MOLECULES

This invention was made with Government support under National Science Foundation under IIP STTR Grant No. 0848528. Therefore, the United States Government may have certain rights in this invention.

Smart polymers, electro-active polymer (EAP) gels, hydrogels, and drag eluting polymer gels have all long been the subject of extensive development and investigation for some time. The majority of the literature today refers to each of these as if they were each a specific discipline or classification of materials wh en, if looked at from a molecular engineering approach, there is a common mechanism of action in many of these materials: a physical change or swelling and de-swelling of a 3-dimensional (3D) polymer matrix on which there are reactive and un-reaciive molecular units or sites. Each of these structures are based on crosslinked gels that have been covalently linked at specific molecular sites using a variety of polymer initiation (e.g. radical, thermal, photo, redox, anionic, cationic, coordination) and propagation (e.g. step, chain, ring-opening) methods as well as different polymer types (e.g. addition, condensation).

The affinity of the 3D polymer networks for a particular solvent or electrolyte enables the gel to swell or de-swell. The network typically is defined as hydrophilic or hydrophobic, and this is determined by the nature of the molecular units on the polymer chains. The hydrophobic or hydrophilic nature is dictated by the thermodynam c parameters of solvent and electrolyte interaction with the molecular units. The thermodynamic parameters can be dramatically altered by the introduction of charge to the polymer matrix thereby changing the nature of the interaction with solvent or electrolyte. The kinetics of uptake or expulsion of a solvent or electrolyte is influenced by the porosity, density and surface area of the polymer 3D network and by laws of diffusion. In general the more porous a structure is the faster it can uptake or expel solvent, but with polymer gels for use as actuators the more porosity there is the less structurally sound the polymer is under load. Notwithstanding, we found that this limitation can be countered, by increasing the surface area of the polymer in the actuator design.

According to the present invention, polymer gels, not usually thought of as structurally strong, can be quite useful as mechanical actuation devices, so much so that the epoxy polymer gels explored, can be engineered to meet a variety of specific performance requirements. More particularly, we have determined that the chemical kinetics, the end products, physical properties including speed and force generated by swelling and de-swelling of a 3D polymer matrix, can be molecularly engineered by varying the quantity, type and ratio of reactive molecules in the gel. Furthermore, polymer chain length impacts the actuator both by the size of the swelling response and by the number of molecules available on the chain to produce force. Our findings more importantly show that the molecular reactive units combined provide the actual driving forces and these reactive molecules can be left im-reacted in the

polymerization process, where the reactive molecules can then be used to effectively increase the speed and force generated by the actuator. As a result the quantity, type and ratio of molecular reactive units can be tailored to produce forces needed for practical applications of the actuators. The ratio and the density of their electrostatic charges, as well as their affinity for the solvent or electrolyte used, play an important role in the gel's actuation performance through protonation or deprotonation, Protonation is the addition of a proton (H ⁺ ) to an atom, molecule, or ion.

Deprotonation is the removal of a proton (H ^f ) from a molecule, forming the conjugate base.

The relative ability of a molecule to give up a proton is measured by its gK _a value, some molecules can have multiple pKa values with dissocation and differing pH levels. Electrolyte molecules or ions can be used according to the performance required. Electrolytes can be mixed according to pKa values and this will modify the performance and swelling range of the polymer.

We also have found that surprisingly, the excess reactive units do not need to be part of the polymer chain, and can be simply placed or entangled within the polymer matrix in order to have significant impact on the gels performance characteristics. Previously we reported ¹ , the results of epoxy-amine hydrogel synthesis and the gel characterization. The present disclosure extends our study on the epoxy polymer gels actuator performance under various conditions that are needed for an actuator to be useful in an industrial or consumer product.

Epoxy amine hydrogels were prepared by reacting branched polyethylene amine polymers such as branched Eihyleneimme or Jeffamine T-403, with the di- epoxide Poly (ethylene glycol) diglyeidvl ether (PEGDGE) compounds are water soluble and readily polymerize.

Further features and advantages of the present invention will be seen from the following detailed description, taken in conjunction with the accompanying drawings, wherein:

Figure 1 illustrates a schematic route of gel synthesis of Jeffamine T-403 polyethylene giycoi deglycidyl ether and water;

Figure 2 illustrates hydration and electrolytic route to gel expansio with sodium acetate:

Figure 3 shows stroke rate comparison, of different electrolytes, using one polymer actuator formulation;

Figure 4 shows passive swelling of NH/OH 1.0 ratio gels with varying molecular weight PEGDGE;

Figure 5 shows hydration of actuator gel formulas with an average PEGDGE chain length with varying ratios of reactive molecular NHNOH ratios;

Figure 6 compares chain linked impact before breakdown of polymers on maximum swelling in ecetic acid;

Figure 7 is a graph showing passive hydration pressure generation comparison of different chain links with NH/OH;

Figure 8 is a graph showing speed, of stroke comparison of different NH/OH ratios;

Figure 9 is a graph showing electrically actuated vertical stroke rate of PEGDGE gels with different NH/OH ratios ;

Figure 10 is a graph showing stroke under low comparison of PEGDGE molecular weight 600 for different NH ratio gels;

Figure 1 1 is a graph comparing 1.75 NH ratio gels stroke rates under load; Figure 12 is a graph comparing electrical activation pressures of different NH ratio gels;

Figure 3 is a graph showing impact of binders in electrical activity' performance at constant amp; and

Figure 14 is a graph comparing a solid disk and Dry Milled Gel. As used herein the term "smart polymers" are stimuli-responsive polymers that change according to the environment they are in. As applied specifically to the present invention, the smart polymers change in polymer chain length, flexibility and may swell or change shape at varying speeds. Typically, the smart polymers may respond non-linearly, and are reversible.

