Signature Redacted Signature of Author

Mechanical Reinforcement of Hydrogels via Bio-Inspired Mineralization

Olivia Saouaf

Submitted to the Department of Material Science and Engineering in Partial Fulfillment of the requirements of the Degree of

Bachelor of Science

at the

Massachusetts Institute of Technology

June 2019

The author hereby grants to MIT permission to reproduce and to reproduce and distribute publicly paper and electronic copies of this thesis document in whole or in part in any medium known or hereafter created.

Signature redacted Signature of Author...... Olivia Saouaf Department of Material Science and Engineering / 47 May 6, 2019

C ertified b y ...... Signature redacted Niels Holten-Andersen Associate Professor of Materials Science and Engineering Thesis Supervisor

A ccepted by...... redacted.Signature 4 --- Juejun H u MASSACHUSETS INSTITUTE Professor of Materials Science and Engineering OF TECHNOWOGY Chair, Undergraduate Committee JUN 1 1 2019 IBRARIES ARCHIVES 1 Mechanical Reinforcement of Hydrogels via Bio-Inspired Mineralization

Olivia Saouaf

Submitted to the Department of Materials Science and Engineering on May 6, 2019 in Partial Fulfillment of the Requirements for the Degree of Bachelor of Science in Material Science and Engineering

Abstract

Nanocomposites made of polymer networks and mineral particles lend great mechanical integrity to biological materials. This study aims to imitate these natural materials by creating a hydrogel mineralized with magnetite particles. We create a hydrogel of polyallylamine crosslinked via tannic acid molecules. Crosslinking is dependent upon pH as well as amounts of periodate and tannic acid. The addition of greater amounts of tannic acid and periodate at higher pH creates a more strongly crosslinked network, shown through rheological measurements as the gel's shear modulus increases. Upon mineralization, a 102-103 order of magnitude increase in shear modulus occurs. This work elucidates a method for nanocomposite hydrogel synthesis that creates a mechanically strong biocompatible material for future applications in bio- interfacing technology and drug delivery.

Thesis Supervisor: Niels Holten-Andersen Title: Associate Professor of Materials Science and Engineering

Acknowledgements

I would like to thank Professor Niels Holten-Andersen for his guidance and during this work. I would also like to thank the Laboratory for Bio-Inspired Interfaces members for their support of my research, especially Jake Song for his mentorship and Sungjin Kim for mineralization of gels.

2 Table of Contents

L ist o f F ig ures...... 4 L ist of T ables...... 5 1. In tro du ction ...... 6 2 . B ackground ...... 8 2.1. H ydrogel N etw ork...... 8 2.2. Polyallylam ine...... 8 2 .3. T annic A cid ...... 9 2.4. P eriodate...... 10 2.5. M agnetite M ineralization ...... 11 2 .6 . R h eo logy ...... 12 3. M aterials and M ethods...... 13 3.1. M aterials ...... 13 3.2. E quipm ent...... 13 3 .3 . P rocedures...... 13 3.4. T estin g ...... 14 4. R esults and D iscussion ...... 16 4.1. Effects of pH on Gel Formation...... 17 4.2. Periodate Amount Variation...... 18 4.3. Tannic Acid Amount Variation...... 20 4.4. Mineralization With and Without Post Treatment...... 21 5. C onclusion ...... 23 6 . F uture W ork s...... 24 R eferences...... 25 A ppendix ...... 27

3 Figure List

Figure 1: A mineralized hydrogel network ...... 6 Figure 2: Molecular structure of polyallylamine...... 8 Figure 3: Structure of tannic acid molecule...... 9 Figure 4: Diagram of pH influenced tannic acid bonding and conjugation...... 10 Figure 5: Sodium periodate structure...... 10 Figure 6: Gel formation and mineralization process...... 11 Figure 7: Example rheological data...... 16 Figure 8: Samples made with different pH...... 17 Figure 9: A pH variation frequency sweep...... 18 Figure 10: Gels formed with 2x and 4x original sodium periodate amount...... 18 Figure 11: Periodate variation...... 19 Figure 12: Tannic acid variation...... 20 Figure 13: A post treated 4x NalO4 2x TA gel...... 21 Figure 14: Post Treatment and Mineralization, 4x Periodate, 2x Tannic Acid...... 22 Figure A 1: pH V ariation...... 27 Figure A 2: Periodate Variation...... 27 Figure A 3: Tannic Acid Variation...... 28 Figure A 4: Post Treatment...... 28 Figure A 4: Mineralization...... 29 Figure B 1: SEM image of mineralized hydrogel, no post treatment...... 30

4 List of Tables

Table 1: Component ratios for hydrogel synthesis ...... 14 Table C. 1: Optimized hydrogel recipe...... 31

5 1. Introduction

Biology has created numerous nanocomposite materials consisting of mineral particles bound in a polymer matrix which exhibit great strength and toughness [1-3].

