eee ee ee ee 61 3.2.1 Synthesis of a,w-terminated vinyl acrylate end-linked networks 613.2.2 Synthesis of a,w-azido polyt-BA model networks .... 33 Modulus and molecular weight between c
Solubility parameter of poly(t-BA) network
Solubility is the maximum amount of material that can dissolve in a given amount of solvent at a given temperature to produce a stable solution [21] This concept can also be thought of through the idea of miscibility If two liquids mix to an
> 92Ƒ e e | @ Networks prepared neat g [ " lì m Netwroks prepared in toluene s 88 Ƒ 7 4 Netwroks prepared in cycohexane
= 84 l | @ Netwroks prepared in butanol s | ¢ | _ Netwroks prepared in methanol
Figure 2.3: Evaporation of precursor solution for crosslinking taking place in open Teflon® molds All precursor solutions contain 6.00 wt% solvent, except for the networks prepared neat. appreciable extent to form a solution, they are said to be miscible [21] The basis for this phenomenon is the weak attractive forces between the two types of molecules. The basic approach to determine miscibility is to add up the energy required to vaporize both materials, the enthalpy change when the two materials are attracted by intermolecular forces in the gas phase, and the enthalpy change when the mixture condenses to form a liquid solution The overall enthalpy change must be exothermic (negative) in order to be product-favored and, hence, have the two materials mix well [21] In polymer physics, the solubility parameter, ð, is what describes polymer solubility The solubility parameter can also be used to describe polymer networks.
32 where U is the internal energy, V is the molar volume, and Aj_,,U is the change in internal energy when two molecules in the liquid phase are brought infinitely far apart in the gaseous phase In order to find the change in internal energy, A;.,U, this term is changed into a heat of vaporization term, as follows:
Aagl © (AH? — RT) (2.9) where AH? is the heat of vaporization, R is the universal gas constant, and T is the temperature.
In certain instances, A”“? for a given compound is not known In these cases, it is appropriate to use a group contribution approach, which assumes that
The group contribution approach breaks the target compound into individual func- tional groups, whose group contribution values are found in reference materials After adding up the respective contribution of each group, a total A* can be determined. This approach was used to calculate the solubility parameters for various solvents by using the Fedor group contribution values, which were then compared with literature values found in the Polymer Handbook [22] The results, shown in Table 2.1, indicate that the Fedor group contribution values can be used to theoretically calculate the solubility parameter of a compound when it is not readily available through litera-
Table 2.1: Comparison of solubility parameters of solvents found in the Polymer Handbook with those determined by Fedor group contribution values.
Solvent Group contributions Polymer Handbook
Methanol 28.2 29.7 ture This is important because the solubility parameter of poly(¢-BA) networks for our system are not known When this method is applied to poly(t-BA) networks, the calculated solubility parameter is 18.1 MPa1⁄2.
Although the solubility parameter of networks can be calculated through group contributions, it would be useful to compare this calcualted value to an experimental value, and, in fact, the solubility parameter for networks can be determined experi- mentally through swelling For this experiment, poly(t-BA) network samples made in both the uncovered and covered molds were cut, placed in excess solvent, and allowed
Swelling ratio, œ nae > @ OH o-oo OH ©
Solubility parameter of solvent, (MPa'”)
Figure 2.4: Solubility parameter of poly(t-BA) networks crosslinked in uncovered and covered molds. to come to their swelling equilibrium The solubility parameter of the solvent that induces the maximum swelling of the network is taken as the solubility parameter of the network The results can be seen in Figure 2.4 The results indicate that the networks made by both methods yield the same solubility parameter results The maximum swelling occurred at the same solubility parameter value of 18.6 MPa1⁄2 for both sets of networks, which corresponds to the solvent THF The experimental value matches well with the calculated solubility parameter of 18.1 MPa!/2 Since this type of experiment utilizes solvents of discrete solubility parameters, the experimen- tal value for the network is only as exact as the solvents used Thus, the calculated solubility parameter is taken as the solubility parameter of the poly(t-BA) networks.
Effect of crosslinker concentration and extraction of poly(t-BA) networks ơơẶaộ 35
The precursor solution that is used to make crosslinked networks can vary widely.