In one aspect, the claimed invention provides a three-dimensional smart polymer matrix formed by a first epoxy chemical reaction linking a linear polymer chain and a branched polymer chain wherein the polymer chains are flexible and one or more reactive molecule species remain un-reacted, at a predetermined

mathematical ratio of reacted to un-reacted molecules and molecule species, and. the un-reacted molecular species are available for further chemical reaction after the first chemical reaction, wherein the smart polymer matrix is hydrophilic and non-soluble to a solvent or electrolyte.

A three-dimensional smart polymer matrix formed by a first epoxy chemical reaction linking a linear polymer chain and a branched polymer chain wherein the polymer chains are flexible and. one or more reactive molecule species are left un- reacted, at a predetermined mathematical ratio of reacted to un-reacted molecules and molecule species, and the un-reacted molecular species are available for further chemical reaction after the first chemical reaction, wherein the smart polymer matrix is hydrophilic and non-soluble to a solvent or electrolyte,

In yet another aspect of the invention there is provided a three- dimensional smart polymer matrix formed by a first and one or more secondary epoxy chemical reactions wherein the polymer chains are flexible and elastic, and one or more reactive molecule species remain un-reacted, at a predetermined mathematical ratio of reacted to un-reacted molecules and molecule species, and available for further chemical reaction after the secondary chemical reactions, wherein the smart polymer matrix is hydrophilic and non-soluble to a solvent or electrolyte, and at least one of the reacted molecule species of the polymer chains and or linkages exhibit a polar attraction to a solvent or electrolyte, and the ratio of reacted to un-reacted molecular species determines the physical performance of the smart material.

In one embodiment the polymer chains and linkages are flexible and elastic and at least one of the reacted molecule species of the poly mer chains and linkage exhibit a polar attraction to a solvent or electrolyte, and the ratio of reacted to un- reacted molecular species determines die physical performance of the smart material.

In another embodiment one or more of the reactive molecule species left un- reacted are available for one or more secondary chemical reactions.

In yet another embodiment one or more of the reacti e molecule species left un-reacted and. available after the first chemical reaction are available for a secondary reversible chemical reaction with a solvent or electrolyte dispersed within and around the polymer matrix

In still another embodiment the solvent or electrolyte contains one or more molecule species that can bond with and reversiblv change the molecular charge of a reactive molecule within or around the polymer matrix.

In one embodiment the solvent or electrolyte contains one or more molecule species that can reversibiy bond with and reversiblv change the molecular charge polarity of a reactive molecule species.

In another embodiment the solvent or electrolyte contains one or more molecule species that can reversibiy bond with and reversibiy change the molecular charge strength of a reactive molecule species.

In yet another embodiment, the soivent or electrolyte contains one or more molecule species that can reversibiy bond with and reversibiy change the molecular charge of a reacted molecule species.

In still yet another embodiment the solvent or electrolyte contains one or more ion species.

In another embodiment, the solvent or electrolyte has more than one pKa value or dissociations.

In another embodiment the chemistry of the solvent and or electrolvie are sensitive to change as a result of an external energy stimuli.

In yet another embodiment the chemistry of the polymer matrix reactive molecule species is sensitive to change as a result of an external energy stimuli.

In still another embodiment the polymer chain lengths of the linear polymer chains are random within a range and determine the maximum swelling size of the polymer matrix,

In another embodiment the polymer chain lengths of the linear polymer chains are random within a range and a mixture of two or more linear polymers, and wherein the linear polymers are hydrophilie or a combination of hydrophilic and hydrophobic polymers.

In yet another embodiment the polymer chain of the linear polymer are assembled block polymers of two or more polymers and are random in length within a range.

In still another embodiment the polymer chain lengths of the branched polymer are not random.

In another embodiment the polymer chain lengths of the branched polymer are random within a range.

In yet another embodiment the polymer chain lengths of the branched polymer are assembled block polymers.

In still yet another embodiment the branched polymer component with one or more reactive molecule species comprises dendrimers, hyperbranched polymers or a combination of one or more of these with branched polymers.

In a preferred embodiment one or more of the components of the first epoxy chemical reaction is selected from the group consisting of a polyethylene glycol digiycidyl ether, polypropylene glycol diglycidyl ether, a polymer having an ionic functional group selected from the group consisting of carboxylic acid, phosphoric acid, sulfonic acid, primary amine, secondary amine, tertiary amine, and ammonium, acrylic acid, mefhacrylic acid, vmylacetic acid, maleic acid, metakuriro yloxy ethylphosphoric acid, virrylsulfonic acid, styrene sulfonic acid, vinylpyridine, vinyl aniline, vinylimidazole, aminoethyl acrylate, methylamino ethyl aerylate, dimethylamino ethyl aerylate, ethylammo ethyl acrylate, ethyl methylamino ethyl acrylate, diethylamino ethyl acrylate, aminoethyl methacryiate, methylamino ethyl methacryiate, dimethylammoethyl methacryiate, ethylammo ethyl methacryiate, ethyl methylamino ethyl methacryiate, eiefhyiamino ethyl methacryiate, aminopropyl acrylate, methylaminopropyl acrylate, dimethylaminopropylacrylate,

ethylaminopropyl acrylate, ethyl methylaminopropyl acrylate, diethylamino propyl acrylate, aminopropyl methacryiate, methylaminopropyl methacryiate, dimethylaminopropyl methacryiate, ethylaminopropyl methacryiate, ethyl methylaminopropyl methacryiate, polymers, such as diethylamino propyl methacrylate, dimethylaminoethyl acryiarnide, dimeihy!aminopropy!acryiarnide, and an akuriroyloxy ethyl trimethylammonium salts, polybutene, a silicone, a

silsesquioxane oligomers, an amino functional silicone or siloxane. a silanol functional polymer, a hydrosiloxane, a carbinol functional silicone, an anhydride functional silicone, a bicycioheptenyl functional silicone, a carboxylate functional silicone, a polymeric metal alkoxide, a hydroxyl functional silicone, a

polysilsequioxane, and a combination thereof w ith a functional epoxy groups.

