flashing) this retroreflected light, conspicuity can be revealed or hidden on demand. The majority of prior work in switchable retroreflectors considers modulation by micro-electro- mechanics [1,2] or quantum wells [3]. These systems exhibit data rates up to 45

Mbps [4] with long distance links of up to 16 km [5], all achieved in a small and highly portable package. However, these previous approaches are not suitable for naked-eye

29 conspicuity applications. The ideal requirements for naked-eye conspicuity are numerous: (1) wide visible spectral range for a bright reflection (full-width-half-maximum of >100 nm); (2) ON/OFF contrast of ideally >10:1; (3) large reflective area (>100 cm2) for ease of viewing at a distance and for ease in aligning interrogating sources such as lasers; (4) flexible tape-like form factors for application to curved surfaces, the flexibility also ideally allowing impact resistance; (5) wide input angle to eliminate the need to precisely orient the retroreflective surface; (6) switching at speeds as fast as the human eye response (<100ms) to maximize recognition (as opposed to a slow gradual switch that does not draw attention); (7) low modulation power, in some cases even battery operation for mobile/remote placement; and (8) high optical efficiency such that retroreflected brightness far exceeds the intensity of reflection from diffusely reflecting surroundings.

3.3 Choosing the Retroreflecting Optical Film

The common optical element across all five investigated platforms is a base retroreflective layer. There are three general types of retroreflectors [6], which chronologically, in order of commercial introduction, are: glass bead, corner-cube, and full-cube. The first task for this work was to determine which base retroreflector layer to utilize. Table 3.1 provides the key parameters considered in the selection. Each type of retroreflector is now briefly introduced.

Table 3.1. Base retroreflector technology comparison. [6,7]

Bead Corner-Cube Full Cube Efficiency 15% 35% 58% Max Input Angle at 30% of Max Refl. ±60° ±45° ±30-38° Internal Switching Integration No Yes No

Arbitrary Orientation Yes Yes No

Glass bead type retroreflectors were the first developed and are also known as

“Engineer Grade Sheeting.” This low cost type of retroreflector uses glass beads with diameters in the micro-scale, partially embedded in a metalized reflective backplane. At the input surface of the bead, the incident light is refracted in a direction path close to the center of the back of the bead. The light reflects back to bead surface and is refracted again upon exiting the bead, parallel to the incident light. Optical efficiency, which is also known as retroreflectivity, is the percent of incident light being reflected back to the source. Glass bead suffers from high surface reflections (loss) and lower overall efficiency of about 15% [6,8]. Glass bead retroreflectors do provide the widest input angle, especially with use of high refractive index glass.

Corner cube retroreflectors [9,10] can be realized as an array of truncated corners of orthogonal cubes. They offer high efficiency while covering half of a hemisphere of input angles. Corner cubes can be inexpensively micro-replicated in large area polymer films. In addition, corner-cubes provide a built-in cavity into which the switching element can be integrated. Their physics of operation will be discussed later in this section.

Full cube retroreflectors are the newest type of retroreflective structure, and were introduced in 2006 by 3M under the name Diamond Grade DG3. These offer the highest efficiency of about 58% [8,11], but are not as widely available as their counterparts.

Furthermore, the retroreflection is not maintained for all orientations (rotations of retroreflecting surface).

After evaluation for efficiency, maximum input angle to the retroreflective surface, the ability for switching integration in the retroreflector, and allowance for arbitrary orientation (Table 3.1), corner-cube retroreflectors were determined to be the platform

31 for this work. Photos and data for a typical corner cube retroreflector are shown in

Figure 3.1. The three sidewalls have a slope of 54.74°, and when incident light reflects off all three sidewalls, it will achieve retroreflection (Figure 3.1(b)). As also seen in

Figure 3.1(a), the optically active aperture is an equilateral triangle, outside of which exist dark areas at the three corners. These dark spots represent ~35% of the total area of the corner cube, and are where incident light reflects off only two sidewalls. A

Figure 3.1. Corner cube retroreflectors: (a) top view of a corner cube

retroreflector which visually reveals the retroreflecting and dead areas; (b) SEM

of corner-cube structures and diagram of an example retroreflection path (not an

actual ray trace); (c) retroreflection vs. wavelength from 350 to 1600 nm; (d)

retroreflection as a function of input angle as measured using the experimental

setup in [19 Figure 3 (b)]. Points are experimental data, trend line is for guidance.

32 commercially available retroreflective film, such as Reflexite V82 (used in this study), exhibits the following properties. The retroreflection of a collimated light reduces to ~50% intensity within a ±0.2° observation cone, and is effectively non-observable (<1%) past

~±2°. The reflection as a function of wavelength covers the entire visible spectrum and is shown in Figure 3.1(c), and spans the visible spectrum and the infrared spectrum input angle is ±45°, defined in this work as the largest angle of incident light before retroreflected intensity reduces to 30% of its maximum (Figure 3.1(d)). There exists more advanced theoretical understanding for retroreflectors, such as bidirectional reflectance distribution function (BRDF), which is included here from reference [12–14].

In this study, two particular corner cube structures were utilized. In some cases,

Reflexite V82 film (175 µm sidewall length) was used behind an optically modulating layer (top-view Figure 3.1(a)). When the optical modulation was to be integrated within the corner-cubes themselves, a custom micro-replicated cornercube surface was also provided by Reflexite (800 µm sidewall length, angled view (SEM image of Figure 3.1(b)).

3.4 Selection of Electrical Modulation Methods

Several reflective optical switching technologies [15] can satisfy the ideal requirements described in the introduction section. Electrophoretic display technology could be used, but only the in-plane variations have the possibility of being transparent, and are too slow to match the human eye response. Electrochromic technology is operable with low voltage, and provides excellent transmission, but commercial devices also switch very slowly (100’s ms to 10’s of s). Cholesteric liquid crystals provide excellent transparency, but cannot modulate over the entire visible spectrum. At the time of this work, it was determined that the most appropriate modulation techniques were based on electrowetting [16] or more conventional liquid crystal technology [17–19].

Both of these technologies can switch quickly, can modulate over the entire visible

33 spectrum and into the infrared, and can operate with little angular dependence.

3.4.1 Electrowetting

Electrowetting will only be reviewed in brief as a complete review can be found elsewhere [16]. Electrowetting uses electromechanical force to reduce liquid contact angle on a dielectric surface. This can be described by the well known electrowetting

equation: where is the voltage induced contact angle, is Young’s angle (at 0V), is the capacitance per unit area of the hydrophobic surface

( ), is the applied voltage, and is the interfacial surface tension ( N/m ) between conducting fluid (often water) and insulating fluid (often oil). The polar fluid serves as one electrode. The dielectric covers a counter electrode, and is always hydrophobic (often an insulating polymer covered with a hydrophobic fluoropolymer). As will be seen in later sections, electrowetting can be applied to change optical transmission [20] or refraction [21]. The speed of operation can be as fast as a few ms for devices of <100 µm in size. Operation voltages can range from 10-100 V, based on materials used (C, ).

3.4.2 Liquid Crystal

A general description of liquid crystal technology can be found elsewhere [17–

19]. Liquid crystals have been used extensively in the display industry as the preferred method of pixel switching in consumer televisions and monitors (LCDs), as well as smart windows with polymer dispersed liquid crystal sheets (PDLC). Liquid crystal systems are highly developed, well understood, and cost is continually decreasing. Some types of liquid crystal devices are highly transparent, and can provide high contrast ratios over a wide spectral range. Liquid crystal devices can operate from low to moderate voltages

(<5 to 40V, based on thickness) and have fast switching speeds (typically 10’s of ms).

3.5 Optical Efficiency Model

For all approaches reviewed herein, the incident light must transmit through the modulation layer, retroreflect, and then pass back through the modulation layer a second time. If the modulation layer is inefficient, the compounded optical losses will significantly reduce the observable contrast ratio.

Optical transmission through each layer can be expressed as , where is the material’s optical absorption coefficient, and is the material’s

Fresnel reflectance due to refractive index mismatch with adjacent layers. The Fresnel

reflectance will be calculated for normal incidence as . The product of adjusted transmissions of all the layers provides a total single optical pass transmittance of , where is the transmittance of each layer. To get a value for the full efficiency, with retroreflection, the light must travel through the modulating device twice ( ). Depending on the modulation device, there is also an associated transmissive aperture ( ), where light outside this transmissive aperture is lost. The retroreflector returns the light parallel to the incident light, but not at the same location, therefore when the modulating layer is external (not integrated inside the corner

cube), the worst-case transmissive aperture should be calculated as ( ). There is also the associated efficiency with the retroreflector itself ( ). For the corner cube retroreflectors tested herein, due to manufacturing imperfections, additional Fresnel losses, 90% reflective Al, and a retroreflective aperture of 65%, a final efficiency of

35% it typically achieved.

The total retroreflective efficiency combines all the losses discussed above, and can be expressed as:

(3.1)

At this point, the reader is now ready for review of the various devices tested in this work.

3.6 Electrowetting Lenslet Scattering

Our research group began investigating switchable retroreflectors in 2007, and therefore this will be the first device topic reviewed. These devices rely on electrowetting lenslets [22] that are integrated into the corner cubes themselves.

3.6.1 Fabrication and Construction

The electrowetting lenslet retroreflector is shown in Figure 3.2. Detailed information on the fabrication and operation of this device is provided elsewhere [22,23].

The device consists of a corner cube retroreflector backplane that has been coated with electrowetting films, and dosed with oil and water. These immiscible fluids create a concave shaped meniscus, which creates a concave lens because the oil has a refractive index of >1.4 and the water has a refractive index of 1.3. To allow ease of fabrication, the larger 800µm corner cubes were utilized.

3.6.2 Electrical Switching

Once the device is fully assembled, and for the case of no applied voltage (off state), the meniscus at the oil and water interface is concave. This can be seen in

Figures 3.2(a) and 3.3(a). In this state, light incident onto the electrowetting retroreflector will be refracted as it passes through the meniscus, and therefore optically scattered.

The device therefore appears as a diffuse reflector (inconspicuous).

Figures 3.2(b) and 3.3(b) show the device when electrically powered.

Electrowetting reduces the water contact angle with the hydrophobic dielectric. With application of ~19V, the water contact angle reaches ~125 degrees, the meniscus

Figure 3.2. Electrowetting lenslet retroreflector in the: (a) scattering (voltage off) state; (b) voltage provided to enable the retroreflecting/on state.

Figure 3.3. Collinear images of the lenslet retroreflectors with: (a) voltage off and retroreflection off; (b) voltage on and retroreflection on. The image intensities are not comparable (the camera sensitivity was adjusted to obtain good images for the off and on states).

37 becomes flat, and the device behaves like a conventional retroreflector. If the applied voltage is increased even more, the meniscus will become convex, creating a second diffusely scattering state. Typically, the voltage is not increased further, but is simply set back to zero to create the diffuse state.

3.6.3 Retroreflection Results

The retroreflection input angle, as a function of the light source’s incident angle, is shown in Figure 3.4. For this device, greater than 10:1 contrast ratio was achieved out to

±30°. The maximum input angle at 30% of maximum reflection is at 30-35°, and can be increased by use of higher refractive index oils with little to no loss to the electrical operation of the device. Uncharacteristic (for a basic retroreflector) dips and peaks in reflection can be seen for ~±10° and ~±20°, which is speculated to be due to thin film interference from the hydrophobic dielectric stack [23].

Figure 3.4. Retroreflection as a function of input angle for integrated

electrowetting lenslets. Both states of operation are plotted. Points are

experimental data, trend lines are for guidance.