Materials like nacre [4] are extremely tough and are also made of cheap, readily available materials. Mimicking these materials gives the potential for new low cost synthetic materials with exceptional mechanical properties as well as biocompatibility. Of particular interest are mineralized hydrogels. Hydrogels consist of a crosslinked hydrophilic polymer network swollen with water [5]. Inorganic nanoparticles are then grown within the network, as in Wang et al. [6].

This study aims to imitate these natural materials by creating a hydrogel mineralized with magnetite particles. We fabricate a hydrogel of polyallylamine hydrochloride crosslinked via tannic acid molecules (Figure 1). Polyallylamine is a polymer known to be biocompatible [7] and iron, the metallic component of magnetite, is found naturally within the body. Tannic acid is a low cost, naturally derived crosslinker

[8]. With this system, we propose a low cost, biocompatible, and magnetically responsive mineralized hydrogel for use in bio-interfacing technology and drug delivery.

Figure 1. A mineralized hydrogel network consisting ofpolyallylamine chains (blue), tannic acidcrosslinks (green) and magnetite particles (red)

6 The system proposed can be manipulated in crosslinking density to allow for greater network stability, but must remain open enough for the growth of magnetite nanoparticles to be possible. Crosslinking is dependent upon pH, and amounts of periodate and tannic acid. The manipulation of pH and amounts of tannic acid and periodate will change the crosslinking density, which will be measureable by changes in the gel's shear modulus. After a gel is created, it will be mineralized in a bath of iron (III) ions and its shear modulus will be evaluated. Through this process of network adjustment and mineralization, this thesis will show an increase in mechanical strength of PAA hydrogels through the addition of magnetite nanoparticles.

7 2. Background

2.1 Hydrogel Network

A hydrogel network consists hydrophilic polymer chains, making up the bulk of the system, which are held together by chemical crosslinkers [5]. Crosslinks can be made in multiple ways- through covalent bonds with another molecule (as will be the case in this work), covalent bonds directly between chains, and ionic coordination with metallic ions, to name a few. The crosslinked network is swollen with water due to its hydrophilicity and the network bunds must be strong enough to hold the chains together while supporting the water. Synthetic hydrogels are of great interest because of their similarity to biological systems, which gives them great flexibility allows them to interact with the body [9].

2.2 Polyallylamine

Polyallylamine (PAA) is a polymer consisting of a carbon backbone with pendant aminomethyl groups on alternating carbon atoms as displayed in Figure 2 [10]. The amino groups make for potential binding sites, but the polymer does not form a gel without an added crosslinker. This makes for easy storage of PAA without risk of self- gelation. The polymer has been used in biological applications such as cell encapsulation

[7, 11]. These works suggest that PAA is a good candidate for biocompatible gels.

NH 2

Figure 2. Molecular structure ofpolyal/ylamine

8 I

2.3. Tannic Acid

Tannic acid (TA) is a polyphenol derived from plants. This compound is a specific form the class of tannins, which can be found in many plants [8]. Tannic acid comes from Tara pods and gallnuts. Other forms of tannins can be found in leaves, wood, and bark of many different plants. The name tannin comes from its use in preserving or

"tanning" animal hides into leather. In this work, tannic acid's ability to bind both metals and organic molecules is of great interest [8].

H H H HO HO/H H

HOH O OH

H H 0 HO H OH

HO O1' 0 OH o OH

HO. 0 0 :__ OH

HO OH O 0 OH

HO "OH OH Figure 3. Structure of tannic acid molecule, exhibiting its five-armed configuration

Previous works have demonstrated tannic acid's ability to form coordination complexes with metals and covalent bonds with organic ligands [12, 13, 14]. This ability arises from the large amount of pyrogallol groups present on tannic acid's five-armed structure (Figure 3). These groups can coordinate with a metal ion or covalently bind with polymer chains once deprotonated [15]. Tannic acid undergoes deprotonation as pH is increased [15]. This creates more reactive binding sites for metal ions or organic compounds and thus tannic acid becomes a more effective crosslinker at higher pH

(Figure 4). The degree of crosslinking in a tannic acid network can be tuned by adjusting pH.

