THE EFFECT OF SOLUTE AND MEMBRANE POLARITY · PDF fileTHE EFFECT OF SOLUTE AND MEMBRANE POLARITY IN CREATING A PSEUDO-ZERO-ORDER CONTROLLED RELEASE PROCESS ... membrane and therefore

[5 -1]

THE EFFECT OF SOLUTE AND MEMBRANE POLARITY IN CREATING A

PSEUDO-ZERO-ORDER CONTROLLED RELEASE PROCESS Ruchita Balasubramanian, Sakshum Chadha, Charmaine Chew, Joseph Da, Sona Dadhania,

Abhinav Karale, Grace Kresge, Meilin Lu, Victoria Ou, Ram Vellanki, Peter Zhou

Advisor: Dr. David Cincotta

Assistant: Tony Chen

ABSTRACT

In a pseudo-zero-order controlled release, the rate of diffusion is independent of time and

concentration. In this experiment, the team analyzed the rate of diffusion for strong electrolytes

(NaCl and CaSO4), weak electrolytes (citric acid, ascorbic acid, and acetylsalicylic acid), and

various alcohols (propanol, ethanol, and methanol) across polymer membranes of varying

compositions of ethylene-vinyl acetate (EVA). The change of concentration in the target area

was recorded using changes in conductivity (strong electrolytes), pH (weak electrolytes), and

gravimetric analysis (alcohols). Increasing the percent composition of EVA in the films

increased the polarity of the film, which should have theoretically increased the membranes’

permeability to polar substances. The experiments showed that no ionic salts tested passed

through the membranes for up to 40% EVA composition, and only citric acid was able to diffuse

for weak electrolytes. All the moderately polar alcohols diffused successfully. These data suggest

that only substances with intermolecular forces comparable to that of EVA can dissolve into the

membrane and therefore diffuse through it.

INTRODUCTION

Zero-Order Controlled Release

Ideally, the controlled release of a specific substance should be a zero-order process,

which describes when the rate at which a substance is transported into its surroundings remains

constant and independent of the concentration. This type of process is often sought after in

pharmacy; medicines should be delivered to the body at a constant rate over a period of time.

However, this is hard to achieve in practice, as the rate of delivery often slows and ultimately

levels off with time as the concentration of the substance in the surroundings increases (1). This

concentration-dependent process is considered first-order. Achieving zero-order controlled

release requires one to disregard the concentration component and release the same amount of

substance at the same rate over time.

Active zero-order controlled release is often achieved using a mechanical pump, as in

intravenous medical fluids. However, there is an increasing urgency to achieve passive zero-

order controlled release that has no mechanical component, because mechanical pumps tend to

be large and inconvenient. Previous mechanisms for such results have included reservoir systems

in which materials diffuse constantly through a polymer matrix when placed in an aqueous

environment (1). Other mechanisms include a pseudo-zero-order release mechanism that relies

on a first-order release kept at a constant concentration through compartmentalization.

[5 -2]

Figure 1: Representation of Uncontrolled Release versus Controlled Release This

representation, produced by Alzet (2), shows the difference between uncontrolled release of

chemicals, as normally done by pills, and the controlled release of chemicals achieved through

pseudo-zero-order diffusion.

Zero-order processes are often sought after in medicine because to be effective, drugs

should be delivered to a patient at a constant rate. Most medicines, however, are given in doses

that create an initial spike in concentration at the time of the dose, and then fall off until the time

of the next dose (Figure 1). Occasionally this dosage system can bring the concentration of the

drug into a cytotoxic range, which can harm the patient. A zero-order controlled release

mechanism is most ideal for such situations in order to keep the medication at a therapeutic

constant (1).

Fick’s Law and Diffusion

Diffusion constitutes the movement of a specific substance through a membrane along a

concentration gradient until the process reaches equilibrium. This process is governed by Fick’s

Law of Diffusion, as stated in the equation

(1)

where J is the flux, D is the diffusion constant, and

is the change in concentration over the

thickness of the membrane. As long as the concentration of the substance remains constant, D is

proportional to the flux, and simplifies to the following equation:

(2)

In this particular equation, J is the diffusion constant, the ∆C represents the concentration

gradient, and the ∆x represents the thickness of the membrane. This equation provides the

relationship between the concentration gradient, the thickness, and the rate of diffusion. By

[5 -3]

maintaining a constant concentration gradient the rate of diffusion remains constant and pseudo-

zero-order controlled release is effectively achieved.