The electrowetting lenslet retroreflector uses a method of modulation that is integrated within the corner cube structure itself. This lends itself to slightly more difficult fabrication, but requires fewer deposited layers (less optical loss), and makes the device thinner (and potentially more flexible as a result). The approximate single pass transmission is calculated to be about T=87.8%. Since there are no layers to limit the active area of the modulation device, the active area for modulation ( ) is 100%. The fixed optical efficiency of the retroreflector with Al reflection and retroreflective aperture is

35%. As a result, the total retroreflective efficiency is =27%.

3.6.4 Discussion

The approach of integrating the electrowetting lenslet inside the retroreflector has several advantages. The operating voltage can be lower than a conventional electrowetting lens [24] because a contact angle change of only 55 degrees is needed to flatten the meniscus. The energy per switch is also very low (~1.8 mJ/m2) because the water only touches ~48% of the total device area for the electrowetting capacitor. Based on the simple materials used, it will be shown that the optical efficiency is the highest of those described herein. As reported previously [22], scaling the corner-cubes down to

10 µm could result in very fast switching speeds of <0.1 ms, which far exceeds the response time of the human eye. The spectral range covers the entire visible spectrum, and the input angle is wide. A large total reflective area is possible (>100 cm2) and the entire device can be thin and flexible. Fast switching speed may be the most significant advantage over the other investigated approaches, the next of which will now be reviewed.

3.7 External Electrowetting Light Valve

Electrowetting light valves were first reported in 2003 [25] and are now close to commercialization for video-rate reflective displays (LiquaVista/Samsung). Electrowetting light valves are highly optically efficient in transmissive or reflective mode [20,25–27], and therefore initially appear as a strong candidate for modulating retroreflection. The simplest way to integrate electrowetting light valves and a retroreflector is to fabricate them as separate components, and simply epoxy/laminate them together.

3.7.1 Fabrication and Construction

The external electrowetting light valve retroreflector is shown in Figure 3.5. The construction is a large pixel array of electrowetting light valves, but uses a common electrode for all pixels, allowing the entire film to switch ON/OFF uniformly. Complete details of our process for pixel array fabrication can be found elsewhere [20].

3.7.2 Electrical Switching

Figures 3.5(a) and 3.6(a) show the zero voltage state of the device. In the off state, the oil (which is hydrophobic) forms a film that fully covers the hydrophobic dielectric. The oil contains ~10 wt. % of black dye [20] which absorbs the incident light, thereby reducing the retroreflection. Most of the light transmitting through the pixel array in the zero voltage state is due to light leakage through the hydrophilic pixel grid. This light leakage can be reduced through use of a black photoresist (not demonstrated herein, but available commercially as MicroChem XP Black SU-8, [27]).

The switching ON of the electrowetting pixels involves a single application of ~15

V, and two resulting fluid mechanisms. First, a vertical electromechanical force overcomes interfacial surface tension force and breaks up the oil film, allowing the water to contact the hydrophobic dielectric. Once in contact, a horizontal electromechanical

Figure 3.5. External electrowetting light valve retroreflector in the: (a) voltage off state and light absorbing state; (b) voltage on state and retroreflecting state

Figure 3.6. Collinear images of the external electrowetting light valve retroreflector in the: (a) voltage off state; (b) voltage on state.

41 force displaces the oil to the ends of the pixels (electrowetting), creating a transmissive aperture, thereby allowing retroreflection (Figure 3.5(b)). A complete study of the predictive nature of the oil film breakup mechanism can be found elsewhere [26]. When the voltage is removed, the oil film suppresses retroreflection by rapidly recovering (10’s ms) the hydrophobic dielectric surface.

3.7.3 Preliminary Results

The retroreflection input angle, as a function of the light source’s incident angle, is shown in Figure 3.7. A 10:1 contrast is not achieved, and about 4:1 contrast is achieved only out to ±15 degrees. This reduction in contrast is due to light leakage through the

2 hydrophilic grid, and the fact that the aMD loss increases significantly with input angle.

The maximum input angle at 30% of maximum reflection ( ) is 30-35°. The external electrowetting light valve retroreflector has several material layers that light must traverse.

This leads to a theoretical single pass transmission T ~ 83.3%. In addition, the aperture

Figure 3.7. Retroreflection as a function of input angle for the external

electrowetting light valve retroreflector. Both states of operation are plotted.

Points are experimental data, trend lines are for guidance.

2 is only ~70% in the on state leading to a maximum value of 49% for aMD . As a result, the total retroreflective efficiency is =12%.

3..7.4 Discussion

The external electrowetting light valve provides sufficiently low operating voltage

(15 V), a low energy per switch (~7 mJ/m2), and rapid switching of ~10-100 ms. The

2 requirement of several material layers and aMD loss leads to an overall low optical efficiency. The off state light leakage provides low contrast (can be improved with a black pixel grid). As with most electrowetting devices, fabrication is scalable to large area on flexible substrates [20]. Because the retroreflective efficiency, contrast, and maximum input angle are rather poor in comparison to other approaches, a preliminary attempt was

2 made to integrate the light valve into the corner cube itself, especially to reduce aMD optical loss (next section).

3.8 Integrated Electrowetting Light Valve

Integration of the light valve into the corner cube structure was investigated to provide lower optical loss and a thinner over-all form factor.

3.8.1 Construction and Fabrication

The integrated electrowetting light valve is shown in Figure 3.8 where the light valve is vertically inverted and placed into the corner cube structure itself. One less substrate is needed, and the device can remain substantially thinner and more flexible.

Construction consists of a glass or polyethylene-terephthalate (PET) substrate coated with a transparent electrode and a hydrophobic dielectric. The bottom substrate is an array of micro-replicated corner cubes coated with a reflective electrode (Al). To allow ease of fabrication, the larger 800 µm corner cubes were utilized. The hydrophilic pixel grid (Figure 3.5(a)) is no longer needed as the top ridges of the corner-cube pixelate the oil into discrete volumes. The technique used for liquid dosing is unique. Water is

Figure 3.8. Integrated electrowetting light valve; (a) voltage off and light

absorbing state; (b) voltage on and retroreflecting state.

Figure 3.9. Collinear images of the integrated electrowetting light valve

retroreflector in its (a) off state and (b) on state. Blue oil was used instead of

black oil to aid visualization of the retroreflector below the oil film. applied onto the corner-cube substrate. The water is slowly advanced over the substrate to avoid air trapping. A vacuum degas can be used to remove any entrapped air bubbles.

Next, a drop of dyed oil and the electrowetting top plate are placed onto the corner cube substrate. The hydrophobic surface of the top plate allows the oil to spread over the entire surface. For this device test, blue oil was used because it also allows visualization

44 of the corner cubes beneath the oil. A more detailed fabrication process can be seen in

Appendix B.

3.8.2 Electrical Switching

Figure 3.8(a) shows the off state of the device whereby the oil film covers the corner cube input aperture and absorbs light (if black oil were used, all visible light would be absorbed). When electrically powered, a charge buildup commences at the top ridges of the hydrophilic Al electrode. The water then wets the hydrophobic dielectric surface of the top substrate, displacing the oil from all directions, forming it into a spherical cap. The oil sphere quickly migrates to one of the corners of the corner cube as can be seen in

Figures 3.8 (b) and 3.9 (b). When the voltage is released, the water recedes and the oil moves back into its original place.

3.8.3 Retroreflection Results

The single biggest advantage for integrating the colored oil into the corner-cubes is the displacement of oil into the corners. The corners already contain the optically dead area (Figure 3.1) and therefore most of the retroreflective aperture is active (aMD≈90%).

2 Furthermore, with a thinner oil film and smaller oil droplets in the ON state, the aMD loss can likely be eliminated. However, the fabrication method does not yet provide a large area with uniform oil dosing, and the available area for testing (Figure 3.8) is too small for angular measurements. The maximum input angle should be only slightly less than a regular corner cube retroreflector film (Fig 1(d)) and the ON/OFF contrast should be

>10:1 because there is no retroreflective light leakage path as discussed for Figure 3.6

(the external light valve).

As this device has very few materials that light must transmit through, its optical efficiency will be very high with a calculated single pass transmission of T=89.8%. Along with an active area ( ) of 90%, the calculated optical efficiency is =23%.

3.8.4 Discussion

The integrated electrowetting light valve has all the advantages of the external light valve, and it has the advantage of high retroreflective efficiency. Furthermore, the integrated approach can lead to a very thin form factor requiring only two substrates.

3.9 Conventional Liquid Crystal Light Valve

Liquid crystal displays are widely available and therefore are highly appropriate for investigation with switchable retroreflection.

3.9.1 Construction and Fabrication

The liquid crystal light valve and retroreflector are shown in Figure 3.10. For this work, we used a standard twisted nematic liquid crystal display (TNLCD) over a corner cube based retroreflective film. Details of the TNLCD fabrication can be reviewed in numerous publications [28–31]. The TNLCD is simply adhered to the retroreflective film.

3.9.2 Electrical Switching

Detailed operation of the liquid crystal light valves can be found elsewhere [28–

31]. At zero voltage the liquid crystal molecules are aligned with each plate and rotate the polarization of light, so that light is able to transmit through the crossed polarizers

(Figures 3.10(a) and 3.11(a) retroreflection). When voltage is applied, the liquid crystal molecules align with the electric field, no polarization rotation occurs, and the light is absorbed due to the crossed polarizers (Figures 3.10(b) and 3.11(b)). Unlike the previously described approaches, this device is retroreflective with zero voltage.

Figure 3.10. Liquid crystal light valve retroreflector in the: (a) retroreflecting state with no voltage applied; (b) light absorbing state with voltage applied.

Figure 3.11. Collinear images of liquid crystal light valve retroreflector in the (a) voltage off and (b) voltage on states.

However, the polarizers can instead be applied in optical alignment, and the device could instead be opaque with zero voltage.

3.9.3 Retroreflection Results

The achievable input angles are plotted in Figure 3.12. This device has a maximum contrast of 600:1 with nearly 100:1 over its entire field of view (though erratic).

The maximum input angle is 35°.

Figure 3.12. Retroreflection as a function of input angle for the TNLCD

retroreflector. Both states of operation are plotted. Points are experimental data,

trend lines are for guidance.

This device has a substantial loss in optical efficiency due to the material layers.

This is mostly due to the polymer-sheet polarizers which typically have a transmittance of only 38% [32]. Also, after retroreflecting, the proper polarization of light for efficient return is lost [33], and will again be filtered out by the polarizer. Polarizer aside, this device also has numerous layers, including three adhesive layers (polarizers and retroreflective film). This leads to a theoretical single pass transmission (T) of about

35.2%, the lowest of all of those reviewed. It does have a 100% device active area

( ). With the standard retroreflective film efficiency, the total efficiency ( ) of this device is about 4%.

3.9.4 Discussion

LCDs are currently widely used in low cost and low power devices and have an operation voltage of <5V. Their spectral range covers the entire visible spectrum and switching can be seen over a wide input angle with high contrast. LCDs are currently manufactured in much larger than 100 cm2 areas. The device has the lowest retroreflection of all those reviewed due to the polarizers. Therefore, an alternate liquid crystal device not utilizing polarizers was investigated (next section).

3.10 Liquid Crystal Scattering

Liquid crystal scattering exploits the birefringent nature of liquid crystals when randomly oriented in a polymer matrix. As a result, an optical diffuser can be electrically switched to be optically transparent, similar to that discussed for the integrated electrowetting lenslet.

