OPTICAL PROPERTIES of the PIGMENTS from CEPHALOPOD CHROMATOPHORES: SOLUTION and PARTICLE DETERMINATION Sean Ryan Dinneen University of New Hampshire, Durham

Home , Camera (cephalopod), Vision in toads

University of New Hampshire University of New Hampshire Scholars' Repository

Doctoral Dissertations Student Scholarship

Fall 2018 OPTICAL PROPERTIES OF THE PIGMENTS FROM CEPHALOPOD CHROMATOPHORES: SOLUTION AND PARTICLE DETERMINATION Sean Ryan Dinneen University of New Hampshire, Durham

Follow this and additional works at: https://scholars.unh.edu/dissertation

Recommended Citation Dinneen, Sean Ryan, "OPTICAL PROPERTIES OF THE PIGMENTS FROM CEPHALOPOD CHROMATOPHORES: SOLUTION AND PARTICLE DETERMINATION" (2018). Doctoral Dissertations. 2423. https://scholars.unh.edu/dissertation/2423

This Dissertation is brought to you for free and open access by the Student Scholarship at University of New Hampshire Scholars' Repository. It has been accepted for inclusion in Doctoral Dissertations by an authorized administrator of University of New Hampshire Scholars' Repository. For more information, please contact [email protected].

OPTICAL PROPERTIES OF THE PIGMENTS FROM CEPHALOPOD CHROMATOPHORES: SOLUTION AND PARTICLE DETERMINATION

SEAN R. DINNEEN B.S., Eastern Nazarene College, 2014

DISSERTATION

Submitted to the University of New Hampshire in Partial Fulfillment of the Requirements for the Degree of

Doctor of Philosophy in Chemistry

September, 2018

This dissertation has been examined and approved in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Chemistry by:

Dr. Margaret E. Greenslade, Associate Professor of Chemistry

Dr. Leila F. Deravi, Assistant Professor of Chemistry and Materials Science, Northeastern University

Dr. Sterling A. Tomellini, Professor of Chemistry

Dr. Arthur Greenberg, Professor of Chemistry

Dr. Richard M. Osgood III, Research Physicist, US Army Natick Soldier RD&E Center

Defended On May 25th, 2018

Original approval signatures are on file with the University of New Hampshire Graduate School

Dinneen | ii

This dissertation is dedicated to my wife, Melanie Catherine Dinneen.

In a world where people often camouflage themselves, my true colors are brighter when we’re together

Dinneen | iii

ACKNOWLEDGMENTS

I would like to thank my fellow members of the Greenslade group and the Deravi group for their support. I specifically acknowledge Tyler Galpin for assistance with the RI and CRD experiments, Jillian Morang for assistance with the CRD experiments, Brent Lawson for assistance with the collection of aerosols, and Christopher DiBona for moral support. I acknowledge the University Instrumentation Center for their assistance with SEM sample preparation, and J. Michel Flores for sharing the refractive index retrieval program. Thanks to Cindi Rohwer and Laura Bicknell for logistical support and scheduling. Thanks to the UNH facilities and custodial staff for keeping our buildings and equipment running smoothly, allowing us to focus on our science.

I would also like to acknowledge my graduate committee members Sterling Tomellini, Arthur Greenberg, Richard Osgood III, and former member Marc Boudreau.

This work was supported by the UNH Department of Chemistry, Northeastern University Department of Chemistry and Chemical Biology, and the Army Research Office (W911NF-16-1- 0455).

Finally, I would like to acknowledge Dr. Margaret Greenslade and Dr. Leila Deravi for being the best co-advisers a graduate student could ask for. Thank you for your support, for giving me unique opportunities and experiences, and for continuing to deal with me even when I don’t respond to emails consistently.

Dinneen | iv

TABLE OF CONTENTS

COMMITTEE PAGE……………………………………………………………… . ii

DEDICATION………………………………………………………………………. iii

ACKNOWLEDGEMENTS…………………………………………………………. iv

LIST OF TABLES……………………………...…………………………………… vii

LIST OF FIGURES……………………………………………..…………………… viii

ABSTRACT…………………………………………………………………………. ix

CHAPTER PAGE

I. INTRODUCTION TO BIOINSPIRED COLORATION SYSTEMS……… 1

Structural coloration…………………………………………………… 2

Pigmented coloration………………………………………………….. 10

Pigmentary and structural coloration in cephalopods…..……………… 14

Bioinspired technologies……..………………………………………… 16

Dissertation aims……………………………………………………….. 23

II. EXPERIMENTAL CONSIDERATIONS………………………………….. 24

Squid dissection and tissue digestion…………………………………… 24

Pigment variability……………………………………………………… 27

Refractometry…………………………………………………………… 29

Cavity ring-down spectroscopy…………………………………………. 30

Experimental vs. calculated extinction values………………………….. 32

III. COLOR RICHNESS IN CEPHALOPOD CHROMATOPHORES ORIGINATING FROM HIGH REFRACTIVE INDEX BIOMOLECULES…………………………………………………………. 33

Dinneen | v

Abstract………………………………………………………………… 33

Introduction…………………………………………………………….. 33

Discussion……………………………………………………………… 35

Experimental…………………………………………………………… 45

IV. OPTICAL EXTINCTION OF SIZE-CONTROLLED AEROSOLS GENERATED FROM SQUID CHROMATOPHORE PIGMENTS………………………………………………………………… 48

Abstract………………………………………………………………… 48

Introduction……………………………………………………………. 48

Results and Discussion………………………………………………… 49

V. AN ITERATIVE CORRECTION APPROACH USED TO RETRIEVE THE REFRACTIVE INDEX OF SQUID PIGMENT AEROSOLS…………… 59

Abstract……………………………………………………………….. 59

Introduction…………………………………………………………… 60

Methods………………………………………………………………. 62

Results………………………………………………………………… 65

Discussion……………………………………………………………. 69

VI. CONCLUSIONS…………………………………………………………. 72

LIST OF REFERENCES…………………………………………………………….. 75

Dinneen | vi

LIST OF TABLES

Table 1.1 Natural pigment characteristics and properties…………………………. 13

Table 3.1 Extrapolated refractive indices of four solvent systems………………… 36

Table 4.1 Size selected and experimentally measured particle diameters with calculated extinction cross-sections……………………………….. 56

Table 4.2 Comparing measured extinction to Mie calculated extinction………….. 57

Table 5.1 Maxwell-Garnett input and output with error…………………………… 68

Dinneen | vii

LIST OF FIGURES

Figure 1.1 Thin-film interference diagrams………………………………………… 2 Figure 1.2 Parotia lawesii iridescent feathers and melanosome TEM images…….. 4 Figure 1.3 Sapphirina metallina guanine crystals and reflectivity spectra………… 6 Figure 1.4 Hibiscus trionum petal structural ridges………………………………… 7 Figure 1.5 Beetle Pachyrhynchus argus photonic crystal scales…………………… 9 Figure 1.6 Common ommochromes and ommochrome precursors………………… 11 Figure 1.7 Pentanin and magnesium-coordinated chlorin of chlorophyll…...……... 12 Figure 1.8 Doryteuthis pealeii optical organ hierarchy…………………………….. 14 Figure 1.9 Doryteuthis pealeii pigments, granules, and absorption……...…...... …. 15 Figure 1.10 Multilayer synthetic melanin nanoparticle films………………………... 17 Figure 1.11 Charidotella egregia humidity-based metallic gold to red transition…… 18 Figure 1.12 Inkjet printed mesoporous silica nanoparticle photonic crystals……….. 20 Figure 1.13 Infrared-reflecting reflectin protein-based films………………………… 21 Figure 1.14 Biomimetic squid granule thin films……………………………………. 22

Figure 2.1 Extracted pigment absorption and TLC separated band absorption……. 26 Figure 2.2 Ommochrome variability and maturity in dragonflies species…………. 28 Figure 2.3 CRD (τ vs. time) and CPC (number concentration vs. time) raw data…. 31

Figure 3.1 Actuation of squid dorsal mantle chromatophores……………………… 35 Figure 3.2 Xanthommatin and decarboxylated xanthommatin…………………….. 36 Figure 3.3 Pigment refractive index as a function of mass fraction of sample…….. 37 Figure 3.4 Pigment imaginary refractive index as a function of wavelength………. 38 Figure 3.5 Scattering and absorption cross-sections of squid pigment…………….. 39 Figure 3.6 FDTD calculated scattering and absorption cross-sections…………….. 40 Figure 3.7 Optical cross sections calculated at different wavelengths for k……….. 41 Figure 3.8 Optical cross sections calculated with ±10% on n and k……………….. 42 Figure 3.9 Optical cross sections calculated over an extended size range…………. 43 Figure 3.10 Compared optical cross sections of squid pigment and melanin……….. 44

Figure 4.1 Atomization scheme and experimental setup for CRD analysis………... 50 Figure 4.2 Particle size distributions for squid pigment aerosols…………………... 51 Figure 4.3 Aerosol collection setup and SEM size analysis………………………... 53 Figure 4.4 Optical extinction of squid pigment versus number concentration……... 55 Figure 4.5 Extinction cross section versus particle diameter for squid pigment…… 57

Figure 5.1 Flow chart describing the iterative doublet correction process…………. 64 Figure 5.2 Mie corrected and uncorrected Qext versus size parameter……………… 66 Figure 5.3 Mie corrected data with Mie traces using n = 1.92 and 1.5606…………. 67 Figure 5.4 Mie corrected data using n = 1.5606 and the percent difference of Mie traces using n = 1.5606 and n = 1.59……………………………. 68

Dinneen | viii

ABSTRACT

OPTICAL PROPERTIES OF THE PIGMENTS FROM CEPHALOPOD CHROMATOPHORES: SOLUTION AND PARTICLE DETERMINATION

Sean R. Dinneen

University of New Hampshire, September, 2018

Cephalopods are arguably one of the most photonically sophisticated marine animals, as they can rapidly adapt their dermal color and texture to their surroundings using both structural and pigmentary coloration. Their chromatophore organs facilitate this process, but the molecular mechanism potentiating color change is not well understood. The central hypothesis of this work is that the pigments, which are localized within nanostructured granules in the chromatophore, enhance the scattering of light within the dermal tissue. To test this, phenoxazone-based pigments were first extracted from the chromatophore and their complex refractive index (RI) from experimentally determined real and approximated imaginary portions of the RI were extrapolated. Mie theory was used to calculate the absorbance and scattering cross sections (cm2/particle) across a broad diameter range at λ = 589 nm, where both portions of RI were determined. The results indicated that the pigments were more likely to scatter attenuated light than absorb it and that these characteristics may contribute to the color richness of cephalopods. Next, pigment-containing nanostructures were fabricated, and their optical extinction and refractive index were experimentally determined. The pigment solution extracted from native squid chromatophores was nebulized, forming aerosols. Their size-dependent optical extinction was measured using cavity ring-down spectroscopy at λ = 532 nm, where a relationship between extinction cross section and particle concentration and size was elucidated. Finally, a detailed analysis of the optical features of the pigment aerosols was developed and optimized. This analysis incorporated corrections that accounted for particle charging and solvent effects that accumulated during the aerosolizing process using an innovative iterative approach tied to retrieved refractive index values. RI retrievals were obtained via the best fit between the corrected, experimentally observed extinction efficiencies compared to those calculated by Mie theory for a specific RI at selected sizes. In addition to these retrievals, the impact of solvent on the particles’ optical properties was also examined via the Maxwell–Garnett mixing rule. Ultimately, an aerosol RI was obtained with a real portion (n) of 1.66 (±0.05) representing a lower limit and an imaginary portion (k) of 0.13 (±0.08)i representing an upper limit for the generated aerosols. Together, the analytical approaches used to retrieve RI values of the squid pigments presented here are advancements to the field of bio-optics that could one day inform the design of new bio-inspired optical materials.

Dinneen | ix

Dinneen | x

CHAPTER 1:

Introduction to Bioinspired Coloration Systems

Nature has an uncanny ability to create complex coloration systems out of a limited range of natural materials. These natural materials, often structures or pigments, are fabricated within the organism itself. Their ability to perform functions such as reflection,1-3 waveguiding,4-6 and interference7-10 with self-assembled structures proves to be challenging to replicate in the lab. What nature can do effortlessly and efficiently has taken years of work to even come close to. For example, light harvesting during photosynthesis has been shown to be over 90% efficient while fabricated photovoltaics can only reach efficiencies of up to 46% under certain conditions.11-12 The light interaction processes that nature can perform have been adapted over millions of years to suit specific needs, and are therefore the best inspiration in designing synthetic versions.

Coloration serves many purposes in the animal kingdom, including sexual attraction, signaling, and even camouflage.13-16 The purpose of the colors, whether it is to be overt or blend in, is to produce either a form of non-verbal communication or provide a defense mechanism. In order to produce these colors, the animals must be genetically disposed to interact with light in some way. In nature, colors are produced using two very distinctive mechanisms: structural coloration or pigmented coloration. By understanding the design rules regulating these processes, we can develop more sophisticated technologies to replicate their abilities.

Dinneen | 1

Structural Coloration: Structural coloration involves the use of ordered micro or nano structures, typically in the size order of the wavelengths of light they interact with, to generate constructive and destructive interference which allows for the reflection of different wavelengths in the visible spectrum.13, 16-17 This reflection is often facilitated by the material’s refractive index, which is the ratio of the speed of light in a vacuum to the speed of light through the material.18 It can also describe the angle at which light will bend, or refract, at the interface of two materials with different refractive indices.18 An increase in refractive index will increase the angle with which the incident light is refracted. An example of structural coloration is the thin-film microstructures found on the wings of butterflies which generate iridescent colors at certain angles of light.17 The term structural color originates in the early 20th century when Lord Rayleigh hypothesized that the shifting colors in iridescence was due to structure and not by dyes or pigments.19 Two of the main mechanisms by which this occurs is interference and diffraction.

Figure 1.1. a. Thin-film interference diagram of a material with thickness d and a refractive index of nB. b,c. Reflectivity of a thin chitin (n = 1.5) film with surrounding medium of b) n = 1.0 and c) n = 2.0. d. Thin-film interference diagram of a multilayer structure with repeating layers A and B.20 Reprinted with permission from Ref. 20. Copyright 2017 American Chemical Society.

Dinneen | 2

Interference is the result of the reflection of light from the interfaces of high and low refractive index materials. The thickness of the material, d, determines how far the wave will travel before being reflected at a lower boundary, causing certain wavelengths to be in (constructive) or out (destructive) of phase with light that was reflected at the upper boundary (Figure 1.1a).20 The conditions by which constructive interference occurs can be described by Equation 1.1,21

푚휆 = 2푛푑 cos 휃푟 (1.1)

where m is an integer or half an integer, λ is the wavelength of light, n is the refractive index of the material, d is the thickness of the layer, and ϴr is the angle of refraction. This equation also suggests that the selectively reflected wavelengths of light will change based on the observed viewing angle. As the incident angle of light increases, the peak reflected wavelength will decrease.20 The refractive index of the mediums will also change the observed reflected wavelengths (Figure 1.1bc).

Interference is enhanced even further when the high and low refractive index materials form a multilayer, alternating structure (Figure 1d). The reflectivity of the reflected wavelength will increase as the number of bilayers increase. For a multilayer system with mediums A and B, constructive interference will occur in the conditions described by Equation 1.2,21

푚휆 = 2(푛퐴푑퐴 cos 휃푟 + 푛퐵푑퐵 cos 휃푟) (1.2)

where na and nb are the refractive indices of mediums A and B respectively, and da and db are the layer thicknesses of mediums A and B respectively. For consistent color across a bulk system, the

Dinneen | 3 layers have to be regularly spaced or it may result in a broadly reflective effect.20-21 Imperfections in the thicknesses of the layers across large spatial areas, or other irregularities in the structure such as divots or holes, will no longer produce uniform coloration. Rather, since localized areas will produce different selectively reflected wavelengths, the bulk color will appear broadly reflective.

(a) (b)

Figure 1.2. a. Broadly reflective occipital feathers at the nape of the bird (scale 1 cm). b. Breast feathers showing a wide range of iridescent colors based on viewing angle (scale 1 cm). c. Transmission electron micrograph of the cross section of an occipital feather barbule (scale 2 µm). d. Transmission electron micrograph of the cross section of a breast feather barbule (scale 5 µm).14 Reprinted with permission from Ref. 14.

An example of interference in nature would be the feather barbules of the bird of paradise

Parotia lawesii, or the common bronzewing Phaps chalcoptera.13, 15 Their occipital feather barbules consist of layers of keratin (n = 1.55, ~400 nm) and melanin nanoparticles called melanosomes (n = 1.72, ~250 nm) (Figure 1.2ac).13-14 Their breast feather barbules also have these layers, but with an overall boomerang shaped microstructure (Figure 1.2bd).14 In addition, both the melanosomes and the keratin layers are smaller (~120 nm and ~230 nm, respectively).14 The alternating layers of differing refractive indices produce thin-film interference and the observed

Dinneen | 4 iridescent colors (oranges, greens, and blues) are dependent on the incident viewing angle. This plays a role in the species’ mating ritual, which involves the male of the species dancing for a female’s attention. The P. lawesii will repeatedly twist its body and therefore change the observed colors of its breast feathers to attract females.14

Another example of structural coloration is in the sapphirinid copepods, where color is used for communication, mate recognition, and possibly for defense by camouflaging. Inside the dorsal cuticle of this crustacean is a multilayer reflector architecture consisting of hexagonal crystals of guanine and cytoplasm (Figure 1.3ac).16 The guanine crystals (~1 µm wide, ~70 nm thick) were found to have one of the highest refractive indices of a biological system at n = 1.83, where the cytoplasm (50-200 nm thick) has a refractive index similar to water at n = 1.33.16 The large difference between the refractive index of the guanine and cytoplasm allows for constructive interference of certain wavelengths of light (often iridescent greens, blues, and purples). Again, much like the butterfly Papilio blumei or the bird of paradise described above, the observed reflected wavelengths are dependent upon the incident viewing angle.13, 16-17 In the case of the copepod, higher tilt angles render the specimen completely invisible, allowing them to escape from predators unseen (Figure 1.3b).16

Dinneen | 5

Figure 1.3. a. Light microscope image of the hexagonal crystals in Sapphirina metallina cuticle. b. Simulated reflectivity spectra at different viewing angles with corresponding images of Copilia mirabilis. c. Cryo-scanning electron microscope (SEM) image of the tightly packed hexagonal guanine crystals.16 Reprinted with permission from Ref. 16. Copyright 2015 American Chemical Society.