Polymer Membrane Films

Ethylene vinyl acetate (EVA) is a copolymer that was used to create the membranes used

in this experiment. It is comprised of polyethylene and vinyl acetate, which can form a uniform

polymer membrane when cast using a solution of toluene (3). This makes a semipermeable

membrane that can filter particles based on a variety of parameters, such as particle size, polarity,

crystallinity, and intermolecular forces between membrane and particle (4). This flexibility

allows us to collect a wide range of data by testing different variables.

Polarity

Polarity describes the presence of a dipole with opposite charges on each end of the

molecule. The difference of polarity between a molecule and the polymer membrane affects the

rate of diffusion across the membrane, because polar substances will only diffuse through polar

membranes and the same with nonpolar substances and membranes. This phenomenon can be

attributed to the solution-diffusion model, where the solute must first dissolve into the membrane

and then move down its concentration gradient (6). Furthermore, when the polarities of

substances are similar, the rate of diffusion across a membrane is faster, and vice versa.

Polyethylene by itself is nonpolar due to its lack of polar groups, such as acetyl or alcohol

groups (3). Increasing the concentration of vinyl acetate in ethylene vinyl acetate, EVA, (Figure

2) increases the polarity of the copolymer. As the polarity of the membrane approaches the

polarity of the solute, the diffusion rate is expected to increase (7). Since electrolytes have a net

ionic charge, a more polar membrane would allow for a greater chance of diffusion, and would

increase the rate of diffusion.

Figure 2: Structure of Polymer Membranes These polymers were used to make semi-

permeable membranes in this experiment. (a) The molecular structure of polyethylene, which is a

nonpolar molecule. (b) The molecular structure of the copolymer ethylene vinyl acetate (EVA).

EVA is made of polyethylene and vinyl acetate. Because vinyl acetate is a polar side chain, as

[5 -4]

the percentage of vinyl acetate increases in an EVA mixture, the polarity of the mixture increases

as well (6).

However, more polar membranes are more difficult to manage because of the adhesive

properties of pure EVA. Therefore, the concentration of EVA must be optimized to allow for the

maximum rate of diffusion without sacrificing manageability.

Adding copolymers such as polyethylene vinyl alcohol or polyethylene glycol can also

vary polarity; this would ideally allow more polar substances to diffuse through them more polar

membrane. However, in order to add these copolymers, they must be compatible with the EVA

solvent (toluene) and with the EVA itself in order to create a uniform, workable membrane (3).

Hansen Solubility Parameters

To predict which polymers could create a solution needed to cast a film, the team referred

to the Hansen Solubility Parameters by Dr. Charles Hansen. The Hansen Solubility Parameters

describes all molecules in terms of their primary intermolecular forces: dispersion forces, dipole-

dipole interactions, and hydrogen bonding. A molecule, with its collection of the three

intermolecular interactions, can be mapped as a spherical volume on a three-dimensional plane

based on the values of the forces (8).

An equation developed by Dr. Klemen Skaarup calculates the solubility difference

between two molecules given their respective solubility parameter components (8).

(Ra)2 = 4(δD1-δD2)

2 + (δP1-δP2)

2 + (δH1-δH2)

2 (3)

The relative energy difference, or RED, is used to predict the solubility of two different

molecules. Using the Ra value calculated above, the RED can be calculated using the equation:

RED = Ra/Ro (4)

where Ro is the experimentally determined radius of the sphere of solubility for the solvating

compound (8). If the RED is less than 1, the two compounds will form a solution. If the RED is

equal to 1, the two compounds will be partially soluble with one another. If the RED is greater

than 1, the two compounds will not form a solution.

The Hansen Solubility Parameters are an attempt to quantify and predict solubility and

may not always be accurate. Solutions were experimentally tested in order to support or refute its

predicted solubility (8).

Previous Research

Team 5 of the New Jersey Governor’s School in the Sciences has been experimenting

with controlled release kinetics for several years. The 2012 team diffused saturated citric acid

through two different types of membranes: 10% EVA and 12% EVA (7). The concentration and

diffusion of citric acid was kept constant due to the continuous dissolution of the extraneous

solid citric acid added to the saturated solution. The solid citric acid continuously dissolved into

the saturated solution as the aqueous solution diffused across the membrane, maintaining a

[5 -5]

constant saturated concentration. Although there were some inconsistencies in their data, the

2012 team showed that it was possible to achieve a controlled release using this method (7).