3.10.1 Construction and Fabrication

Liquid crystal scattering is performed by use of a polymer dispersed liquid crystal

(PDLC) shutter adhered to a retroreflective film as can be seen in Figure 3.13. PDLC shutters have been extensively researched and fabrication can be reviewed in numerous publications [34–37]. A more detailed fabrication process can be seen in Appendix B.

Figure 3.13. Polymer dispersed liquid crystal retroreflector in the: (a) voltage off state and scattering state. (b) Voltage on and retroreflecting state.

Figure 3.14. Collinear images of the polymer dispersed liquid crystal scattering retroreflector in its (a) voltage off state and (b) voltage on state.

3.10.2 Electrical Switching

Detailed description of operation for the switching mechanisms of the PDLC modulator can be found elsewhere [34–37]. With no voltage applied, the polymer suspended liquid crystal droplets are randomly oriented with respect to the primary optical axis, as can be seen in Figure 3.13(a). This creates a random difference in refractive index that incident light will encounter, causes the light to scatter, and the device appears optically diffuse (no retroreflection, Figure 3.14(a)). When the device is electrically powered, the liquid crystal droplets orient themselves to be aligned with the electric field

(Figures 3.13(b) and 3.14(b)). This allows for a match in refractive index with the suspending polymer (often poly-methyl-methacrylate) and the PDLC layer becomes transparent, thus allowing retroreflection. When voltage is turned off, the device reverts to its diffuse state.

3.10.3 Retroreflection Results

The PDLC device has excellent switching capabilities as a maximum input angle

of 40° can be seen in Figure 3.15. A contrast of >10:1 is also maintained out to

~±40.° There are several irregular features in the angular dependence that are currently not understood, but overall the retroreflective performance is comparable to a bare retroreflective film. The liquid crystal scattering retroreflector has few material layers and subsequently lower optical loss with a single pass transmission (T) of about 85%.

The device has no loss in modulation device active area (aMD=100%) and with the retroreflective film efficiency applied, comes out to a total efficiency of =25%.

Figure 3.15. Retroreflection as a function of input angle for the polymer-

dispersed liquid crystal retroreflector. Both states of operation are plotted. Points

are experimental data, trend lines are for guidance.

3.10.4 Discussion

Operating voltage greatly depends on the PDLC’s composition and thickness, and can be as low as 5V [36]. Energy per switch (~1.3 mJ/m2) is low and switching speeds of

< 2ms [37] have been observed. The films tested herein switched with a rise time of <2 ms with a fall time of <30ms. The spectral range covers the entire visible spectrum and into night-vision infrared, and boasts a wide input angle with good contrast. A large reflection area is also achievable on thin and flexible materials. Optical efficiency is high because no polarizers are used. The required fabrication process is also one of the simplest of those reviewed herein.

3.11 Discussion, Conclusions, and Demonstration

This chapter has shown that the devices discussed herein are all viable technologies to be used for switchable retroreflectors for human eye conspicuity applications. TNLCD and the external electrowetting light valve, however, do not meet

Table 3.2. Comparison of key attributes of the five types of switchable retroreflectors. External EW Integrated EW EW Lenslet Light Valve Light Valve TNLC PDLC

Efficiency 27% 12% 23% 4% 25% Maximum Input Angle ±35 ±30 not measured ±35 ±38 Contrast >10:1 5:1 not measured >100:1 40:1 Operating Voltage 5 10 10 < 5 25 Energy Per Switch (mJ/m2) ~1.8 ~7.0 ~9.5 ~1.2 ~1.3

all of the ideal performance metrics set in the introduction. Table 3.2 provides a summary of performance for all tested devices. The electrowetting lenslet retroreflector provides the highest optical efficiency, followed closely by the PDLC retroreflector, and then the integrated electrowetting light valve. The overall conclusions are as follows.

The integrated electrowetting lenslet has the highest performance of all approaches tested, including the potential for the fastest switching speed. Higher speed may be useful for additional data transfer or lock-in amplification of the detected retroreflection signal. The PDCL device is second in overall performance, mainly because of slower switching speed. However, PDLC films can be easily obtained commercially and therefore PDLC is initially a more attractive approach for switchable retroreflectors.

PDLC films are currently operable from -10 to 60 degrees Celsius, however with alternate chemistries can be extended to -40 to 100 degrees Celsius [38]. The use of

PET substrates with proper sealing and electrode isolation allows for a high impact resistant device that is operable in harsh environments including rain and dust.

With the PDLC chosen over the other switching mechanisms, a full night and day demonstration with a polymer dispersed liquid crystal retroreflector was performed. As shown in Figure 3.16(a,b), a ~5x15cm device is easily seen at 200m at night using conventional night-vision devices and a 28 mW 850 nm source spread to an

Figure 3.16. Field demonstration of a switchable retroreflector using night vision

through binoculars at a range of 200m with (a) no voltage applied/tag off, and (b)

voltage applied/tag on. The author J. Heikenfeld is holding and modulating the

tags while authors P. Schultz and B. Cumby illuminate with a 28 mW source

spread to 2 m diameter and take photographs through night vision goggles. (c)

Image of a ~5 x 15 cm switchable retroreflector being bent to about a 10 cm

diameter.

~2m diameter. Night vision was tested, as it is emerging in automotive vision and safety systems. During the day, 400 m viewing distance has also been obtained with a 30 mW

532nm laser pointer. The fabricated devices are highly flexible (Figure 3.16(c)), with bend radii of <5 cm demonstrated. For visible and night-vision conspicuity applications, switchable retroreflectors are now proven as feasible based on use of electrowetting lenslets or polymer-dispersed liquid crystals.

3.12 References

[1] A. Lundvall, F. Nikolajeff, and T. Lindstrom, “High performing micromachined retroreflector.,” Opt. Express, vol. 11, no. 20, pp. 2459–73, Oct. 2003.

[2] J. B. Stewart, D. Freedman, S. Cornelissen, M. N. Horenstein, P. Woskov, and J. Tang, “Low power MEMS modulating retroreflectors for optical communication,” Library (Lond)., vol. 7595, pp. 759505–759505–13, 2010.

[3] G. C. Gilbreath, W. S. Rabinovich, T. J. Meehan, M. J. Vilcheck, R. Mahon, R. Burris, M. Ferraro, I. Sokolsky, J. a. Vasquez, C. S. Bovais, K. Cochrell, K. C. Goins, R. Barbehenn, D. S. Katzer, K. Ikossi-Anastasiou, and M. J. Montes, “Large-aperture multiple quantum well modulating retroreflector for free-space optical data transfer on unmanned aerial vehicles,” Opt. Eng., vol. 40, no. 7, p. 1348, 2001.

[4] W. S. Rabinovich, P. G. Goetz, R. Mahon, L. Swingen, J. Murphy, M. Ferraro, H. R. Burris, C. I. Moore, M. Suite, G. C. Gilbreath, S. Binari, and D. Klotzkin, “45- Mbit/s cat’s-eye modulating retroreflectors,” Opt. Eng., vol. 46, no. 10, p. 104001, 2007.

[5] M. Plett, W. S. Rabinovich, R. Mahon, M. S. Ferraro, P. G. Goetz, C. I. Moore, and W. Freeman, “Free-space optical communication link across 16 kilometers over the Chesapeake Bay to a modulated retroreflector array,” Opt. Eng., vol. 47, no. 4, p. 045001, 2008.

[6] J. Lloyd, “A brief history of retroreflective sign face sheet materials,” 2008. [Online]. Available: http://www.rema.org.uk/pdf/history-retroreflective-materials.pdf .

[7] S. Winburn, A. Baker, and J. Leishman, “Angular response properties of retroreflective screen materials used in wide-field shadowgraphy,” Exp. Fluids, vol. 20, no. 3, pp. 227–229, 1996.

[8] MinimumReflectivity.org, “No Title,” 2009. [Online]. Available: http://www.minimumreflectivity.org/retroreflective.asp.

[9] H. D. Eckhardt, “Simple model of corner reflector phenomena.,” Appl. Opt., vol. 10, no. 7, pp. 1559–66, Jul. 1971.

[10] Reflexite, “Reflexite Corporation,” 2011. [Online]. Available: http://www.reflexite.com.

[11] 3m Corporation, “No Title,” 2011. [Online]. Available: http://www.3m.com.

[12] F. E. Nicodemus, “Directional Reflectance and Emissivity of an Opaque Surface,” Appl. Opt., vol. 4, no. 7, p. 767, Jul. 1965.

[13] S. C. Pont and J. J. Koenderink, “Bidirectional reflectance distribution function of specular surfaces with hemispherical pits.,” J. Opt. Soc. Am. A. Opt. Image Sci. Vis., vol. 19, no. 12, pp. 2456–66, Dec. 2002.

[14] A. Marchant, “Conspicuity tape for enhanced laser range finding,” Opt. …, vol. 49, no. April, pp. 1–7, 2010.

[15] J. Heikenfeld and T. Drzaic, Paul, Yeo, Jong-Souk, Koch, “Review Paper: A critical review of the present and future prospects for electronic paper,” J. Soc. Inf. Disp., vol. 19, no. 2, pp. 129–156, 2011.

[16] F. Mugele and J.-C. Baret, “Electrowetting: from basics to applications,” J. Phys. Condens. Matter, vol. 17, no. 28, pp. R705–R774, Jul. 2005.

[17] P. G. De Gennes and J. Prost, The Physics of Liquid Crystals, vol. 46, no. 4. Oxford University Press, 1995, p. 597.

[18] V. G. Chigrinov, Liquid Crystal Devices: Physics and Applications. Artech House Optoelectronics Library, 1999, p. 366.

[19] S.-T. Wu and D.-K. Yang, Fundamentals of Liquid Crystal Devices. Hoboken, NJ: John Wiley, 2006, p. 394.

[20] K. Zhou, J. Heikenfeld, K. a Dean, E. M. Howard, and M. R. Johnson, “A full description of a simple and scalable fabrication process for electrowetting displays,” J. Micromechanics Microengineering, vol. 19, no. 6, p. 065029, Jun. 2009.

[21] N. R. Smith, D. C. Abeysinghe, J. W. Haus, and J. Heikenfeld, “Agile wide-angle beam steering with electrowetting microprisms.,” Opt. Express, vol. 14, no. 14, pp. 6557–63, Jul. 2006.

[22] M. K. Kilaru, B. Cumby, and J. Heikenfeld, “Electrowetting retroreflectors: Scalable and wide-spectrum modulation between corner cube and scattering reflection,” Appl. Phys. Lett., vol. 94, no. 4, p. 041108, 2009.

[23] M. K. Kilaru, J. Yang, and J. Heikenfeld, “Advanced characterization of electrowetting retroreflectors.,” Opt. Express, vol. 17, no. 20, pp. 17563–9, Sep. 2009.

[24] S. Kuiper and B. H. W. Hendriks, “Variable-focus liquid lens for miniature cameras,” Appl. Phys. Lett., vol. 85, no. 7, p. 1128, 2004.

[25] R. A. Hayes and B. J. Feenstra, “Video-speed electronic paper based on electrowetting,” Nature, vol. 425, no. 6956, pp. 383–385, 2003.

[26] B. Sun and J. Heikenfeld, “Observation and optical implications of oil dewetting patterns in electrowetting displays,” J. Micromechanics Microengineering, vol. 18, no. 2, p. 025027, Feb. 2008.

[27] A. Schultz, J. Heikenfeld, H. Kang, and W. Cheng, “1000: 1 Contrast Ratio Transmissive Electrowetting Displays,” Disp. Technol. J., vol. 7, no. 11, pp. 583– 585, 2011.