Another mechanism of structural coloration is based on diffraction. Diffraction occurs when light reaches an object or slit on the same size order of the wavelength of light and bends around it.20 This causes a diffraction pattern, which is when certain wavelengths of light will interfere to produce patterned areas of alternating light and dark spots, usually in the same shape as the aperture or object. To increase the brilliance of the effect, the micro or nano structures must have a high level of periodicity, meaning they need to have consistent size and spacing, or it will result in a more diffuse reflection. Diffraction can be described using the conditions in Equation

1.3,21

Dinneen | 6

푚휆 = 푑(sin 휃퐷 − sin 휃퐼) (1.3)

where m is the diffraction order, λ is the wavelength of light, d is the periodicity of the grating or structures, ϴD is the angle of diffracted light, and ϴI is the angle of incident light.

Figure 1.4. a. Photo of the Hibiscus trionum flower. b. SEM of a H. trionum petal, where the top half of the image spans the smoother ridges of the white portion of the petal and the bottom half spans the striated ridges of the pigmented portion of the petal (scale 50 µm). c. SEM image of the ridges of Tulipa kolpakowskiana petal, resembling a diffraction grating (inset shows its optical appearance in transmission, scale 1 µm).22 From Ref. 22. Reprinted with permission from AAAS.

One example of diffraction in nature is the structural ridges found on the petals of the

Hibiscus trionum flower. The petals of the H. trionum flower have a red pigmented base that transitions into a diffuse white appearance (Figure 1.4a).22 However, the pigmented patch also displays iridescent blues, greens, and yellows depending on the viewing angle.22 This is due to the longitudinal ridges found on the surface (Figure 1.4b). The ridges act like diffraction gratings to incident light, and together with the pigment beneath it, produce iridescence in the visible region.22 Though, there are some species, like the Tulipa kolpakowskiana where iridescence can only be observed when the petal surface is separated from the pigment (Figure 1.4c). To further prove that the structures were responsible for the iridescence, researchers at the University of

Dinneen | 7

Cambridge made epoxy castings of the petals and achieved diffraction patterns in its optical appearance in transmission.22

Diffraction can also be caused by naturally occurring photonic crystals. Photonic crystals are regularly arranged particles that form a crystal structure which will reflect light in specular and diffuse directions.21, 23-24 The crystal periodicity, or the spacings of the particles, is typically within the same size scales (or in some cases half) of the wavelengths that are diffracted, like with most structural coloration.21 One such occurrence of photonic crystals in nature is the metallic colored scales in the shells of beetle Pachyrhynchus argus (Figure 1.5a). The scales contain tightly packed chitinous spheres, similar to the nanostructure of an opal (Figure 1.5bc).2, 25 In this case specifically, the photonic crystals are considered three-dimensional, as the periodicity can be defined in all three dimensions by the packing of the spheres. The selectively reflected wavelength of the photonic crystals in this system can be described by Equation 1.4,25

2 2 휆푚푎푥 = 2푑 ∗ 0.816√푛 − 푠푖푛 휃 (1.4)

where λmax is the wavelength of maximum reflectance, d is the diameter of the spheres, n is the average refractive index of the system, and ϴ is the angle of incidence. In Figure 1.5d, the reflectance spectrum shows a max reflectance wavelength in the green of the visible spectrum.

This matches the calculated max reflectance wavelength of 573 nm when the spheres are said to be 250 nm in diameter and the refractive index of the chitin and water are 1.56 and 1.33, respectively.

Dinneen | 8

Figure 1.5. a. Photo of the beetle Pachyrhynchus argus, showing the metallic coloration on its dorsal side. b. SEM image of the overlapping scales responsible for the coloration (scale 100 µm). c. SEM image of a section of a single scale, showing its opal-like structure (scale 1 µm). d. Reflectance spectra of the scale surface with incident white 20˚ to the normal of the surface and measured at 20˚ to the other side of the normal.25 Reprinted with permission from Springer Nature. Ref. 25, Copyright 2003.

There are also instances where both interference and diffraction will occur in the same animal. For example, the wings on Morpho rhetenor and Morpho didius butterflies are covered with overlapping, microstructured scales.9 On top of these scales are even smaller features that form branching ridges resulting in a multilayer structure (500 nm to 5 µm layer spacing).9 The multilayer structure is the basis for the iridescent colors found in these butterflies. Furthermore, the overlapping nature of the scales produces strong diffraction, even when the iridescent scales are overlapped by transparent scales.9 It is believed that this effect increases the range of angles over which light can be reflected.9

Dinneen | 9

Whether the structures cause interference, diffraction, or both, structural coloration can be advantageous for animals. They would not need to synthesize pigments biologically, which often requires a special diet. This is the case for pink flamingos, whose color is a result of eating shrimp rich in carotenoids.20 Rather, species who use structural coloration typically use materials such as chitin or protein platelets. These often optically-transparent materials can create vivid, reflective colors. However, the use pigments may have an advantage as they can provide additional utility at a molecular level.

Pigmented coloration: While structural coloration is dependent on the size, spacing, and refractive index of materials, pigment-based coloration relies on chemical contribution. Unlike structural coloration, pigments do not typically change the wavelengths of light they reflect unless they undergo a chemical reaction, like oxidation or reduction. This means that if an animal’s coloration is used for a purpose such as camouflage, the species must be in complimentary natural surroundings in order to become covert.

In fact, a majority of the broad range of visible color we see in nature and in commercially available materials is due to pigmented coloration. Organic pigments absorb light due to their conjugation, or the overlapping of p orbitals resulting in delocalized pi electrons.26 As conjugation increases, the energy needed to excite an electron decreases, corresponding to longer wavelengths of light. This principle applies to other ends of the electromagnetic spectrum as well, such as ultraviolet (UV) and infrared (IR) wavelengths. The selective absorption of light can lead to color, heat dissipation, or more complex functions including light harvesting, showing that pigments can serve other functions in nature.20

Dinneen | 10

In humans, the pigment melanin provides protection from UV irradiation while also offering a multitude of services in other natural systems, such as redox activity and conductivity.28-

30 Melanin is also found in cephalopod cuttlefish Sepia officinalis ink, which is used to distract their predators if they feel threatened.31 The conjugated nature of the melanin in the ink results in a dark, broadly absorptive color that obstructs the vision of sharks or dolphins who try to eat them.

Crustaceans and insects are also well known for their use of pigments. While humans have melanin, arthropods such as the crab spider, brine shrimp, and the buckeye butterfly all contain conjugated absorbing molecules called ommochromes.32-34 Ommochromes are a class of polycyclic aromatic pigments (Figure 1.6) that are derived from tryptophan in a known pathway and is found in animals across many species.27, 35-37 One of the defining characteristics of ommochromes is their absorption peaks centered around ~360 and ~480 nm.37-39

Pigment absorption peaks can also shift based on changes in solution pH. As acidity could affect the conjugation in the chromophore of pigment molecules, the wavelengths of light that can

Dinneen | 11 be absorbed will change. Petanin (Figure 1.7a), a naturally occurring anthocyanin found in blue

40 potatoes, can shift its λmax by up to 50 nm within a pH range of 1-9. The color intensity at petanin’s λmax also changes with pH. As pH increases, the molar absorptivity decreases until pH =

5, then will increase as it reaches its maximum molar absorptivity at pH = 8.40 These pH properties have interesting applications in areas such as pH indicators and food colorants in products with a wide range of acidity.

(a) (b)

Pigments can also be found as metal-coordinated compounds in nature. For metal containing compounds (most commonly transition metals), the observed colors come from excitations in d-orbital energy levels.20 Organisms that use metals in pigments can take advantage of bioavailable metals to generate strong absorption as well as use them for natural functionality in their respective biochemistries. Visible color is not the only use of absorption by pigments in a natural system. For example, chlorophyll, a magnesium-coordinating pigment found in algae and the leaves of plants, plays a role in harvesting the sun’s energy to turn into chemical energy (Figure

1.7b).41 In fact, metal-coordinated pigments have a slight utility advantage by being able to bind

Dinneen | 12 oxygen (such as the heme in blood), stabilize protein, or even perform respiration in addition to being responsible for color.42

Table 1.1. Table of natural pigments classified by origin, appearance, and chemical properties.

Natural Molecular Absorbance Origin Color Pigments Weight (g mol-1) peaks (nm) Insects, Xanthommatin43 Red/Yellow 424.07 230, 450 Cephalopods Blue potatoes, Petanin40 Blue/Purple 933.84 ~300, 577 berries Chlorophyll a44 Plant leaves, algae Green 893.51 613, 659 Carrots, pumpkin, β-Carotene45 Orange 536.89 440 cantaloupe Cephalopod eyes, Ommin A43 Purple 650.16 520 silkworm Indigo46 Flowering plants Blue 262.27 437

In nature, pigments are also often found in self-assembled nanoparticle forms. This was shown previously in Figure 1.2, where melanin nanoparticles populate the interior layers of feather barbules in the birds of paradise.14 These melanin nanoparticles, called melanosomes, are responsible for the storage, transport, and even synthesis of melanin within the animal and are found in many animal species.47 For example, the microbe Streptomyces glaucescens produces extracellular melanin nanoparticles that contain anticancer and antioxidative properties.48

Ommochromes have also been reported to be found in natural nanoparticles. Bombyx mori, a species of silkworm, was found to have dark red pigment granules in its central nervous system using light microscopy.49 The pigment, after UV-vis, IR, and TLC analysis, was determined to be mostly ommin, and that an ommochrome-binding protein was also found in the pigment granule.49

At the time of the study, the purpose of these pigment granules found in the central nervous system of silkworms was unknown. However, it has been reported that 3-hydroxykynurenine, a precursor

Dinneen | 13 to ommochromes, is neurotoxic to mammalian brains and therefore these granules may be the product of a natural pathway that removes these harmful compounds.50-51

Structural and pigmented coloration each have their own advantages. While structural color can provide a broad range of reflected colors, pigments can provide a deep and consistent color across a large area of the species. Both aspects are important in dynamic camouflage processes and are therefore more efficient when paired. There are very few examples in nature where an organism can enable both structural and pigmented coloration effectively, and perhaps the most popular of which are cephalopods.

Figure 1.8. Mechanisms of coloration span multiple spatial scales. Hierarchical optical organs in the squid skin allow for rapid color change across the dermal tissue. The chromatophore actuates to control the area with which the pigments can facilitate absorption and scattering.

Pigmentary and structural coloration in cephalopods: Cephalopods, such as squid, octopus, and cuttlefish, are among the most widely cited inspirations for new camouflaging and

Dinneen | 14 adaptive coloration technologies.52-55 Their unparalleled ability to match their surroundings is highly sought after and is the reason we choose to study them. Doryteuthis pealeii, the longfin inshore squid, can camouflage to their environment within less than one second. This fast response time is initiated by changes in their highly organized dermal tissue which contains a layer of pigmented chromatophore organs above reflective iridophore cells (Figure 1.8).56 The iridophores, the lowermost portion, are platelets containing assembled reflectin protein which act as Bragg stack reflectors.57 By changing the spacing between the protein platelets, the squid are able to change which wavelengths of light are selectively reflected. This structure accounts for the iridescent colors (blues and greens) observed in the animals and has been widely studied.58-61

Figure 1.9. a. Doryteuthis pealeii. b. Microscopic image of expanded and punctate chromatophores (scale 1 mm). c. Squid pigment granules before and after pigment extraction with acidic methanol. d. Scanning electron micrographs of i) pigment granules and ii) post-extraction granules (scale bars at 300 nm). e. Bar graph of average pigment granule diameter versus treatment process (n = 75, error bars are standard error of the mean). f. UV-vis absorbance of pigment granule suspensions in water before (black) and after (blue) extraction.62 Reprinted with permission from Ref. 62. Copyright 2016 American Chemical Society.

Dinneen | 15

The chromatophore organs are commonly described as color filters to the light that is reflected off the iridophores.57 They are anchored by radial muscle fibers which contract and relax to expand and collapse the pigment sac (Figure 1.8, 1.9b), contributing to the dynamic optical properties of the skin in response to visual cues.63 Nanostructured pigment granules populate the interior of the chromatophore organ (Figure 1.8). These granules are comprised of protein and pigment molecules that are tethered together within the cytoelastic sac.64

In our previous work, we have outlined a method by which to extract and isolate the pigments from the granules using a series of buffer washes (Figure 1.9c).65-66 Using mass spectrometry, we were able to identify the pigments as xanthommatin and decarboxylated xanthommatin.65 When extracted, the pigment granules experience a 71% decrease in diameter confirmed by scanning electron microscopy (Figure 1.9e).65 This drastic change in volume infers that the pigments comprise a significant portion of the composition of the granule. In addition, it was also confirmed that the pigment was responsible for the color of the granules, as pigment- extracted granules no longer absorbed in the visible when measured by ultraviolet-visible (UV- vis) spectrophotometry (Figure 1.9f).65

Bioinspired technologies: There have been many advances in recent years trying to develop technologies to replicate nature-inspired mechanisms for self-assembly and/or reliable color change.67-70 Applications for these systems range from solar efficiency (antireflection with increased transmission),71-72 to low powered reflective displays.73-75 These techniques draw inspiration from nature’s limited materials to produce brilliant colors.

Xiao and coworkers recently developed optical materials inspired by the layers of melanosomes in the feather barbules of the birds of paradise (Figure 1.10a). Seeking to replicate its ability to self-assemble nanoparticles to form the structural coloration, the group used an

Dinneen | 16 evaporative method to produce multilayer films of synthetic melanin nanoparticles (SMNPs)

(Figure 1.10b).76 Their particles, made of polydopamine, broadly absorb in the visible and have a high refractive index of 1.74.76 The evaporative method works by capillary action, where the particles are pulled by the surface tension of the solvent onto the substrate as the solvent evaporates. The evaporation rates and the concentration of the particles in solution will affect the periodicity and thickness of the films.77-78

Figure 1.10. a. Structural coloration in nature with iridescent bird feathers. b. Synthetic melanin nanoparticles (SMNP). c. SMNP films where i+ii are different regions of the same film and iii+iv are different regions of another film using starting SMNP concentrations of 0.6 mg/mL and 1.0 mg/mL, respectively. d. SEM image of the SMNP films showing the multilayer structure (top) and film cracking behavior (bottom).76 Reprinted with permission by Ref. 76. Copyright 2015 American Chemical Society.

It was found that the color of the films (Figure 1.10c) were based on the thickness of the

SMNP layer, and that the reflectance of the films were well fit to a thin-film interference model.76

Each bulk film produced multiple colors on a single substrate, which was explained by the differences in evaporation rates and the resulting change in SMNP concentration as the process proceeded.76 This makes it hard to generate a consistent color across a large surface area. These films also have a tendency to crack from shrinkage during the drying process (Figure 1.10d), but

Dinneen | 17 the cracks do not have any effect on the coloration.76 While this system excels in the area of self- assembly, it cannot produce dynamic color change.

Figure 1.11. Photos of the beetle Charidotella egregia during its transition from a metallic gold to a diffuse red. This transition signifies evacuation of liquid from the porous layers of its shell.79 Reprinted with permission from Ref. 79. Copyright 2007 by the American Physical Society.

Alternatively, Bai et al. developed a method for dynamically tuning the color reflected by photonic crystals through the adsorption of ethanol vapors.80 This work was inspired by beetles whose shell can change color based on whether liquid filled the porous layers of the elytron (Figure

1.11).79 In its humid state, the beetle displays a metallic gold color due to interference in its multilayer patches.79 When this liquid dissipates, the light mostly scatters and the multilayers become translucent, allowing for the pigmented layer beneath to be seen.79 By using mesoporous silica nanoparticles (MSNs) in their inkjet printed photonic crystals, which have strong vapor adsorption capabilities, ethanol can be adsorbed in their porous structure which will change the effective refractive index of the particles.80

By using Bragg’s law under normal incidence, Equation 1.5 can be used to calculate the peak wavelength (λ) of the mesoporous colloidal photonic crystals.80

Dinneen | 18

2 2 2 2 휆 = 1.633푑√푓푣표푖푑 푛 푣표푖푑 + 푓푐표푟푒푛 푆푖푂2 + 푓푚푝표푟푒푛 푚푝표푟푒 + 푓푚푆푖푂2 푛 푆푖푂2 (1.5)

The diameter of the mesoporous silica nanoparticles is represented by d; fvoid, fcore, fmpore, and fmSiO2 are the volume ratio of empty space, the SiO2 core of the particles, the mesopores of the shell, and the SiO2 in the mesopores shell, respectively. The values of nvoid, nSiO2, and nmpore are the refractive indices of air, SiO2, and the mesopores, respectively. When these particles adsorb the ethanol (n =

1.36) in their pores while replacing air (n = 1), they are increasing the effective refractive index of the mesopores shell.80 This change in refractive index is responsible for the color shifts observed by the photonic crystals (Figure 1.12). By designing specific core/shell diameters of the mesoporous nanoparticles, the peak wavelength can be predicted using the Bragg’s law equation so that specific colors can be generated when placed under ethanol atmospheres. The range of colors observed in this study spanned almost the entire visible spectrum.