However, during their first experiment, the team found that NaCl and maleic solutions were

unusable as solutes due to their irregular diffusion rates through the 10% EVA film. In 2013, the

team started to produce EVA films as opposed to the factory-made ones in 2012 (5). EVA was

dissolved into solutions; 12% EVA was dissolved in xylene while 25% and 40% EVA were

dissolved in toluene. Films were casted onto silicon paper attached to glass, and a doctor blade

was used to ensure consistency for width of the film. The 2013 team performed experiments that

tested the effects of polarity and crystallinity; they concluded that non-polar substances will

diffuse through polymer membranes using a vapor pressure analysis (5).

Hypothesis

A semi-permeable polymer membrane can be used to model the diffusion of a substance

(electrolytes or alcohols) with a controlled constant concentration at pseudo-zero-order

controlled release. Furthermore, increasing the membrane polarity will allow more polar

molecules to dissolve into the membrane and therefore diffuse down the concentration gradient

at a faster rate.

METHODS & MATERIALS

Membranes

For this experiment, it was essential to adjust the polarity of a membrane in order to

control the rate of diffusion. EVA is a copolymer composed of two monomers, ethylene and

vinyl acetate. Ethylene is a nonpolar molecule, while vinyl acetate is a polar molecule. The

varied monomers give EVA an interesting property; the regions with ethylene are nonpolar while

the regions with vinyl acetate are polar. By increasing the percentage of vinyl acetate in a

membrane, the polarity of the membrane increases. Therefore, in order to change the polarity of

the membranes, the percentage of vinyl acetate can be increased or decreased (5).

Several semipermeable EVA membranes were created with percentages of vinyl acetate

varying from 25 to 40%. In order to increase polarity, there was speculation about the addition of

99%+ Polyvinyl Alcohol (PVA) in low concentrations to allow electrolytes to diffuse through

the membrane at a greater rate. However, a proper solvent for both 40% EVA and 99%+ PVA

was not found and as a result, films were not made with PVA additive. The application of

Hansen Solubility Parameters further confirmed that PVA and EVA are incompatible in terms of

solubility. In addition to the potential of PVA, there was also potential for polyethylene glycol

(PEG) to serve as an additive to the membrane. Both a 1:2 mass ratio of PEG to EVA in addition

to a 50% solid solution of PEG and a lower percentage by solids of EVA have been tested for

membrane integrity and evenness. All films containing PEG had more cracks than membranes

containing only EVA, and also displayed clusters of PEG due to its low molecular weight. As a

result, varying concentrations of membranes containing only EVA were used to create

membranes.

Several techniques exist to create membranes. In this experiment, a mixture of 30% EVA

by mass was dissolved with toluene in order to create a gel-like substance. This solution was

[5 -6]

loaded into a Doctor Blade, an apparatus used to cast the film at a uniform thickness. The Doctor

Blade could produce films at 25, 30, 40, and 50 thousandths of an inch in thickness, which

allowed the team to create versatile membranes.

Prior to being loaded, however, it was determined through trial and error that the mixture

of EVA and toluene had to be stirred rapidly and subjected to heat to prevent the mixture from

solidifying. This was accomplished by occasionally heating the covered solution.

In order to cast the film, a sheet of release paper was clipped onto a sheet of glass to

provide a flat surface. The Doctor Blade was then set to the appropriate width, and the solution

was slowly poured into the Doctor Blade as it was dragged across the release sheet to create an

even film. Through experimentation, it was also determined that the Doctor Blade and release

paper needed to be heated to prevent the liquid from solidifying as it was poured into the Doctor

Blade. This was accomplished by placing the Doctor Blade and release paper into an oven that

was heated to approximately 80 degrees Celsius before use.

The membranes themselves had two properties that could be tested; polarity and

thickness. However, dealing with two variables in the membrane proved to be difficult in

experimentation, and thicknesses lower than 40 thousandths of an inch proved to exhibit more

pinholes and more stickiness than 40 thousandths of an inch. Therefore, thickness was kept

constant at 40 thousandths of an inch. Only membranes of various polarities (30%, 35%, or 40%

EVA) were tested with various solutes.

Solutions

Choosing Electrolytic Solutions

In order to test the capacity of a membrane to achieve pseudo-zero order kinetics, it is

important to use a solute that allows for a slow release mechanism. The experiment required the

creation of a saturated solution in the petri dish. The following solutes were chosen and tested in

the apparatus:

NaCl: This highly soluble ionic compound is relatively small in size and has some

medical applications in that 0.9% sodium chloride is the most common intravenous

medical infusion solution.