[28] M. Schadt, “Voltage-Dependent Optical Activity of a Twisted Nematic Liquid Crystal,” Appl. Phys. Lett., vol. 18, no. 4, p. 127, 1971.

[29] J. Fergason, “Display Devices Utilizing Liquid Crystal Light Modulation,” US3,731,9861973.

[30] M. Gu, “The World of Liquid Crystal Displays,” 2006. [Online]. Available: http://www.personal.kent.edu/~mgu/LCD/tn.htm. [Accessed: 06-Jan-2012].

[31] G. W. Gray and S. M. Kelly, “Liquid crystals for twisted nematic display devices,” J. Mater. Chem., vol. 9, no. 9, pp. 2037–2050, 1999.

[32] E. Optics, “Visible Linear Polarizing Laminated Film,” 2011. [Online]. Available: http://www.edmundoptics.com/products/displayproduct.cfm?productid=1912.

[33] S. E. Segre and V. Zanza, “Mueller calculus of polarization change in the cube- corner retroreflector.,” J. Opt. Soc. Am. A. Opt. Image Sci. Vis., vol. 20, no. 9, pp. 1804–11, Sep. 2003.

[34] P. S. Drzaic, “Polymer Dispersed Nematic Liquid Crystal for Large Area Displays and Light Valves,” J. Appl. Phys., vol. 60, no. 6, pp. 2142–2148, 1986.

[35] J. W. Doane, A. Golemme, J. L. West, J. B. Whitehead, and B. G. Wu, “Polymer Dispersed Liquid Crystals for Display Application,” Mol. Cryst. Liq. Cryst. Inc. Nonlinear Opt., vol. 165, no. 1, pp. 511–532, 1988.

[36] G. Spruce and R. D. Pringle, “Polymer dispersed liquid crystal (PDLC) films,” Electron. Commun. Eng. J., vol. 4, no. 2, pp. 91–100, 1992.

[37] L. Petti, P. Mormile, and W. J. Blau, “Fast electro-optical switching and high contrast ratio in epoxy-based polymer dispersed liquid crystals,” Opt. Lasers Eng., vol. 39, pp. 369–377, 2003.

[38] J. West, “Extended temperature range polymer dispersed liquid crystal light shutters,” US Pat. 5,004,323, 1991.

* A majority of Chapter 3 (text and figures) was re-published with permission from the Optical Society of America (OSA) with the following citation:

P. Schultz, B. Cumby, and J. Heikenfeld, "Investigation of five types of switchable retroreflector films for enhanced visible and infrared conspicuity applications," Appl. Opt. 51, 3744-3754 (2012).

Chapter 4 – Switchable Retroreflector and Interrogation System

4.1 Introduction

This chapter takes the chosen PDLC-based switchable retroreflector (from

Chapter 3) and integrates it with several other components created in this dissertation.

These components include electronic interrogation, receiver, and driver for an optical discrimination system. This chapter also provides system-level characterization of the switchable retroreflector in the important context of the electronic system built around it.

It then provides a system level theoretical optical model of maximum operating range for the gamut of possible ambient light conditions and interrogating optical powers. Lastly, this chapter presents, and discusses, field demonstration results of the naked-eye optical conspicuity system.

4.2 Background

Range-finding [1] and general conspicuity [2] have long exploited the efficient and predictable optical properties of retroreflectors. By reflecting incident light back to the source in a narrow beam, retroreflectors offer substantially higher contrast compared to diffusely reflecting surroundings. Adding a means to switch on or off the retroreflected light provides additional advantages including: (1) increased conspicuity [3], (2) ability to allow or disallow retroreflection as needed, and (3) the potential for free-space optical communications. Free space optical communications has been the field most heavily investigated for switchable retroreflectors. Prior approaches include microelectromechanical systems (MEMS) [4] and multiple quantum wells (MQW) [5–7].

Though these previous approaches do exhibit high switching speeds [8], they suffer from narrow spectral band operation, low contrast, small retroreflective area, and challenging optical system alignment. These challenges are particularly undesirable for applications

58 in naked eye conspicuity. Our group had previously reported an electrowetting light scattering switchable retroreflector [9–12] to overcome these challenges, which we recently improved upon through a liquid crystal modulator [13]. Our new approach does not provide the high switching speeds of the previous MEMS and MQW methods, but provides superior performance in nearly all other metrics of interest to conspicuity applications: high contrast of >2000:1 at 635nm, and >400:1 at 850nm; large area of 75 cm2; visible and infrared spectrum from 400 - 1600nm; and thin-flexible construction at

<0.6 mm thick.

We report here a system-level demonstration, which is the culmination of our continuum of effort to create an enhanced optical discrimination system based on switchable retroreflective films. The system was developed collaboratively with Northrop

Grumman Corporation’s (NGC) Xetron division, in which NGC provided as needed tools, hardware, and design support of the electronic components. In this chapter, physics, design, and characterization of the retroreflective films are first discussed, followed by a description of the electronic interrogation and response system. An optical model is provided showing theoretical long distance viewing of the films with both naked eye and night vision viewing for day and nighttime conditions. The model is then verified with various field demonstrations, confirming the ability to integrate such a system into conspicuity applications.

4.3 System Description

A laser interrogation and response system was integrated with the switchable retroreflective films to provide a way to remotely actuate them from a distance by means of a single operator.

Figure 4.1. Block diagram of the interrogator and receiver/retroreflector

4.3.1 Introduction

The basic operational goals for the system were to provide a means of interrogating and actuating the switchable retroreflective films with only specific light sources, and to limit access to the reflected signal. The block diagram in Figure 4.1 shows the general system solution utilized to achieve these goals.

Referring to Figure 4.1, the system can be realized with two main blocks: an interrogator and a receiver. The interrogator is user enabled by depressing a push button. The interrogator has two basic functions: (a) provide a light source that can be viewed by an individual (naked eye or night vision) after retroreflection; and (b) provide an encoded signal to the receiver to tell it to actuate (retroreflect). These were achieved by use of laser diodes with an encoded data/pulse stream.

The receiver block functions include: (a) receive the interrogating optical signal

(wavelength and code), (b) decode the signal and process it to determine if it matches the stored code, and (c) drive the switchable retroreflective film if matching is successful.

The interrogating light source is retroreflected back to the interrogator when the retroreflective film is electrically powered, allowing the signal to be viewed through the interrogator’s observation optics.

In order to choose the appropriate parameters for the full system, the retroreflective films first needed to be fully characterized and modeled. This then allows predictive system-level design, especially when attempting long-distance operation.

4.3.2 Retroreflector Film Characterization

The primary optical component of the switchable retroreflective film is a micro- replicated array of corner cube retroreflectors [14, 15]. As shown in Figure 4.2, each corner-cube is comprised of three orthogonally connected mirrors forming an inverted trihedral reflector. Incident light enters the aperture and reflects off all three surfaces, the last one of which reflects the light back in the direction of, but parallel to, the incident light in a narrow cone of about ±0.2o. This divergence is caused be diffraction effects due to wavelength ( λ ) and size of the corner cube aperture ( sidewall, a ), approximated by

[14].

Figure 4.2c shows the normalized retroreflected intensity response due to incident angle for the bare retroreflective film. A line fit was provided by the following heuristic model [16]:

(4.1)

where RMAX is the maximum retroreflected relative intensity in units of retroreflected percent (at θ=0) and γ is the heuristic parameter for best line fit. It was determined that γ

= 5.25 agrees well at all wavelengths measured (λ=400-1000 nm).

As shown in Figure 4.3, the electro-optical switching is performed by adding a polymer dispersed liquid crystal (PDLC) shutter, whose fabrication is discussed elsewhere [17–19]. PDLC consists of several randomly oriented droplets of liquid crystal

(location and angle) in a suspending polymer. In the off state, incident light scatters, causing a diffuse reflection (no retroreflection). When voltage is applied, the liquid crystal

Figure 4.2. Fundamentals of corner cube retroreflectors: (a) collinear top view of a corner cube array with on-axis incident light showing the active and dead areas of a single cube. (b) SEM of a corner cube with an example ray trace. Corner cube sidewall lengths, x, are typically about 150-800 microns for sheeting materials. (c) Retroreflection intensity vs. input angle of the retroreflective film measured using the characterization setup of Figure 4.5 for 635nm with the points being measured data and the line being a heuristic fit curve from equation

(1).

Figure 4.3. Polymer dispersed liquid crystal switchable retrorefelective film in its

(a) voltage OFF/scattering state and (b) voltage ON/retroreflecting state. droplets align with the electric field, making the films optically clear, allowing retroreflection.

The inorganic layers in the device are very thin, specifically ~100 nm In2O3:SnO2 electrodes (ITO) and ~100 nm of reflective Al. All other materials are polymers, and therefore the entire device can be bent to a radius of curvature of <5 cm. Images of the films in operation can be seen in Figure 4.4 with the on state showing high retroreflected intensity when observation is performed in line with the interrogating light, and attenuation

63 of the retroreflected light when observed slightly off axis (~1.5 degrees).

Figure 4.4. Images showing the operation of the polymer dispersed liquid crystal

switchable retroreflector in its voltage OFF/scattering state, voltage

ON/retroreflecting state with viewing slightly off axis to show the color and

attenuation, and retroreflecting with viewing on axis with the light source.

The setup shown in Figure 4.5 was used to characterize the on/off retroreflection response versus wavelength and angle for the switchable retroreflective films. It consists of a 150W MR16 halogen lamp (Osram 54732) coupled into an optical fiber (Ocean

Optics P200-2-VIS-NIR). The output was collimated with a 200mm bi-convex lens placed

200mm from the end of the fiber. The desired spot diameter was acquired by placing an aperture just before the lens. Collimated light transmitted through a beam splitter adjusted to allow the light to be incident on the switchable retroreflector. Retroreflection transmitted light back through the beam splitter to be incident on the collimator attached to a fiber optic cable (Ocean Optics QP600-1-SR) located ~31cm from the retroreflector.

Figure 4.5. Spectral response characterization setup for the switchable retroreflector.

Figure 4.6. (a) Retroreflection intensity vs. input angle for the PDLC retroreflective film in its on and off states with data collected for 635nm with the characterization setup shown in Figure 4.5. Points are measured and the “on” line is the heuristic fit curve using equation (1) with γ = 6.85. (b) ON/OFF percent of retroreflection versus wavelength from 400 to 1000nm, calibrated to a bare retroreflector film. The black line is the measured data with the red overlay being a trend curve.

The collimator collected the characteristic ±0.2o cone of retroreflection. The collected light was then guided into a spectrometer (Ocean Optics HR4000-CG-UV-NIR) in which data was analyzed using Ocean Optics SpectraSuite version 1.4.2. As a reference, a bare retroreflector film was used, angled by ~2 degrees to eliminate specular reflection.

The maximum retroreflected reading was then found by adjusting the collimator and 0.6 mm fiber.

The results of the setup from Figure 4.5 are plotted in Figure 4.6. As shown in

Figure 4.6a, comparing the on/off retroreflection response as a function of incident angle yields a maximum reflection that is 57% of the response of the bare reference retroreflective film. The spectral response was also measured (Figure 4.6b) at 0o incident angle and shows high contrast over a broad spectrum from 400-1000nm (encompassing visual and night vision wavelengths). The non-uniformity of the collected data is

Figure 4.7. Diffraction intensity at various input angles as a function of percent

retroreflection vs. wavelength as measured using the characterization setup

shown in Figure 4.5 for (a) a bare retroreflector film and (b) the switchable

retroreflector film.

66 suspected to be due to several thin film interference effects.