Solid silica nanoparticles (SSNs) can also be used to create portions of the printed design where ethanol vapors would not change the observed color (Figure 1.12c).80 When measuring the change in peak reflection wavelength from N2 atmosphere to saturated ethanol atmosphere, the mesoporous silica nanoparticles shifted by up to ~50 nm while solid silica nanoparticles only shifted by 7 nm (Figure 1.12d).80 Because there is less surface area for these vapors to adsorb to, as expected, it is observed that the reflectance shift would not be significant for the solid nanoparticles. Overall, this system can produce self-assembled structures with dynamic color change, but the reversibility of the color change is dependent on the evaporation of the ethanol from the mesopores.

Dinneen | 19

Figure 1.12. a. TEM image of MSNs (scale 200 nm). b. TEM image of SSNs (scale 200 nm). c. (i) Printed colloidal photonic crystal pattern in N2 atmosphere, and (ii) same crystal under saturated ethanol atmosphere (leaves and fruit are mesoporous silica nanoparticles while the stalk is solid silica nanoparticles). d. Reflectivity spectra showing the peak wavelength shifts of the stalk, leaves, and fruit 80 patterns under N2 atmosphere versus an ethanol atmosphere. Reprinted with permission from Ref. 80. Copyright 2014 American Chemical Society.

Finally, actuation in the chromatophore organs of cephalopods served as inspiration for another mechanism of color change, as they can expand and contract to control the coloration of the animal. Rather than focus solely on the visible spectrum, Phan et al. used the protein reflectin to make films that can modulate their reflectance from the visible (orange) into the near-infrared

(NIR) region of the spectrum by stretching them.81

The films are made by coating fluorinated ethylene propylene (FEP) tape with graphene oxide to form a negatively charged surface to promote adhesion of the reflectin protein.81 Then an even layer of reflectin is applied using a doctor blade technique (Figure 1.13a).81 One of the most unique features of this film is that it can be applied to essentially any surface to create IR reflective patterns. With the naked eye, the films remain transparent and look no different than uncoated FEP tape (Figure 1.13b).81 Under an infrared camera, the IR reflective tape and the leaf look no different

Dinneen | 20 from each other except for shape. This technology could have potential applications in obscuring or camouflaging objects in the infrared at a low cost.

Figure 1.13. a. Reflectin film fabrication procedure. b. Photo of uncoated FEP tape, five pieces of reflectin coated FEP tape, and a leaf. c. Infrared camera image of the same objects showing the IR reflectivity of the leaf and the reflectin coated tape.81 Reproduced from Ref. 81 with permission of The Royal Society of Chemistry.

Recently our group sought to mimic the expansion of cephalopod chromatophores by fabricating films with different thicknesses of pigment granules (Figure 1.14ab).82 Instead of building actuating systems to mimic the organ, the films served as snapshots of different stages in the process. The films were made using a drop casting method with varying concentrations of granule suspension.82 The films were then measured using a total integrating sphere, which allows the user to quantify forward and backward scattering, and therefore more accurate values for transmission and specular reflection.82 In all films, transmission is decreasing as the thickness of

Dinneen | 21 the granules increases (Figure 1.14c).82 This can be explained by reabsorption, since light would have much farther to travel to pass through the film. Overall, it was found that the pigment granule layers can modulate reflected light from the visible into the short-wave infrared by changing the granule layer height and choosing to include a back-reflector.

Figure 1.14. a. Photos of six films representing the stages of chromatophore actuation. b. SEM cross- sections of films G4-G6 to show layer height. c. Integrated transmission and reflectivity for all films.82 Reprinted with permission by Ref. 82. Copyright 2018 John Wiley and Sons.

While chromatophores have served as the inspiration for camouflaging systems, its optical properties had yet to be quantitated. Until recently, chromatophores were simply described as color filters, but their contribution to camouflage is more important and deserving of optical analysis.

The network of granules within the chromatophore are believed to increase the absorbance of light in the dermal tissue; however, the mechanism of coloration remains unknown. The current hypothesis is that the pigment granules promote the scattering of light within the chromatophore to enhance the efficiency of coloration due to reabsorption, which is facilitated by high refractive index pigment molecules within the granules. This allows for the squid to maintain color richness within the dermal tissue even through the expansion of its chromatophores. The findings could make advancements in the areas of flexible displays and synthetic camouflaging systems.

Dinneen | 22

Dissertation aims: To better understand the mechanism by which these animals can adaptively camouflage on a molecular level, the optical properties of the squid pigment were evaluated in the following chapters. The pigments’ interaction with light by absorption and scattering should be fully understood before it is applied to a materials system. This is especially important for systems that need to be modeled, as many of the variables needed to calculate its interactions are measured experimentally, such as refractive index. Chapter 2 prefaces the following chapters with additional experimental details that were not included in the published works, such as squid dissection, skin digestion, pigment extraction, refractometry, and cavity ring- down spectroscopy. There is also a comment about the variability of the pigments within the animals. Chapter 3 focuses on obtaining the complex refractive index of squid pigment in solution. We measured the pigment on a refractometer and used mass fraction analysis to determine the real portion, n, at 589 nm.83 We then used UV-vis spectrophotometry to calculate the imaginary portion of the refractive index, k, at 589 nm using proxies for the density of pigment, as it is unknown.83 Then we calculated the optical cross-sections of the pigment using Mie theory to determine its contribution to both scattering and absorption. Chapter 4 details the procedure by which the pigments were aerosolized to be measured by a cavity ring-down spectrometer (λ = 532 nm) in order to ultimately retrieve a refractive index to compare to our previous answer and models.84 The size dependent extinction of the pigments were compared to our measured refractive index. Lastly, in Chapter 5, solvent on the measured particles needed to be accounted for as it caused some discrepancy in our measurements. We employed SEM analysis of the generated aerosols in tandem with Maxwell-Garnett mixing rules to account for the solvent and retrieve an updated refractive index value for the pigment.85

Dinneen | 23

CHAPTER 2:

Experimental Considerations

Squid dissection and tissue digestion

The longfin inshore squid, Doryteuthis pealeii, were shipped overnight from the Marine

Biological Laboratory in Styrofoam-lined boxes with either multiple ice packs or crushed ice to keep the specimens chilled throughout the transit. They are shipped with their mantles separated from their heads, both of which are included in the shipment. In a silicon-lined dissection tray, the ventral side of the mantle was cut across its whole length using dissection scissors, resulting in a flayed sample.65-66 The inner organs were then removed by cutting the connective tissue holding them to the inside of the skin and were disposed of in a plastic bag.66 Then, the squid skin was fixed tightly to the silicon mat in the dissection tray with T-pins, with the dorsal side was facing upward.66 Filtered seawater (collected on the New Hampshire seacoast then filtered through 0.2

µm pore size filters) was used to keep the skin from drying out during the dissection process.66

Using tweezers, the translucent epidermal layer was carefully removed, making sure to keep the chromatophore layer beneath intact.66 The chromatophore layer was then removed in smaller sections (1-4 cm2) with scissors and placed in 1.5 mL microcentrifuge tubes, filling them halfway.66

To digest the tissue, we used a series of buffer washes with repeated vortex, sonication, and centrifuging steps at regular intervals.65 First a 10 mL solution of collagenase and papain in water was prepared using 30 mg and 5 mg, respectively.66 500 µL of the collagenase/papain

Dinneen | 24 solution was then added to each of the tubes containing chromatophore tissue.66 The samples were vortexed for 1-2 minutes to mix, then sonicated for 5 minutes to dissociate the cells from the surrounding tissue.66 To remove the collagenase/papain solution, the samples were centrifuged at

14,000 x g for 5 minutes and the supernatant was removed and discarded.66 The collagenase/papain solution steps are repeated two times, and any larger pieces of flesh that had not been digested were removed.66

The pigment granules were then isolated by using a homogenization buffer made with 100 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid, 10 mM magnesium chloride, 50 mM potassium aspartate, 1 mM dithiothreitol, and a mini tablet of protease inhibitor in water.65-66 500

µL of the homogenization buffer was added to the centrifuged samples and vortexed to mix for 1-

2 minutes.66 The samples were then sonicated for 30 minutes, followed by centrifugation at 14,000 x g for 5 minutes.66 The supernatant was discarded, and the homogenization buffer steps were repeated 2-3 more times.66 The resulting sample was a red pellet consisting of isolated pigment granules, separated from a yellow supernatant.

To extract the pigment from the granules, we used acidified methanol (0.5% w/ HCl-

MeOH). 500 µL HCl-MeOH was added to the granule pellet and vortexed for 1-2 minutes to mix.66

The sample was sonicated for 10 minutes and then centrifuged for 5 min at 14,000 x g.66 The colored supernatant was collected making sure not to disturb the granule pellet. The acidified methanol extraction process was repeated until no more visible color was observed in the supernatant.66 In order to prepare solutions with known concentrations, the solvent from the pigment solution was dried using a rotary evaporator. The dried pigment was then scrapped off the bottom of the flask and used to prepare pigment solutions of specific concentrations.

Dinneen | 25

Throughout the following chapters, the dissection process, as well as the pigment isolation, was performed multiple times using many squid over the course of three years (three squid fishing seasons). Because the experimental techniques used in this work use destructive processes, the pigment samples are not always able to be recovered. This, along with the fact that yields from the granules are low (58%), results in the need to collect pigment often to restore available sample.66

In our previous work, we characterized the isolated pigments by thin layer chromatography (TLC),

UV-vis spectroscopy, fluorimetry, and mass spectrometry.65 Historically, isolated ommochromes have been difficult to characterize with nuclear magnetic resonance (NMR) spectrometry because of their similarity between structures or redox states within mixtures and their limited solubility.36,

86 Often the only reported characterization methods are UV-vis and infrared (IR) spectroscopy, chromatography (paper and column), and degradation (acid and alkaline).86

Figure 2.1. a) UV-vis absorption profile of extracted squid chromatophore pigment in acidic methanol, inset is a photo of a normal-phase TLC separation of the pigment resulting in four distinct bands. b) UV- vis absorption profiles of the separated and purified TLC bands, inset is a photo of the separated bands in solution. Reprinted with permission from Ref. 62. Copyright 2016 American Chemical Society.

TLC revealed that the isolated pigment separates into four distinctly different colored bands (Figure 2.1). This was further characterized by UV-vis spectroscopy, where each of the

Dinneen | 26 colored bands contained unique absorption profiles (Figure 2.1). While UV-vis itself is not sufficient to characterize the pigments, the absorbance peaks match what has been previously characterized for ommochromes.37-38 We have also shown the pigment’s UV-vis profile in three publications, all of which match very well and originate from different extractions.65-66, 83 Mass spectrometry further confirmed the presence of ommochromes in these bands, and by matching likely ommochrome fragments we identified the presence of xanthommatin and decarboxylated xanthommatin in varying amounts in each band.65

Pigment variability

While the extraction method has been held constant, there is a question of whether the pigment varies within each animal or even by their maturation. Since pigment extraction yield is low, it would be difficult to study the pigment differences between single animals. In cases like this, the pigment would need to be studied as a sample pool to have enough material for quantitative studies. The pigment in squid chromatophores have only recently been identified by our group, so there have not been any pigment maturation studies performed on Doryteuthis pealeii. If the pigment within the chromatophore changes throughout the lifetime of the squid (less than one year), then animal maturation could be a factor in pigment characterization.

This very phenomenon has been observed in the ommochromes found in male dragonflies as they reach sexual maturity. Many young male and female Crocothemis and Sympetrum dragonflies start life as a yellow/orange color.87 Upon maturity, the male dragonflies will turn red while the females remain yellow. This process occurs through redox chemistry, where the reduced form of the ommochromes (mostly decarboxylated xanthommatin with xanthommatin in some species) in mature males are found in greater ratios than in mature females.87 This results in a drastic and noticeable change in color throughout the lifetime of male dragonflies (Figure 2.2).87

Dinneen | 27

Figure 2.2. a) Age- and sex-specific photos of C. servilia. b) Comparison photos of immature and mature male and female dragonflies of C. servilia, S. darwinianum, and S. frequens. Copyright 2012 National Academy of Sciences.87

While this is true for dragonflies, the visible colors of cephalopod chromatophores do not change once they have fully developed in the early stages of their life.88-89 This could indicate that the levels of ommochromes present in chromatophores do not change once they develop, and that the relative age of the squid, once dissected, will not greatly affect the chemical makeup of the extracted pigment solution. In addition, the squid received from our source at the Marine

Biological Laboratory are past the juvenile stage of their life, which further supports our notion that the maturity of the squid will not affect our pigment solution as we do not sample across the different stages of maturity of the squid.

Dinneen | 28

As there are multiple components to the pigment solution, it is indeed important to stress that the optical measurements performed in the following works represent the values of a mixture, and that the error associated with these values encompasses multiple measurements with possible variations in the pigment solution. Besides the xanthommatin and decarboxylated xanthommatin, there is a possibility that proteins could be found in the mixture in the form of ommochrome binding proteins (OBP’s).43 While we acknowledge this is a possibility, the existence of protein in the mixture would only lower the already uniquely high refractive index (RI) of the pigment we measured. The protein component, which is thought to be reflectin or crystallin, would have an average RI of ~1.5, meaning that the pigments themselves would contribute more to the linear combination of the RI of the mixture.60, 90-91

Refractometry

The solution phase refractive index measurements were carried out using an Abbe refractometer. Before each use, the refractometer was calibrated using a standard solid test piece with a refractive index of 1.5133 along with a RI-matching liquid (1-bromonapthalene) to ensure full contact with the prism with no air bubbles. Once the dark and light interfaces of the refractometer are within the cross-hairs, the scale can be adjusted to match the RI of the calibration test piece using an adjustment screw. However, calibrations only last for as long as the temperature remains constant since refractive index is temperature dependent. The calibration can then be checked with distilled water or a standard sugar or salt solution. Though a one-point calibration is often all that is needed, it can also be checked against multiple solutions with known, repeatable values. There is no universal standard for refractive index, though it is accepted by definition that the refractive index of a vacuum is exactly 1.000. Throughout the course of the refractive index experiments, the blank solvents were measured in order to calculate the contribution of the pigment

Dinneen | 29 itself to the solution’s value. In all cases, the refractive indices of the blank solvents have percent errors of less than half a percent when compared to their accepted RI values.83, 92

To measure solutions on the refractometer, it must first be set to a constant temperature using a circulating water bath. All values were measured at 25 ˚C, allowing each sample to equilibrate for about 30 seconds. Each solvent type was measured at least three times using four different concentrations to create a concentration curve. This was repeated for each solvent condition, and uncertainties are reflected as variabilities of multiple measurements.

Cavity ring-down spectroscopy

The uncertainties associated with values obtained by the cavity ring-down spectrometer

(CRD) also reflect variabilities from multiple measurements. For every measured extinction value from the spectrometer, it is the result of an average of multiple laser shots corresponding to an average of simultaneous concentration measurements taking place within a five-minute window.

This five-minute window translates into different numbers of averaged values when comparing

CRD to CPC data. The CRD will take a measurement for each laser shot, but the CPC will take a measurement each second.

Raw data from a cavity ring-down experiment is plotted in Figure 2.3ab in the forms given by the CRD and the condensation particle counter (CPC), respectively. The max values from the

CRD plot within the first 500 seconds correspond to a τ0 value where there are no particles present in the system, which can also be seen in first 500 seconds of the CPC plot (Figure 2.3ab). Using both the τ0 and τ values, and the concentrations at those times, extinction cross sections can be calculated.

Dinneen | 30

a) 120 110 100 90

80 Tau (µs) Tau 70 60 50 0 1000 2000 3000 4000 Time (s)

b) 250

) 3 200

150

100

50 Number (#/cm Number Concentration 0 0 1000 2000 3000 4000 Time (s)

Figure 2.3. a) Tau versus time for a full experimental CRD trial including switching particle diameters at certain intervals. b) Number concentration of pigment aerosols versus time from a corresponding CPC trial, showing a clearer indication of when the particle diameters were switched as indicated with changing number concentration.

Dinneen | 31

Experimental vs. calculated extinction values

CRD values are plotted as data points with error bars because they need to be measured for single sizes at a time due to extinction being a size dependent value. When continuous lines are plotted in the same graph as CRD data points, this is an indication that the line was calculated with

Mie theory using refractive index, wavelength, and a size range as inputs. The calculated Mie theory traces are useful for comparing against measured data points. The closer the measured points are to the calculated trace, the closer in refractive index the sample is to the Mie inputs. The refractive indices used to calculate the Mie traces come from an experimentally determined value or are the result of a retrieval program that finds a best-fit value for experimental data. This is a useful way to retrieve the refractive index of a sample if it is unknown.

Dinneen | 32

CHAPTER 3:

Color Richness in Cephalopod Chromatophores Originating

from High Refractive Index Biomolecules

Abstract:

589 nm. We observed that the pigments were more likely to scatter attenuated light than absorb it and that these characteristics may contribute to the color richness of cephalopods.