CaSO4: Although it is still considered a strong electrolyte, it is less soluble and has less

intramolecular ionic attractions as most highly soluble electrolytes.

Acetylsalicylic Acid (Aspirin): This molecule has a lower solubility and polarity that

should make it diffuse more easily. It also has medical applications.

Ascorbic Acid: This is highly polar and more soluble in comparison to other organic

weak electrolytes, but should still diffuse efficiently through a polymer membrane.

Citric Acid: This is a much more polar weak organic electrolyte, so this should display

the variation between molecules of varying polarity.

[5 -7]

All the weak organic acids have a similar size and molecular mass, so any change in diffusion

rate should be strictly based on solute and membrane polarity.

After creating saturated solutions from these solutes, the solid solute particles in the

saturated solution effectively replenished the diffused solute particles by consistently dissolving

into the solution until the saturation point was reached. This system maintained a constant

concentration gradient in the petri dish and thus achieved pseudo-zero-order kinetics in

accordance with Fick’s Law.

Choosing Alcohols to test with Gravimetric Analysis

In a test to determine if mostly nonpolar substances would diffuse through a polymer

membrane, certain alcohols with high vapor pressures and low boiling points were also tested

using a gravimetric analysis of vapor pressures. Since the vapor pressure of a liquid in a closed

container is kept constant at a constant temperature, it will still establish an analogous

concentration gradient as that of a saturated aqueous solution, in which the evaporated vapor will

diffuse through the polymer membrane at a constant rate. The following organic alcohols were

tested using vapor pressure gravimetric analysis:

Methanol: Methanol is a very small and polar organic alcohol that should diffuse through

a polymer membrane rather easily. It has a high vapor pressure (13.02kPa at 20°C), and a

low boiling point such that it can form a distinct concentration gradient.

Ethanol: Although it is still polar, this molecule is much larger and has a lower vapor

pressure (5.95kPa at 20°C). It has a lower polarity, so it should diffuse more easily

through less polar membranes.

Propanol: This is a mostly non-polar molecule with a slightly polar hydroxyl group on a

larger molecule. It also has a much lower vapor pressure, so it should diffuse faster

through less polar membranes.

Because the team used pure alcohols for diffusion, vapor pressures were kept constant by

Raoult’s Law. Thus, any amount that diffuses out of the container through the membrane will be

replaced by the evaporation of the liquid, keeping a constant concentration gradient of the

gaseous alcohol across the membrane. This will correlate to pseudo-zero-order controlled release

according to Fick’s Law.

Apparatus

The stable apparatus suspended the petri-dish/membrane mechanism in the reservoir

beaker at a chosen height. Using two straight wires, the team created a cage for the petri-dish

(Figure 3). The wire segments were intentionally longer than the diameter of the reservoir to

bend them over the reservoir edges and maintain further stability. Also, the hanging wire could

be adjusted for any height in the reservoir. Then, initially with silicon glue, the team fastened the

membrane to the petri-dish. However, it was later determined that the silicon glue released

ammonia, which affected the pH readings. As a result, the team substituted the silicon glue with

the 40% EVA polymer as a glue. The part of the cage that was below the petri-dish was slightly

[5 -8]

depressed, so that it did not touch the membrane itself. The entire mechanism, pictured below,

was inserted into the reservoir. To create the concentration gradient between the solution in the

petri dish and the reservoir, the apparatus was adjusted such that the mechanism was barely in

contact with the water. Conductivity/pH probes were placed in the side of the reservoirs to take

measurements. Water was poured in until both the petri-dish and reservoir had the same water

level to avoid the effects of hydrostatic pressure.

Figure 3: Aqueous Diffusion Apparatus The membrane was glued to a petri dish using excess

40% EVA solution (a), and then submerged into a beaker and suspended by metal wires such

that the membrane is in contact with deionized water (b). Conductivity and/or pH probes were

placed into the beaker and connected to a Vernier device (c) so data could be collected over time.

Gravimetric Vapor Pressure Apparatus

A similar apparatus was also used to prepare the alcohols for gravimetric vapor pressure

analysis. About 25 mL of the selected alcohol was sealed into a plastic petri dish using liquid

EVA as an adhesive, thus creating a closed container with a constant vapor pressure. The petri

dishes were kept in an incubation oven at a constant temperature of 30°C. The entire apparatus

was massed periodically so as to plot the change in mass in the container, which directly

correlates to the rate of diffusion over time.