Also measured with the setup from Figure 4.5 was the angular wavelength response due to incident angle. The inverse proportionality of the retroreflection percentage due to the increase in incident angle remained similar throughout the spectrum. The main difference was with respect to the diffraction angles, in which a select few can be seen in Figure 4.7. When white light is incident at larger angles onto an array of corner cubes, the input apertures and reflection areas become small enough to where the array can act as a reflective diffraction grating. For the bare retroreflective film, these incident angles were between 71-77 degrees with relatively consistent intensities between ~11-13% of the max retroreflection for any given wavelength. The switchable films showed a substantial decrease and shift of the diffraction spectral envelope with angles between 57-59 degrees. The intensity was not consistent however with the near infrared wavelengths (~715-900nm) showing a higher intensity (~slope linearity from 13-

22% respectively). The visible wavelength intensities, however, were lower from ~3-13%

(~linear slope) of the max bare film retroreflection for 400-715nm.

4.3.3 Electronics and Optics Design and Integration

The following electronics and optics for the interrogation and response prototype were developed and assembled with electronic components from NGC Xetron

(Cincinnati, OH), and optical components (interrogator optics and retroreflective films) from the University of Cincinnati.

Interrogator: The interrogator had the requirement to be able to modulate a light source and provide a small packet of data and relatively uniform illumination of a target from distances between 10 m out to as much as 1 km. A block diagram and optics train

Figure 4.8. Interrogator electronics block diagram feeding into its optics train. A

bare laser diode has an elliptical output, which requires beam shaping along

each axis as shown with the different ray trace colors (red vs. yellow). The optics

provide variable divergence as shown with the dashed secondary location of f4

(f4B) providing the subsequent dashed ray trace. of the pursued solution for the interrogator is shown in Figure 4.8.

The interrogator is a microcontroller driven laser diode with a push button user interface. A bare laser diode output has a highly divergent elliptical shape (±5-10o or ±15-

30o axis based) that requires beam shaping along each axis to circularize it. Shaping was implemented by use of plano-concave cylindrical lenses oriented based on the axis they needed to shape. The higher diverging axis uses two lenses (f=15 mm and 25.4 mm) spaced 8 mm and 15 mm (respectively) from the laser output to provide a ~12.7 mm diameter collimated beam. The beam trace for this path is shown in dark red in Figure

4.8, prior to the aperture. The lower divergence axis uses only one f = 40 mm lens placed 40 mm from the source for collimation. This trace is shown in yellow of Figure 4.8 prior to the aperture. The beam then passes through 12.7 mm diameter circular aperture for final beam shaping. The output lenses (f4 = -25 mm and f5 = 50 mm) provide the controlled expansion/collimation for a ~25.4 mm output diameter beam (bright red trace post aperture). Variable divergence is accomplished by adjusting f4 from its collimating

o position (f4A) closer to f5 (position f4B). This provides a maximum divergence of ~±5 . The output trace adjustment for this is shown with the dashed orange lines.

Power consumption is approximately 30 mW with the laser off (standby) and 300 mW with the driven laser. It currently runs on two CR123A batteries in series. Simple relocation of the push-button switch can allow for zero battery drain, which would limit the battery life to the number of interrogations at the expense of increasing the system response time.

Receiver: The basic function of the receiver was to take in the optical signal

(data/code) from the interrogator, process it, confirm the code is correct, and then drive the switchable retroreflective film. The receiver also needs to function in a broad range of ambient light conditions (night and daylight) and with multiple laser diode wavelengths. the receiver also should be able to operate with low enough power to be able to use small, lightweight batteries (coin cell). Figure 4.9 shows the block diagram of the electronic solution used for the receiver.

Figure 4.9. Receiver/switchable retroreflector driver electronics block diagram.

The optical reception and analog circuitry are shown in red while the digital

portion is shown in blue.

The front end consists of a Si PIN photodiode. Typical PIN photodiodes have spectral sensitivities from ~200-1100 nm, with their highest sensitivities (typ.) in the 850 nm range, making them ideal for this system. The received optical signal is converted to a current by the photodiode, which is fed into a transimpedance amplifier (TIA) for conversion and amplification. Received ambient light passes the TIA as direct current. A high pass filter (HPF) filters the DC component. A low pass filter, integrated into the feedback of the TIA, filters out the higher frequencies, further reducing the processed bandwidth. The signal is then shifted up to modulate around a reference voltage, provided by the Reference Buffer, and further amplified with a Voltage Summer amplifier.

This provides a signal large enough to recreate the received data using a comparator with the same reference voltage as used in the summing stage. The reconstituted data is input directly into a microcontroller where it is read and compared to the stored code. If the codes match, then the microcontroller signals the switchable retroreflector’s driver to modulate the incident light.

The receiver/driver present standby power consumption is <4 mW and can run on two CR2032 coin cell batteries for ~100 hours. With further optimization,

Figure 4.10. Custom optical bandpass filter used for daytime demonstrations.

Table 4.1. Laser diode limits for receiver detectable spot sizes and associated irradiances with/without a filtered photodiode (see Fig. 4.10). Max Spot Min Spot Laser Output Diameter Diameter Min Irradiance Max Irradiance Wavelength Power Power No No (nm) Filter Filter Filtered No Filter Filtered No Filter (mW) (mW) Filter Filter 2 2 2 2 (m) (cm) (nW/cm ) (nW/cm ) (mW/cm ) (mW/cm ) (m) (cm) 5 2.5 1.4 1.7 1.6 1.8 635 10 4.5 1.9 2.3 2.1 2.4 153.5 104.4 1.25 1.02 20 10 2.9 3.5 3.2 3.5 5 1.5 - 1.5 - 1.5 850 - 87.55 - 0.85 10 3.2 - 2.2 - 2.2

consumption can be closer to 0.3 mW and would be able to operate on one CR2032

coin cell battery (225 mAh) for >1500 hours. Operation in daylight conditions does

require optical filtration on the photodiode, of which the transmission spectrum shown in

Figure 4.10 is for the custom assembled filter used in this study. Nighttime conditions do

not require such a filter. The receiver’s limit on optical irradiance and the associated

Figure 4.11. Image of the (a) interrogator and optics, (b) receiver and tag, and (c)

example interrogator setup with a spotting scope. The ruler show at the bottom

of (a) and (b) is 6” long, for reference.

71 spots sizes for the various laser diodes used are shown in Table 4.1 (and will be discussed in more detail in Section 4.4).

System: The completed system can be seen in Figure 4.11. The interrogator, shown in Figure 4.11 (a), is set up to be able to mount to a spotting scope for simple alignment/zeroing with viewing optics, as shown in Figure 4.11 (c).

The setup shown in Figure 4.12 (a) was used to characterize the system temporal response from the initial push button press to 90% of the maximum retroreflected response. The setup uses a 0.5 mW HeNe laser (Melles Griot 05-SRP-

810). The HeNe beam is split and directed to the switchable retroreflector with a beam splitter. The photodetector (Newport 918C-SL-OD3) was aligned to receive the retroreflected signal using a base retroreflective film. The analog output of the optical

Figure 4.12. a) Characterization setup to measure the temporal response. b)

Oscilloscope screen shot of the system temporal response showing both the

push-button depression (top black line) and the total response time of the

receiver as it turns on the retroreflector film response (bottom red line).

72 power meter (Newport 1918-C) is connected to an oscilloscope along with the output of the push button. The oscilloscope is set to trigger with the rising edge signal created by the depression of the push button. The fastest theoretical response of the system is

8.35ms. The fastest achieved response was 9 ms, however 10-15 ms is more typical as shown in the oscilloscope screen shot of Figure 4.12(b).

4.4 Optical Model for Maximum Distance

Understanding the maximum distances at which the system can operate has obvious value from an applied perspective, and is interesting theoretically in terms of the optics/physics involved. From an optical perspective, the system can be viewed as two divergent radiators separated by a certain distance in free space with air in between. In this case, we can use a geometric means to calculate the returned optical power when the switchable retroreflector is oriented at zero degrees. A diagram of this optical model is shown in Figure 4.13, which can be described as follows.

Figure 4.13. Diagram of the optical model of the system.

The only relevant light coming into the system is from the interrogating light source. All other variables in the system will affect this to determine what is actually detected. Incident irradiance (EI) on the retroreflector can be expressed as:

(4.2)

where PI is the light source’s output optical power ( mW ), α is the ambient attenuation coefficient ( dB/km ), x is the distance between the light source and film, and r is the radius of the incident spot on the retroreflector ( cm ).

The effect the switchable retroreflector has on the incident irradiance to provide the return beam can be expressed as:

(4.3)

2 where ASR is the area of the switchable retroreflector ( cm ), ηSR is the retroreflector efficiency coefficient, θR is the retroreflected divergence, and HSR is the height of the film

( cm ). A second attenuation is introduced due to the retroreflected beam transmitting through the same free space distance as the incident light.

The observation optics and detector have an associated optical acceptance area

2 (AD in cm ) and efficiency (ηD). Combining this with equations 4.2 and 4.3 provides a complete expression for the detected optical power (PD in mW) for a given detector area:

(4.4)

By setting AD to a unitless value of “1”, one can calculate the detectable irradiance in units of mW/cm2.

Equation (4.4) can be alternatively expressed with the light source’s divergence

(θI) instead of incident spot radius by setting r = x tan θI. Equation (4) assumes that the incident light is uniform and fully illuminating the switchable retroreflector, the incident and retroreflected beams are circular, and that the light source/laser is a point source.

Example curves (irradiance vs. distance) can be seen in Figure 4.14 for various spot diameters, laser outputs, and wavelengths on log-log plots. Angular orientation can also be taken into account by adjusting equation (4):

(4.5) with γ = 6.85 for these films.

The irradiance plots in Figure 4.14 are only meaningful from an applications perspective if minimum detection power can be understood on the interrogator end of the system. Therefore, a model is also provided to determine the irradiance threshold levels necessary for naked eye visualization. This is based on the background surrounding the retroreflective films having an associated reflection coefficient (albedo) of sunlight. This can be modeled by assuming the surroundings are diffuse reflectors [20-21], using a form of Lambert’s Cosine Law [22] for the received optical power on the detector due to ambient reflections (PDA):

(4.6)

where PAI is the incident ambient light on the surface of the background, AD is the area of the detector aperture, θO is the observation angle between the ambient light source on the retroreflector and the observing optics, and x is the distance between the film and the observing optics.

For human eye observation, alternative criteria need to be considered when using equation (6). In high ambient light (daytime) conditions, the human eye is in its highest acuity photopic response. This response relates to the use of the color sensitive cones only, which are physically located in the fovea of the eye. The fovea’s focusing field of view is ~2 degrees, which provides the subtended background area at the

75 location of the retroreflecting film that determines the reflected ambient saturation levels that need to be overcome to identify the retroreflection with the naked eye. This leads to the following equation for the incident background illumination (PAI) from equation (6) being reflected back to the viewing optics (eye):

(7)

where EV is the ambient illuminance (lux), AFOV is the area subtended by the foveal field of view, and θFOV is the foveal field of view angle. This assumes that the focal area of the eye is circular in shape.

The final threshold flux, ϕV-Thres (lumens), that the eye needs to see over the ambient conditions can be expressed as:

(8) which takes into consideration the ambient attenuation (α), the observation optics and detector efficiency coefficient (ηD), the reflection coefficient (R) of the surrounding surface, and the contrast change (C) needed for a given wavelength for the eye to perceive a difference in of an object compared to its surroundings. The threshold flux can then be converted to its radiometric unit for the specific wavelength used for interrogation (watts). Example plots of this can be seen in Figure 4.14 showing the threshold irradiances needed for visualization in various visible lighting conditions and wavelengths. Lighting conditions ranging from direct sunlight (30k lux) to overcast sunset (22 lux) are plotted. All other variables used can be seen in Table 4.2.