9 a HO OH HO OHO OH b

NH o OH 2 0 0 OH

HOOHO HO 0 HHO HO HHO 0 OH OH

c HO OH OH OH -NH OH

RO RN ichaelrp addition R O OH 0 0

R R R d HO-01 pH HO pHt HOA HO OH HO OH - 0--- Fe HO OH O---Ee----O OH O-----O OH R- -6 HO - R- R0 b-

OH MOno R OH Bis R OH Tris R

Figure 4. Diagram ofpH influenced tannic acid bonding and conjugation. The pyrogallol groups are deprotonatedas pH increases and then can covalently bond with amino groups or chelate an iron ion [15]

2.4 Periodate

Periodate is an anion with a charge of -1. In this work periodate is used in the form of a salt, sodium periodate (NalO 4) as shown in Figure 5. Periodate is an oxidizer of tannic acid [16], allowing tannic acid to function more readily as a crosslinker once hydrogen is removed from hydroxyl groups. NaIO 4 degrades when exposed to light, so steps involving sodium periodate must be performed quickly and allowed to complete reaction in a dark environment. 0 Na' -:

Figure 5. Sodium periodatestructure, ions dissociate in water

10 2.5 Magnetite Mineralization

The final step in synthesis of our mechanically reinforced gels is mineralization with magnetite. This process utilizes tannic acid's ability to form catechol complexes with metal ions, as demonstrated in Figure 6. The already formed gel network still has phenol groups remaining open on tannic acid molecules. These sites can bind Fe3 ions from solution. Exposure to a bath of iron ions allows magnetite (Fe304) crystals to form.

a C

~ PAA *Non-Oxidized TA *Oxidized TA

3 0 Fe + Ion & Magnetite

Figure 6. Gelformation and mineralizationprocess (a) unconnected PAA chains (b) PAA chains still unconnected when TA is first added (c) TA is oxidized after exposure to periodateand raisedpH(d) oxidized TA covalently crosslinks PAA (e) an open site on a TA molecule chelates an iron ion (f) a magnetite particlegrows from the ion

Magnetite has a Mohs hardness of 5-6 [17] and crystals embedded in a hydrogel network have the potential to greatly increase the gel's strength. Additionally, the magnetite particles are magnetic. Magnetite is the most magnetic of minerals found

11 naturally on Earth [17]. Magnetic hydrogels have been pursued in previous works [18].

They are of interest because of their applications in drug delivery [19], flexible electronics, and biomedical engineering [20]. Magnetic hydrogels can be mechanically actuated without physical contact through the use of a magnetic field. With this ability, a hydrogel could release a drug upon magnetic stimulation only when it has reached the desired location in the body, or a robot could move without need for connecting wires.

Mineralization of hydrogels with magnetite gives them improved mechanical properties and new magnetic responsiveness.

2.6 Rheology

Rheology measurements study the flow of matter, whether solid, liquid, or somewhere in between as is the case with hydrogels. [21] To perform rheological measurements, a sample is subjected to a rotational force, resulting on some strain of the material [22]. How the material deforms and then returns to its initial state gives a measure of its mechanical properties. One type of test performed with a rheometer is a frequency sweep. In this test, the probe contacting the gel or "geometry" is oscillated at a range of frequencies. The shear storage and loss moduli are recorded. If the material is more solid like, the modulus will remain mostly constant over the frequency range. A higher average modulus indicates a stronger material that is less easily deformed.

12 3. Materials and Methods

3.1 Materials

Polyallylamine (Sigma Aldrich) is used as the network polymer, with tannic acid

(Sigma Aldrich, gallotannin) as the crosslinker. Sodium periodate (NaIO 4) solution provides periodate anions (104) which deprotonate pyrogallol groups on tannic acid.

Sodium hydroxide (NaOH) is used to raise pH of the system. A bath of FeCl3 donates

Fe3 ions for mineralization.