In another preferred embodiment one or more of the components of the first chemical reaction is selected from the group consisting of a branched polyethylene oxide, polyethylene glycol, polypropylene oxide, polypropylene glycol, silicon, a polysilsequioxane. a miiliiiunciional siloxane, polybutene or another flexible polymer with a terminal amine, carboxyl, hydroxyl or other functional unit.

In another embodiment the poiymer gelation of the reacted polymer is below 98%, more preferably below 90% and even more preferably between 75 and 85% In yet another embodiment one or more non linking polymers or

macromolecuies with reactive molecule species are entangled, trapped within and around the polymer matrix and not linked to the polymer matrix after polymerization and are available for chemical reaction.

In still another embodiment one or more non linking polymers or

macromolecuies with reactive molecule species are entangled, trapped within and around the polymer matrix and partially linked at one or more molecuiar sites to the poiymer matrix after polymerization and are available for chemical reaction,

In still yet another one or more non linking polymers or macromolecuies with reactive molecule species are entangled, trapped within and around the polymer matrix and partialv linked at one or more molecular sites to the poiymer matrix after polymerization and are available for chemical reaction.

In another embodiment one or more reactive molecule species available for chemical reaction determine the pressure generation of the poiymer material.

In yet another embodiment one or more reactive molecule species available for chemical reaction determine the pressure generation of the polymer actuator and the surface area determines the speed.

In yet another embodiment one or more reactive molecule species available for chemical reaction determine the pH range of one or both of the passive and act ve swelling generation of the polymer material.

Amine-cured epoxy networks are formed from the amine hydrogen— epoxide reaction which is shown in Figure 1. I. The CH ₃ group closest to the NH ₂ group in T-Jeffamine will produce a steric interference to slow the reaction rate of the secondary amines as compared to the non-reacted primary amines ^"' , ⁴ , ⁵ . Therefore, the polymer matrix ends up with a ratio dependent (amine/epoxide) combination of primary and secondary amine hydrogens on the final network.

A. Equilibrium Swelling and Electrolytic Actuation Mechanics

of Amine Epoxy Gel

It is important to understand the role the electrolyte or a solvent plays in a hvdrogel swelling process. Due to the interconnections between the polymer chains, the crosslinked polymers are insoluble but the structure will uptake a solvent or electrolyte and expand the polymer matrix to the extent that it will allow. If the affinity of the electrolyte or solvent for the molecular sites is stronger than the polymer chains or crosslink bonding then the structure can fracture or decompose redu cing the force of the actuator material.

The solvent or electrolyte is attracted to the reactive molecular units on or within the polymer matrix and can remain within the polymer structure in what is typically cited as three states; bound, unbound and free. The strength of each of these interactions varies according to the molecule and its interaction with the solvent or electrolyte, so it is possible to have a combination of several variants of the bound unbound and free states occurring in simultaneous dynamic action ⁶ . When charge is introduced to the matrix by protons generated by electrolytic breakdown of the solvent at the electrodes, ihe solvent or electrolyte in and around the polymer gel will be attracted to, or repulsed from, the polymer matrix according to the strength and density of the charge. Since the molecular sites in the matrix have differing strengths. and therefore different protonated states at equilibrium, it is possible to regulate the volume, speed and force of the swelling /deswelling process via proton injection or ion concentration.

For this study of the epoxy amine gel we used only one supporting electrolyte, aqueous sodium acetate, and looked at the impact that electrolyte concentration had on the swelling characteristics with each change in the polymer/molecule matrix. From this we developed a linear model of the amine epoxy gel's mass volume change in relation to the salt concentration. To reach swelling equilibrium upon hydration the polymer- water interaction of the mixing uptake phase generates osmotic pressures P _n ii _x acting expansively. Due to the polymer-polymer interactions the polymer network counteracts this expansion by an elastic force respected by P _e iast- The contribution of molecular electrostatic interaction to balance the osmotic pressure has to be represented as an expansive pressure Pj _on . The epoxy-amine gels obtain swelling equilibrium at the balance of three pressures ⁷ , ⁸ which can he described by Equation 1.

Equation I The epoxy-amine gels comprise weak base groups such as Nil and ¾ that can be protonated. The polymer gels also contain hydroxy] (OH) and ether (0) units that form hydrogen bonds easily with water (Equation III ^■ · H ₂ 0 route). As water enters the matrix and creates hydrogen bonds at the 0 and OH and NH sites the H bonded water molecules also contribute to this process by providing another three, possible H bonding sites ⁹ , ¹⁰ each, and each of these H bonded waters can perpetuate three more H bonds with each new water molecule (Formula I). Therefore, the density of the water charged groups within the network strongly increases upon aqueous hydration. Once the gel reaches equilibrium of hydration it can be further expanded by an adequate generation of mobiie counter- ions inside the gel (Equation II- NaAc current route), which induces the phase transition due to electrostatic repulsion and the polymer gel is forced to increase its swelling volume. k ,.,H n >

Formula I. Multiple hydrogen water bonding

The primary and secondary amine groups and the hydroxy! groups are electron rich and easily form hydrogen bonds with water molecules, this is demonstrated in hydration study results later in the paper. In an aqueous media an equilibrium exists (Equation II) between the free amino groups and their protonated forms.

R-N¾ + H ₇ 0 ^ZT fR- N¾l ⁺ OH ^" o Equation II

9 When applying a DC current to a hydrated gel the region near the anode, in an

10 aqueous electrolyte cell, becomes more acidic as oxygen is released and protons are

1 1 formed in solution, according to Equation III.