Figure 4.14. Theoretical plots of retroreflected irradiance (I) as a function of distance (x) for various interrogator spot sizes, laser powers, and laser wavelengths at the tag. The horizontal red plots show human eye photopic threshold levels for various ambient illuminance (Ev) levels along with an example night vision (NVIS) threshold.

Table 4.2. Values used for calculations

Retroreflected Irradiance Description Variable Value Units Attenuation Coefficient α 0.1 dB/km

2 Switchable Retroreflector Area ASR 77.4 cm

Switchable Retroreflector Efficiency Coefficient ηSR 0.25 -

Switchable Retroreflector Height HSR 7.62 cm

o Retroreflected Beam Divergence Angle θR 0.2

Observation Optics and Detector Area AD 1 -

Observation Optics and Detector Efficiency ηD 1 -

Naked Eye Thresholds Description Variable Value Units

Ambient Illuminance EV 22-30k lux 2 Detector Area AD 0.0001 m o Observation Angle θO 80 o Field of View Angle θFOV 1 Reflection Coefficient R 0.3 - Contrast Change C 0.006 -

4.5 Demonstration

Demonstrations were performed in a location consisting of grassy fields and wooded deciduous backgrounds. Daytime conditions were clear (no fog/airborne particulate) of a mostly cloudy/nearly overcast sky with illuminance levels at ~18 klux as measured with a Circuit Specialists MS8209 Multimeter. Interrogated distances ranged from ~50 m to 400 m, the latter of which is shown in Figure 4.15 using a 10 mW output

635nm light source. Viewing and imaging were performed through a Vortex Viper HD

80mm spotting scope. The switchable retroreflector showed sufficient contrasts for visual conspicuity. Successful electronic interrogation response was performed from

~50-200 m using a 2.5 mW output light source with spot sizes up to ~80 cm, compared to lab condition (~5 m distance) spot sizes of ~86 cm.

Nighttime demonstrations were performed in the same location over the same distance range. A night vision monocular device was used for viewing and imaging purposes. The light source used was a 1.5 mW output laser with peak wavelength at

850 nm. Electronic response was achieved with a spot size of ~104 cm (diameter). In comparison, lab condition measurements provided an equivalent spot

Figure 4.15. Example demonstration of a switchable retroreflector (~10 x 7.5c m)

interrogated at a distance of 400 m with a 10 mW output 635 nm laser and a spot

size of ~50 - 75 cm. Image was taken through a Vortex Viper HD 80mm spotting

scope with a Canon Power Shot S3 IS camera. The original image has been

modified from its original state.

Figure 4.16. Example demonstration of a switchable retroreflector (~10 x 7.5 cm)

interrogated at a distance of 100 m with a 1.5 mW output 850 nm light source

with an elliptical divergence of ~5.5 degrees and ~4.5 degrees being viewed with

night vision through a spotting scope. The spot that shows up below the

highlighted area in (b) is the reflection off the surface of the retroreflector being

reflected off the surface of a pond. The smaller spots to the left are reflections

from external lights. size of ~147 cm. Optical response (manual switching) was achieved over the full range with the highest divergence setting (elliptical shape ~4-5.5 degrees). This can be seen in Figure 4.16.

4.6 Discussion and Conclusions

The system discussed and demonstrated in this chapter provides proof-of- concept results for an optical discrimination system with PDLC based switchable retroreflective films. The switchable retroreflective film’s high optical efficiency, contrast, and broad-spectrum operation makes it an ideal solution for enhanced conspicuity for signage, range finding, and tagging, among other possible applications. We have also shown that the switchable retroreflective films can operate over a temperature range of -

15 to +95 degrees Celsius, fully submerged underwater, and with several induced film piercings/tears lending them to be useful even in harsh environments.

The discussed setup was for proof of concept demonstrations with ease of alignment, interrogation, and observation in mind, and may not necessarily be useful to all industries. However, the simplicity of the system allows it to be easily integrated with existing systems by replacing static retroreflective or light emitting tags with the switchable receiver/films we report here, and by retrofitting light sources (LEDs/Lasers) with simple encoding circuits. The size, weight, and power consumption can be substantially reduced with further optimization and the use of application specific integrated circuits allowing further ease of mobility and decrease in frequency of battery change.

4.7 References

[1] E. Laskowski, “Range finder wherein distance between target and source is determined by measuring scan time across a retroreflective target,” US Patent 4,788,441, 1988.

[2] G. V. L. Jr, “Reflectors used in highway signs and warning signals, Parts I, II, III,” Journal of the Optical Society of America ( …, pp. 462–487, 1940.

[3] S. J. Gerathewohl, “Conspicuity of steady and flashing light signals: variation of contrast.,” Journal of the Optical Society of America, vol. 43, no. 7, pp. 567–71, Jul. 1953.

[4] D. Gunawan, L. Lin, and K. Pister, “Micromachined corner cube reflectors as a communication link,” Sensors and Actuators A: Physical, vol. 7, pp. 5–8, 1995.

[5] W. S. Rabinovich, S. R. Bowman, D. S. Katzer, and C. S. Kyono, “Intrinsic multiple quantum well spatial light modulators,” Society, vol. 66, no. August 1994, pp. 1044–1046, 1995.

[6] S. C. Binari and G. C. Gilbreath, “A Cat ’ s Eye Multiple Quantum-Well Modulating,” Technology, vol. 15, no. 3, pp. 461–463, 2003.

[7] W. S. Rabinovich, “Free-space optical communications link at 1550 nm using multiple-quantum-well modulating retroreflectors in a marine environment,” Optical Engineering, vol. 44, no. 5, p. 056001, May 2005.

[8] W. S. Rabinovich, P. G. Goetz, R. Mahon, L. Swingen, J. Murphy, M. Ferraro, H. R. Burris, C. I. Moore, M. Suite, G. C. Gilbreath, S. Binari, and D. Klotzkin, “45- Mbit/s cat’s-eye modulating retroreflectors,” Optical Engineering, vol. 46, no. 10, p. 104001, 2007.

[9] M. G. Pollack, R. B. Fair, and N. Carolina, “Electrowetting-based actuation of liquid droplets for microfluidic applications,” Applied Physics Letters, vol. 77, no. 11, pp. 1725–1726, 2000.

[10] F. Mugele and J.-C. Baret, “Electrowetting: from basics to applications,” Journal of Physics: Condensed Matter, vol. 17, no. 28, pp. R705–R774, Jul. 2005.

[11] M. K. Kilaru, B. Cumby, and J. Heikenfeld, “Electrowetting retroreflectors: Scalable and wide-spectrum modulation between corner cube and scattering reflection,” Applied Physics Letters, vol. 94, no. 4, p. 041108, 2009.

[12] M. K. Kilaru, J. Yang, and J. Heikenfeld, “Advanced characterization of electrowetting retroreflectors.,” Optics express, vol. 17, no. 20, pp. 17563–9, Sep. 2009.

[13] P. Schultz, B. Cumby, and J. Heikenfeld, “Investigation of five types of switchable retroreflector films for enhanced visible and infrared conspicuity applications.,” Applied optics, vol. 51, no. 17, pp. 3744–54, Jun. 2012.

[14] H. D. Eckhardt, “Simple model of corner reflector phenomena.,” Applied optics, vol. 10, no. 7, pp. 1559–66, Jul. 1971.

[15] Reflexite, “Reflexite Corporation,” 2011. [Online]. Available: http://www.reflexite.com.

[16] A. Marchant, “Conspicuity tape for enhanced laser range finding,” Optical …, vol. 49, no. April, pp. 1–7, 2010.

[17] P. S. Drzaic, “Polymer Dispersed Nematic Liquid Crystal for Large Area Displays and Light Valves,” Journal of Applied Physics, vol. 60, no. 6, pp. 2142–2148, 1986.

[18] J. W. Doane, A. Golemme, J. L. West, J. B. Whitehead, and B. G. Wu, “Polymer Dispersed Liquid Crystals for Display Application,” Molecular Crystals and Liquid Crystals Incorporating Nonlinear Optics, vol. 165, no. 1, pp. 511–532, 1988.

[19] G. Spruce and R. D. Pringle, “Polymer dispersed liquid crystal (PDLC) films,” Electronics & Communication Engineering Journal, vol. 4, no. 2, pp. 91–100, 1992.

[20] L. Mandelstam, “Light scattering by inhomogeneous media,” Zh. Russ. Fiz-Khim. Ova, 1926.

[21] M. Kerker, “The scattering of light, and other electromagnetic radiation,” 1969.

[22] J. Lambert, Lamberts Photometrie:(Photometria, sive De mensura et gradibus luminis, colorum et umbrae)(1760). 1892.

Chapter 5: Applications and Future Work for Switchable

Retroreflector Films and Visual Identification Systems

5.1 Introduction to this Chapter

The previous chapters have provided motivation, theory, and device investigations of various technologies used for switchable retroreflectors. They then provided design, assembly, and operational results for a rapid visual identification system developed around large area switchable retroreflective tags. Such devices and systems have several potential commercial applications. This chapter will describe these applications and provide details on the future work needed to achieve further integration and improvements.

5.2 Retroreflective Tag

5.2.1 Tag Specific Applications

The retroreflective tags discussed and developed herein have several potential applications alone. The ability to scale the tag size up in area, its flexibility, and its low power operation makes it a suitable device for increasing the conspicuity of road signs, road markings, vehicular markings, and pedestrian safety gear. Figure 5.1 shows a few examples. Such devices can be set up with a simple electronic driver that will continually switch from diffuse to retroreflecting at a certain interval, powered by a hybrid solar/battery supply. The flashing action, much like that of LED lined signs, helps draw attention, increasing awareness, and safety. However, since it is not an emissive technology like LED, it would not be causing a nuisance flooding of light into areas it is not needed. This would be especially beneficial in a residential neighborhood where individuals may be sensitive to such issues.

Figure 5.1. Examples of enhancing conspicuity by way of integrating a flashing

The tags can also be used for dynamic signs to control traffic flow on demand. A large scale active matrix version [1] could replace the large message signs found over top expressways or in construction zones. They could also be more simplistic such as re-routing/detour arrows or temporary road closures.

Some of the potential issues that may arise would be the decrease in optical efficiency as discussed in Chapter 3, and the increase in power consumption when scaling such devices up. If placed on a vehicle or anything with a continually rechargeable power system, it is a non-issue. However, it may be an issue when there is heavy reliance on standalone battery power. The tags tested here, operated at

85 approximately 100-400 µW per square centimeter, but in conjunction with the electronic driver, operated at approximately 2.04mW per square centimeter (3VDC Supply to driver, tag driven at 50Hz +/-90V square, ~80% duty) with a linear increase in consumption with increase in active area. These noticeable inefficiencies are detrimental, leading to a substantial load increase on the power system. Adjustments can be made to the driver for lower consumption. This will be discussed in Section 5.3.2.

5.2.2 Tag Specific Future Work

Materials

Several improvements can be considered with the tag fabrication process, mostly with respect to materials used. One of the biggest issues with the use of PDLC, with the current electrode materials and substrate, is with the bending radius. When bent too much, the PDLC layer becomes transparent, allowing retroreflection without an applied bias. This is mostly due to the rigidity of the substrate materials and stress on the suspension polymer of the PDLC network as it is comparatively softer and more flexible.