However, it is important to understand how each constituent affects the final network structure In the case of crosslinker concentration, it has been shown that one can expect to see an increase in modulus with crosslinker amount, since increasing the crosslinker concentration in the precursor solution results in a greater opportunity for creating crosslinks Usually, the final crosslinked network has a lower molecular weight between crosslinks and a higher crosslink density The effect of crosslinker concentration was tested for our poly(t-BA) network system The crosslinker con- centrations that were investigated included 0.21, 0.52, 1.02, and 5.08 wt% EGDMA crosslinker The networks were crosslinked in the uncovered molds and tested on the rheometer in the parallel plate mode The results are shown in Figure 2.5 The mod- ulus increases with crosslinker concentration, which agrees with what was expected.
As crosslinker concentration increases, the amount of crosslinks formed increases and a tighter network is obtained.
The effects of extraction were also tested Extraction was done in toluene for one week, and the solvent was exchanged daily in order to draw out any unreacted reagents and sol fraction in the networks As Figure 2.5 shows, upon extraction the modulus decreases This can be attributed to the fact that the residual reactants and sol fraction have an increasing effect of the modulus of the networks When the sol
0 1 2 3 4 5 Amount crosslinker in precursor solution (wt%)
Figure 2.5: Modulus of poly(t-BA) networks crosslinked in uncovered molds for dif- ferent crosslinker concentrations in precursor solution Extracted and non-extracted networks samples shown. fraction is removed from the network, the pores of the crosslinked network become vacant and the network itself is more deformable Because of this, we see a decrease in modulus after extraction.
Solvent quantity and quality influences on poly(t-BA) network Structure Q Q Q Q HQ HQ HQ ng Q k k k v k k xa 36
An important aspect of this project was to determine how different precursor solutions and curing methods influenced the final network structure of poly(¢-BA) networks It was investigated how solvent quality and quantity affected network structure Specifi- cally, we report on how the use of solvent in the precursor solution affects the modulus,swelling, and glass transition temperature of the poly(t-BA) networks.
The two curing methods investigated were the uncovered Teflon® molds, which allowed the precursor solution to evaporate uncontrollably, possibly changing the concentrations of the reagents, and the covered stainless steel molds, which prevented evaporation through use of a quartz plate cover that enclosed the precursor solution inside of the mold Another variable was the temperature at which crosslinking occurred The temperature of the covered stainless steel molds was controlled through a thermocouple that was attached to a cold plate, providing a real-time feedback loop in order to precisely control the temperature of the mold The temperature of the uncovered molds were not controlled For this setup, the UV lamp in the AtmosBag caused a rise in air temperature as the crosslinking reaction progressed for 45 minutes This rise in air temperature was uncontrolled and unrecorded, and subsequently caused an inevitable rise in precursor solution temperature as well as the immediate surrounding environment In this way, the crosslinking reaction may have occurred at varying temperatures, as well as the possibility of a temperature increase as UV curing progressed In addition to these curing differences, various amounts of solvent was added to the precursor solution The addition of solvent may have altered the glass transition temperature of the poly(t-BA) networks, possibly influencing the modulus results of such networks All of these variables were considered for how they effected the final network structure.
@ Network prepared neat @ Networks prepared neat m= Networks crosslinked in toluene @ Networks prepared in toluene
@ Networks crosslinked in butanol! 12-4 Networks prepared in cyclohexane 4
@ Networks prepared in butanol 14L ° 4 v_ Networks prepared in methanol s a Ll| a i? © + 3
Figure 2.6: Modulus results of poly(t-BA) networks as a function of volume fraction of ¢-BA in precursor solution a) Networks prepared in uncovered molds and rheo- logically tested in parallel plate mode b) Networks prepared in covered molds and rheologically tested in torsion mode.
Modulus and molecular weight between crosslinks was determined for both curing methods The networks crosslinked in uncovered molds were rheologically tested using high temperature parallel plates, while the networks crosslinked in covered molds were tested in torsion mode Both types of rheology tests were performed in the ETC The results can be seen in Tables 2.2 and 2.3 and in Figure 2.6.
For the most part, modulus was not affected by cure procedure However, the modulus results for the networks tested in parallel plate mode may have been influ- enced by the controlled normal force throughout the course of the experiment For these experiments, the network was subjected to a constant force of 2 N, essentially continuously compressing the network This process may have negated any solvent effects that may have been present.
The difference in swelling due to the choice of curing method can be seen in Fig- ure 2.4 The swelling that was capable of being reached from the poly(t-BA) net- works crosslinked in the uncovered molds exceeds that which was achieved with the networks prepared in an enclosed environment This trend was seen in each sol- vent that was chosen for swelling One explanation may be due to the fact that the crosslinking reaction that occurred in the uncovered molds did not have a means of temperature control during polymerization As the crosslinking reaction progressed, the UV lamp became hot and, as a result, heated up the local environment, includ- ing the mold, precursor solution, and air immediately surrounding the mold This increase in temperature may have resulted in the evaporation of precursor solution in an uncontrolled manner as the network formed, which may have left voids in the final network structure Once placed in solvent, these voids would be able to intake much more solvent than those networks crosslinked in a controlled-temperature en- vironment, which discouraged evaporation of the precursor solution through use of both a cold plate and quartz cover.