12 2¾0 → O2 + 4 H ⁺ + 4e

] 3 Equation III

14 The region near the cathode becomes more basic according to Equation IV,

15

16 2 H2O 2 ^ H2 20H Equation IV

17 During the water electrolysis from the applied current, localized pH gradients

18 are generated near both of the electrodes. At low pH, nearby the anode, the protons

19 enter the matrix, making positively charged (~NH ₃ ^÷ ), this disturbs the electro- neutrality of the polymer gel. To balance the charge, anions enter the polymer gel matrix along with the solvent (Figure 2 - the current route, which shows hydration and electrolytic route to gel expansion with sodium acetate. It is important to note that a different electrolyte such as sodium bicarbonate will produce negative electrostatic force repulsion). This causes increased swelling, added to this a large amount of ionic repulsion occurs at a higher concentration of MH ₃ ions, thereby physically forcing the gel matrix to open and swell allowing an increase in water to water bonding. The epoxy-araine polymer gel contraction or expulsion of fluid occurs at higher H, in the vicini ty of the cathode a reversible process takes place wi ere the hydroxyl ions are created, and the gel shrinks. Simply explained the swelling/shrinking of these epoxy-araine gel actuators works by changing the pH concentration of the solution. This is determined by the electrolyte (sodium acetate), current charge, time and polarity of the electrode,

The results of our study indicate that the amine epoxy gels are highly conducive to materials engineering. The performance of the polymer gel actuator can be incrementally adjusted by changing the following formulation parameters; polymer crosslink density, polymer chain length, degree of polymer gelation, ratio of reactive molecular units, and binding materials. Solvents and electrolytes also play a very important role in the polymer actuation as shown in Figure 3, so much so that for the purpose of this study the electrolyte concentration was fixed at one concentration in aqueous solution so that the impact of each change made to the polymer material showed as a direct change in the performance of the actuator. Electrical actuation experiments where conducted using an aqueous sodium acetate electrolyte at salt concentration of 0.05m and passive swelling tests were conducted, using 3 aqueous solutions, 0.05m acetic acid, 0.05m sodium acetate and DI water. All gels were tested by mechanical tensile tests, hydration swelling change of mass and swelling performance via electrical current over time. Rates of actuation strain both free standing and weight loaded, as well as force generation over hydration time and actuation time were recorded. The formulation was then altered to generate a range of reactive H to OH molecule ratios from an excess of 2.50H / 1.0 NH to 2.5 H / 1.0 OH according to the following studies and results. Passive Aqueous, Electrolyte and Acidic Swelling and Diffusion of Amine Epoxy Gel

Passive hydration tests were accomplished using cured cylindrical polymers that were cast in machined molds to provide a uniform sample size of the polymers. Polymer gels were formulated with an equal ratio of NH to OH molecules or an f ratio of 1.0 to 1.0 using Jeffamine and PEGDGE 200, 400, 526, and 600 molecular weights. The gels were allowed to swell for 72 hours. The gel cylinders showed expected results, with one exception, as shown in Figure 4, demonstrating incremental increases in swelling sizes that corresponded, with increases in the chain length of the PEGDGE component. The 200 mw geis at the f L0 ratio showed increased swelling over the rest of the higher molecular weight geis. It is not understood why there was this difference with the 200mw gel

When a comparison of how the NH / OH ratio affects the passive swellmg of actuator gels with Jeffamine and a single PEGDGE chain length was performed, the data supports the correlation of NH / OH ratios impacting swelling results as shown in Figure 3. Actuator gels were formulated with PEGDGE 526mw and the ratio changes produced higher overall swelling with increased, ratios. The study also importantly shows that the water swelling of the gels increased significantly with an increase in OH units, where the higher NI-I ratio gels showed much less swelling in water and a much higher 4x difference between the water and acid, swelling at the higher NH ratio suggesting less water bonding until the acid, protonated state of the NH units. This clearly demonstrates the high degree of impact a change in reactive molecular ratios can accomplish in the overall swelling performance of an actuator material.

B, Polymer Crosslink Density

It has been demonstrated in other studies that dendritic components lend certain desirable characteristics to polymeric networks within which they are incorporated. For example, (a) dendrimers can increase the permeability of a dense crosslinked network, and this can be exploited for sensor applications;"' ¹⁸ (b) the mechanical properties and water content of a hydro gel can easily be tuned by changing the demirimer concentration;' ² (c) PAMAM dendrimers in hydro gels impart increased, hydrophiiicity ⁹ which may increase proton conductance through the hydro gels; and (d) properties of hydro gels that incorporate PAMAM are largely dependent on the properties of the other component of the hydro gel polymer/ Literature suggests that hydro gels crossiinked with dendrimers do not show a significant decrease in their ability to swell as a function of increased crosslinking.

We investigated the mechanical performance effects both dendrimers and hyperbranched polymers had when incrementally added into the amine epoxy formulation. In addition, using dendrimers or hyper branched polymers allowed measured increase in concentration of ionizable amine groups to the amine epoxy gels. We theorized this would increase the performance aspects of the amine epoxy gel actuators,

Dendrimers are symmetric, highly branched, synthetic macromoiecuies as shown in Formula II. Advantages of dendrimers for hydrogel formation include high crosslinking densities at low polymer concentration, systematic control of physical properties through structural variables provided by the dendrimer and low viscosity of precursor soiutions for pre-processing. For this portion of the study, we incorporated both PAMAM and DAB dendrimers from generations 0, 2 and 3.

Formula II. Symmetrical dendrimer structure with both PAMAM and DAB end groups shown.

We prepared amine-epoxy hydrogels fabricated from GE-PEGsoe-GE, Jeffamine T-403 and dendrimers DAB-Am-8, DAB-Am-16, P.AMAM-8, PAMAM- 36, SP-012 and XTJ-582 in separate formulations. The ratio of NH to epoxide previously determined to give optimal performance and stability characteristics to our amine epoxy standard gel formulation without dendrimer is 2,67: L Hydrogels having this formulation are called the standard gels (gel- 1), Amine epoxy gel formulations where the NH to epoxide ratio remains 2.67: 1 but the amount of water used is doubled and tripled are the X ₂ and. the X ₃ gels respectively (gels 2-3). Two generations of DAB and PAMAM dendrimers were investigated, in the formulations for the hydrogel actuators. They are generation 2.0 and 3.0 with 8 and 16 amine residues on the periphery respect ely. The volume fraction of Jeffamine T-403 was systematically replaced with dendrimer over several, increments, whereas the ratio of NH to epoxide was maintained at 2.67: 1 (gels 4-1 1), The critical parameter that distinguishes these dendrimers is the number of amino residues on the surface, which is indicated in the designation of these compounds. We have also characterized these gels in terms of equilibrium swelling characteristics and gelation degree. The dendrimer formuiations studied during this period probed the increase of dendrimer composition with a decrease in Jeffamine T-403 composition by one third, while NH to epoxide ratio was maintained at 2,67: 1 thro ughout the en tire formula ti o .