3.2 Equipment

For rheological frequency sweep data, an Anton Paar 302 Rheometer was used.

Force measurement error for the machine is +/- 0.001 Newtons and displacement measurement error is +/- 1 tm.

3.3 Procedures

The first step in creating our mechanically reinforced hydrogels is forming the polymer network of polyallylamine and tannic acid. First, a measured amount of PAA on the scale of milligrams is added to a vial. Water is then pipetted into the vial and the contents of the vial are mixed using a vortex machine. The water at this step dilutes the viscous PAA so that future components can mix more easily into the solution. Tannic acid powder is measured by weight and then deposited into the vial. More water is added to further hydrate the system and to rinse tannic acid powder from the sides of the vial, then the vial contents are mixed again. A 0.5M solution of sodium periodate (NalO 4) is added in small quantity. The solution is mixed and then deposited quickly, still a liquid, onto a piece of Parafilm. A IM NaOH solution is then added to raise gel pH and the components are gently mixed by hand until a gel is formed. The gel is then placed in a

13 container covered in aluminum foil to prevent exposure of periodate to light and allowed to set overnight. Gel formation component ratios are varied between samples to manipulate pH, tannic acid amount, and sodium periodate amount. The original network component ratios are shown in Table 1.

Component Polyallylamine Sodium Periodate Mol component per mol 0.041 0.625 tannic acid Mol monomer per tannic 13.16 0.125 acid ligand Table 1. Component ratiosfor hydrogel synthesis: ratios of tannic acid, polyallylamine, and sodium periodate in the base version of the gel before variation (Ix TA, Ix PAA)

If a post treatment is added, the procedure is as follows. The gel is placed in a larger vial filled with water and allowed to swell overnight. After swelling, the gel is placed in a solution containing tannic acid, equal in amount to the TA used to form the gel. Periodate (the same amount as used in gel formation) is then added to the vial and gently mixed. NaOH is added (same amount), mixed, and the gel is allowed to sit overnight. This process may be repeated for a second post treatment.

Mineralization takes place once the gel is formed. Whether post treated or not, the gel is allowed to swell fully in water. The gel is then placed in a bath of FeC 3 and allowed time to grow magnetite particles.

3.4 Testing

Mineralized and nonmineralized gels are tested via rheometer to find their shear moduli. A disk-shaped piece of gel is placed onto the rheometer stage. The geometry of the rheometer is lowered to contact the gel and the sample is trimmed to fit the surface of the geometry. The gel is kept in a humid environment at 25'C for the duration of the test.

A frequency sweep is performed on the gel, during which the sample is subjected to a

14 range of oscillation speeds. Shear storage and loss moduli are recorded and plotted against oscillation frequency.

15 4. Results and Discussion

To measure the strength of each gel's structure, rheological tests were performed using a frequency sweep at constant oscillation amplitude. The shear storage (G') and loss (G") moduli were recorded from each test. Gels with stronger network bonding exhibit higher shear moduli. Most gels measured, once a stable gel-forming system had been achieved, had relatively constant shear moduli with storage modulus greater than loss modulus. Constant moduli over a range of oscillation frequencies indicate that the gels are more solid-like in structure, and having G' > G" means that the gel can return close to its original position after deformation. Data throughout this section will be presented with storage moduli only for ease of reading because largely the same trends were observed in both sets of moduli. (For storage and loss plots, consult appendices.)

-M ONNOaMENSEEm M E6m '6im m m mm 13 100:

:, 10 D Storage Modulus

-0 3 Loss Modulus

0.10

0.01,, 0.1 0.5 1 5 10 50 100 Angular Frequency (rad/sec)

Figure 7. Example rheologicaldata from a pH 6.22 gel shows constant shear moduli until highfrequency and storage modulus (G') greaterthan loss modulus (G').

16 4.1 Effects of pH on Gel Formation

ab C d

Figure 8. Samples made with different pH (a) pH 1.72 sample remains liquid and light brown (b) pH 4.35 sample still liquid-like but more viscous, a darker brown color (c) pH 4.90 sample is a cohesive dark brown gel but breaks aparteasily (d) pH 6.22 dark brown gel with structuralintegrity

Manipulating the pH of the gel at formation allows the formation of different amounts of covalent crosslinking. Tannic acid's pyrogallol groups are oxidized when exposed to periodate and the amount of groups oxidized depends on pH. Oxidized groups can then bond with PAA amine groups. Figure 8 shows samples made at four different pH levels. The first sample had no NaOH added and remained a light brown liquid.