The results of how solvent quantity affected the swelling capability of poly(t-BA) networks can be seen in Figure 2.7 The results show that for networks crosslinked in uncovered molds the swelling ratio increases with decreasing volume fraction of t-BA in the precursor solution This is because the additional solvent acts as a diluent,effectively creating a looser final network that is capable of increased swelling.
Figure 2.7 also indicates that the quality of the solvent used in the precursor solution is equally important As discussed in Section 2.3.2, the solubility parameter can describe polymer solubility In a good solvent, a polymer will essentially swell, surrounding itself with as much solvent as it can, since it likes the solvent better than it likes itself In this case, x, as defined in 1.1.6, is found to have a low value On the other hand, a polymer placed in a poor solvent coils up upon itself, trying to minimize the contact it has with the surrounding solvent and maximize the contact that it has with itself In this case, the x parameter is high There is a crossover x value, which turns out to be 1/2 At this value, the polymer-solvent interactions disappear, and the polymer is in what is known as its theta solvent.
Toluene is a good solvent for t-BA, with solvent quality decreasing in cyclohexane, butanol, and methanol This fact can be interpreted in Figure 2.7 For a given volume fraction of t-BA in the precursor solution, the swelling ratio increases with solvent quality Again, this can be explained through the x parameter Since toluene is a good solvent for ¢-BA, it essentially swells the monomer, causing an elongated configuration As the crosslinking reaction takes place, the polymer and network that forms is also in an elongated configuration, yielding fewer trapped entanglements in the final network On the other hand, methanol, being a poor solvent for ¢-BA, causes a collapsed configuration of the chains during crosslinking, which may lead to an increase in the number of entangled chains Since trapped entanglements can act as crosslinks, an increase in the number of trapped entanglements decreases swelling capacity.
The quality of solvent used to swell the crosslinked network influences the swelling ratio, as well As mentioned in Section 2.3.2, the solvent that matches the solubility parameter of the gel best will make the network swell the greatest With this under consideration, we would expect that poly(t-BA) networks should swell the most in toluene and THF, and swell much less in benzene and acetone The results support this notion, with the swelling ratio of poly(t-BA) networks in toluene surpassing the swelling ratio of the networks in acetone The same trends can be seen with the poly(¢-BA) networks crosslinked in the covered molds utilizing the temperature controlled environment These results are shown in Figure 2.8.
The sol fraction, as defined in Section 1.1.2, can be indicative of the extent of the crosslinking reaction Thus, it is important to analyze the amount of sol fraction that is extracted from the networks during the course of the swelling experiment After the initial network weight was recorded, the swelling experiment was started, with the chosen solvent being changed daily After the final dried weight of the extracted network was recorded, the two dried weights, before and after extraction, were com- pared, and a sol fraction weight was determined The sol fraction results can be seen in Figure 2.9 for the poly(t-BA) networks crosslinked in both the uncovered and covered molds In all cases, the amount of sol fraction that is extracted increases
Figure 2.7: Swelling ratios of poly(t-BA) networks prepared in uncovered molds as a function of volume fraction of t-BA in precursor solution Legend indicates the
42 solvent that the poly(¢-BA) networks were prepared in Networks were swollen in a) toluene, b) THF, c) benzene, or d) acetone for one week. a) 6 ă 1 "ấn ĩ T U b) 6 + Ls T 1 T
@ Networks prepared neat @ Networks prepared neat
@ Networks crosslinked in toluene m Networks crosslinked in toluene
@ Networks crosslinked in butanol @ Networks crosslinked in butanol
@ Networks prepared neat @ Networks prepared neat m@ Networks crosslinked in toluene m Networks crosslinked in toluene
5+@ Networks crosslinked in butanol 5- @ Networks crosslinked in butanol
Figure 2.8: Swelling ratios of poly(t-BA) networks prepared in covered molds as a function of volume fraction of ¢-BA in precursor solution Legend indicates the solvent that the poly(t-BA) networks were prepared in Networks were swollen in a) toluene,b) THF, c) benzene, or d) acetone for one week.