As we studied, the effect of dendrimer or hyperbranched polymer crosslinking agents on amine epoxy gels performance, we assayed actuation response, increased hydrolytic stability, and longer cyclical lifetimes. In general, higher crosslink density leads to enhanced, mechanical properties and significa t reduction in network breakdown. The well-defined structure of dendrimers, with the number (density) of the junctions and their chemical composition easily controlled, should lead, to more reproducible properties and responsive behavior of the resulting hydrogels, allowing a systematic study of structure-property relationships and subsequent materials optimization.

Our study indicates that introducing the dendrimers in the PEGDGE/ jeffamine formulation also increases mechanical stiffness thereby reducing the gels flexibility as shown in Table ί . The dendrimer structures also play an important role in improving the gels strength at a higher N¾ to epoxide ratio. Shaped gels were formulated and mechanically tested, and the results are reported below in Table 1.

PoiyEDGE+0307Ethyleneimine 2.84 no break no break 1300+7.16 Carbon+H ₂ 0

PolyEDGE+JT- 3.40 no break no break

403+0.363Etbyleiieiraine-1300 -7.16

Carbon+F Q Table 1. Mechanical test results showing effects of adding dendrimers.

hyperbranched, elastomers and binders to the base gel formulation. Gels that remained intact with no break either had elastomers or binders added, over 36 formulations were tested.

The addition of dendrimers do impact the gels total size swelling ability as shown data presented in Figure 4. This made sense when put into the following mechanical metaphor, a cross section of organized dendrimers are similar to the way a bicycle wheel is. where the spokes create stiffness and hold the shape of the circular wheel, just like a wheel the dendrimer imparts strength, but not flexibility. Hyper branched polymers on the other hand are typically non symmetric molecular constructions having a branched structure, generally around a core. Their structure is typically fan shaped and the base units or monomers used to construct the hyper branched polymer can vary and their distribution is non-uniform. The branches of the polymer can be of different natures and. lengths therefore imparting flexibilit)', we also studied integration of several molecular weights of poly ethyieneimme hyperbranched polymers such as SP-012 shown in Formula III,

Formula III, Hyperbranched Ethyleneimine SP-012 Hyper branched polymer SP-012 is not entirely linear polymer but partly branched polymer (Formula VI), containing primary, secondary and tertiary amine residues, it is a water soluble polymer which by polymerization of Jeffamine with poiyethyleneimine produces a hydrogel. The hydrogel formulation wherem 10% of the Jeffamine T-403 weight was replaced by hyper branched dendrimer SP-012 was formulated. The molecular weight of the SP-012 is 1200 and it has an average of 19 NH residues on its periphery. The epoxide NH ratio for the above formulation is calculated to be 1 :3.4. The hydrogels with hyper branched dendrimer SP-012 incorporated demonstrated good swelling properties when compared to the dendrimer containing formulations of the hydrogel using DAB and PAMAM as shown in Table 1 , They also showed, a fairly good tensile strength in comparison to the standard gels, Table 1 shows the tensile test results of samples of the gels made from the hyper branched and dendrimer formulations in comparison to the standard gel. The following trends were observed when the amount of SP-012 was increased systematically. With the increase in SP-012 content in the gel, the elastic modulus and fracture strength increases and. Strain to Break decreases. Incorporation of 10% XTJ-582, a linear te!eche c diamine, in the hyper branched, formulations greatly reduces the elastic modulus and increases the elasticity.

C. Polymer Chain Length

The polymer chain length and elasticity sets the stage for the polymers ultimate strength, flexibility and swelling size of the polymer gel, we theorized that adding a longer chain elastomer would improve the higher crosslinked dendrimer- containing gels' degree of swelling and mechanical properties of the hydrogels. This proved to be successful XTJ-582 is a polyetherdiamine of about the same weight as JT-403 with a lower level of unconverted hydroxy! groups than the standard

Jeffamine D-400 polyetheramine. The pol ether backbone contributes flexibility to the hydrogel. We systematically replaced 10% of the JT-403 with XTJ-582 in both base gel and dendrimer-con taming gels. Figure 6 shows a comparison of the water swelling properties of the dendrimer hydrogels with incorporation of XTJ-582. With the increase in D.AB-8, DAB-16, PAMAM-8, PAMAM-16 and SP-012 dendrimer content in the gel, the elastic modulus and fracture strength increases and Strain to Break decreases. The higher the generation of dendrimer, greater is the increase in the elastic modulus. Our results show thai incorporation of 10% XTJ-582 elastomer in the dendrimer formulations greatly reduces the elastic modulus and increases the elasticity as shown in Table 1. XTJ elastomer, added incrementally also impacted the degree of actuation performance in the gel Incorporation of 6% XTJ -582 produces a relatively soft highly elastic gel. Several other elastomers could be considered in future in order to improve the mechanical properties without compromising the elasticity of the polymer matrix. Gel composition additi ves including the water amount, curing temperature and time all impacted the crosslink density and the gel elasticity,

In the passive swelling test comparing the jeffamine and PEGDGE gels, the data for the acetic acid the trend shows that the longer the chain length the larger the swelling. The data shows that the reactive molecule ratio has significant impact and. needs to be taken into account as the force generated internally can eventually degrade the integrity of the polymer matrix as shown in the maximum s well rates in Figure 6 many of the gels broke apart after just a few hours limiting the results, the 200, 400 and 600 PEGDGE accounting for the different scales per chain length. The 200, 400 and 600 are from one supplier and the 526 is from another, this probably accounts for the different stability results at maximum swelling the 526 proves to be extremely stable in all hydration states.