With multi-layered devices, each layer will have a different radius of curvature, the inner most of which will have the smallest. With the more rigid substrates, this inner most layer is going to shift up closer to the outer substrate at the apex, creating a thinner

PDLC layer (at the apex) and shear stresses on the rest of PDLC, affecting its optics.

There is also a limit on the bending radius of the ITO electrode material, typically between 1-2 cm [2].

Bending radius issues can be addressed with the use of more flexible substrates and electrode materials. A combination of a thinner polymer substrate (PET, PEN, etc.) and a more flexible transparent electrode material will allow greater overall flexibility, decreasing the PDLC stresses. Potential electrode materials would include PEDOT, ITO

86 nana-wires [2], or metallic nano-wires [3], which have shown to improve the bending radius to better than 2.5 mm.

Fabrication

Structure - One improvement that can be pursued with respect to structure is taking out the bottom electrode substrate of the PDLC stack. The top side of the retroreflective film would then need to be coated with a transparent electrode and the

PDLC would be fabricated on top. This can be seen in Figure 5.2 (b).

Figure 5.2. Potential improvements to the structure of the PDLC based

switchable retroreflector. Please note that these figures are not to scale for

visualization purposes only. The corner cube facets for retroreflective films

typically have aperture openings around 200 microns whereas the liquid crystal

droplets of the PDLC are typically < 5 microns diameter.

The reason behind taking out layers is to improve the optical efficiency of the tags. The more layers taken out, the less reflective and absorptive losses can occur as incident light transmits through the stack. This allows for the use of lower power light sources and tags that function much more closely to the original retroreflective films. It would also improve the overall flexibility and lead to lower weight.

Droplet Size – Liquid crystal droplet size in the PDLC is the main specification that determines several operating parameters [4]. Larger droplets provide better on state transmission efficiency with lower operating voltage, but suffers from lower scattering efficiencies in the off state. Smaller droplets allow for higher speeds and better off state scattering efficiencies at the expense of lower on-state transmission and higher operating voltages [5–7]. Droplet size control is a common practice [8–10] to obtain specific operating conditions needed for a device in which a balance is needed based on the application at hand.

Localized Switching – Currently the entire area of the tag is actuated regardless of the incident light direction or tag orientation. This means a higher power consumption when the tag is driven and the potential for an unintended off axis observer to see the tag. One option that could be pursued would be to segment the electrodes on the tag.

This would require additional photodiodes per segment with the risk of potentially increasing standby power consumption.

Surface Reflection – Currently, the tags possess a semi-specular surface reflection due to the materials being used. It is also necessary to have conformal/non- diffuse surfaces to allow retroreflection to occur properly without diffusing the returned beam. This surface reflection may not be desirable for certain applications that need to be fully inconspicuous in the off state.

Some early work has been started on thermoforming macro-sized features into a fabricated tag to provide a continuous wavy surface. Ideally, this would allow the surface reflections to be spread out in several different directions, while on the micro- scale, allow interrogating light to transmit through and allow retroreflection to occur normally. This idea can be seen in Figure 5.3 showing the reflection differences in the planar and formed tags.

Visual and ImageJ software analysis [11] have shown this effect from a near field perspective as shown with the comparison in Figure 5.4. This also has shown to be effective in far field observation as can be seen in Figure 5.5. From a short distance (~2 meters) the reflection intensities are relatively similar though noticeably attenuated

Figure 5.3. The difference in surface reflections between (a) a planar

retroreflective tag, which, in the far field, would be viewed as a bright flash, and b)

a thermoformed tag that reflects light in all directions, spreading out the intensity.

Figure 5.4. Thermoforming data with ImageJ analysis showing differences in the near field surface reflection.

Figure 5.5. Far field (~10 m) observation of a planar and thermoformed tag.

The image taken with direct sunlight behind with a Canon Rebel T3I set to Auto

White Balance with 1/4000 s exposure time and f/45 aperture. Notice the

specular reflection of the planar tag is still bright compared to the thermoformed

tag. with the thermoformed tag. However, the farther out the observer is, the more the tag appears to become diffuse. The only downside to this approach is the fact that the surface reflection can now be seen from a wider field of view with continual attenuation the farther off the normal axis (from the surface) the observer is. In the sense of conspicuity, this would be somewhat beneficial since it is not a high intensity flash, but rather a subtle change with change in angle.

Additional work is needed with this with variations in wave radius and in conjunction with different material changes and device stack as mentioned previously

(Figure 5.2).

5.3 Visual Identification System

5.3.1 System Level Applications

The rapid visual identification system, being an application itself, can be used for more specialized applications. They would be beneficial where lightweight, low power, and relatively discrete visual indicators are needed. These applications include vehicular monitoring, surveying, and object marking among others.

5.3.2 System Level Future Work

Optics

The light source transmitting optics have a couple potential areas of improvement.

They are currently glass optics. By changing them to polymer, they will provide lower weight and increased impact resistance to the overall system while lowering cost in the end. With design improvements, including more specialized optics, they can have a tighter layout allowing for a decrease in system size and simpler assembly.

Electronics

Improvement of the low signal intensity reception on the tag would be desirable.

This would allow for large area/flood interrogations for easier aiming, and longer distance electronic reception. This can be improved simply by the use of additional photodiodes and/or focusing lenses. Another method would be to integrate a structure similar to an LCD backlight diffuser into certain areas of the tag to couple the interrogated light into a waveguide directed to a single photodiode. This would allow for the photodiode to be mounted closer to the electronic board to help prevent noise injection, as this may be one of the issues causing higher standby power consumption in the receiver board. This structure would have to be surrounding the active retroreflection area to prevent the retroreflected light from being diffused.

An automatic gain control (AGC) circuit on the input signal of the receiver would provide improvements. It would improve the allowable incident interrogator intensity

(high and low end), and it would broaden the range of ambient lighting conditions the system can operate in without additional external components (optical filters, etc.). This circuit would need an ambient light sensor along with the ability to determine the incident interrogator intensity. These would be used to provide automatic gain adjustment of the transimpedance amplifier and the summing amplifier stages to get a usable signal for the data regeneration. This can be implemented with the on board microcontroller, analog to digital convertors, and voltage controlled resistors. Such a system would allow operation during the day or at night with no manual modification. It would also allow simpler integration of different wavelengths as photodiodes have wavelength dependant sensitivities. However, this would increase power consumption and board space.

An improved electronic tag driver would be beneficial. As mentioned in Section

5.2.1, scaling up the tag causes a substantial load increase on the current battery powered electronic system, mostly due to the driver inefficiencies. The driver output frequency can be adjusted to about 25 Hz (flicker limit), in which this can cut the power down to less than 1 mW/cm2. The output voltage of the driver can also be adjusted.

Tag switching speed and transmission efficiency are based on applied voltage. The higher the voltage, the faster the tag can switch; however, higher speeds may not be necessary based on the application. In such a case, the voltage output from the driver can be lowered to about 55 V (tested tag transmission saturation). Further tag improvements can lower this to < 20V for alternative and more efficient drive circuit architectures. For the most efficient drive circuit, ideal PDLC can be driven at 5V [6], which would lead to a tag load of approximately 1-4 µW/cm2.

The size and power consumption of the electronic boards can be substantially improved with the use of a custom designed application specific integrated circuit (ASIC).

In other words, it would be a single chip for all of the functionality of the current boards.

This would also allow for mounting the electronics directly to the back of the retroreflective tag itself allowing a smaller, lighter weight system with lower battery change frequency.

Two-Way Communications

Two-way communications would be the next step after the aforementioned improvements. It would however be limited to relatively low speeds (~1 kHz) with the use of PDLC in which better chemistries and alternative switching methods would be needed. The use of the electrowetting lenslet based switchable retroreflector would be ideal with respect to a large areas and higher speed with theoretical speeds of ~10kHz.

This device would need additional work, most notably with better materials to make it flexible and provide long life stability. These are the two greatest hurdles with respect to most electrowetting devices to make them commercially viable.

On the other end, high gain/sensitive electronic circuitry and/or an optical signal intensifier tube would be needed for the returned signal reception. Long-range interrogation and retroreflected signals will be heavily attenuated and have low intensity considering the divergent nature of the retroreflected signal. These high performance circuits would add cost and weight to the system that some users may not be able to afford.

5.4 Conclusion

In this chapter, several potential applications were introduced for the switchable retroreflective tags alone and when integrated in the visual identification system built around it. All of these applications are based on the improvement of conspicuity in a

94 given area (large area flashing) along with providing the means to disable retroreflection when desired. Other proposed improvements would provide additional functionality making it more desirable from a commercial standpoint.

5.5 References

[1] C. Sheraw and L. Zhou, “Organic thin-film transistor-driven polymer-dispersed liquid crystal displays on flexible polymeric substrates,” Applied Physics …, 2002.

[2] H. Wu, L. Hu, T. Carney, and Z. Ruan, “Low reflectivity and high flexibility of tin- doped indium oxide nanofiber transparent electrodes,” Journal of the …, 2010.

[3] S. De, T. Higgins, and P. Lyons, “Silver nanowire networks as flexible, transparent, conducting films: extremely high DC to optical conductivity ratios,” ACS …, vol. 3, no. 7, pp. 1767–1774, 2009.

[4] G. Spruce and R. D. Pringle, “Polymer dispersed liquid crystal (PDLC) films,” Electronics & Communication Engineering Journal, vol. 4, no. 2, pp. 91–100, 1992.

[5] G. Montgomery, “Light scattering from polymer-dispersed liquid crystal films: Droplet size effects,” Journal of applied …, pp. 1605–1612, 1991.

[6] D. Coates, “Polymer-dispersed liquid crystals,” Journal of Materials Chemistry, vol. 5, no. 12, p. 2063, 1995.

[7] V. G. Chigrinov, Liquid Crystal Devices: Physics and Applications. Artech House Optoelectronics Library, 1999, p. 366.

[8] R. H. Chen, Liquid Crystal Displays: Fundamental Physics and Technology. Wiley, 2011, p. 350.

[9] A. Lackner, “Droplet size control in polymer dispersed liquid crystal films,” OE/LASE’89, …, 1989.

[10] J. a. Ferrari, E. a. Dalchiele, E. M. Frins, J. a. Gentilini, C. D. Perciante, and E. Scherschener, “Effect of size polydispersity in polymer-dispersed liquid-crystal films,” Journal of Applied Physics, vol. 103, no. 8, p. 084505, 2008.

[11] Rasband, Wayne and National Institute of Health (NIH), “ImageJ Image Processing.” [Online]. Available: http://rsbweb.nih.gov/ij/.

Chapter 6: Summary and Conclusion

6.1 Introduction to this Chapter

This dissertation described, in detail, switchable retroreflectors and their integration into an optical discrimination system. It provided background information, design, fabrication, and demonstration results of all components. It also went into performance results and described some limitations, potential improvements, and some additional applications. This chapter summarizes the important aspects of this dissertation (performance, limitations, etc.) and provides a statement of separation of work between the different parties involved in the project.

6.2 Summary of System Performance and Limitations

Individual component and full system performance specifications and results are shown in Table 6.1 below. The table shows what was achieved and, for some of the specifications, the theoretical best with various improvements.

Overall system limitations mostly come in the form of operation in high ambient light intensity and operational power of the receiver/tag driver, however are also seen in material properties and circuit component choices.

In high ambient light conditions, receiver signal reception and visual observation are easily saturated. In its present form, optical filters, lower divergence, and higher optical power light sources are used to achieve the desired range. Optical filters increase the cost and weight of the retroreflective tag. Lower divergence makes it more difficult to aim the interrogator, and higher optical powers increase power consumption and bring about eye safety concerns. A potential system improvement to help with this is an automatic gain control (described in Chapter 5) to broaden the allowable interrogator signal intensity, and broaden the operational ambient light intensity.