Introduction:

A broad range of visible color observed in nature comes from pigment-based coloration that originates from biomolecules. For instance, chlorophyll, an inorganic, magnesium-

Dinneen | 33 coordinating chlorin (2,3-dihydroporphine) pigment, is abundant in chloroplasts and is responsible for the green color of plants and algae.41, 93-94 Melanin, an organic pigment composed of polymerized networks of 5,6- dihydroxyindole, is responsible for the dark pigmentation in humans,95-96 squid ink,97 and insects.98 And the deep red and yellow colors observed in crab spiders, brine shrimp, buckeye butterflies, dragonflies, and other arthropods stem from phenoxazone-based pigments derived from tryptophan, known as ommochromes.32-34, 37-38

Recently, the ommochromes xanthommatin and decarboxylated xanthommatin were spectrometrically identified as the source of visible color in squid Doryteuthis pealeii chromatophore organs.62 Chromatophores are commonly described as pigmentary soft actuators which function together with the underlying, reflective iridophore organs to change the visible color displayed on the cephalopod dermis in the presence of different environmental cues.57, 63

Nanostructured pigment granules populate the interior of the chromatophore organ.57, 63-64, 99 These granules are tethered together within an elastic sacculus that is anchored by radial muscle fibers.64,

100 When the muscle fibers contract, the network of granules expands, increasing the surface area of the organ, as it absorbs, reflects, and scatters light from its surrounding environment (Figure

3.1).58 While it is known that the tethered granule network regulates the absorption of light during actuation, the role of the pigments contained within the granules remains unknown. The goal of this work is to elucidate their role in the scattering and absorbance of light to better understand how they potentiate the adaptive coloration.

Dinneen | 34

Figure 3.1. Bright-field microscopic image of squid dorsal mantle chromatophores illustrating the dynamic range of size achieved during actuation. Scale bar is 1 mm.

Discussion:

Pigments confined within the chromatophore granules were first isolated using solutions of acidified methanol (HCl-MeOH) and were separated from the remaining insoluble, colorless pellet prior to use.62, 66 The resultant extract was a deep red color containing combinations of xanthommatin and decarboxylated (DC) xanthommatin (Figure 3.2).62 We asked how these extracted biomolecules contribute to optical properties of the granules. To test this, we first experimentally measured the solution refractive index (RI) at 589 nm. We began with a concentrated 1.00 % (w/w) of pigments in 0.50 % (v/v) of acidified methanol (HCl-MeOH) (Figure

3.3). At this concentration, the pigment solution exhibited a measured RI of 1.3327 ± 0.0003. To determine the dependence of the measured RI on pigment concentration, the original 1.00 % (w/w) solution was serially diluted to a final concentration of 0.13 % (w/w), where the solution RI was measured as function of concentration (Figure 3.3). A linear correlation between concentration and measured RI was observed (R2 = 0.8798 ± 0.0006), suggesting the pigments are responsible for the recorded values.

Dinneen | 35

Figure 3.2. Pigments identified in squid chromatophore pigment granules include a) xanthommatin and b) DC xanthommatin.

Table 3.1. Extrapolated refractive indices from 4 solvent system. The HCl-MeOH, N= 5 samples, measured in triplicate for 15 total measurements; MeOH, N = 2 samples, measured in triplicate for 6 total measurements; MeOH/H2O, N =2, measured in triplicate for 6 total measurements; NaOH, N = 1, measured in triplicate for 3 total measurements. The 0.2M NaOH system followed an alternate procedure where the solvent was also used to directly extract the pigment from the granules, rather than extracting with acidic methanol and resuspending in an alternate solvent.

Solvent system Extrapolated Refractive Index Standard Deviation

HCl-MeOH 1.92 0.11

Methanol 1.95 0.32

50/50 MeOH/H2O 1.88 0.26

0.2M NaOH 1.73 0.15

At these concentrations, the extrapolated real RI (n) for the pigment was then determined

(Table 3.1). Solution-based extrapolation of n from a component in a mixture is most commonly calculated using the component’s volume fraction or density (Equation 3.1).101-102 Because the densities of xanthommatin and DC xanthommatin are unknown, their n was instead calculated

Dinneen | 36 using a mass fraction analysis and their experimentally determined solution RI (see details in the experimental section). We found that the pigments have an n of 1.92 ± 0.11 (N=5 samples). The reliability of these solution-based RI measurements has previously been validated using dicarboxylic acids, where experimentally extrapolated n values matched with reported literature values.103 These measurements also agreed with the calculated molar refraction for xanthommatin

(114.1) and DC xanthommatin (110.4),104 which were within the range of the experimentally measured xanthommatin (130) and DC xanthommatin (120) calculated using their RI of 1.92 and a density that was approximated from the values in Figure 3.4, of 1.5 g cm-3. These similarities suggest good agreement between theory and the measured experimental values.

Figure 3.3. Solution refractive index as a function of mass fraction of one sample of pigments in HCl- MeOH solution. Error bars are sample standard deviation (1σ, 3 repeat measurements). Weighted linear trendline is fit with a slope of 0.0040 ± 0.0004 and an intercept of 1.3290 ± 0.0002. Inset pictures of pigment solutions used at their respective concentrations labeled as percent values.

To determine the role of solvent on the measured RI, we next varied the solution conditions.

Because the solubility of the chromatophore pigments is limited to MeOH, we used mass percentages of 0.28 %, 0.17 %, and 0.07% of pigment in MeOH (no HCl) and comparable mass

Dinneen | 37 percentages 0.28, 0.18, and 0.09 % of pigment in 50/50 (v/v) MeOH/ H2O. We found that regardless of the solvent used, the n of the soluble pigment extracted from the chromatophore granules was calculated as 1.92 ± 0.23 (N = 9 samples measured in triplicate for a total of 27 measurements across three solvent conditions). Larger standard deviations were observed from the MeOH and 50/50 MeOH/H2O conditions compared to HCl-MeOH, raising the total error from

~6% with HCl-MeOH measurements only to ~11%. These deviations are attributed to the limited solubility of the pigments in the absence of acidic medium. HCl-MeOH is the most favorable medium, yielding the smallest error. Regardless of the solvent conditions used, the extrapolated n is high for biological molecules; where, the closest comparators are experimentally derived and approximated natural and synthetic melanin with an RI range of 1.7-2.076, 105-106 and naturally occurring guanine crystals in copepods, which have an RI of 1.83.16

Figure 3.4. Imaginary refractive index as a function of wavelength. Each line represents results for a different density (blue = 1.20 g cm-3, red = 1.47 g cm-3, green = 1.54 g cm-3). Inset is the UV-vis absorption plot of 0.24 g L-1 pigment in HCl-MeOH, which is used to calculate the k.

We next calculated the imaginary portion of the RI (k) using Equation 3.2 (details in experimental section).107-108 Because the densities (ρ) of xanthommatin and DC xanthommatin are unknown, other known aromatic heterostructures with structural similarities including Nile red (ρ

Dinneen | 38

= 1.20 g cm-3)109 and γ- and β-quinacridone (ρ = 1.47 g cm-3 and ρ = 1.53 g cm-3, respectively)110 were used to bound the results. The absorbance (A) of the pigments in HCl-MeOH at a known concentration (c = 0.24 g L-1, Figure 3.4, inset) and optical path length (L = 1 cm) was used with these densities to estimate the absorption coefficient (α) and thus the k of the pigments across the ultraviolet-visible (UV-vis) spectrum (Figure 3.4). We found that regardless of the densities used, k values measured at 589 nm ranged from 0.011 to 0.014, making our complex RI 1.92 + 0.014i.

Figure 3.5. Optical cross-sections of the pigment particles calculated with an n of 1.92 and k of 0.014 as a function of the particle diameter. Each line represents a different cross-section (red = scattering, blue = absorption).

Because the pigments are localized within nanospherical granules inside the chromatophore organ, we next asked how particles containing these high RI pigments contribute to the absorption and scattering of light. We used Mie theory to calculate the absorbance and scattering cross-sections (cm2/particle) across a 470 - 696 nm diameter range (Figure 3.5) at a wavelength of 589 nm. The diameter range selected in this calculation was based on the experimental average diameter previously measured for pigment granules extracted from the squid chromatophore.62 We found that throughout this size range, the pigments were more likely to scatter attenuated light than absorb it in a mechanism that is principally enabled by the high real

Dinneen | 39

RI of the pigment (Figure 3.5). These analytical calculations agreed with Finite Difference Time

Domain (FDTD) simulations of the scattering and absorption cross-sections, further supporting this finding (Figure 3.6).

Figure 3.6. Calculations of scattering and absorption cross-sections, using the Finite Difference Time Domain (FDTD) algorithm implemented by commercial software, for pigment particles as a function of the particle diameter representing the granule size range, at a wavelength of 589 nm. Each line represents a different cross-section (red=scattering, blue=absorption). Indices of n = 1.99 and k = 0.014 were used. A “total field - scattered field” light source, enclosing the particle, was employed, and fields were assumed converged at a relative amplitude of 1e-6. Similar FDTD simulations, from a baseline case of a 1 µm diameter water droplet in air, were verified to reproduce results in the literature (Fig. 4.6 on p.105 of Ref. 149).

Because variations in and values of k appeared larger at wavelengths other than 589 nm, we also calculated the absorption and scattering cross-sections at additional wavelengths (Figure

3.7).

Dinneen | 40

Figure 3.7. Calculated optical cross-sections of pigment particles as a function of particle diameter using imaginary RI values from different wavelengths where k was extrapolated at a) 300 nm, b) 400 nm, c) 500 nm, and d) 600 nm with a constant real RI of 1.92. Here, the red lines are scattering cross-sections and the blue lines are absorption.

We tested k values at 600 nm (k=0.010), 500 nm (k=0.041), 400 nm (k=0.024), and 300 nm (k=0.028). At these wavelengths, k was never close enough to 0.1, where absorbance would matter. We next asked how variations in both n and k influence scattering and absorption at 589 nm. To test this, the k was first varied by 10% with a constant n (Figure 3.8a-b), then n was varied by 10% at a constant k (Figure 3.8c-d), where 10% was selected based on the error associated with the experimentally determined n from Table 1. Under each condition, optical cross-sections of the pigments were calculated (Figure 3.8). We observed that variations of k did not significantly impact the shape or magnitude of the scattering profile at this wavelength. However, when the n was varied by 10%, the scattering and absorbance spectra appeared to either approach each other

(+10% n), or separate further (- 10% n), suggesting that variations in n impact the profiles of the

Dinneen | 41 scattered cross-sections. In spite of these differences, scattering remained dominate over absorption.

Figure 3.8. Calculated optical cross-section of pigment particles as a function of particle diameter based on measured complex RI 1.92+0.014i, where red lines are scattering cross-sections and the blue lines are absorption. Optical cross-section based on complex refractive index was changed so that a) is constant n with +10% k, b) is constant n with -10% k, c) is constant k with +10% n, and d) is constant k with -10% n.

To determine how variations in particle diameter affected optical cross-sections, we next varied the theoretical particle diameter from 5 - 1000 nm at 589 nm using an n of 1.92 and k of

0.014, and the optical cross-sections were calculated. We observed that scattering predominated at diameters greater than 100 nm (Figure 3.9). Below this size, the difference in scattering and absorbance cross-sections became less distinguishable. Above this size, two regimes persist: a plateau from 400-750 nm and a sharp increase from 750-1000 nm (Figure 3.9). In the native chromatophore, pigment granules have particle diameters of 583 ± 113 nm.62, 66, 99 This range also coincides with the observed plateau that emerges at 400 - 750 nm particle diameters in our

Dinneen | 42 calculations, suggesting an optimized range for these pigments to effectively scatter light. Beyond this range, scattering exponentially increases, indicating a high degree of variability dependent on particle size.

Figure 3.9. Calculated optical cross-section of pigment particles as a function of particle diameter, where red lines are scattering cross-sections and the blue lines are absorption cross-sections. Inset is expanded view of particle diameters 5 – 200 nm.

To determine whether these scattering profiles are unique to the xanthommatin pigments of squid chromatophores, we compared their optical cross-sections with those reported for melanin. We chose the n (1.74) and k (0.188) values reported for synthetic melanin nanoparticles76 and the n (1.72) and k (0.0632) from melanin measured from the occipital feathers of the bird of paradise105 both at 589 nm (Figure 3.10). In both cases, we observed that the calculated absorbance cross-sections were 2 -4x larger than those calculated for the xanthommatin pigments, suggesting that melanin is a better light absorber when assembled as nanoparticles in this size range. In the case of the synthetic melanin nanoparticles with the k over 0.1, the scattering and absorption contributions were closer in magnitude, suggesting a more equal contribution of the two in attenuating light in a manner that was significantly different from the xanthommatin pigments.

Dinneen | 43

Figure 3.10. Calculated optical cross-sections of xanthommatin pigments (red, scattering; blue, absorption), melanin nanoparticles using n = 1.74 and k = 0.188 from Xiao and coworkers76 (black solid line, scattering; black dotted line, absorption) and soluble melanin using n = 1.72 and k = 0.0632 from Stavenga and coworkers105 as a function of particle diameter, where green lines are scattering (top) and absorption (bottom) cross-sections absorption.

Collectively, our data suggest that both the particle size and the presence of high n and low k pigments enable light to be primarily scattered in the chromatophore. However, it is important to note that the composition of the insoluble granule (~170 nm diameter62) that is separated from the soluble pigment during the initial extraction stage remains unknown, and thus its effects on the granule RI are not well understood. A previous study on cuttlefish Sepia officinalis chromatophore granules has shown that the full granular structure can be disrupted in concentrated sodium hydroxide, where this structural denaturation is correlated with a decrease in the abundance of crystallin and reflectin proteins.99 The RI of these proteins have previously been reported between

1.44- 1.59 and 1.55 for reflectin60, 90 and crystallin,91 respectively. Thus, if these proteins were also present in the extracted pigment, then we would have observed even lower RI’s then those reported here. We tested this hypothesis by first denaturing the pre-extracted granules in concentrated sodium hydroxide then measuring its RI at 589 nm (Table 3.1). We found that the suspension had an n of 1.73 ± 0.15, indicating a lower RI than those reported for the MeOH-based studies that

Dinneen | 44 may be a result of the lower n values associated with the structural proteins. Future work will focus on a compositional analysis of the remaining, insoluble portion of the granules to correlate with the observed decreases in the NaOH extracted RI.

In this report, we have experimentally determined the n and approximated the k of the soluble pigments isolated from the nanostructured granules that populate cephalopod chromatophores, representing the first RI measurements on this class of biomolecules. While solution-based RI measurements at one wavelength do not fully capture the dynamic range of color the cephalopods experience in the ocean, we believe that our measurements at 589 nm – the center of the solar spectrum – represent a critical piece of data that will help inform future experiments designed to investigate the solid-state granule RI. With a more complete understanding of how these granules scatter and absorb light, we may also better understand the fast (hundreds of milliseconds99) response time afforded to the chromatophores during actuation. For instance, the use of such high RI biomolecules could potentially enable light to refract closer to the normal of the surface of the granules, thus facilitating light penetration to the underlying iridophore

(reflector) organs, even at higher incident angles experienced in the ocean. Collectively, the properties inferred from the complex RI of the soluble biomolecules suggest that color-richness in cephalopod dermal tissue originates from the molecular level assembly of high RI pigments localized within nanostructured granules in the chromatophore organs.

Experimental:

Chromatophore pigment granules are isolated and purified using previously published procedures.62, 66 The soluble pigments are extracted using acidic solutions of methanol (0.50% concentrated HCl/methanol by volume). The real refractive index, n, of the pigments in solution were measured in serial dilutions using three different solvents on a Bausch and Lomb Abbe-3L

Dinneen | 45 refractometer at 589 nm at 25 ˚C. The solvents used were acidic methanol, methanol, and 50/50 methanol in water (v/v). Each concentration of each solution was measured three times, and the measured refractive index of the solution (nmix) and solvent (nsolvent) was used to extrapolate the refractive index of the pigments (npigment) using the mass fraction of the pigments (wpigment) and solvent (wsolvent) in solution using this equation:

nmix = nsolvent*wsolvent + npigment*wpigment (3.1)

The equation for the mix is a linear combination of each component’s refractive index multiplied by the mass fraction of the component in solution. The mass fraction was found by evaporating the solvent from a known mass of solution and weighing again to find the mass of pigments remaining. The pigment’s refractive index is averaged from the values extrapolated from each dilution.

The imaginary portion, k, of the refractive index was calculated using a range of densities of molecules with similar structures, since the density of the pigments are unknown, in addition to

UV-vis absorbance data of the pigments in solution where the solvent absorbance was subtracted as background. The absorption coefficient was solved for using measured absorption data and assumptions of the pigment’s density using Equation 3.2, where α is the absorption coefficient, ρ is the density in g cm-3, A is absorbance, c is concentration in g L-1, and L is the optical path length in cm:107

훼(휆) 퐴(휆) = 1000 ln (10) (3.2) 휌 푐퐿

The absorption coefficient can then be related to the imaginary portion of the refractive index using the relation α = 4πk/λ.107

Dinneen | 46

The median value of k at 589 nm, as well as the average value of n at 589 nm, was used to calculate the absorption and scattering cross-sections of pigment particles with diameters between

470 and 696 nm to simulate approximate sizes of pigment granules. The calculations were performed with a FORTRAN program using Mie theory at 589 nm.

Dinneen | 47

CHAPTER 4:

Optical Extinction of Size-controlled Aerosols Generated

from Squid Chromatophore Pigments

Abstract:

Nanophotonic granules populate the interior of squid Doryteuthis pealeii chromatophores, contributing to dermal color richness by selectively absorbing and scattering light. Inspired by the performance of these granules, we fabricated nanostructured aerosols by nebulizing a pigment solution extracted from native squid chromatophores. We determined the extinction of these aerosols using cavity ring-down (CRD) spectroscopy and show how extinction cross-section is dependent on both particle concentration and size. This work not only advances the fundamental knowledge of the optical properties of chromatophore pigments, but also serves as a proof-of- concept method that can be adapted to develop coatings derived from these materials.