D. Polymer Gelation

One methodology of assessing the integrity and quality of the hydrogel is the measurement of the gelation degree (G) and degree of swelling (Q), which was conducted as follows. The mass of the actuator was taken after it had been cured and air-dried to constant mass (ni). The actuators were then placed in distilled water and allowed to soak until they reached an equilibrium and constant mass (n¾). The swollen actuators were then removed from the water bath and

allowed to air-dry until they reached, constant mass (n¾). Gelation degree (G) and swelling degree (Q) were then measured as follows:

Equation V where p _ge i is the density of the gel after re-drying and pmo is the density of water. Table 2 shows the gelation degree and Table 3 shows degree of swelling in water for several of the actiiators studied. Some significant trends are evident, increasing the amount of water in the formulation (base gel -→X2→Xi) results in a dramatic increase in Q. Changing the nature of the dendrimer from DAB to PAMAM and also the generation from 2 to 3 results in an increase in G and a decrease in Q, presumably because of higher crosslinking density introduced with increasing generation. Also, for the same generation dendrimer, when the ratio of Jeffamine to dendrimer decreases, we see an increase in G and decrease in Q for the same reason.

The amine epoxy gels formulated with the hyper branched ethyleneimine also show the impact of gelation on the gels overall swelling characteristics. From both the dendrimer and. hyperbranched polymer data we can conclude that increasing the crosslink density strengthens the gel significantly. This strengthening impacts the ability of polymer matrix to physically expand, and the higher the gelation coefficient, the less the hydrogei is able to freely expand during hydration, as evident when comparing tables 2 and 3 where the lowest gelation degree shows the highest swelling.

Table 2: Gelation degree (G) of hydrogels. t mi ^ ά ^ά^ the

1 SP-312

^& 0.0¾:[¾S¾>XT¾S2

09¾0ΑΒ δ

05?¾0ΑΒ δ

0<¾0ΑΒ δ

022S0AB S

O.D¾2DAB- i SS2

1.S33AB-S

1.3523AB-S

052DAB-8 s¾?id÷iO¾i ;-5B2

8 20 ® 6i> si) Table 3: Swelling degree (Q) of hydrogels.

Csm&lwi of s s¾»elfi¾g degee (¾f sf sll fe mrn studied

1.44ISP--012 Q36iS?-32

0.451 ? m-

S. Si A&-i6*XTJ'S¾3

G.67SDS 6

I ½ DA

Ώ.

Bass ¾^ ^† i0 r ^j -s¾

500 E, Type and Density of Reactive Units

To see the impact of the reactive molecule groups we conducted two avenues of study, since the swelling kinetics on the amnine epoxy gels is determined by the molecular charge attraction and the density of those charges, the amine to epoxide ratio was incrementally increased and decreased to show what effect a higher density of primary amines or hydroxy! units would have on the actuation performance, both by stroke performance under a set load of 0.7psi and then by speed of total force generated. Another aspect we show is the impact chain length has on pressure generated, since there are water attachment 0 sites on PEGDGE chains there should be a change in pressure this is seen when two different chain lengths are used with an equal NH/OH ratios as shown in Figure 7.

In order to see the impact of changing the ratio of the hydroxy! units to amine units by increasing the density of hydroxy! units 'within the gel we used a polyester hyper branched chain with 32 hydroxy! units on the periphery with a ratio of amine to epoxide units of and collected passive as well as electrically activated swelling data on each. First, the impact of the NH epoxide ratios on passive swelling and. electrical actuation, we started the study using PEGDGE 526, at the 2.67 ratios, but found when we tried to incorporate the other molecular weight PEGDGE's i.e. 200, 400, 600mws the gels would not polymerize at the higher amine ratios. We then incrementally reduced the NH/epoxide ratios to reach the 2.5 NH/OH, i ,75ΝΉ/ΌΗ, and 1.0/1.0 MH/OH. The trend indicated that the higher OH gels initial hydration was faster but produced less force overall while the higher NH gels initial hydration was slower but produced higher forces overall and were able to test force generation over time, as shown in Figure 8.

To allo w for an apples to apples comparison of the NH and OH ratios, we formulated the five series of gels shown as 2.5 NH/i.O OH, 1.75 NH/1 .0 OH, 1.0 NH/1 .0 OH, ! ,75 OH/1.0 NH and 2.5 OH/1.0 NH. Positive results of higher and faster electrical actuation under load according to NH/OH ratios where obtained from the Jeffamine and PEGDGE 526 gels having incremental ratio increases of the excess reactive molecules as shown in Figure 9,

Actuation stroke rates of the PEGDGE 526 gels in Figure 9 showed that at each of reacti ve molecule ratio increments the gels performance aligned with the other corresponding molecular ratios. The 1.0 NH/1.0 OH demonstrated lowest stroke while the 2.5 NH./ 1.0 OH and 2.5 OH/1.0 OH ratios showed the highest.

When incorporating higher NH ratios into the PEGDGE 600 gels, passive swelling produced very large hut not structurally sound material and it become more gelatinous, this had a very negative impact on the stroke performance under load and as shown in Figure 8 when tested the 2.5NH gels did not perform nearly as well as the 1.75NH 600 gels.

The same held true for the PEGDGE 400 gels and both of the PEGDGE 400 and 600geis performed, best under electrical activation in the 1.75NH formulations as shown in F igure 9.

Pressures generated under electrical activation also sho agreement with the same electrical stroke trends with the higher NH ratio gels attaining the higher forces in line with the results in Figure 9. Figure 12 shows maximum pressure generated under electrical current to date on the PEGDGE 526 eels with the 2.5 NH/OH gels attaining the highest pressure and the 1 .0 NH/OH gels. Thus, it would appear that eventually the electrically activated gels should be able to match the force generated. by the passive swelling gels.