Subsequent gels became darker upon crosslinking and more solid-like. Figure 9 shows that with increasing pH, the storage modulus of the gels increases. The first sample created had no NaOH added, resulting in a pH of 1.72. This sample did not form a solid- like gel; it was mostly liquid in character as evidenced by its modulus changing throughout the frequency sweep. pH was increased in the next samples, which became more solid-like though with decreasing returns for amount of NaOH added. From this data, a pH of 6.22 was selected for further testing and of the system because it exhibited a high modulus while still remaining gel-like.

17 1000

A A AA A-A %.m 1 m pH 1.72 A pH 4.35 U) AtE 0.001 * pH 4.90 0 M ME.. age * pH 6.22 M No EE ME MMEMEM 10-6 0.1 0.5 1 5 10 50 100 Angular Frequency (rad/sec)

Figure 9. A pH Variationfrequency sweep shows that storage modulus increases with increasingpH. Samples with pH 1.72 and 4.35 display the liquid-like tendency of modulus changing with oscillationfrequency whereas gels with pH 4.90 and 6.22 have more solid-likeflat curves.

4.2 Periodate Amount Variation

2x NaIO 4 Pre-Swell 4x NalO 4 Pre-Swell

2x NalOA.woIen,

Figure 10. Gels formed with 2x and 4x originalsodium periodate amount before and after swelling with water. 2x gel is lighter in color and becomes more fragile when swollen

18 Periodate (NaIO 4) is the crosslinking agent in this system. It facilitates the oxidation of tannic acid's pyrogallol groups. The amount of NaIO 4 added was varied from a standard (Ix) of 1.6 mol TA per mol NaIO4, or 8.0 TA ligands per NaIO 4 molecule (TA is a 5-armed molecule, this calculation assumes one bonding site per arm).

It should not be assumed that each 104- oxidizes exactly one pyrogallol, or that each group on a tannic acid can bond with PAA (steric hindrance makes this unlikely). Gels were subjected to swelling in a water bath (Figure 10) and it was found that gels with 4x

NaIO 4 maintained structural stability best when swollen. Figure 11 shows that the shear modulus of the gel increases initially with increasing periodate amount, but reaches a peak with 2x and 4x the initial ratio. At 6x the original periodate amount, the gel undergoes a dramatic drop in modulus. This may be because the gel formed too quickly while mixing and as a result became more particulate and broken, therefore exhibiting a lower shear modulus in testing. From this apparent peak in gel forming ability, 2x and 4x formulations were chosen for further testing in the creation of a strengthened gel network.

1000-

10=- 10 mix0.x

*2x o 0.100- A 4X

4t v 6x

0.001 - V

0.1 0.5 1 5 10 50 100 Angular Frequency (rad/sec) Figure 11. Periodatevariation (unswollen gels) shows that storage modulus increases with increasingNaIO 4 until 6x

19 4.3 Tannic Acid Amount Variation

Tannic acid (TA) is the crosslinker for our polyallylamine network. The original

stoichiometry (lx) of the system had 1 TA ligand per 13.16 allylamine monomers. By

increasing this ratio, we expect to see a higher density of crosslinking throughout the gel.

Figure 12 shows a progression through 0.5, 1, 2, and 4 times the original amount of

tannic acid for gels using 4x the original amount, as determined by previous experiment.

The 0.5x sample was a liquid, and samples after that exhibited increasing moduli.

However, the gel with 4x TA exhibited stiff fracture upon mixing and handling and so

was not used further. These experiments showed the most strengthened gel of our system

resulted from 4x original periodate and 2x original tannic acid amounts.

S100 - E E E E E E . iiia iiaa S* 0.5x

0.1 m1x -o set +2x 0 10-4 A 4x

SI I r I 0.1 0.5 1 5 10 50 100 Angular Frequency (rad/sec)

Figure 12. Tannic acid variation (unswollen gels) shows that storage modulus increases with increasingTA up to 4x the originalamount. However, the 4x TA sample was particulatedue to fracture and not a good composite candidate. With half the originalTA amount, the sample remains liquid (non-constantshear modulus).