Introduction:

Nanostructured materials that exhibit size and shape dependent optical properties for sensing, visualization, and photo-protection are observed in many natural systems. For instance, the bird-of-paradise, Parotia lawesii, can selectively reflect light across the visible spectrum to attract mates due to constructive interference propagated from the ordered layers of melanin nanoparticles confined within their feather barbules.14 Opals, which are historically regarded as

Dinneen | 48 good luck stones, similarly reflect visible light; however, they do so using a tightly packed arrangement of silicon dioxide nanoparticles.111 In cephalopods, protein-based nanostructured granules populate the chromatophore and leucophore (cuttlefish and octopus only) organs in their dermal tissue and contribute to pigmentary color and diffuse scattering, respectively.62, 83, 99-100, 112

While many of the components integral to the optical function of these organs have been identified, the relationship between chemical composition and structural morphology remains largely unknown. This is especially true for the pigmented granules in the chromatophore. Even though the abundance of crystallin and reflectin proteins and xanthommatin and decarboxylated (DC) xanthommatin pigments are observed to be associated with the spherical architecture and volume of the granules, it is still unclear whether the pigments and proteins coordinate together to form each nanoparticle. It is also unknown how this combination may contribute to absorption and scattering of light, and the sum of these, extinction.62, 99 What is known is that when the pigments are selectively removed from the chromatophore granules, the particles reduce by ~70% in diameter and become devoid of color, suggesting an important structural and optical role.

Results and Discussion:

To investigate how the pigments alone contribute to structural morphology and optical function of the chromatophore granules, we first set out to manufacture pigment nanoparticles. We adapted a protocol commonly used to generate environmental aerosols and measure their optical extinction.113-116 In many ways, aerosols can be described as nanoparticles present in the gas phase, of which a major fraction are known to contain organic carbon.117-118 Because the squid pigments contain xanthommatin and DC xanthommatin, both rich in carbon, we asked whether it is also feasible to generate aerosols using a solution of pigments. We describe a method for producing

Dinneen | 49 and collecting select sizes of pigment aerosols and show how experimentally measured optical extinction is dependent on both particle concentration and size.

Figure 4.1. Atomization scheme and experimental set-up for the optical measurement of squid pigment aerosols. The pigment solution is drawn into the atomization block where it is nebulized under N2. The particles pass through two diffusion dryers and are size separated by the differential mobility analyzer (DMA). The particles are then measured optically by cavity ring-down spectroscopy (CRD) and are counted by the condensation particle counter (CPC). The ratio of B (optical path length) to A (sample path length) in this experiment is 1.17.

Chromatophore pigments are extracted from D. pealeii using previously described methods.62,66 The extracted solution is dried in a vacuum desiccator overnight and used to make a

1.00% (w/w) solution in 0.20 M hydrochloric acid (HCl). The acidic pigment solution is then placed in the intake tube of a Collison-type atomizer (Figure 4.1), where a 3.0 L/min stream of N2 dispersed the solution into solvated aerosol droplets producing a log normal distribution of sizes

Dinneen | 50

(Figure 4.2).119 To mitigate the effects associated with solvent inclusions during aerosol formation, we employed two diffusion dryers in series prior to analysis. The dryers are effective at achieving a low relative humidity (<9%, Figure 4.2). Once dried, the aerosol stream is directed through a differential mobility analyzer (DMA, TSI model 3080L) which separated the particle diameters by their electrical mobility. The charged particles are entrained in a flow under an electric field where particle geometry and charge are combined to produce an aerosol stream nearly monodisperse (~

±10%) in diameter.120

Figure 4.2. Particle size distributions for pigment solution and solvent blank.

The separated pigment aerosols are collected on straight-through pore polycarbonate membrane filters (Whatman, 200 nm pore size) using an in-line filter holder121 with a custom- designed bypass system that allowed for the filters to be changed without ceasing aerosol flow

(Figure 4.3a). Once collected on the filters, the size-selected aerosols are analyzed using scanning

Dinneen | 51 electron microscopy (SEM, Tescan Lyra3-GMU), and diameters are measured using ImageJ software.122 The average measured diameters of the particles exhibited some variance to the selected DMA size (Figure 4.3b). For instance, we observed that for particle sizes up to 240 nm, the measured diameters trended larger than the ones specified by the DMA, and that for diameters greater than 300 nm, the measured values trended smaller. These variances are likely a result of the charging effects, where larger, multiply-charged particles pass through the DMA with the same electrical mobility as the singly charged particles, effectively increasing the average measured diameter.123-124 This is more likely to occur at the smaller sizes which are closer to the peak of the particle distribution (Figure 4.2). Beyond this point, the aerosols exhibited a smaller physical diameter than those specified with the DMA. For particle sizes greater than 350 nm, we also observed the emergence of a heterogenous surface topology (e.g. non-spherical). The appearance of collapsed structures suggested either that the pigments are not capable of forming homogenous nanospheres larger than 350 nm or that at these larger diameters, the solvent gets trapped in the forming aerosols. If it is the latter, then the spherical aerosols are likely to deform as the solvent inclusions dry under the high vacuum of the SEM, reducing the overall sizes of the selected particles (Figure 4.3b, inset).

Dinneen | 52

Figure 4.3. a. Aerosol collection setup. b. Average measured diameters from SEM images plotted versus DMA selected diameters. Error bars indicate standard error, where N = 82 – 135. The inset includes representative SEM micrographs of pigment aerosols (scale bar = 500 nm).

Because many nanostructured materials have optical properties that are dependent on size, we next measured the effects of size on optical extinction. Extinction cross-sections are extrapolated from the measured optical extinction of particles at increasing number concentrations using our custom-built cavity ring-down (CRD) cell. Generated aerosols are exposed to 20 Hz pulsed 532 nm light within a ~70 cm cell from a neodymium-doped yttrium aluminum garnet

(Nd:YAG) laser source (Quantel USA, Ultra). Highly reflective concave mirrors (R>0.999) on either end of the cell reflected the light multiple times to increase the effective path length to ~30 km, albeit with an exponential decay of intensity. During this process, a small amount of light leaks out of the cell opposite to the source, and its intensity is measured by a photomultiplier tube

114, 125 (PMT). In a particle-free CRD cell, light intensity decays at a time constant τ0; whereas, the cell that is filled with an aerosol stream absorbs and scatters more light, decaying with a shorter time, τ. These decay times are measured and used to determine the extinction of the sample with units of inverse length, αext, per equation 4.1:

Dinneen | 53

푅퐿 1 1 훼푒푥푡 = ( − ) (4.1) 푐 휏 휏0

where c is the speed of light, and RL is the ratio between optical path length, B, and sample path

114, 125-126 length, A (Figure 4.1; RL = 1.17 in this case). A condensation particle counter (CPC, TSI model 3775, ±10%) is then used to determine the number concentration (N, particles/cm3) of the aerosol stream that passed through the CRD. The CPC makes a measurement once every second, and these measurements are averaged over the same range of time used for the CRD (typically 60-

80 seconds). Based on the measurements from the CPC and CRD, the extinction cross-sections

2 (σext, cm /particle) are calculated (Equation 4.2).

휎 = 훼푒푥푡 (4.2) 푒푥푡 푁

Three number concentrations varying in magnitude are achieved for the seven diameters (216 –

545 nm) selected by the DMA using increasing atomization pump speeds, and their optical extinctions are measured using the CRD. As expected, the concentration-dependent extinctions are found to increase linearly with particle number concentration, indicating that the generated pigment aerosols are indeed responsible for extinguishing light. The size dependent extinction cross-sections are calculated from the slope of the linear least squares fit (illustrated in Figure 4.4; values in Table 4.1).116, 127

Dinneen | 54

Figure 4.4. Optical extinction of squid pigment particles based on CRD measurements versus number concentration as counted by the CPC. Symbols indicate the DMA size selected diameter. Y-error propagated from error on τ values. X-error is simple standard deviation of average number concentration (1σ, N=81 concentration measurements).

In each case, the error in the slope is propagated due to the high linearity (R2) portrayed by the regression at each aerosol size. We observed an increasing extinction cross-section that is dependent on the geometric cross-sectional area of the aerosols. For example, the extinction for a given number of 300 nm particles (solid upside-down triangles) is equivalent to approximately twice that number of 216 nm particles (solid circles) per unit volume, indicating that the larger particles extinguished more light than the smaller particles. Number concentrations also appeared to decrease significantly for the larger (480 and 545 nm) diameter particles under the same atomization conditions, indicating that high concentrations of pigment particles in these diameter ranges are less likely to be generated at these pump speeds (Figure 3.4). This observation is also in agreement with the particle distribution, as shown in Figure 3.2, where diameters over 400 nm are at the tail end of the curve. Collectively, these data indicate that pigment particles have a

Dinneen | 55 predictable optical extinction behavior that is highly tunable over a wide range of sizes and number concentrations.

Table 4.1. The size-selected and experimentally measured particle diameters and calculated extinction cross-sections of the squid pigment aerosols.

DMA SEM measured N of SEM Extinction cross-section R2 of linear

selected sizes with measured with error from regression fit

sizes (nm) standard error particles regression (cm2/particle)

(nm)

216 240 ± 10 100 9.58±0.09x10-10 0.9999

240 258 ± 8 119 1.09±0.04x10-9 0.9983

300 289 ± 5 121 1.77±0.01x10-9 1.0000

351 334 ± 7 115 2.40±0.06x10-9 0.9994

396 375 ± 5 135 3.44±0.06x10-9 0.9997

480 422 ± 11 82 6.66±0.07x10-9 0.9999

545 489 ± 10 96 9.22±0.21x10-9 0.9995

We also compared our experimental data to the extinction cross-sections calculated using

Mie theory at a wavelength of 589 nm from our previous report (Table 4.2).83 We observed 5 -

30% differences in extinction cross-sections between the two data sets, where the largest errors occurred at sizes that reached the highest optical extinctions (240-396 nm). These variances are not all too surprising given the differences in reference wavelengths used in the two studies and

Dinneen | 56 the heterogeneity potentially persisting in the pigment aerosol structures at diameters above 350 nm. Despite these discrepancies, we observed good agreement between theoretically extrapolated extinction cross-sections and the measured values here (Figure 4.5), suggesting that the pigment assembled as theses nanostructures is responsible for the optical extinction.

Table 4.2. Comparison of our measured extinction cross-sections measured at 532 nm to previously reported Mie calculated extinction cross-sections of squid pigment at 589 nm using a complex RI of 1.92 + 0.014i.

Extinction cross-sections (cm2/particle) DMA selected SEM shifted Measured Mie calculated at 589 Percent size (nm) size (nm) data at 532 nm with SEM shifted difference (%) nm sizes 216 240 9.58x10-10 8.30x10-10 14.3 240 258 1.09x10-9 1.32x10-9 19.1 300 289 1.77x10-9 2.38x10-9 29.6 351 334 2.40x10-9 3.19x10-9 28.2 396 375 3.44x10-9 4.58x10-9 28.4 480 422 6.66x10-9 7.03x10-9 5.45 545 489 9.22x10-9 7.69x1009 18.1

Figure 4.5. Extinction cross-section versus particle diameters for measured pigment aerosols at 532 nm (black dots) with error reported in Table 1 in text, and calculated Mie extinction cross-sections at 589 nm (solid line) and at 532 nm (dashed line) using the complex RI of 1.92 + 0.014i for both wavelengths.

Dinneen | 57

In this report, we show that we can successfully aerosolize squid pigments and evaluate their number concentration and size-dependent optical properties. Even though these measurements do not account for the full makeup of the native chromatophore granules, the pigment is a major component,62 and therefore a dominant contributor to the optical behavior. By exploiting their ability to assemble into nanostructures in the gas phase, we also show a new method for processing these pigments that can conceivably be scaled up for spray-on coatings. We show that particle diameters can be separated by their electronic mobility, suggesting a control over selected size parameters that makes this process highly tunable. Overall these properties, along with the pigment refractive index at a specific wavelength or range of wavelengths, are important considerations in manufacturing pigment particles for future materials or optical applications.

Dinneen | 58

CHAPTER 5

An Iterative Correction Approach Used to Retrieve the Refractive Index

of Squid Pigment Aerosols

Abstract:

Pigments localized within cephalopod chromatophores are important for dermal coloration. When isolated and used as materials outside of the animal, the pigments can be processed as aerosols, illustrating a potential application for spray-on-coatings. The optical features of the pigment aerosols are difficult to analyze and require a method to correct for the particle charging and solvent effects accumulated during the aerosolizing process. We describe a method to account for these effects using an innovative iterative approach tied to retrieved refractive index (RI) values.

RI retrievals were obtained via the best fit between the corrected, experimentally observed extinction efficiencies compared to those calculated by Mie theory for a specific RI at selected sizes. In addition to these retrievals, the impact of solvent on the particles’ optical properties was also examined via the Maxwell-Garnett mixing rule. Ultimately, we obtained a pigment RI with a real portion (n) of 1.66 (±0.05) representing a lower limit and an imaginary portion (k) of 0.13

(±0.08)i representing an upper limit for the generated aerosols. Combined, this approach advances techniques used to retrieve RI values that benefits both atmospheric chemistry and bio-inspired materials.

Dinneen | 59

Introduction:

Particles dispersed in gaseous medium, also referred to as aerosols, have been known to impact the natural world in many applications for over the past 400 years. One illustration of this is in the use of colors that represent changing atmospheric aerosol concentrations by painters. A recent statistical study of paintings from 1506-1892 showed a correlation between red-to-green ratios associated with aerosol rich sunsets and what is known as the dust veil index– a metric of the atmospheric aerosol concentration as related to volcanic activity.128-129 The effect of aerosols on the transmission of light has other important features and applications. For instance, the use of aerosols to disrupt visibility has been incorporated in strategic defense for many years. A specific early example is the 1701 use of wet hay to generate large quantities of smoke aerosol allowing

Swedish troops under Charles XII to obscure their river crossing to advance their position.130 In addition to use as obscurants, aerosols have been used for signaling, where specific chemicals when ignited are used to produce different colored smoke transmitting information over long distances.130

Since the 1960s when scientists predicted that the scattering of light by certain types of aerosols could provide an offset for some of the climate warming due to carbon dioxide, methane and other gas phase species, the atmospheric chemistry community has been advancing techniques to study aerosols in the laboratory and field. Later, in 1982, the American Association of Aerosol

Research (AAAR) was founded to promote the study of particulates as related to air pollution, industrial hygiene, atmospheric sciences, clean room technology, and nuclear safety. To better understand how aerosols impact environmental and climate systems, techniques have been developed to quantify their optical properties. One such property is their composition-determined, wavelength-dependent refractive index (RI), which, along with the size or size distribution of Dinneen | 60 particles, governs how electromagnetic radiation interacts with a specific particle or group of particles. In the atmospheric chemistry community, techniques to obtain RI of aerosols include creating thin films from aerosol samples to be studied by ellipsometry131-132 and in situ examination of extinction, scattering and/or absorption by aerosol particles and combining such measurements with retrieval via model matching.113, 116, 133-143 Recent studies have advanced the later techniques via careful data analysis with calibration and a combination of techniques.144-145

The RI of chemicals can vary significantly with wavelength of light, and this effect is also observed frequently in naturally occurring pigments found in both plants and animals. Brightly colored flowers attract animals and insects for pollen to be distributed.146-147 Poison dart frogs use their long-wavelength pigment as a signal to predators that they are poisonous.148 Cephalopods, including the squid Doryteuthis pealeii, have dermal pigments localized within chromatophore organs that are comprised of xanthommatin and decarboxylated (DC) xanthommatin.62 These conjugated heterostructures contribute to the absorption and scattering of light when assembled as nanostructures.62, 83 We recently extrapolated the real, n, (1.92 ± 0.23) and approximated the imaginary, k, (0.011-0.014) portions of the pigment solution RI and used Mie Theory to calculate the absorbance and scattering cross-sections.83 We observed that the chromatophore pigments were more likely to scatter light than absorb it across a broad diameter range, suggesting an important role of the pigments in maintaining color fidelity during camouflage. In an effort to move away from solution-based experiments, we also manufactured squid pigment aerosols. This process enabled us to directly measure the size-dependent optical extinction of pigment particles suspended in nitrogen gas using a cavity ring-down (CRD) spectrometer, a technique common in atmospheric chemistry.84 Our data suggested good agreement between extinction data measured at 532 nm and the calculated Mie values; however, we still observed a 5-30% error between the

Dinneen | 61 experimental and calculated methods. In this work, we ask if we can account for this error using doublet (i.e. larger, doubly charged particles with the same electrical mobility) corrections and a refractive index retrieval program to return a value within our originally reported range of the pigment RI at 589 nm based on Abbe refractometry and UV-vis spectroscopy.83 We examine the possibility that solvent remained in the particles during optical interrogation by using the Maxwell-

Garnett mixing rule and spherical volume differences calculated between the diameter set by the differential mobility analyzer (DMA) and those observed by scanning electron microscopy (SEM).

This examination of our initial extinction data reveals both the capabilities and the limitations of using aerosol methods to study particles generated from squid pigment solutions.

Methods: All analysis is done using the experimental data set reported by Dinneen and coworkers in a previous report,83-84 where we determined extinction cross-sections of atomized squid pigment aerosol from 1% by weight solutions in 0.2M hydrochloric acid in water. Seven aerosol mobility diameters from 216– 545 nm were selected using a differential mobility analyzer (DMA, TSI model 3080L), and interrogated with a custom cavity ring-down (CRD) spectrometer (<9% relative humidity) to measure extinction; separately a scanning electron microscope (SEM) was used to validate the aerosol diameters. Prior to making measurements, the accuracy of the CRD was examined using squalene with a flow of 0.30 L min-1 at the condensation particle counter

(CPC); the same flow is used for these experiments.144 Abbe refractometer experiments were performed at 25 ˚C and CRD experiments were performed between 20-22 ˚C.83-84 Here, we describe methods used for the Mie correction, refractive index retrievals, and solvent mixing calculations of the data.