F, Binders Role in the Polymer Matrix The standard amine epoxy gel without a binder is quite fragile upon hydration and the stress of the actuation strains over 200% can cause gel fracture. To eliminate this problem several materials with binding properties were investigated. At first a fe types of fiber fillers were tried such as glass and chopped, carbon all of which resulted in very poor results of passive swelling. Then we tried high surface area milled carbons which produced very acceptable swelling results as shown in Figure 33. The binder does not produce a negative force such as an elastomer but it does act as a static force that has to be overcome by the forces of swelling, the increase in binder equals a stronger material that takes more force to actuate. Therefore it is possible to find an optimum balance for activation response and structural integrity for a polymer actuator that is dependent on the crosslink density, elastic forces, density' of reactive molecules, ion concentration and current density used to actuate the material and the desired structural configuration of the actuator. The binders static force does not seem to be countered by adding elastomer as shown in Figure 13 as the percentage of carbon binder was increased we systematically added elastomer this negatively impacted the performance in the higher binders concentrations were in the lower binder concentrations the elastomer did not make as large a difference.

The invention will now be illustrated with reference to the following working examples. 1. Materials

Poly (ethylene glycol) diglycidyl ether PEGDGE 200,400 and 600mw available from Poly Sciences; Poly (ethylene glycol) diglycidyl ether PEGDGE 526 the DAB and PAMAM dendrimer, Ethyleneimine, hyperbranched polyester,

Nanoclay, Manomer 1.30E, clay surface modified available from Sigma --Aldrich; JEFFAMINE T403 available from the Huntsman Chemical Company; and Printex conductive carbon black available from Degussa were assembled. Deionized water was used for all experiments. Teflon gel molds for gel casting were made in house. II. Preparation of the gels

The hyrogels were prepared by reacting the crosslinking agent JEFFAMINE T-403 with PEGDE. The binders and water were added to the mixture subsequently. These materials were kept at the same percent for all gels. The JEFFAMINE T403 and the POLYEDGE percentage varied depending on the studied ratios between them. The standard gel of H 2,67 contained JEFFAMINE T-403 and PEGDGE 526. This composition was used for some of the presented data. After the gels are cast they were cured for 5 hours at 60C and dried for 48 hours at room temperature.

Solid gel discs and dry milled (DMG) high surface area gels were tested for stroke rate and force generation, and the results shown in Figure 12. which indicate the pressure produced is the same by both only the speed is affected by the change in surface area. All pressure and stroke data was performed with the Dry Milled Gel (DMG).

To formulate the high OH ratio actuator gels hyperbranched polyesters with 32 OH units on the periphery of the polyester chain (Formula VII) were incorporated into the gel formulation. Hyperbra ched. polyesters are related to dendrimers in their fundamentally branched architecture. However, the branching is not perfectly regular, and the materials have a structural and molecular weight polydispersity more typical of conventional polymers.

Formula VII Hyperbranched polyester with 32 OH units The hyperbranched polyesters were added to the standard gel formulation as an additive in two amounts to adjust the NH/OH ratios to an excess of ί .75 OH units to NH and an excess of 2.5 OH units to NH. We believe the water soluble polymer becomes entangled within the amine epoxy polymer matrix as the epoxy reaction takes place. To determine this we conducted FT-IR analysis of 5 different formulations, incorporating the hyperbranched dendrimer HP- 16 and HP-32, which did not gel when mixed wi th JT-403 or with Poly EDGE. The FT-IR of ail the neat samples were recorded on polyethylene films except the base gel which was recorded (as a reference) with KBr, Table 4 shows the major IR peaks of 5 different formulations including the polymerized gel without the HP- 16 or HP-32.

Table 4; IR values of HP 16 aad HP-32 f rmislatiosis In the above table, IR values in the range of 3300 to 3500 cm ^"1 indicates the 0- H stretch of water and N-H stretch of primary amines, 2800-2900 cm ^"1 indicates the C-H stretch of the aliphatic chain. 1640 cm ^{" 1} is the C=0 stretch of the hyper branched dendrimer. 1100-1200 cm ^"1 is the region of C-0 stretch of JT-403, XTJ-582 and Poly EDGE, 1109 m is the C-N stretch of JT-403 and 947 is the N-H wagging frequency. For comparison. Table 4 shows the IR values of the individual components JT-403, Poly EDGE and hyper branched polyesters 16 and 32. The IR values of the above 5 samples show that there is no evidence of any amide bond formation involving the hyper branched polymer. However, we see a peak at 1730 cm ^{" 1} which we believe correlates with the polyester chain in the Poly EDGE and HP combination formulations. The remaining peaks are in agreement with the structure of the dendrimer and other reactants.

ill, Equilibrium swelling

Cast cylinders were allowed to hydrate in deionizer water, aqueous sodium acetate .05m amid acetic acid 0.05m for 72 hours. Gel samples and 150 cc of a solutiorni were used for all measurements. The hydrated samples were removed from the bath, blotted dry, and weight to obtained W _t . The Wo and W _t , were average of 3 samples. Percent change mass of hydrogei samples H, was calculated according to Equation VI: H = (W _t — Wo)/ W ₀

Equation VI

where Wo is the weight of the dried sample.

IV. Electrical Actuation

The gels were cast in 20ml batch sizes and cast as one disc discs were then dry milled to achieve high surface area. Samples were weighed and allowed to reach equilibrium in electrolyte solution and then loaded into a test apparatus.

V. Force and stroke testing

Force and stroke testing was conducted, using 1 gram of dry milled gel subjected to hydrating solutions in force sensing fixtures and stroke displacement fixtures, and the results recorded.

The amine epoxy gels elasticity is governed, by the cross— linked density and the polymer backbone of the crosslinking di-epoxide. The speed and force of actuation strongly depends on the amine— epoxide group ratio [6, 7]. Typically we found that at NH ratios higher than 2.5, the gels are very sticky and. easily break under further hydration. An improvement of the mechanical gel properties was achieved by different fillers like active carbons, clays [7] and other high surface additives. The polymer-polymer affinity is due to chemical interaction between the polymer and the solvent [8]. Finally, the change of pH solution by the electrolysis impacts the hydrogen ion pressure. These three forces contribute to the osmotic pressure, which determines the equilibrium state of the gel and creates the volume change and total osmotic pressure [8-1 1 ] as long as it is determined by the reactive molecule species available on the polymer chains and within the polymer matrix.