20 4.4 Mineralization With and Without Post Treatment

Figure 13. A post treated 4x NaIO4 2x TA gel exhibits dark brown color, strong cohesion, and solid like properties

After optimizing the initial gel formation, experiments were conducted on how to best mineralize the gels. For the rest of this study a network of 4x NaIO 4 and 2x TA was used. To further strengthen the gel, a "post treatment" was added. This consists of the same amount of tannic acid, periodate, and NaOH as previously used in the formation of the gel being added to the sample. Samples were allowed to fully hydrate after this process. Magnetite (Fe304) was then grown in samples treated twice with this process, once, and without any treatment. Figure 14 shows that post treatment led to a higher modulus. After mineralization, both Ox and Ix post treated samples showed large moduli increases. However, 2x post treated gel did not exhibit a modulus increase. Here we predict that the post treatment has created a covalent network so tightly bound that magnetite particles cannot infiltrate it.

21 U 2xPostTreat .gua.m ... m EEEUSEEEE 105 Mineralized CU A AA AAAAAiA AA AA AA AA AA A A LA 03 2x Post Treat A 1000 .0000 0 0 0 0o0OOO 0 0 0 0Oo A 1x Post Treat 10- Mineralized 0 A 1x Post Treat 0.1 - - Ox Post Treat 0 00 Mineralized

0.1 0.5 1 5 10 50 100 0 Ox Post Treat Angular Frequency (rad/sec)

Figure 14. Post Treatment andMineralization, 4x Periodate, 2x Tannic Acid - Post treatment increasedshear modulus by at least an order of magnitude between Ox, lx, and 2x treatments. Mineralization increasednetwork modulus by two orders of magnitude on an untreatedsample but this increase lessened on post treatedgels.

22 5. Conclusion

In this study, a method for forming nanocomposite hydrogels made of polyallylamine, tannic acid, and magnetite was developed. With this method came an understanding of how the network operates. In forming the initial hydrogel network, pH was increased to 6.22 in order to establish tannic acid crosslinking. Crosslinking density was further increased by adding 4x the original amount of sodium periodate, which helps to deprotonate tannic acid. The final adjustment in initial gel formation was increasing tannic acid amount to 2x the original amount. These adjustments created a network with strength derived from crosslink density that is able to support mineralization with magnetite. The crosslink density was further increased with post treatments of tannic acid, periodate, and NaOH. With one such post treatment, the unmineralized gel increased in modulus to over 1000 Pa. After mineralization, the modulus jumped to over

105 Pa, a maximum of the gels tested here. The gel given two post treatments exhibited a high modulus, but did not increase with mineralization leading to the conclusion that an overly densely crosslinked PAA network does not allow for the growth of magnetite particles. Through this work, it is clear that an in-situ mineralization of PAA hydrogels leads to a composite with greatly improved mechanical properties.

23 6. Future Works

A system of nanocomposite hydrogel formation lends itself to further exploration of its applications. This study focused on the creation and mechanical characterization of the composite, but has not explored the magnetic properties of the gel. In future research, the mineralized hydrogel should be tested to see how it responds to a magnetic field. It is important that the structural integrity of the gel remains intact- the magnetite particles should not become dislodged form the network or cause it to fracture while bending in response to the magnetic field. The mineralized hydrogels should also be tested with respect to tensile response. This testing will tell us if the hydrogel has become more brittle as a result of mineralization. For flexible electronic applications, it is important that the hydrogel remains bendable and can recover its original shape.

In addition to exploring these properties of the hydrogel, the system can be tested for biocompatibility. This exploration will enable our nanocomposite's expansion into the fields of bio-interfacing technology and drug delivery. It is important that the magnetite particles are well secured within the network and will not easily escape into the body.

The network should also be tested for response to the body's different pH environments.

Our hydrogel should remain intact through pH changes because it was formed with permanent covalent crosslinks and magnetite particles grown into the network, but this must be verified. Studies such as these will open up new fields of application for the magnetite mineralized hydrogel.

24 References

I. Zhao, H., Yang, Z., & Guo, L. (2018). Nacre-inspired composites with different macroscopic dimensions: Strategies for improved mechanical performance and applications. NPG Asia Materials, 10(4), 1-22.