20 [4 Cyclohexane "“ $4 20 Fe Networks prepared neat 21
0 L® Butanoi | @ Networks prepared in toluene v Methanol 1 @ Networks prepared in butanol
Figure 2.9: Sol fraction of poly(t-BA) networks as a function of volume fraction of t-
BA in precursor solution Column i) Networks prepared in uncovered molds Column ii) Networks prepared in covered molds Legend indicates the solvent that the poly(t- BA) networks were prepared in Networks were swollen in row a) toluene, row b) THF, row c) benzene, or row d) acetone for one week. with decreasing volume fraction of t-BA in the precursor solution This can be at- tributed to the looser final network that is a result of more solvent present during the crosslinking reaction, which will allow the passing of more sol fraction during the extraction process Furthermore, at a certain critical volume fraction of t-BA in the precursor solution, the sol fraction begins to decrease This indicates that there must be an optimal volume fraction of t-BA in the precursor solution that allows for the best networks to be formed.
The incorporation of solvent in the precursor solution may have had an effect on the glass transition temperature, T,, of the poly(t-BA) networks The addition of increasingly more solvent during crosslinking often leads to a systematic decrease in
T, In order to determine if this was the case for our system, the Gordon-Taylor Equation [23,24] was used:
9 wi + kua (2.11)2.11 where w; are the weight fractions, T,, are the glass transition temperatures of the blend components in degrees Kelvin, and the subscript 2 refers to the component with the higher glass transition temperature The model specific parameter k is defined as p= Pitot PP (2.12)+ where /ứĂ are the densities of the blend components.
Normal force effects in parallel plate rheology
Rheological testing involves a precise manner in which the sample must be tested.
Over or underfilling results in edge effects Overfilling can result in the sample ma- terial migrating to the top of the geometry, effectively changing the inertia of the geometry If the sample is of low enough viscosity, however, this effect is less likely to happen Underfilling effectively changes the diameter of the plate geometry, and thus adversely affects the sheer stress factors, leading to large errors Slip is another worry for the experimenter Certain samples, especially hydrogels, are particularly susceptible to slippage The water in hydrogels can migrate to the surface, causing a layer of water between the bulk material and the geometry This can cause slipping at the interface As a result, crosshatched geometries are often used in order to cre-
52 ate more of a grip between the geometry and sample through a slightly roughened surface.
Another variable that the experimenter needs to consider is the normal force exerted on the sample The normal force can be controlled throughout the experiment in order to constantly adjust the normal force that the sample experiences throughout the total experimental time If this option is elected, however, the gap may change with experiment time Although much care is taken by the software to minimize the extent of head movement, it is unavoidable to completely take out any risk of gap change unless controlling the normal force option is turned off We tested the effect of normal force control throughout the experiment in order to assess the possibility of slippage.
Because the poly(t-BA) networks are tested in the dry state, the possibility of slippage is decreased It is a possibility that residual monomer, crosslinker, solvent, and sol fraction migrate to the interface of the geometry and the bulk network, causing a layer to be present at the surface of the network in contact with the upper geometry which may cause some slippage However, due to the elevated temperatures in which the rheological experiment is conducted, any such sol fraction most likely becomes vaporized It is more of a possibility, though, that the uneven nature of the surface of the network samples result in poor contact between the network surface and the geometry plate, especially when small normal forces are used In this case, the small force normal exerted on the network may not be great enough to deform the surfaces of the network to conform to the geometry.
It turns out that the difference in normal force control is very prominent in rheo- logical testing and can alter the modulus results based on what normal force value is chosen, so it is important to understand how to correctly choose which normal force to operate the rheometer at This is especially important in rubbery materials since these materials are deformable and have a definite force they can withstand In other words, too large of a normal force exuded onto the rubbery material may deform the sample in such a way as to force the sample to act as an overfilled sample, bulging from the sides of the geometry and possibly oozing onto the top of the geometry, which, as explained previously, could introduce error into the measurement Thus, it is important to understand how the normal force affects the sample being tested.