Dinneen | 62

To start, the number of doubly charged particles was calculated and subtracted from the total particle concentration for each size that was measured using a bipolar charge approximation.123 In addition, the contribution to extinction from that same number of doubly charged particles was subtracted from the total extinction. To accomplish this, the extinction cross- section at that size was calculated using Mie theory with our previously reported complex RI (m = n + ki) where the real portion, n, was measured at 589 nm and the imaginary portion, k, was calculated at 532 nm,83 then multiplied by the concentration of multiply charged particles. This doublet extinction contribution was then subtracted from the total extinction measured to obtain the extinction for only the singlet sized particles.

Since using Mie theory for corrections may bias the outcome,143 we elected to use an iterative process for our Mie doublet corrections, where our input complex RI was averaged until the input complex RI matched the best fit retrieved RI value (Figure 5.1). Following each iterative

Mie correction, the data was entered into a custom Igor coded retrieval program in the form of the extinction cross-section (σext) divided by the geometric cross-sectional area of the particle, also known as the extinction efficiency, Qext.

푄 = 휎푒푥푡 (5.1) 푒푥푡 휋푟2

When plotted against size parameter (χ, defined by πd/λ) both axes are dimensionless. The retrieval program performs a best fit between the experimentally measured Qext and a Mie calculated Qext varied over a range of complex RIs input by the user. A complex RI with error on both n and k based on a least squares fit is then retrieved. The input range can then be constrained with a higher resolution until a value is given with minimized error. The retrieved RI was averaged with the original complex RI that was used to make the corrections. The averaged value was then used to

Dinneen | 63 make a new set of corrections, and this process continued until the RI used for corrections matched the retrieved value (Figure 5.1).

Figure 5.1. Flow chart for the iterative doublet correction process used to remove bias from the initial RI input. To account for solvent effects on the generated squid pigment aerosols in our experiments, we utilized the Maxwell-Garnett mixing rule,149

휖−휖 3푓( 푚 ) 휖+2휖푚 휖푎푣 = 휖푚 [1 + 휖−휖 ] (5.2) 1−푓( 푚 ) 휖+2휖푚

where f is the volume fraction of the pigment component, and ϵav, ϵm, and ϵ are the complex dielectric constants of the mixture, the solvent matrix, and the pigment component, respectively.

3 푑푆퐸푀 푓 = 3 (5.3) 푑퐷푀퐴

The pigment volume fractions (spherical volumes simplify to yield equation 5.3) were calculated by assuming the SEM diameters, dSEM, measured in our previous report account for the volume of the pigment in the particles. The DMA diameters, dDMA, account for the total volume of the particles in the optical experiments, and any particles that measured larger than the DMA selected size (considering ±10% error)120 were assumed to be multiplets, where their DMA sizes were

Dinneen | 64 adjusted accordingly. Therefore, the difference between the two volumes was attributed to the volume of solvent in the particle when the optical cross-section was measured by the CRD. Each particle imaged with the SEM was included in these calculations and the average of each selected size was used.

To make quantitative comparisons of the fits of our data to retrieved RIs, we used merit function values (χ2/N2) where

2 2 푁 (푄푒푥푡 푚푒푎푠푢푟푒푑−푄푒푥푡 푟푒푡푟푖푒푣푒푑)푖 휒 = ∑푖=1 2 (5.4) 휀푖

N is the number of diameters measured and ε is the simple standard deviation of repeated measurements at each particle diameter.116 A minimized value of χ2/N2 indicates the best fit, and is helpful for comparing RI against measured data. Merit values depend on experimental parameters such that increasing the number of diameters measured will decrease the value, and therefore should only be compared within data sets.

Results: The differential mobility analyzer (DMA) uses a krypton source to charge the particles, which leads to an inherent problem where larger particles can be doubly charged and end up with the same electrical mobility as the smaller, singly charged particles.123-124 Therefore, the measured extinction was likely biased high due to contributions from the doublet particles in the sample. To account for this in the data, we used Mie corrections to subtract out the contribution from the doublet particles. Because of DMA limitations, only the first five measured diameters could be corrected in this way, as instrument limitations prevent particle concentrations from being measured at the doublet sizes of the largest two diameters (Figure 5.2). We found that indeed our original RI of 1.92 + 0.039i83 biased larger corrections, producing percent differences of up to

Dinneen | 65

~200% (Figure 5.2). After eight correction iterations our program retrieved a complex RI of 1.5606

+ 0.0128i, which produced percent differences of up to ~48% for the smallest particle diameter.

The magnitude of these corrections decreased as particle diameter increased due to the log normal size distribution created by the atomization method.124 Smaller number concentrations of aerosols at large sizes resulted in a reduced amount of doubly charged particles, where the size parameter of 2.34 had a Qext correction of only 4%.

Figure 5.2. Qext versus size parameter for uncorrected CRD data, Mie corrected data using the complex RI of 1.92 + 0.039i, and Mie corrected data using the complex RI of 1.5606 + 0.0128i. Error bars are simple standard deviation based on experimental CRD data (1σ, N = 3). Based on statistics, these error bars are quite large and in some cases negative (the negative error bars are only found in the most extreme corrections); however, these negative values are not physically meaningful.

As seen in Figure 5.3, the corrected data no longer matches as strongly to our originally reported complex RI of 1.92 + 0.039i. The Mie corrected data compared to its retrieval produces a merit function value of 0.012 while the uncorrected data compared against our reported complex

RI produces a value of 0.794, indicating that our correction produces a better fit. However, the discrepancy between our original RI and our new retrieved RI prompted us to look at other possible sources of error. In our previous report, we concluded that it was possible that the aerosols had

Dinneen | 66 solvent inclusions or that solvent was mixed in with the pigment while they were being measured by the CRD but had evaporated while under the high vacuum of the SEM. This was evident in the difference between the DMA selected size of the particles and what was measured by the SEM.84

Thus, to account for solvent effects within the generated pigment aerosols, we employed the

Maxwell-Garnett mixing rule. We chose to use Maxwell-Garnett because the simple volume weighted linear mixing rule, while often used for mixtures, was shown to not work well for absorbing aerosols.116, 150

Figure 5.3. Mie corrected data overlaid with Mie calculated traces using our previously reported complex RI of 1.92 + 0.039i and the fit retrieved complex RI of 1.5606 + 0.0128i as Qext vs size parameter. Error bars are simple standard deviation (1σ, N = 3).

In Table 5.1 we show how accounting for the solvent changes the effective refractive index.

We started by using our original pigment complex RI and the RI of acidic water, which produced a high merit value when its Maxwell-Garnett result of 1.743 + 0.0089i was compared to our Mie corrected Qext data. Next, we input the lower end of our reported real RI based on the error, 1.69.

This produced a much better fit with a merit value in the same order of magnitude of the one given by our corrected data compared to its retrieved RI. We then asked what Maxwell-Garnett input

Dinneen | 67 pigment RI would give an output that matches our retrieved RI, and so by trial and error, we found an input complex RI of 1.66 + 0.13i, where mixing was applied to the real and the imaginary portions of the RI.

Table 5.1. List of pigment complex RIs input in the Maxwell-Garnett equation with their calculated output RIs. Errors on n and k are simple standard deviation of the average of each selected size. Merit values are based on equation 5.4 for the output RIs when used to calculate Mie theory compared to Mie corrected experimentally obtained extinction efficiency data. Input pigment complex RI Output n n error Output k k error Merit value 1.92 + 0.039i 1.74 0.035 0.0089 0.0002 1.6 1.69 + 0.039i 1.59 0.021 0.0087 0.0002 0.070 1.66 + 0.13i 1.5606 0.0189 0.0128 0.0003 0.012

We next compared the complex RI of the Maxwell-Garnett mixing rule output (using the low end of our reported RI as the input) to our retrieved RI from our corrected data (Figure 5.4a).

Both Mie traces fit within the error bars of the corrected data, showing that the mixing rule provides a good estimate of the effective RI that was measured by the CRD. Despite correlating to a value on the lower end of our originally reported real RI, our data never exceeded a 20% difference with our retrieved value obtained via optical extinction measurements (Figure 5.4b), resulting in a better fit than our uncorrected data compared to calculated Mie values.84

Figure 5.4. a. Mie corrected data overlaid with Mie calculated traces as Qext vs size parameter using our retrieved complex RI of 1.5606 + 0.0128i and the output complex RI of 1.59 + 0.0087i found using the Maxwell-Garnett mixing rule. b. Percent difference of the Mie calculated traces of 1.5606 + 0.0128i and 1.59 + 0.0087i.

Dinneen | 68

Discussion:

Most aerosol generation methods create a polydisperse sample. This size variability should be constrained to interpret the optical measurements. Typically, DMA control is used to do this; however, doubly charged particles can contribute to the measured extinction to bias the data. It has been reported that an ammonium sulfate solution, at a concentration of 2.0 g L-1 and size selected at 200 nm using a DMA, produces a distribution where only 71% of the particles are singly charged.143 To reduce this error, a common procedure is to subtract out this contribution using values calculated from Mie theory.151-153 While this method uses an assumed RI out of necessity, our new method only starts with an assumed RI, then is iteratively processed until the RI used for corrections matches the retrieved RI. This removes any bias, whether low or high, that can be introduced through this step because of an unknown RI. In addition to the removal of doublet contribution to extinction, it has also been reported that increasing the number of diameters used in retrievals will decrease the range of n and k’s obtained, and that the ranges do not decrease significantly beyond the inclusion of six diameters.145 As our experiment contains seven diameters, we are within the range where retrieved values become more precise.

Though our extinction data was optically measured at 532 nm and our original n was measured at 589 nm, it is not expected that these values should differ significantly in this wavelength range.143 However, as k is based on a material’s absorbance, we used the squid pigment’s k at 532 nm.62 This wavelength region is especially important for the pigment, as its absorbance maximum occurs at 507 nm, resulting in a 94% difference in k at 589 and 532 nm.62

Though we are retrieving at 532 nm and using the k value at 532 nm, we can expect that the largest discrepancy in our retrieval will lie in the k because we are measuring and retrieving in a region close to an inflection point in the absorption profile. We find that indeed our percent difference in

Dinneen | 69 k between our originally reported k of 0.039 and our retrieved k of 0.13 is 110%, while the percent difference between our originally reported n of 1.92 and our retrieved n of 1.66 is only 15%.

Since the retrieval program seeks an RI that best matches the Qext data, the extinction values between what is measured and what is retrieved are very similar. However, when retrieving a higher k value than expected, the contribution of absorbance to extinction is higher than expected as well. Since absorbance is historically difficult to pull out of extinction data,145 the only metric we can use to determine if a k value is reasonable is by relating it back to a density value.83 Based on data published in a previous report and a k value of 0.13, we back-calculated a density value of

4.81 g cm-3.83 This is not a realistic value for an organic compound, so it is possible that error is introduced when using the Maxwell-Garnett mixing rule on k.150 This observation is not all too surprising considering the recent theoretical examination by Zarzana and coworkers, which reported that constraining the diameter range when retrieving RI values could return inaccurate complex RIs due to the fact that extinction cross-sections may converge in certain regions despite varied k values.145 They returned low n values for ammonium sulfate when they retrieved in the

300-450 nm particle diameter range. Approximately half of our measured data points are within this diameter range, so it is possible that our n value should be considered a lower limit, as it may contain an artifact of the diameters used in the retrieval. This could also explain our high k value, which could have been artificially inflated due to an increase in the absorption contribution to obtain the same extinction as if the n was larger. To retrieve more accurate n values additional diameters should be included in the 150-300 nm range, where extinction cross-sections do not converge with varied k values.145

Future work will also involve separate absorption measurements using a photoacoustic spectrometer or other techniques, such that contributions due to absorption can be used to more

Dinneen | 70 accurately define the k value for these pigments. In turn, this will also help narrow the currently reported range on n. Measuring additional particle diameters, especially in the 150-300 nm diameter region, may further improve the retrieval precision. Overall, these future experiments are aimed at further defining pigment n and k values.

This study was an exploration converging two seemingly disparate fields of atmospheric chemistry and bio-inspired materials, as CRD spectroscopy and retrieval programs (mainly used as an atmospheric technique) have been applied to a biological derived material. We advance the atmospheric chemistry field by using an innovative iterative approach tied to retrieved RI values to correct for doubly charged particles present in the optically interrogated sample. Using the

Maxwell-Garnett mixing rule and volume fractions calculated from the differences in our SEM and DMA data, we showed how solvent affects the optical properties of squid pigment aerosols.

With this combination of tools, we retrieved a complex RI of 1.66 + 0.13i that best fits our experimental data. While this n value is at the lower end of the error reported from the complex

RI extrapolated for the pigment solution,83 it is still high for biological materials,60, 90-91, 105 further supporting its important role in modulating the scattering and absorption of light within cephalopod dermal tissue.

Dinneen | 71

CHAPTER 6:

Conclusions

The unique photonic nature of cephalopods has garnered a lot of attention from scientists; and while they are widely studied, little was known about the molecular source of color in the granules. Our group specifically took an interest in how the fully expanded chromatophore organ was able to maintain a rich color despite being ~3 granules thick.99 We hypothesized that the pigment granules contain high refractive index biomolecules that promote light scattering while also contributing to absorbance.

To elucidate the camouflaging mechanism on a molecular level and model its optical properties, we measured the refractive index of the squid pigment. Starting with the squid, we dissected its skin and isolated the pigment granules from which we can extract the pigment. We have previously identified the pigments as xanthommatin and decarboxylated xanthommatin using mass spectrometry. Using mass fraction analysis over multiple solvent conditions, we were able to extrapolate the real portion of the refractive index (n = 1.92±0.23) of pigment in solution with an Abbe refractometer at 589 nm. To model the pigments’ optical properties, we also needed its imaginary refractive index (k). Using UV-vis spectroscopy, we calculated the pigment’s k over a wide range of wavelengths (0.014 at 589 nm) based on density, absorption, and concentration.

Using Mie theory to model the scattering and absorption cross-sections, we found that the pigment will scatter more attenuated light than absorb it. In addition, its high refractive index would bend

Dinneen | 72 light farther into the animals’ skin, possibly causing reabsorption and explaining how chromatophores maintain their color.

We have previously shown that pigment is a significant portion of the makeup of the granules, and while they have been identified, the relationship between the pigments’ composition and structural morphology is largely unknown. To further determine the pigments’ contribution to overall coloration, we generated squid pigment aerosols and size-selected them using a differential mobility analyzer, as most aerosols have size dependent optical properties. SEM analysis was performed on the sizes later measured by the CRD. Discrepancies between the DMA selected sizes and the measured SEM sizes were believed to be solvent remaining on the aerosols from the drying process. The size dependent optical extinctions of the aerosols were measured and found that they had good agreement (5-30% error) with theoretically extrapolated extinction cross sections using our measured refractive index.

Lastly, to further elucidate the discrepancies in the measured extinctions with the theoretical model, we developed an iterative correction procedure that reduced the bias of the initial guess used in Mie theory doublet corrections. Once the data was doublet corrected, we accounted for the difference in the selected sizes of the aerosols versus the SEM measured sizes by assuming it is the volume of solvent left on the aerosol. Using the Maxwell-Garnett mixing rule, we found a refractive index that would return the value we retrieved from our optical measurements. Ultimately, we retrieved a complex refractive index of 1.66 + 0.13i, where the real portion, n, describes a lower limit and the k, an upper limit.

The works described here to experimentally determine the refractive index of squid chromatophore pigment in solution and aerosol form have advanced the fields of solvent-generated Dinneen | 73 aerosols and bio-optics. With a novel approach to correct optical measurements of aerosols that have not been fully dried, retrieved refractive indices will be more accurate when not accounting for the volume fraction of solvent, which would lower its value. Further, with the optical properties of squid pigment characterized, it may inspire a new wave of bioinspired coloration mechanisms that take advantage of the high refractive index properties of these pigment molecules.

Dinneen | 74

LIST OF REFERENCES

Dinneen | 75

1. Denton, E. J., Review lecture: on the organization of reflecting surfaces in some marine animals. Philos Trans R Soc Lond B Biol Sci 1970, 258 (824), 285-313.

2. Parker, A. R., A geological history of reflecting optics. Journal of the Royal Society Interface 2005, 2 (2), 1-17.

3. Land, M. F., The physics and biology of animal reflectors. Progress in Biophysics and Molecular Biology 1972, 24, 75-106.

4. Kujala, S.; Mannila, A.; Karvonen, L.; Kieu, K.; Sun, Z., Natural Silk as a Photonics Component: a Study on Its Light Guiding and Nonlinear Optical Properties. Scientific Reports 2016, 6, 22358.

5. Monks, J. N.; Yan, B.; Hawkins, N.; Vollrath, F.; Wang, Z., Spider Silk: Mother Nature’s Bio-Superlens. Nano Letters 2016, 16 (9), 5842-5845.

6. Applegate, M. B.; Perotto, G.; Kaplan, D. L.; Omenetto, F. G., Biocompatible silk step- index optical waveguides. Biomedical Optics Express 2015, 6 (11), 4221-4227.

7. Yoshioka, S.; Matsuhana, B.; Tanaka, S.; Inouye, Y.; Oshima, N.; Kinoshita, S., Mechanism of variable structural colour in the neon tetra: quantitative evaluation of the Venetian blind model. Journal of The Royal Society Interface 2011, 8 (54), 56-66.

8. Sharma, V.; Crne, M.; Park, J. O.; Srinivasarao, M., Structural origin of circularly polarized iridescence in jeweled beetles. Science 2009, 325 (5939), 449-51.