2. Holten-Andersen, N., Harrington, M. J., Birkedal, H., Lee, B. P., Messersmith, P. B., Lee, K. Y., & Waite, J. H. (2011). PH-induced metal-ligand cross-links inspired by mussel yield self- healing polymer networks with near-covalent elastic moduli. Proceedings of the National Academy of Sciences, 108(7), 2651-2655.

3. Mayer, G. (2005). Rigid Biological Systems as Models for Synthetic Composites. Science,310(5751), 1144-1147.

4. Finnemore, A., Cunha, P., Shean, T., Vignolini, S., Guldin, S., Oyen, M., & Steiner, U. (2012). Biornimetic layer-by-layer assembly of artificial nacre. Nature Communications, 3(1).

5. Ahmed, E. (2013). Hydrogel: Preparation, characterization, and applications: A review. Journal ofAdvanced Research, 6(2), 105-121.

6. Wang, Y., Li, B., Zhou, Y., & Jia, D. (2009). In Situ Mineralization of Magnetite Nanoparticles in Chitosan Hydrogel. Nanoscale Research Letters,4(9), 1041-1046.

7. Antipov, A. A., Sukhorukov, G. B., Donath, E., & M6hwald, H. (2001). Sustained Release Properties of Polyelectrolyte Multilayer Capsules. The Journal of Physical Chemistry B, 105(12), 2281-2284.

8. Quideau, S. et al. (2011). Plant Polyphenols: Chemical Properties, Biological Activities, and Synthesis. Angewandte Chemie InternationalEdition, 50(3). 586-621.

9. Picchioni, F., & Muljana, H. (2018). Hydrogels Based on Dynamic Covalent and Non Covalent Bonds: A Chemistry Perspective. Gels,4(1), 21.

10. Carretti, E., Dei, L., Baglioni, P., & Weiss, R. G. (2003). Synthesis and Characterization of Gels from Polyallylamine and Carbon Dioxide as Gellant. Journalof the American Chemical Society,125(17), 5121-5129.

11. Antipov, Alexei A, et al. (2003). Carbonate Microparticles for Hollow Polyelectrolyte Capsules Fabrication. Colloids and Surfaces A: Physicochemicaland EngineeringAspects, 224(1-3), 175-183.

12. Ejima, H., Richardson, J. J., Liang, K., Best, J. P., Koeverden, M. P., Such, G. K., Caruso, F. (2013). One-Step Assembly of Coordination Complexes for Versatile Film and Particle Engineering. Science, 341(6142), 154-157. 13. Lee, J., Yeo, M., Kim, W., Koo, Y., & Kim, G. H. (2018). Development of a tannic acid cross-linking process for obtaining 3D porous cell-laden collagen structure. International Journal of Biological Macromolecules, 110, 497-503.

25 14. Korey, M., Mendis, G. P., Youngblood, J. P., & Howarter, J. A. (2018). Tannic acid: A sustainable crosslinking agent for high glass transition epoxy materials. Journal ofPolymer Science PartA: Polymer Chemistry, 56(13), 1468-1480.

15. Krogsgaard, M., Andersen, A., & Birkedal, H. (2014). Gels and threads: Mussel-inspired one- pot route to advanced responsive materials. Chem. Commun.,50(87), 13278-13281.

16. Allais, M., Meyer, F., & Ball, V. (2018). Multilayered films made from tannic acid and alkaline phosphatase with enzymatic activity and electrochemical behavior. Journal of Colloid and Interface Science, 512, 722-729.

17. Page, L. W. (1951). Minerals and How to Study Them. Edward Salisbury Dana. The Journal of Geology,59(6), 624-626.

18. Tang, J., Tong, Z., Xia, Y., Liu, M., Lv, Z., Gao, Y., Wang, T. J. (2018). Super tough magnetic hydrogels for remotely triggered shape morphing. Journal of Materials Chemistry B,6(18), 2713-2722.

19. Satarkar, N., & Hilt, J. (2008). Magnetic hydrogel nanocomposites for remote controlled pulsatile drug release. Journalof ControlledRelease, 130(3), 246-251.

20. Kim, J. I., Chun, C., Kim, B., Hong, J. M., Cho, J., Lee, S. H., & Song, S. (2012). Thermosensitive/magnetic poly(organophosphazene) hydrogel as a long-term magnetic resonance contrast platform. Biomaterials,33(1),218-224.