Rheological experiments were first done with a normal force of 2 N continuously applied during the course of the time sweep The results showed a general trend of increasing modulus, which is commonly seen in literature and is often referred to as a plateau This plateau is where the modulus values are taken from, since it is at these constant values that a network has formed However, it was thought that controlling the normal force may have been contributing to this gradual and constant increase in modulus For this particular experiment, as time increases, the geometry is lowered or raised in order to keep the normal force at a constant value of 2 N More often than not, the head is lowered, causing the sample to be compressed One thought was that this compression was actually deforming the sample, causing errors to arise in
100000 T T T T T y T T T T T T T T T 7 2.0 ơ wis pOOOOOOGOOOOOOOOOOOOOOOCOOOOCOOCOCOCCQd ® "NNnnnnnnu © 10000 + Xm= T 1° & z li NAAN And WVAAAWAAWAAVAAAAWVAWWAAVWAWAAAAWAAAVWAVAXAXN
Figure 2.14: Modulus results of poly(t-BA) networks crosslinked in uncovered molds and rheologically tested in parallel plate mode Normal force was not controlled during the time sweep, resulting in a modulus plateau value. the experiment A more realistic explanation is that the network itself may not have been strong enough to withstand this amount of normal force, causing the modulus to rise with time since the sample was continuously being compressed To see what the modulus results would look like if the normal force was not controlled, a time sweep experiment was performed without controlling the normal force throughout the duration of the experiment The results can be seen in Figure 2.14 and indicate that modulus decreased with time, and eventually found a real constant plateau value after the normal force ceased to change significantly These results demonstrate that when the normal force is uncontrolled during parallel plate rheological experiments for this network system, a definite plataeu can be attained It is also apparent from these results that a lack of normal force control for this network system does not induce slippage of any kind Thus, it is recommended that for the testing of this system of networks in parallel plate mode that the normal force be uncontrolled.
Table 2.2: Modulus and molecular weight between crosslinks for poly(t-BA) networks prepared with varying amounts of solvent present during crosslinking in uncovered molds Modulus was determined in parallel plate mode. do Eary (10° Pa) M, (g/mol)
Table 2.3: Modulus and molecular weight between crosslinks for poly(t-BA) net- works prepared with varying amounts of solvent present during crosslinking in covered molds Modulus was determined in torsion mode. o Eary (10° Pa) Mẹ (g/mol)
Model poly(tert-butyal acrylate) networks
There is a need for homogeneous polymer networks with controlled molecular struc- tures and mechanical properties However, this remains one of the most elusive goals to date The importance of uniformity throughout a crosslinked polymer network is immense, since they are used for many applications that depend upon a coordinated, controlled response, such as in drug delivery systems Such uniform networks are often dubbed model networks These types of networks are discussed in Section 3.1.1 Two different approaches to end-linking were explored First, the synthesis of a, w- vinyl-terminated telechelic macromonomers based on poly(tert-butyl methacrylate
(poly(t-BMA)) and poly(methacrylic acid) (poly(MAA)) was studied with the aim of preparing end-linked gels and hydrogels Second, an azide telechelic macromonomer based on poly(¢-BA) with a cleavable functionality at its center was synthesized in order to produce degradable model networks.
Introduction © cv và kg kg va 58
Model Networks 0 20.0.0 0000000000 0G 58
Free-radical crosslinking, being a random process, cannot provide precise control over the microscopic structure of the networks formed Furthermore, free-radical crosslink- ing fails to provide any accurate knowledge of the resulting network structure The crosslinks in such structures are formed in an uncontrolled manner and ultimately results in a heterogeneous pore size The final result is that direct comparisons be- tween experimental data and those obtained from theoretical models described in literature are difficult to achieve In light of these difficulties, it is desirable to pro- duce networks of more controllable properties That is why model networks are of great interest, since they allow for the testing of the validity of the various developed theories by establishing precise relationships between the structure of the network and its properties.
One method used to synthesize model networks involves the concept of endlinking well-defined linear polymer precursor chains, called a model macromonomer [25].
Figure 3.1: Types of network defects a) Dangling chains b) Closed loops c) Entanglements.
This macromonomer is of known length and, ideally, narrow polydispersity It also has known functionality, and, hence, the pore size of the network can be chosen in advance Under these assumptions, the network structure should be homogeneous.One sustaining disadvantage of model networks, however, is that they are not entirely void of defects Common network defect are shown in Figure 3.1.