9. Vukusic, P.; Sambles, J. R.; Lawrence, C. R.; Wootton, R. J., Quantified interference and diffraction in single Morpho butterfly scales. Proceedings of the Royal Society of London. Series B: Biological Sciences 1999, 266 (1427), 1403-1411.

10. Kinoshita, S.; Yoshioka, S.; Kawagoe, K., Mechanisms of structural colour in the Morpho butterfly: cooperation of regularity and irregularity in an iridescent scale. Proceedings of the Royal Society of London. Series B: Biological Sciences 2002, 269 (1499), 1417-1421.

11. Caffarri, S.; Broess, K.; Croce, R.; van Amerongen, H., Excitation Energy Transfer and Trapping in Higher Plant Photosystem II Complexes with Different Antenna Sizes. Biophysical Journal 2011, 100 (9), 2094-2103.

Dinneen | 76

12. Green, M. A.; Hishikawa, Y.; Dunlop, E. D.; Levi, D. H.; Hohl‐Ebinger, J.; Ho‐Baillie, A. W. Y., Solar cell efficiency tables (version 51). Progress in Photovoltaics: Research and Applications 2018, 26 (1), 3-12.

13. Stavenga, D. G.; Leertouwer, H. L.; Marshall, N. J.; Osorio, D., Dramatic colour changes in a bird of paradise caused by uniquely structured breast feather barbules. Proceedings of the Royal Society of London B: Biological Sciences 2011, 278 (1715), 2098-2104.

14. Wilts, B. D.; Michielsen, K.; De Raedt, H.; Stavenga, D. G., Sparkling feather reflections of a bird-of-paradise explained by finite-difference time-domain modeling. Proc. Natl. Acad. Sci 2014, 111 (12), 4363-4368.

15. Xiao, M.; Dhinojwala, A.; Shawkey, M., Nanostructural basis of rainbow-like iridescence in common bronzewing Phaps chalcoptera feathers. Optics Express 2014, 22 (12), 14625-14636.

16. Gur, D.; Leshem, B.; Pierantoni, M.; Farstey, V.; Oron, D.; Weiner, S.; Addadi, L., Structural Basis for the Brilliant Colors of the Sapphirinid Copepods. Journal of the American Chemical Society 2015, 137 (26), 8408-8411.

17. Tada, H.; Mann, S. E.; Miaoulis, I. N.; Wong, P. Y., Effects of a butterfly scale microstructure on the iridescent color observed at different angles. Optics Express 1999, 5 (4), 87-92.

18. Hecht, E., Optics. Addison-Wesley: 2002.

19. Kinoshita, S., Structural Colors in the Realm of Nature. World Scientific: 2008.

20. Tadepalli, S.; Slocik, J. M.; Gupta, M. K.; Naik, R. R.; Singamaneni, S., Bio-Optics and Bio-Inspired Optical Materials. Chemical Reviews 2017, 117 (20), 12705-12763.

21. Kinoshita, S.; Yoshioka, S.; Miyazaki, J., Physics of structural colors. Reports on Progress in Physics 2008, 71 (7), 076401.

22. Whitney, H. M.; Kolle, M.; Andrew, P.; Chittka, L.; Steiner, U.; Glover, B. J., Floral Iridescence, Produced by Diffractive Optics, Acts As a Cue for Animal Pollinators. Science 2009, 323 (5910), 130-133.

Dinneen | 77

23. Zhao, Y.; Xie, Z.; Gu, H.; Zhu, C.; Gu, Z., Bio-inspired variable structural color materials. Chemical Society Reviews 2012, 41 (8), 3297-3317.

24. Nam, H.; Song, K.; Ha, D.; Kim, T., Inkjet Printing Based Mono-layered Photonic Crystal Patterning for Anti-counterfeiting Structural Colors. Scientific Reports 2016, 6, 30885.

25. Parker, A. R.; Welch, V. L.; Driver, D.; Martini, N., Opal analogue discovered in a weevil. Nature 2003, 426, 786.

26. Mulliken, R. S., Overlap Integrals and Chemical Binding1. Journal of the American Chemical Society 1950, 72 (10), 4493-4503.

27. Romero, Y.; Martínez, A., Antiradical capacity of ommochromes. Journal of Molecular Modeling 2015, 21 (8), 220.

28. Liu, G. Y.; Nizet, V., Color me bad: microbial pigments as virulence factors. Trends Microbiol 2009, 17 (9), 406-13.

29. Solano, F., Photoprotection versus photodamage: updating an old but still unsolved controversy about melanin. Polymer International 2016, 65 (11), 1276-1287.

30. Michaela, B.; J., H. V., The Protective Role of Melanin Against UV Damage in Human Skin†. Photochemistry and Photobiology 2008, 84 (3), 539-549.

31. Bush, S. L.; Robison, B. H., Ink utilization by mesopelagic squid. Marine Biology 2007, 152 (3), 485-494.

32. Kiyomoto, R. K.; Poon, M.-C.; Bowen, S. T., Ommochrome pigments of the compound eyes of Artemia salina. Comparative Biochemistry and Physiology 1969, 29 (3), 975-984.

33. Nijhout, H. F., Ommochrome pigmentation of the linea and rosa seasonal forms of Precis coenia (Lepidoptera: Nymphalidae). Archives of Insect Biochemistry and Physiology 1997, 36 (3), 215-222.

34. Riou, M.; Christides, J. P., Cryptic color change in a crab spider (Misumena vatia): identification and quantification of precursors and ommochrome pigments by HPLC. J Chem Ecol 2010, 36 (4), 412-23.

Dinneen | 78

35. Aubourg, S. P.; Torres‐Arreola, W.; Trigo, M.; Ezquerra‐Brauer, J. M., Partial characterization of jumbo squid skin pigment extract and its antioxidant potential in a marine oil system. European Journal of Lipid Science and Technology 2016, 118 (9), 1293-1304.

36. Branden, C. V. D.; Decleir, W., A Study Of The Chromatophore Pigments In The Skin Of The Cephalopod Sepia Officinalis L. . Biol. Jb. Dodonaea 1976, 44, 345-352.

37. Linzen, B., The Tryptophan → Ommochrome Pathway in Insects1. In Advances in Insect Physiology, J.E. Treherne, M. J. B.; Wigglesworth, V. B., Eds. Academic Press: 1974; Vol. Volume 10, pp 117-246.

38. Hase, S.; Wakamatsu, K.; Fujimoto, K.; Inaba, A.; Kobayashi, K.; Matsumoto, M.; Hoshi, M.; Negishi, S., Characterization of the pigment produced by the planarian, Dugesia ryukyuensis. Pigment Cell Research 2006, 19 (3), 248-249.

39. Insausti, T. C.; Casas, J., The functional morphology of color changing in a spider: development of ommochrome pigment granules. J Exp Biol 2008, 211 (Pt 5), 780-9.

40. Fossen, T.; Cabrita, L.; Andersen, O. M., Colour and stability of pure anthocyanins influenced by pH including the alkaline region. Food Chemistry 1998, 63 (4), 435-440.

41. Pelletier, J. C., J. , Notice sur la matière verte des feuilles. Journal de Pharmacie 1817, 3, 486-491.

42. Holm, R. H.; Kennepohl, P.; Solomon, E. I., Structural and Functional Aspects of Metal Sites in Biology. Chemical Reviews 1996, 96 (7), 2239-2314.

43. Figon, F.; Casas, J., Ommochromes in invertebrates: biochemistry and cell biology. Biological Reviews 0 (0).

44. Shipman, L. L.; Cotton, T. M.; Norris, J. R.; Katz, J. J., An analysis of the visible absorption spectrum of chlorophyll a monomer, dimer, and oligomers in solution. Journal of the American Chemical Society 1976, 98 (25), 8222-8230.

45. Sen Gupta, S.; Ghosh, M., In Vitro Antioxidative Evaluation of - and -Carotene, Isolated from Crude Palm Oil. 2013; Vol. 2013, p 351671.

Dinneen | 79

46. Basuki; Suyitno; Kristiawan, B., Absorbance and electrochemical properties of natural indigo dye. AIP Conference Proceedings 2018, 1931 (1), 030067.

47. Wasmeier, C.; Hume, A. N.; Bolasco, G.; Seabra, M. C., Melanosomes at a glance. Journal of Cell Science 2008, 121 (24), 3995-3999.

48. El-Naggar, N. E.; El-Ewasy, S. M., Bioproduction, characterization, anticancer and antioxidant activities of extracellular melanin pigment produced by newly isolated microbial cell factories Streptomyces glaucescens NEAE-H. Sci Rep 2017, 7, 42129.

49. Sawada, H.; Nakagoshi, M.; Mase, K.; Yamamoto, T., Occurrence of ommochrome- containing pigment granules in the central nervous system of the silkworm, Bombyx mori. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology 2000, 125 (3), 421-428.

50. Okuda, S.; Nishiyama, N.; Saito, H.; Katsuki, H., Hydrogen peroxide-mediated neuronal cell death induced by an endogenous neurotoxin, 3-hydroxykynurenine. Proc Natl Acad Sci U S A 1996, 93 (22), 12553-8.

51. Okuda, S.; Nishiyama, N.; Saito, H.; Katsuki, H., 3-Hydroxykynurenine, an endogenous oxidative stress generator, causes neuronal cell death with apoptotic features and region selectivity. J Neurochem 1998, 70 (1), 299-307.

52. Morin, S. A.; Shepherd, R. F.; Kwok, S. W.; Stokes, A. A.; Nemiroski, A.; Whitesides, G. M., Camouflage and Display for Soft Machines. Science 2012, 337 (6096), 828-832.

53. Yu, C.; Li, Y.; Zhang, X.; Huang, X.; Malyarchuk, V.; Wang, S.; Shi, Y.; Gao, L.; Su, Y.; Zhang, Y.; Xu, H.; Hanlon, R. T.; Huang, Y.; Rogers, J. A., Adaptive optoelectronic camouflage systems with designs inspired by cephalopod skins. Proceedings of the National Academy of Sciences 2014, 111 (36), 12998-13003.

54. Wang, Q.; Gossweiler, G. R.; Craig, S. L.; Zhao, X., Cephalopod-inspired design of electro-mechano-chemically responsive elastomers for on-demand fluorescent patterning. Nature Communications 2014, 5, 4899.

55. Chou, H.-H.; Nguyen, A.; Chortos, A.; To, J. W. F.; Lu, C.; Mei, J.; Kurosawa, T.; Bae, W.-G.; Tok, J. B. H.; Bao, Z., A chameleon-inspired stretchable electronic skin with interactive colour changing controlled by tactile sensing. Nature Communications 2015, 6, 8011.

Dinneen | 80

56. Kreit, E.; Mäthger, L. M.; Hanlon, R. T.; Dennis, P. B.; Naik, R. R.; Forsythe, E.; Heikenfeld, J., Biological versus electronic adaptive coloration: how can one inform the other? J R Soc Interface: 2013; Vol. 10.

57. Sutherland, R. L.; Mäthger, L. M.; Hanlon, R. T.; Urbas, A. M.; Stone, M. O., Cephalopod coloration model. I. Squid chromatophores and iridophores. Journal of the Optical Society of America A 2008, 25 (3), 588-599.

58. Mäthger, L. M.; Denton, E. J.; Marshall, N. J.; Hanlon, R. T., Mechanisms and behavioural functions of structural coloration in cephalopods. Journal of The Royal Society Interface 2009, 6 (Suppl 2), S149-S163.

59. Crookes, W. J.; Ding, L. L.; Huang, Q. L.; Kimbell, J. R.; Horwitz, J.; McFall-Ngai, M. J., Reflectins: the unusual proteins of squid reflective tissues. Science 2004, 303 (5655), 235-8.

60. Kramer, R. M.; Crookes-Goodson, W. J.; Naik, R. R., The self-organizing properties of squid reflectin protein. Nat Mater 2007, 6 (7), 533-8.

61. Sutherland, R. L.; Mäthger, L. M.; Hanlon, R. T.; Urbas, A. M.; Stone, M. O., Cephalopod coloration model. II. Multiple layer skin effects. Journal of the Optical Society of America A 2008, 25 (8), 2044-2054.

62. Williams, T. L.; DiBona, C. W.; Dinneen, S. R.; Jones Labadie, S. F.; Chu, F.; Deravi, L. F., Contributions of Phenoxazone-Based Pigments to the Structure and Function of Nanostructured Granules in Squid Chromatophores. Langmuir 2016, 32 (15), 3754-3759.

63. Florey, E., Ultrastructure and function of cephalopod chromatophores. Am Zool 1969, 9 (2), 429-42.

64. Cloney, R. A.; Brocco, S. L., Chromatophore Organs, Reflector Cells, Iridocytes and Leucophores in Cephalopods. American Zoologist 1983, 23 (3), 581-592.

65. Williams, T. L.; DiBona, C. W.; Dinneen, S. R.; Labadie, S. F.; Chu, F.; Deravi, L. F., Contributions of Phenoxazone-Based Pigments to the Structure and Function of Nanostructured Granules in Squid Chromatophores. Langmuir 2016, 32 (15), 3754-9.

66. DiBona, C. W.; Williams, T. L.; Dinneen, S. R.; Jones Labadie, S. F.; Deravi, L. F., A Method for Extracting Pigments from Squid Doryteuthis pealeii. J. Vis. Exp. 2016, 117 (117), e54803.

Dinneen | 81

67. Ye, B.-F.; Zhao, Y.-J.; Li, T.-T.; Xie, Z.-Y.; Gu, Z.-Z., Aptamer-based suspension array indexed by structural color and shape. Journal of Materials Chemistry 2011, 21 (46), 18659- 18664.

68. Vinayak, R.; Sonia, M.; G., C. O.; A., G. A.; Manuel, M.; D., V. O., Synthesis of Light‐ Diffracting Assemblies from Microspheres and Nanoparticles in Droplets on a Superhydrophobic Surface. Advanced Materials 2008, 20 (22), 4263-4268.

69. Millman, J. R.; Bhatt, K. H.; Prevo, B. G.; Velev, O. D., Anisotropic particle synthesis in dielectrophoretically controlled microdroplet reactors. Nature Materials 2004, 4, 98.

70. Ge, J.; Lee, H.; He, L.; Kim, J.; Lu, Z.; Kim, H.; Goebl, J.; Kwon, S.; Yin, Y., Magnetochromatic Microspheres: Rotating Photonic Crystals. Journal of the American Chemical Society 2009, 131 (43), 15687-15694.

71. Zhu, J.; Hsu, C.-M.; Yu, Z.; Fan, S.; Cui, Y., Nanodome Solar Cells with Efficient Light Management and Self-Cleaning. Nano Letters 2010, 10 (6), 1979-1984.

72. Hiller, J. A.; Mendelsohn, J. D.; Rubner, M. F., Reversibly erasable nanoporous antireflection coatings from polyelectrolyte multilayers. Nature Materials 2002, 1, 59.

73. Shin‐Hyun, K.; Seog‐Jin, J.; Chan, J. W.; Sung, P. H.; Seung‐Man, Y., Optofluidic Synthesis of Electroresponsive Photonic Janus Balls with Isotropic Structural Colors. Advanced Materials 2008, 20 (21), 4129-4134.

74. Junhoi, K.; Younghoon, S.; Le, H.; Hyoki, K.; Howon, L.; Wook, P.; Yadong, Y.; Sunghoon, K., Real‐Time Optofluidic Synthesis of Magnetochromatic Microspheres for Reversible Structural Color Patterning. Small 2011, 7 (9), 1163-1168.

75. Cun, Z.; Wenyu, X.; Liangshui, C.; Wendong, Z.; Hua, X.; Zhong‐Ze, G., Magnetochromatic Microcapsule Arrays for Displays. Advanced Functional Materials 2011, 21 (11), 2043-2048.

76. Xiao, M.; Li, Y.; Allen, M. C.; Deheyn, D. D.; Yue, X.; Zhao, J.; Gianneschi, N. C.; Shawkey, M. D.; Dhinojwala, A., Bio-Inspired Structural Colors Produced via Self-Assembly of Synthetic Melanin Nanoparticles. ACS Nano 2015, 9 (5), 5454-5460.

77. Wang, J.; Wang, L.; Song, Y.; Jiang, L., Patterned photonic crystals fabricated by inkjet printing. Journal of Materials Chemistry C 2013, 1 (38), 6048-6058.

Dinneen | 82

78. Sowade, E.; Blaudeck, T.; Baumann, R. R., Inkjet Printing of Colloidal Nanospheres: Engineering the Evaporation-Driven Self-Assembly Process to Form Defined Layer Morphologies. Nanoscale Res Lett 2015, 10 (1), 362.

79. Vigneron, J. P.; Pasteels, J. M.; Windsor, D. M.; Vértesy, Z.; Rassart, M.; Seldrum, T.; Dumont, J.; Deparis, O.; Lousse, V.; Biró, L. P.; Ertz, D.; Welch, V., Switchable reflector in the Panamanian tortoise beetle Charidotella egregia (Chrysomelidae: Cassidinae). Physical Review E 2007, 76 (3), 031907.

80. Bai, L.; Xie, Z.; Wang, W.; Yuan, C.; Zhao, Y.; Mu, Z.; Zhong, Q.; Gu, Z., Bio-Inspired Vapor-Responsive Colloidal Photonic Crystal Patterns by Inkjet Printing. ACS Nano 2014, 8 (11), 11094-11100.

81. Phan, L.; Ordinario, D. D.; Karshalev, E.; Walkup Iv, W. G.; Shenk, M. A.; Gorodetsky, A. A., Infrared invisibility stickers inspired by cephalopods. Journal of Materials Chemistry C 2015, 3 (25), 6493-6498.