21. Ramachandran, V. S., & Beaudoin, J. J. (2001). Handbook of analytical techniques in concrete science and technology: Principles, techniques, and applications.Norwich, NY: Noyes Publications.

22. Witten, T., & Pincus, P. A. (2004). Structuredfluids. Polymers, colloids, surfactants. Oxford: Oxford University Press.

26 Appendix A:

Frequency sweep graphs with storage and loss moduli

Filled markers indicate storage, empty markers indicate loss

pH Variation 1000

AAAAAAA~AI~A %No 1 * pH 1.72 -- nO U) 0 0 0 000 A pH 4.35 mass 0 pH 4.90 0 0.001 - .1* *EEA pH 6.22 0 E +

*.EEEEEEEE 10~6 0.1 0.5 1 5 10 50 100 Angular Frequency (rad/sec) FigureA 1

Periodate Variation 4I flV~IUU I

100 * 0.5x 10 * IX 1 MIII *2x 0.100 0 A 4x 0.010[ va yE v 6x 0.001

0.1 0.5 1 5 10 50 100 Angular Frequency (rad/sec) Figure A 2

27 Tannic Acid Variation

100 AAAAAAAAAAAAAAAAAeflAAQ C. 0 .5x

ZL 0.1 Aix *2x 0 a0 *4x 10-4 Em...... i i. Ii i | | r rA

I -

0.1 0.5 1 5 10 50 100 Angular Frequency (rad/sec) FigureA 3

Post Treatment, 4x Periodate, 2x TA

105 F0 _3 03 Li AA U~AAA AAAAAAAAAAAAAA 1000 - * 2x Post Treat

.50 10 ooooooooooooooooooooo A 1x Post Treat 0 0 * Ox Post Treat 0.1 0* 0 0.1 0.5 1 5 10 50 100 Angular Frequency (rad/sec) FigureA 4

28 Mineralization, 4x Periodate, 2x TA

1 x 105 m 2x Post Treat "Nw5x 104 A 55 A Mineralized 0 6 3 10 A 030000 0a A 1x Post Treat A A A 0 1"x104 A A913 n Mineralized 00 00 0 AO * Ox Post Treat 5000 Mineralized 000000000000000

0.1 0.5 1 5 10 50 100 Angular Frequency (rad/sec) FigureA 5

29 Appendix B:

Figure B 1. SEM image of mineralizedhydrogel, no post treatment. Ridged topography indicates the growth of magnetite nanocrystals

30 Appendix C: Hydrogel formation specifics

Polyallylamine 35.6 mg Water 53.4 [L (initial addition) Tannic Acid 7.87 mg Water 80.77 L (second addition) Sodium Periodate 11.57 [L (Na1O 4) Sodium Hydroxide 26.83 VL (NaOH) Table C. 1. Optimized hydrogel recipe consisting of 4x NalO4, 2x TA as adjustedfrom original ratios (before post treatment and mineralization)

Calculation of ratios

Allylamine hydrochloride C3HgC1N 93.554 g/mol Poly(Allylamine hydrochloride) 150,000 g/mol, 40 wt% aqueous solution Tannic Acid 1701.20 g/mol, 5 ligands per TA molecule NaIO 4 0.5 M

Initial amounts used (1 x TA, lx NaIO 4 ) 19.9 mg PAA 40% 2.2 mg TA

1.617 [tL NaIO 4

150,000 gimol PAA 5 monomer AA 93.554 glmol monomer AA 1603.3" ' chain P AA

0.0199 g PAA (40%) X 0.4 g PAA X 1 mol PAA X 1701.2 g TA 0 0 41 mol PAA 0.0022 g TA 1 g PAA (40%) 150000 g PAA I mol TA mol TA

0.0041 mol PAA X 1603.35 monomer AA X 1 TA _ 13. 1 6 monomer AA 1 mol TA 1 chain PAA 5 TA ligand TA ligand

0.0022 g TA 1 mol TA X 1 L Na1O4 5 TA ligand TA ligand

1.617 x 106 L NalO4 " 1701.2 g TA 0.5 mol Na1O4 1 TA ' NaIO 4

AA 13.16'"on"m'""^^x 8.0 TA ligand = 10 5 .3monomer TA ligand NaIO4 NaI04