Click Chemistry 2 2 ee àya 59
“Click chemistry” is a term that was introduced in 2001 by K Barry Sharpless and describes chemistry that emulates nature by joining small units together quickly and reliably This chemistry relies on a modular combinatorial set of reactions that are
“wide in scope, give very high yields, generate only inoffensive byproducts that can be removed by nonchromatographic methods, and [are] stereospecific” [26] Further- more, click reactions are selective, exothermic, and occur under mild conditions where no solvent, or at the very worst a benign solvent (such as water), is used Although
Figure 3.2: The copper(I) catalyzed Huisgen 1,3-dipolar cycloaddition reaction. several reactions are categorized as click reactions, it is the Huisgen 1,3-dipolar cy- cloaddition reactions, in which an azide and a terminal alkyne undergo cycloaddition to yield a 1,2,3-triazole, that are often referred to as the best example, as seen in Fig- ure 3.2 Even better are the copper(I) catalyzed 1,3-dipolar cycloaddition reactions, since the polarization of the terminal triple bond by the covalently bound copper(I) catalyzes the cycloaddition [27] An advantage of this reaction is the fact that the azide is easy to introduce and is nearly devoid of side reactions [27] Although 1,3- dipolar cycloaddition of azides and alkynes was discovered in the 1960s, it has been heavily underutilized possibly due to the fact that azides can be extremely dangerous to work with However, it is currently receiving renewed interest due to the vast possibilities that click chemistry brings.
The appeal of click chemistry to making end-linked networks is already under- way Ossipov and co-workers [28] introduced azide and alkyne pendant groups onto poly(vinyl alcohol) (PVA) and poly(ehtylene glycol) (PEG) in order to investigate the possibility of hydrogel formation by click chemistry It was found that hydro- gels did form by the chemoselective 1,3-cycloaddition between the alkynyl and azido functional groups of PVA yielding the formation of triazole crosslinks.
Experimental Procedure 2 0.0 eee ee ee ee 61
Synthesis of a,w-terminated vinyl acrylate end-linked networks 61
The a, w-terminated vinyl acrylate macromonomers were synthesized via atom trans- fer radical polymerization (ATRP) by Dr Lucy Vojtova Benzene was dried by
CaH; and sealed under argon gas The precursor solution ratio for making end- linked poly(t-BMA) crosslinked networks consisted of 0.4 g a,w-terminated vinyl poly(t-BMA) macromonomer, 0.0053 g photoinitiator 2,2-dimethoxy-2-phenyl ace- tophenone (DMPA), and 2 mL dry benzene The precursor solution ratio for mak- ing end-linked poly(MMA) crosslinked networks consisted of 0.4 g a,w-terminated vinyl acrylate poly(MMA) macromonomer, 0.0053 g photoinitiator DMPA, and 2 mL methanol/water solvent (1:1 ratio of methanol to water) These ratios of reactants remained constant although the absolute values may have varied In all cases, the reactants were placed in a sealed vial and flushed with argon gas Within an inert argon or nitrogen environment, the monomer solution was dropped onto a 1 cm x 1 cm silicon wafer or an aluminum dish and exposed to 365 nm for 45 minutes using aBlak-Ray B-100A/R UV lamp.
Synthesis of a,w-azido poly(t-BA) model networks
The a,w-azido poly(t-BA) macromonomer was synthesized via ATRP by Jeremiah Johnson The macromonomer was placed in a vial and the copper catalyst (CuBr, Cul, CuSO,, or CuBr(PPh3)3, ten equivalents to alkyne) was added The vial was evacuated for five minutes and backfilled with argon Degassed solvent (DMF or toluene, 0.5 mL per gram of macromonomer) was then added, followed by crosslinker
(0.10 Min DMF) Sodium ascorbate (4 M in HạO, twenty equivalents) and alkyl amine (PMDETA or DIEA, ten equivalents) were then added, if necessary The vial was immediately placed in an oven set at a temperature of 80°C.
Cleavage of tert-butyl ester group on the a, w-terminated vinyl
A 0.2% (wt/volume) of photoacid generator (PAG) in acetone was applied via a pipette to the surface of a prepared a,w-terminated vinyl poly(t-BMA) network.
The PAG solution was allowed to dry, and then the network sample was exposed to 254 nm UV light for twenty minutes, then washed with acetone and DI water.
The chemical modification yielded a final product that was an a,w-terminated vinyl poly(MAA) network.
Swelling of a,w-terminated vinyl acrylate end-linked networks 63
The networks were characterized by experimentally determining the swelling ratio. Prepared networks were placed in a 20 mL vial The vials were filled with acetone,
THF, or water and the networks were allowed to swell The excess solvent was removed from the gel using filter paper, and the swollen weight was measured (Mettler, Model AE100) The networks were then dried overnight in a vacuum oven at room temperature (Fisher Scientific Isotemp Vacuum Oven, Model 280A) The weight of the dried networks was then measured This procedure was repeated until there was no longer a change in both the swollen and dry weight Swelling ratio experiments were done in triplicate, with + SD.