Table 6.1. Specifications and results for visual identification system and its various components

Retroreflective Tag Result / Specification Comments Size - Demo / Theory 1 cm2 - 155 cm2 / > 1 m2

Optical Efficiency 25% Chapter 3 Input Angle ±38 Chapter 3 Contrast 40:1 Chapter 3 Operating Voltage - Demo / Theory 55V:AC Square / 5V:DC

Switching Speed - Demo / Theory 2ms rise : 30ms fall / 1kHz 2 Weight ~80 mg/cm Adhesive Backing only Thickness 0.6 mm, ~3 mm Adhesive Backing, Velcro Backing Spectral Range 400 - 1600 nm Minimum 2:1 contrast Attributes Flexible, cut to shape, impact resistant, custom graphics

Interrogator Result / Specification Comments Wavelengths Demonstrated 635, 850 nm

Output Optical Power 1-10 mW

Supply Voltage 6V 2 x 3V CR123A Batteries (Series) Supply Power 30 mW, 300 mW Standby, Active Max Divergence 5.5o x 4.5o Elliptical, typ. Weight 131 g, 300 g Electronics, Mounted Optics Packaged Electronics Dimensions 120 mm x 62 mm x 32 mm L x W x H Mounted Optics Dimensions 120 mm x 105 mm x 50 mm L x W x H

Receiver Result / Specification Comments Supply Voltage 3V 2 x 3V CR2032 Coin Cell (Parallel) Supply Power < 4mW, 45mW Standby, Driving Tag (65cm2) Weight 64 g Packaged Electronics Size 88 mm x 62 mm x 22 mm L x W x H Max Tag Size Demonstrated 77 cm2

Spectral Range 400-1100 nm, ~610-660 nm Full Range, Filtered

System Result / Specification Comments Demonstrated Range 1 - 400 m Chapter 4 - Night and Day Theoretical Range 1 m - >10 km Chapter 4 - Figure 4.14 Response 9 ms, 10-15 ms Chapter 4 - Low, Typ.

Another improvement would be the use of alternate wavelengths not affected as much by ambient conditions (1550nm is ideal). This would require changing the circuit and introducing or changing out the chosen Si photodiode to an indium gallium arsenide based photodiode, but would also need specialty-viewing equipment to observe. Human observation improvements would include the use of magnifying or alternative specialty optics/viewing systems, focused on the tag, limiting the affects of ambient light saturation.

Improved optical efficiency of the tag by ways of decreasing the number of layers and a more efficient base retroreflector would also help, as there would be more optical power available for viewing.

Power consumption of the receiver/tag driver circuit is a limiting factor for long term/high interrogation rate use. This is mostly due to the standby power consumption and choice of battery when operating in low interrogation applications, but increasingly becomes a problem with tag power consumption when going to high frequency interrogations. Improvement of the standby power consumption, as described in

Chapter 5, can be done with more efficient components, circuit duty cycling, and an application specific integrated circuit (ASIC). Also described in Chapter 5 are various methods to improve the tag power consumption including improved PDLC chemical composition, optimized droplet size, PDLC layer thickness, and localized switching.

6.3 Separation of Work

Certain aspects of this project were developed collaboratively with Northrop

Grumman Corporation’s (NGC) Xetron division of Cincinnati, OH. The author performed all of the work presented. Retroreflective tag and tag driver development was performed solely at the University of Cincinnati. The interrogator optics were designed and developed at the University of Cincinnati, with final assembly at Xetron. Xetron's main

98 contribution was providing, as needed, space, tools, software, hardware, and design support of the electronic components (interrogator and receiver).

Of the various switchable retroreflector devices mentioned in Chapter 3, all devices were at minimum assembled at the University of Cincinnati. The Electrowetting

Lenslet retroreflector was developed at the University of Cincinnati by M. Kilaru et. al.

The External Electrowetting Light Valve retroreflector was fabricated using processes from K. Zhou et. al. with the integration of the retroreflector in house. Development of the Integrated Electrowetting Light Valve was performed by the author. The

Conventional Liquid Crystal Light Valve retroreflector was a commercial product with the integration of the retroreflector performed in house. The Liquid Crystal Scattering (PDLC) retroreflector was fabricated in house with processes from various sources (P.S. Drzaic et. al., J. West et.al.). Commercial PDLC films were also used with retroreflector integration performed in house.

6.4 Conclusion

In conclusion, this dissertation has demonstrated several methods to switch a retroreflected signal based on electrowetting and liquid crystal display technologies for human eye response applications. The choice of PDLC as the base retroreflected signal switching method provides key attributes over the current switching technologies (MEMS and MQW) including broad spectrum operation (visible to near infrared), large area, simple and low cost fabrication, flexible, impact resistance, light weight, low power, and high optical efficiency.

This dissertation has also demonstrated the integration of a large area switchable retroreflector, based on PDLC, into a laser interrogation and visual identification system, having potential applications in personnel, vehicular, and object tracking, among others.

Through device testing and various demonstrations, the system has proven to be viable for a dynamic identification system for various day and night applications.

Overall, the system is presently operational for lower switching speed applications, beyond naked eye response times, moving into the realm of electronic detectors. Additional improvements mentioned in this chapter (and the previous) need to be made to the tag and the system to make it more attractive from a commercial standpoint, including smaller electronic size and lower power consumption. Integrating two-way communications would be possible, but ideally would be based on the electrowetting lenslet retroreflector for its higher speeds.

100

Appendix A – Fabrication Sheets

A.1 Integrated Electrowetting Light Valve Retroreflector

a. Top Plate

i. Cut ITO coated PET to desired size

ii. Attached to rigid carrier (glass) with one of the following:

- Full layer transfer tape or double sided tape, remove ITO side

liner

- Remove ITO side liner, then top side perimeter tape down

iii. Coat with 1 gram Parylene C (~1 micron)

iv. Spin coat Fluoropolymer with one of the following:

- 0.5wt% Fluoropel

- 1.0wt% Cytop

v. Air Dry for 20 minutes

vi. Bake at 100 C for 1 hour

vii. Remove from carrier, if necessary remove Parylene C and fluoropolymer

with razor for electrical connection

b. Retoreflective substrate

i. Cut Reflexite supplied uncoated corner cube retroreflective sheet/film to

size

101

ii. Rinse with IPA, water, IPA

iii. Dry with N2 dry

iv. Place in vacuum chamber/over (no heat) - 1-2 Hours

v. Sputter deposit reflective metal electrode (aluminum, silver, etc.) c. Dosing and Assembly

i. Prepare one of the following fluid systems (keep separate)

1. Aqueous/Polar - 1wt% SDS in DI or RO Water, Non-Polar - 3:1

Dodecane (C12):OS20 + Keystone Dye of choice

2. Aqueous/Polar - DI or RO Water, Non-Polar - Dodecane (C12)

+ Keystone Dye of choice

ii. Using a pipette - drop and spread polar fluid over sputtered retroreflector

- Ensure full coverage and fill

- If desired, place in vacuum to ensure voids are filled

iii. Using another pipette - drop 1-2 drops of non-polar fluid on top of polar

fluid in retroreflector

iv. Place top plate, coated side down, on top of dosed retroreflector leaving

at least 1 edge of retroreflector exposed for electrical contact

v. Attach binder clips on the edges

vi. Attach electrodes - alligator clip, conductive epoxy, silver paste

102

Notes: When dosing and assembling, it may be beneficial to drop the non-polar fluid towards one end of the retroreflector and put to top plate down starting from that end

(clam shell method) to spread the non-polar fluid over the entire corner cube array.

A.1 Polymer Dispersed Liquid Crystal Retroreflector

a. Prepare PDLC with one of following methods:

1. Self Made

i. Mix 1:1 E7 Liquid Crystal (Merck) to 10wt% PMMA in Toluene

ii. Spin coat or drop mixture onto ITO coated PET substrate

iii. Place top ITO coated PET substrate (ITO Down), leaving an edge

exposed on both substrates for electrical contacts

iv. Cure on hot plate at 60 C for ~2 hours

2. Purchase commercially available PDLC sheeting

i. Cut to size using laser cutting system (ULS 3.50 - Vector cut

Speed = 100, Power = 80, PPI = 1000)

b. Attach Retroreflector

i. Cut retroreflective sheeting to size (smaller than PDLC) using laser

cutting system (ULS 3.50 - Vector cut Speed = 100, Power = 60, PPI =

1000)

ii. Rinse with IPA, N2 dry

iii. Tape one edge of PDLC to laminator carrier

103

iv. Tape one edge of cut retroreflective sheet, facing down, align with same

taped edge of PDLC - check length alignment

v. Lift retroreflector up, apply flexible optical adhesive (Dymax OP-30) in "T"

shape with the top of the "T" at the taped edges

vi. Let retroreflector drop, fold over carrier, and run through laminator at

20psi, any speed, <50 C (or no heat).

vii. Check adhesive coverage - edge spot apply or lift retroreflector up and

apply more - run through laminator again if necessary

viii. Run through UV curing conveyor system (Dymax UVCS-F-1-230) at 20

ft/min, retoreflector down

ix. Allow to cool - 30 seconds

x. Run through UV curing conveyor system (Dymax UVCS-F-1-230) at 20

ft/min a second time, retroreflector down c. Graphics

i. Using graphics editing software - edit desired graphic - adjust

"Brightness," "Lightness," "Color Balance," print size, etc. Ensure graphic

is mirrored.

ii. Print onto a transparency with a color laser jet printer - Samsung CLP-

550N was used here.

iii. Follow steps v-vii. from section b. "Attach Retroreflector" attaching printed

side of transparency to top side of PDLC.

104

iv. Allow to sit for 1-2 minutes

v. Run through UV curing conveyor system (Dymax UVCS-F-1-230) at 20

ft/min, graphics side up

vi. Allow to cool - 30 seconds

vii. Run through UV curing conveyor system (Dymax UVCS-F-1-230) at 20

ft/min a second time, graphics side up

viii. Pull off transparency - Graphic is transferred to surface.

ix. If a final tag shape cut is necessary, use the laser cutting system (ULS

3.50 - Vector cut Speed = 100, Power = 80, PPI = 500) d. Attach Electrodes

i. On top side of one end of the tag, cut one ITO coated PET layer using

laser cutting system (ULS 3.50 - Vector cut Speed = 100, Power = 20,

PPI = 500) - expose ~5mm of the other ITO coated PET substrate

ii. Flip tag over and do the same on the opposite end - if retroreflector and

ITO coated PET need to be cut: Vector cut Speed = 100, Power = 35, PPI

= 500) - expose ~5mm of the other ITO coated PET substrate

iii. If necessary - scrape off any residual PDLC to get a clean electrode

surface. Cotton swap soaked in acetone works well.

iv. Apply UV epoxy (Dymax OP-30 used here) to the newly cut edges - do

not cover much of the exposed electrode

v. Apply colloidal silver paste to both exposed electrodes

105

vi. Wrap both electrode ends with appropriately sized copper tape vii. Solder wires to copper tape - 315C viii. For strain relief, route/snake wires up back of retroreflector until they join

in the center, bring down to bottom. Strategically hold wires in place on

back of tag with epoxy (Loctite 352, Dymax OP-30, etc)

ix. For water proofing, dip electrode ends in Dymax OP-30 or similar to

entirely coat exposed copper tape, dip side edges to provide a lip, cure

with UV curing light wand

x. If desired, place adhesive backed velcro to retroreflector side of tag.

106