The incorporation of the dendrimers into the hydrogeis was investigated with zero, second and third generation of DAB and PAMAM dendrimers. Initial studies with incorporation of hyperbranched dendrimer SP-012 and elastomer ΧΤ.Ϊ-582 were also investigated. These gels were characterized in terms of swelling properties, mechanical properties and time dependent pH volume changes. Techniques like FTIR, DSC and TGA were used to characterize some of these hydrogeis and additional hyperbranched polymers. We observed some significant trends in the properties of these actuators when the dendrimers and hyperbranched materials are incorporated into the gels and thus motivate us to carry further study on them.

The addition of OH groups gives the actuator gel an initial faster swelling response under electrical activation in the first 5 hrs over other versions of the amine epoxy gels. A similar trend of hydration swelling to activation swelling was shown in the dendrimer formulations of the gels,

We have shown the materials to produce >500% volume/weight gain in 1 hr under electrical activation. Previous to this work we were only able to get 100% volume/weight gain over 24 hours of acitvaiion and retain an intact gei actuator. We have successfully demonstrated that incorporation of dendximers and hyperbranched materials result in pH responsive hydrogels that demonstrate (1) increased stability; (2) increased reproducibility of actuation; and we have demonstrated that the reacti ve molecule ratio can be manipulated to increased actuation response both in speed and pressure generation due to pronation or deprotonation of the molecules.

REFERENCES [1] Plata, E.R,; Rogers, 1-1.; Banister, M.; Vohnoist, S.; McGrath, D. EAP Hydrogels for Pulse-Actuated Cell System (PACS) Architectures, Proceedings-SPIE The International Society for Optical Engineering, 2007, 6524, 6524 IT.1-6524 ΓΤ.8.

[2] Calvert, P.; Patra, P.; Duggal, D, Epoxy Hydrogels as Sensors and Actuators, Proceedings-SPIE The International Society for Optical Engineering, 2007, 6524, 65240M.1-65240M.6.

[3] Calvert,?. Elecetroactive Polymer Gel, in Eletroactive Polymer (EAP) Actuators as Artificial Muscles- Reality, Potential and Challenges, ed. Y.Bar-Chen, SPIE Press. Vol. PM98, 123-138, 2001.

[4] Yoshtoka, Y.; Calvert, P. Epoxy-Based Electroactive Polymer Gels, Experimental Mechanics, 2002, 42, 404-408.

[5] Samui, A.B.; Jayakumar, S.; Jayalakshmi, C.G.; Pandey, K.; Sivaraman, P.

Electroactive polymer gels based on epoxy resin, Smart Material and Structures, 2007, 16, 237-242.

[6] Richter, A.; Paschew, G.; Klatt, S.: Lienig, J.; Arndt, K-F. ; Adler, H.-J.P. Review on Hydrogel-based pH Sensors and Microsensors, Sensors, 2008, 8, 561-581.

[7] Omidian IE; Hashemi S.-A.; Askari, F.; Nafisi, S. Swelling and Crosslink Density Measurements for Hydrogels, Iranian Journal of Polymer Science and Technology, 1994, 3, 115- 119,

[8] Choe, K.; Kim, K.; Doyeon, K.; Manford , C; Heo, S.; Shahinpoor, M,

Performance Characteristics of ElectroChemically Drivers Polyaerylonitrile Fiber Bundle Actuators, journal of Intelligent Material Systems and Structures, 2006, 17, 563-576.

[9] Richter, A.; Paschew, G.; Klatt, 8.; Lienig, J.; Arndt, K-F. ; Adler, H.-J.P. Review on Hydrogel-based pH Sensors and Microsensors, Sensors, 2008, 8, 561-581 ,

[10] Schneider, H.-J.; Kato, K.; Strongin, R. Chemomechaiiical Polymers as Sensors and Actuators for Biological and Medicinal Applications, Sensors, 2007, 7, -578-161 1.

[I IjOmidian H.; Hashemi S-A.; Askari, F.; Nafisi, S, Swelling and Crosslink Density Measurements for Hy drogels, Iranian Journal of Polymer Science and Technology, 1994,3, 115-1 19.

[12]Bajpai, S.K.; Duhey, S. Synthesis and Swelling Kinetics of a H-Sensitive i ^' e poi) rnerie Hydrogel System, Iranian Polymer Journal, 2004, 13, 189-203,

[13] Sraydin, D.; Karadag, E.; Oztop, H. N.; Guv en, O. Adsorption of Bovine Serum Albumin onto Acrylamide-Maleic Acid Hydrogels, Biomaterials, 1994, 15, 917-920,

[14] Skaarup, S. Polypyrrole for Artificial Muscles: Ionic Mechanisms, Solid State Ionics: Advanced Materials for Emerging Technologies, 2006, 10, 768-779.

[15] Velmurugu, Y,; Skaarup, S. Ion and Solvent Transport in Polypyrrole:

Experiment Test of Osmotic Model, Ionics, 2005, 11,370-374.

[ 16] Gabrielli EC; Keddam, M. Non-Steady State Techniques, Comprehensive Treatise of Electrochemistry, ed. E. Yeagar, J.O'M. Bockris, B. Conway, S.

Sarangaparmi, Plenium Press. Vol. 9, 62-172 (1984).

[17] Tanaka, T.; Fillmore, D. Kinetics of swelling of gels, J. Chem. Phys., 1979, 70, 1214- 1218

[1 8] Xiang, Y.; Peng, Z„,° Chen. D. A new polymer/clay nano-composite hydrogel with improved response rate and tensile mechanical properties, European Polymer Journal, 2006, 42, 2125 -2132.