Modulated differential scanning calorimetry of a, w-azido poly(t- BA) macromonomer 0000s eee eee 63
Modulated differential scanning calorimetry (DSC) (TA Instruments, New Castle,DE) was performed using a Q100 The DSC was calibrated according to the operating manual prior to any experiments 8.5 mg of a,w-azido poly(¢-BA) macromonomer was weighed (Mettler, Model AE100) and placed in an aluminum DSC pan The pan was then crimped and placed in the DSC cell on the front stage, along with a crimped reference pan, which was placed on the rear stage A temperature ramp was
64 performed on the sample The sample was allowed to equilibrate to -50°C, and was held at that temperature for five minutes Then, a temperature ramp was applied at a rate of 5°C/min to a final temperature of 100°C The sampling rate was 0.10 sec/pt and the temperature modulation was +1.00°C every 60 seconds The gas purge was nitrogen at a rate of 50 mL/min.
Swelling of a,w-azido poly(t-BA) model networks
The networks were characterized by exerimentally determining the swelling ratio. Prepared networks were cut into smaller pieces using a razor blade Each piece was placed in a 20 mL vial The vials were filled with acetone and the networks were allowed to swell for one week The solvent was changed daily After one week, the excess solvent was removed from the gel using filter paper, and the swollen weight was measured (Mettler, Model AE100) The networks were then dried overnight in a vacuum oven at room temperature (Fisher Scientific Isotemp Vacuum Oven, Model 280A) The weight of the dried networks was then measured Swelling ratio experiments were done in triplicate, with + SD.
Rheology of a, w-azido poly(t-BA) model networks
Rheometric testing was done using an AR2000 (TA Instruments, New Castle, DE).
The Peltier plate was used in conjunction with a 20 mm stainless steel parallel plate.The solution of a, w-azido poly(t-BA) macromonomer, CuBr, DMF, crosslinker, sodium ascorbate, and alkyl amine was pipetted onto the Peltier plate Time sweeps of one hour were performed at a strain % of 1 and a frequency of 1 Hz.
End-linked networks were formed from a, w-terminated vinyl poly(¢-BMA) and poly(MAA).
These networks were characterized by measuring the swelling ratio, defined as:
:=— 3.1 Yi 8 where Q, is the swelling ratio, W, is the weight of the swollen network, and W, is the weight of the dried network Table 3.1 gives the results of the swelling experiments. The results show that the end-linked poly(t-BMA) networks swell greater in THF than in acetone This corroborates the fact that THF is a better solvent for ¿-BMA than acetone, causing the poly(t-BMA) networks to swell to a greater extent in THF. The results show that the poly(MAA) end-linked networks are, indeed, hydrophilic.
Furthermore, the swelling ratio value of the end-linked poly(MAA)networks matches the swelling ratio of the poly(t-BMA) networks that were treated with PAG, which cleaves the tert-butyl group The fact that these two swelling ratio values are so similar gives tremendous weight to the ability to chemically modify this system of networks from hydrophobic materials to hydrophilic materials In this way, PAG can
Table 3.1: Swelling ratios of a,w-terminated vinyl acrylate networks.
Macromonomer Swelling solvent Q, poly(¢-BMA) Acetone 4,42 poly(t-BMA) THF 10.2 poly(MAA) Water 32.7 poly(MAA) (after PAG treatment of Water 34.1 poly(¢-BMA) network) be use to pattern these networks, and create areas of hydrophobicity and areas of hydrophilicity to meet any design parameters desired.
The fact that all networks were insoluble in their respective solvents indicate that end-linking did occur, although more experimental experiments would be required in order to assess the exact nature of the structure of each network.
3.3.2 Modulated differential scanning calorimetry of a, w-azido poly(£-BA) macromonomer
Modulated DSC was performed on the a,w-azido poly(t-BA) macromonomer The results can be seen in Figure 3.3 Modulated DSC deconvolutes the total heat flow into reversible heat flow and non-reversible heat flow In the reversible heat flow curve, the glass transition temperature can be seen more clearly due to the subtraction of the
" Reversing Tg H Lo 3 Want re eee ee eee L 0.05 + 2
Exo Up Temperature (°C) Universal V3.7A TA Instruments
Figure 3.3: Modulated differential scanning calorimetry of a,w-azido poly(t-BA) macromonomer. enthalpic relaxations, whose fingerprint can be seen in the non-reversible heat flow curve In this way, it can be more clearly seen that the glass transition temperature of the a, w-azido poly(t-BA) macromonomer is 43.4°C.