82. Kumar, A.; Osgood, R. M.; Dinneen, S. R.; Koker, B. D.; Pang, R.; Deravi, L. F., Natural Light‐Scattering Nanoparticles Enable Visible through Short‐Wave Infrared Color Modulation. Advanced Optical Materials 2018, 6 (8), 1701369.

83. Dinneen, S. R.; Osgood, R. M.; Greenslade, M. E.; Deravi, L. F., Color Richness in Cephalopod Chromatophores Originating from High Refractive Index Biomolecules. Journal of Physical Chemistry Letters 2017, 8 (1), 313-317.

84. Dinneen, S. R.; Greenslade, M. E.; Deravi, L. F., Optical extinction of size-controlled aerosols generated from squid chromatophore pigments. APL Materials 2017, 5 (10), 104802.

85. Dinneen, S. R.; Deravi, L. F.; Greenslade, M. E., An iterative correction approach used to retrieve the refractive index of squid pigment aerosols. Journal of Optics 2018, 20 (3), 034003.

86. Engineers, N. B. C., The Complete book on Natural Dyes & Pigments. Asia Pacific Business Press: 2005.

87. Futahashi, R.; Kurita, R.; Mano, H.; Fukatsu, T., Redox alters yellow dragonflies into red. Proceedings of the National Academy of Sciences 2012, 109 (31), 12626-12631.

88. Messenger, J. B., Cephalopod chromatophores: neurobiology and natural history. Biological Reviews 2001, 76 (4), 473-528.

Dinneen | 83

89. Packard, A., Size and distribution of chromatophores during post-embryonic development in cephalopods. Vie et Milieu 1985, 35.

90. Ghoshal, A.; DeMartini, D. G.; Eck, E.; Morse, D. E., Experimental determination of refractive index of condensed reflectin in squid iridocytes. J R Soc Interface 2014, 11 (95), 20140106.

91. Sweeney, A. M.; Des Marais, D. L.; Ban, Y. E.; Johnsen, S., Evolution of graded refractive index in squid lenses. J R Soc Interface 2007, 4 (15), 685-98.

92. Iglesias, M.; Orge, B.; Tojo, J., Refractive indices, densities and excess properties on mixing of the systems acetone + methanol + water and acetone + methanol + 1-butanol at 298.15 K. Fluid Phase Equilibria 1996, 126 (2), 203-223.

93. Marker, A. F. H., The use of acetone and methanol in the estimation of chlorophyll in the presence of phaeophytin. Freshwater Biology 1972, 2 (4), 361-385.

94. Willstätter, R., Zur Kenntniss der Zusammensetzung des Chlorophylls. Annalen der Chemie 1906, 350, 48-82.

95. Jimbow, K.; Quevedo, W. C.; Fitzpatrick, T. B.; Szabo, G., Some Aspects Of Melanin Biology: 1950–1975. Journal of Investigative Dermatology 1976, 67 (1), 72-89.

96. Hearing, V. J.; Tsukamoto, K., Enzymatic control of pigmentation in mammals. FASEB 1991, 5 (14), 2902-9.

97. Prota, G.; Ortonne, J. P.; Voulot, C.; Khatchadourian, C.; Nardi, G.; Palumbo, A., Occurrence and properties of tyrosinase in the ejected ink of cephalopods. Comparative Biochemistry and Physiology Part B: Comparative Biochemistry 1981, 68 (3), 415-419.

98. Sugumaran, M., Unified Mechanism for Sclerotization of Insect Cuticle. In Advances in Insect Physiology, Evans, P. D., Ed. Academic Press: 1998; Vol. Volume 27, pp 229-334.

99. Deravi, L. F.; Magyar, A. P.; Sheehy, S. P.; Bell, G. R. R.; Mathger, L. M.; Senft, S. L.; Wardill, T. J.; Lane, W. S.; Kuzirian, A. M.; Hanlon, R. T.; Hu, E. L.; Parker, K. K., The structure-function relationships of a natural nanoscale photonic device in cuttlefish chromatophores. Journal of the Royal Society Interface 2014, 11 (93).

Dinneen | 84

100. Bell, G. R. R.; Kuzirian, A. M.; Senft, S. L.; Mäthger, L. M.; Wardill, T. J.; Hanlon, R. T., Chromatophore radial muscle fibers anchor in flexible squid skin. Invertebr. Biol. 2013, 132 (2), 120-132.

101. Taylor, S. D.; Czarnecki, J.; Masliyah, J., Refractive index measurements of diluted bitumen solutions. Fuel 2001, 80 (14), 2013-2018.

102. Szewczyk, J.; Sangwal, K., Density, Refractive-Index, And Specific Refraction Of Aqueous Lithium Iodate Solutions. Journal of Chemical and Engineering Data 1988, 33 (4), 418-423.

103. Pollock, J. J. Optical characterization techniques for common components of aerosols. University of New Hampshire, University of New Hampshire, 2009.

104. Dean, J. A., Lange's Handbook of Chemistry. 13 ed.; McGraw-Hill: New York City, 1985.

105. Stavenga, D. G.; Leertouwer, H. L.; Osorio, D. C.; Wilts, B. D., High refractive index of melanin in shiny occipital feathers of a bird of paradise. Light Sci Appl 2015, 4, e243.

106. Eliason, C. M.; Bitton, P.-P.; Shawkey, M. D., How hollow melanosomes affect iridescent colour production in birds. Proceedings of the Royal Society B: Biological Sciences 2013, 280 (1767).

107. Sun, H.; Biedermann, L.; Bond, T. C., Color of brown carbon: A model for ultraviolet and visible light absorption by organic carbon aerosol. Geophysical Research Letters 2007, 34 (17), L17813.

108. Chen, Y.; Bond, T. C., Light absorption by organic carbon from wood combustion. Atmos. Chem. Phys. 2010, 10 (4), 1773-1787.

109. Yaws, C. L., Thermophysical Properties of Chemicals and Hydrocarbons. 2 ed.; Gulf Professional Publishing: Houston, 2014.

110. Paulus, E. F.; Leusen, F. J. J.; Schmidt, M. U., Crystal structures of quinacridones. CrystEngComm. 2007, 9, 131-143.

111. Sanders, J. V., Colour of Precious Opal. Nature 1964, 204 (4964), 1151-1153.

Dinneen | 85

112. Mäthger, L. M.; Senft, S. L.; Gao, M.; Karaveli, S.; Bell, G. R. R.; Zia, R.; Kuzirian, A. M.; Dennis, P. B.; Crookes-Goodson, W. J.; Naik, R. R.; Kattawar, G. W.; Hanlon, R. T., Bright White Scattering from Protein Spheres in Color Changing, Flexible Cuttlefish Skin. Advanced Functional Materials 2013, 23 (32), 3980-3989.

113. Kim, H.; Barkey, B.; Paulson, S. E., Real refractive indices of alpha- and beta-pinene and toluene secondary organic aerosols generated from ozonolysis and photo-oxidation. Journal of Geophysical Research-Atmospheres 2010, 115.

114. Attwood, A. R.; Greenslade, M. E., Optical Properties and Associated Hygroscopicity of Clay Aerosols. Aerosol Sci. Tech. 2011, 45 (11), 1350-1359.

115. Attwood, A. R.; Greenslade, M. E., Deliquescence Behavior of Internally Mixed Clay and Salt Aerosols by Optical Extinction Measurements. J. Phys. Chem. A 2012, 116 (18), 4518- 4527.

116. Abo Riziq, A.; Erlick, C.; Dinar, E.; Rudich, Y., Optical properties of absorbing and non- absorbing aerosols retrieved by cavity ring down (CRD) spectroscopy. Atmospheric Chemistry and Physics 2007, 7 (6), 1523-1536.

117. Valsaraj, K. T.; Kommalapati, R. R., Atmospheric Aerosols. American Chemical Society: Washington, DC, 2009; Vol. 1005, p 216.

118. Kanakidou, M.; Seinfeld, J. H.; Pandis, S. N.; Barnes, I.; Dentener, F. J.; Facchini, M. C.; Van Dingenen, R.; Ervens, B.; Nenes, A.; Nielsen, C. J.; Swietlicki, E.; Putaud, J. P.; Balkanski, Y.; Fuzzi, S.; Horth, J.; Moortgat, G. K.; Winterhalter, R.; Myhre, C. E. L.; Tsigaridis, K.; Vignati, E.; Stephanou, E. G.; Wilson, J., Organic aerosol and global climate modelling: a review. Atmos. Chem. Phys. 2005, 5 (4), 1053-1123.

119. May, K. R., The collison nebulizer: Description, performance and application. J. Aerosol Sci. 1973, 4 (3), 235-243.

120. McMurry, P. H., A review of atmospheric aerosol measurements. Atmospheric Environment 2000, 34 (12–14), 1959-1999.

121. Baron, P. A.; Willeke, K., Aerosol Measurement: Principles, Techniques, and Applications. Wiley: Hoboken, 2005.

Dinneen | 86

122. Schneider, C. A.; Rasband, W. S.; Eliceiri, K. W., NIH Image to ImageJ: 25 years of image analysis. Nat. Meth. 2012, 9 (7), 671-675.

123. Wiedensohler, A., An approximation of the bipolar charge distribution for particles in the submicron size range. J. Aerosol Sci. 1988, 19 (3), 387-389.

124. Hagen, D. E.; Alofs, D. J., Linear Inversion Method to Obtain Aerosol Size Distributions from Measurements with a Differential Mobility Analyzer. Aerosol Sci. Tech. 1983, 2 (4), 465- 475.

125. Baynard, T.; Lovejoy, E. R.; Pettersson, A.; Brown, S. S.; Lack, D.; Osthoff, H.; Massoli, P.; Ciciora, S.; Dube, W. P.; Ravishankara, A. R., Design and Application of a Pulsed Cavity Ring-Down Aerosol Extinction Spectrometer for Field Measurements. Aerosol Sci. Tech. 2007, 41 (4), 447-462.

126. Pettersson, A.; Lovejoy, E. R.; Brock, C. A.; Brown, S. S.; Ravishankara, A. R., Measurement of aerosol optical extinction at 532nm with pulsed cavity ring down spectroscopy. J. Aerosol Sci. 2004, 35 (8), 995-1011.

127. Miles, R. E. H.; Rudic, S.; Orr-Ewing, A. J.; Reid, J. P., Measurements of the wavelength dependent extinction of aerosols by cavity ring down spectroscopy. Phys. Chem. Chem. Phys. 2010, 12 (15), 3914-3920.

128. Zerefos, C. S.; Gerogiannis, V. T.; Balis, D.; Zerefos, S. C.; Kazantzidis, A., Atmospheric effects of volcanic eruptions as seen by famous artists and depicted in their paintings. Atmospheric Chemistry and Physics 2007, 7 (15), 4027-4042.

129. Zerefos, C. S.; Tetsis, P.; Kazantzidis, A.; Amiridis, V.; Zerefos, S. C.; Luterbacher, J.; Eleftheratos, K.; Gerasopoulos, E.; Kazadzis, S.; Papayannis, A., Further evidence of important environmental information content in red-to-green ratios as depicted in paintings by great masters. Atmospheric Chemistry and Physics 2014, 14 (6), 2987-3015.

130. Ensor, D. S., Aerosol Science and Technology: History and Reviews. RTI International: Research Triangle Park, NC, 2011.

131. Liu, P. F.; Abdelmalki, N.; Hung, H. M.; Wang, Y.; Brune, W. H.; Martin, S. T., Ultraviolet and visible complex refractive indices of secondary organic material produced by photooxidation of the aromatic compounds toluene and m-xylene. Atmospheric Chemistry and Physics 2015, 15 (3), 1435-1446.

Dinneen | 87

132. Liu, P. F.; Zhang, Y.; Martin, S. T., Complex Refractive Indices of Thin Films of Secondary Organic Materials by Spectroscopic Ellipsometry from 220 to 1200 nm. Environmental Science & Technology 2013, 47 (23), 13594-13601.

133. Bluvshtein, N.; Flores, J. M.; Riziq, A. A.; Rudich, Y., An Approach for Faster Retrieval of Aerosols' Complex Refractive Index Using Cavity Ring-Down Spectroscopy. Aerosol Science and Technology 2012, 46 (10), 1140-1150.

134. Dinar, E.; Riziq, A. A.; Spindler, C.; Erlick, C.; Kiss, G.; Rudich, Y., The complex refractive index of atmospheric and model humic-like substances (HULIS) retrieved by a cavity ring down aerosol spectrometer (CRD-AS). Faraday Discussions 2008, 137, 279-295.

135. Erlick, C.; Haspel, M.; Rudich, Y., Simultaneous retrieval of the complex refractive indices of the core and shell of coated aerosol particles from extinction measurements using simulated annealing. Applied Optics 2011, 50 (22), 4393-4402.

136. Flores, J. M.; Washenfelder, R. A.; Adler, G.; Lee, H. J.; Segev, L.; Laskin, J.; Laskin, A.; Nizkorodov, S. A.; Brown, S. S.; Rudich, Y., Complex refractive indices in the near- ultraviolet spectral region of biogenic secondary organic aerosol aged with ammonia. Physical Chemistry Chemical Physics 2014, 16 (22), 10629-10642.

137. Freedman, M. A.; Hasenkopf, C. A.; Beaver, M. R.; Tolbert, M. A., Optical Properties of Internally Mixed Aerosol Particles Composed of Dicarboxylic Acids and Ammonium Sulfate. Journal of Physical Chemistry A 2009, 113 (48), 13584-13592.

138. Lack, D. A.; Lovejoy, E. R.; Baynard, T.; Pettersson, A.; Ravishankara, A. R., Aerosol absorption measurement using photoacoustic spectroscopy: Sensitivity, calibration, and uncertainty developments. Aerosol Science and Technology 2006, 40 (9), 697-708.

139. Lang-Yona, M.; Rudich, Y.; Segre, E.; Dinar, E.; Abo-Riziq, A., Complex Refractive Indices of Aerosols Retrieved by Continuous Wave-Cavity Ring Down Aerosol Spectrometer. Analytical Chemistry 2009, 81 (5), 1762-1769.

140. Mack, L. A.; Levin, E. J. T.; Kreidenweis, S. M.; Obrist, D.; Moosmuller, H.; Lewis, K. A.; Arnott, W. P.; McMeeking, G. R.; Sullivan, A. P.; Wold, C. E.; Hao, W. M.; Collett, J. L.; Malm, W. C., Optical closure experiments for biomass smoke aerosols. Atmospheric Chemistry and Physics 2010, 10 (18), 9017-9026.

141. Miles, R. E. H.; Rudic, S.; Orr-Ewing, A. J.; Reid, J. P., Influence of Uncertainties in the Diameter and Refractive Index of Calibration Polystyrene Beads on the Retrieval of Aerosol

Dinneen | 88

Optical Properties Using Cavity Ring Down Spectroscopy. Journal of Physical Chemistry A 2010, 114 (26), 7077-7084.

142. Spindler, C.; Abo Riziq, A.; Rudich, Y., Retrieval of aerosol complex refractive index by combining cavity ring down aerosol spectrometer measurements with full size distribution information. Aerosol Science and Technology 2007, 41, 1011-1017.

143. Washenfelder, R. A.; Flores, J. M.; Brock, C. A.; Brown, S. S.; Rudich, Y., Broadband measurements of aerosol extinction in the ultraviolet spectral region. Atmospheric Measurement Techniques 2013, 6 (4), 861-877.

144. Toole, J. R.; Renbaum-Wolff, L.; Smith, G. D., A Calibration Technique for Improving Refractive Index Retrieval from Aerosol Cavity Ring-Down Spectroscopy. Aerosol Science and Technology 2013, 47 (9), 955-965.

145. Zarzana, K. J.; Cappa, C. D.; Tolbert, M. A., Sensitivity of Aerosol Refractive Index Retrievals Using Optical Spectroscopy. Aerosol Science and Technology 2014, 48 (11), 1133- 1144.

146. Weiss, M. R., Floral colour changes as cues for pollinators. Nature 1991, 354 (6350), 227-229.

147. Yan, J.; Wang, G.; Sui, Y.; Wang, M.; Zhang, L., Pollinator responses to floral colour change, nectar, and scent promote reproductive fitness in Quisqualis indica (Combretaceae). Sci Rep 2016, 6, 24408.

148. Bonansea, M. I.; Heit, C.; Vaira, M., Pigment composition of the bright skin in the poison toad Melanophryniscus rubriventris (Anura: Bufonidae) from Argentina. Salamandra 2017, 53 (1), 142-147.

149. Bohren, C. F.; Huffman, D. R., Absorption and Scattering of Light by Small Particles. Wiley-VCH: Weinheim, Germany, 1998.

150. Ouimette, J. R.; Flagan, R. C., The extinction coefficient of multicomponent aerosols. Atmospheric Environment 1982, 16 (10), 2405-2419.

151. Fierz-Schmidhauser, R.; Zieger, P.; Wehrle, G.; Jefferson, A.; Ogren, J. A.; Baltensperger, U.; Weingartner, E., Measurement of relative humidity dependent light scattering of aerosols. Atmos. Meas. Tech. 2010, 3 (1), 39-50.

Dinneen | 89

152. Cappa, C. D.; Che, D. L.; Kessler, S. H.; Kroll, J. H.; Wilson, K. R., Variations in organic aerosol optical and hygroscopic properties upon heterogeneous OH oxidation. Journal of Geophysical Research-Atmospheres 2011, 116 (D15), D15204.

153. Flores, J. M.; Bar-Or, R. Z.; Bluvshtein, N.; Abo-Riziq, A.; Kostinski, A.; Borrmann, S.; Koren, I.; Koren, I.; Rudich, Y., Absorbing aerosols at high relative humidity: linking hygroscopic growth to optical properties. Atmos. Chem. Phys. 2012, 12 (12), 5511-5521.

Dinneen | 90