3.3.3 Formation of model networks via copper(I)-catalyzed azide-alkyne cycloaddition
As mentioned in Section 3.1.2, the most powerful click reaction is the copper(I)- catalyzed 1,3-dipolar cycloaddition between an azide and an alkyne It is this reaction
68 that was exploited to make novel model poly(t-BA) networks The general scheme is shown in Figure 3.4 Specifically, the copper(I)-catalyzed azide-alkyne cycloaddition (CuAAC) reaction was employed between an a,w-azido poly(t-BA) macromonomer and either a tri- or tetra-functional alkyne [29].
The original reaction conditions involved CuBr / PMDETA / NaAsc / DMF.
However, a variety of CuAAC reaction conditions were evaluated The time required for insoluble gel material to form at 80°C was evaluated Ultimately, the system CuBr / PMDETA / NaAsc / DMF gave the fastest crosslinking and the best results The
CuBr / DMF (no PMDETA) and Cul / N,N-diisopropylethylamine / DMF systems yielded insoluble networks, but required an overnight reaction The use of toluene as a solvent for the poly(t-BA) macromonomer as opposed to DMF proved equally effective, but gave longer reaction times of hours instead of minutes Table 3.2 shows the various CuAAC reaction conditions that were investigated [29].
3.3.4 Swelling of copper(I)-catalyzed azide-alkyne cycloaddi- tion model networks
The swelling ratios of the gels were calculated using the Equation 3.1 Table 3.3 shows the different molecular weights of poly(t-BA) macromonomer that were synthesized via ATRP and whose swelling ratios were investigated The macromonomer was | crosslinked with either a tri-functional acetylene or a tetra-functional acetylene The molecular weight of the macromonomer and the crosslinker functionality are expected
Ng WK >A ° AY CuBr, PMDETA o^o ° ©0.4 M sodium ascorbate/H,O
Figure 3.4: The copper(I) catalyzed Huisgen 1,3-dipolar cycloaddition reaction be- tween a, w-azido poly(t-BA) macromonomer and a tetra-functional alkyne. to alter the swelling ratios of each network As can be seen, the results indicate that the swelling ratio does, in fact, increase with macromonomer molecular weight. This increase in swelling confirms that the pore size in the resulting model network structure becomes larger with higher molecular weight macromonomer Also, the swelling ratio decreases with increasing functionality of the crosslinker, indicating that tighter networks are formed when higher functionality crosslinkers are used These results indicate that this system of networks can be engineered based on specific design parameters for any application.
Table 3.2: Conditions for a,w-azido poly(t-BA) network synthesis.
M, ross- Catalyst Solvent Alkyl Reducing Material linker Amine Agent Appearance
14,400 3 CuBr DMF PMDETA Na Ase Hom. 14,400 4 CuBr DMF PMDETA Na Asc Hom. 10,600 4 CuBr DMF none Na Asc Hom. 10,600 4 CuBr DMF PMDETA Na Asc Hom. 10,600 4 CuBr(PPh3)3 DMF DIEA Na Asc Het.
10,600 4 Cul DMF DIEA Na Asc Hom. 10,600 4 CuSO,.H2,0 DMF none Na Asc Het.
10,600 4 CuBr DMF none none Hom. 10,600 4 CuBr DMF PMDETA none Hom. 10,600 4 CuBr(PPh3)3 DMF DIEA none No reaction 10,600 4 Cul DMF DIEA none Hom. 10,600 4 CuBr Toluene none none Hom. 10,600 4 CuBr Toluene PMDETA none Hom. 10,600 4 CuBr(PPh3)3 Toluene DIEA none No reaction 10,600 4 Cul Toluene DIEA none Hom. 5,780 3 CuBr DMF PMDETA Na Asc Hom. 5,780 4 CuBr DMF PMDETA Na Asc Hom. 8,790 3 CuBr DMF PMDETA Na Asc Hom. 8,790 4 CuBr DMF PMDETA Na Asc Hom.
3.3.5 Gel point of copper(T)-catalyzed azide-alkyne cycload- dition model networks
Copper(T)-catalyzed azide-alkyne networks were found to also form by simply mixing the reactants together at room temperature in a vial In light of this discovery,rheology was used to study the real-time gelation of the CuAAC networks A time sweep was performed on the a,w-azido poly(t-BA) macromonomer / CuBr / DMF
Table 3.3: Swelling ratios of a,w-azido poly(t-BA) networks Networks were com- posed of various molecular weight macromonomers, and crosslinked with either a tri-functional or tetra-functional crosslinker.
Mạ tri-functional tetra-functional
/ crosslinker / Na Asc / PMDETA system, and the crossover point was monitored.