[5 -9]

Measurements

Conductivity/pH

Because the dissociated ions of the electrolytic solutes had charge, their presence after

diffusing through the membrane could be detected by a Vernier conductivity probe. A maximum

of four probes were connected to a Vernier, which was set to record conductivity and/or pH

readings every 5 minutes for a maximum of 100 hours. By recording the change in conductivity

at five-minute intervals, the team could track the changing concentration of the electrolytic

solution in the reservoir. If the change in conductivity was linear, the solute diffused at a pseudo-

zero-order rate. In the case of weak electrolytic solutions (including organic acids), conductivity

could not be accurately measured because the organic molecules did not dissociate into ions to a

significant extent. Thus, pH was measured using a Vernier pH probe, because pH directly

correlated to an increase in internal concentration of the acid.

The probe had to be calibrated to the team’s electrolytic solutions because expected

reading values found online did not account for the team’s specific lab conditions. First, the pH

and conductivity probes were calibrated with stock solutions of known pH/conductivity to

maintain consistency among all of the probes. Then, the team performed serial dilutions for the

various electrolytic solutions. The maximum concentration used was 0.1M because the probe

was unable to accurately measure concentrations greater than 0.15M. The team checked the

readings for various concentrations, from 0.1M to 0.0001M. Conductivity readings in

microSiemens per centimeter were plotted against molarity on Microsoft Excel, and a line of best

fit was derived. A similar test was done for pH.

(a) Calibration Curve for Sodium Chloride

[5 -10]

(b) Calibration Curve for Calcium Sulfate Dihydrate

Figure 4: The Calibration Curves for Sodium Chloride and Calcium Sulfate Dihydrate These calibration curves for sodium chloride NaCl (a) and calcium sulfate dihydrate CaSO4

CaSO4 2H2O (b), which were made prior to the experiments, relate concentration of solution to

conductivity in microSiemens per centimeter. If there was an increase in concentration, the

conductivity should increase linearly.

Vapor Pressure Gravimetric Analysis

In a separate experiment, nonpolar alcohols had also been tested to diffuse through the

polymer membrane using a vapor pressure gradient. By Raoult’s Law, the vapor pressure of a

liquid in a closed container will stay constant at a constant temperature, thus creating a

concentration gradient of gaseous alcohols across the polymer membrane. The vapor will diffuse

through the polymer membrane at a constant rate, thus losing mass to the atmosphere. Hence, a

gravimetric method was used to affirm that the vapor diffused at zero-order controlled release.

The sample was massed periodically and plotted on an Excel graph to display the decrease in

mass. If the average change in mass was linear, the process occurred at zero-order controlled

release.

Technical Limitations

Membranes

Errors during the making of the membranes may lead to unexpected failures during the

experiment. Since the EVA solutions used to cast the films solidify easily at room temperature,

the solutions, along with the Doctor Blade and the glass plates, had to be kept warm at all times.

If anything was cold, the film turned out uneven or solidified during the casting process.

Sometimes, holes appeared in the films; these holes could be caused by several factors, including

drawing the films on uneven release paper. If there were air bubbles trapped in the EVA

solutions, those bubble sometimes were visible in the finished films. Also, if the films were too

[5 -11]

thin or sticky, they caused leakages in the apparatus. Membrane makers were very careful to

avoid these problems.

Apparatus

The wires that constructed the cages of each apparatus were not completely straight due

to the limitations of mechanical equipment. Also, the glue that held the membrane to the petri

dish did not always make the seal airtight, so some of the solution may have diffused out through

a leak. Another error in the conductivity readings could have resulted from the fact that the

apparatus could not be sealed off from its surroundings. The minimal increase in the

concentration, despite the lack of diffusion of solute through the membrane, may be a result of

the CO2 in the air reacting with water to create H2CO3.

For the pH tests, the main problem was that the pH increased, which contradicted the

predicted decrease in pH. The team later discovered that the silicon glue that bound the

membrane to the petri dish was increasing the pH of the solution by releasing NH3 into the

solution.

Thickness

Even though each slide was put into the oven with the same thickness, each was removed

from the oven at varying thicknesses because each stock solution of 30%, 35%, and 40% EVA

had slightly different amounts of toluene added. Since varying thicknesses affect the diffusion

rate of molecules, thickness is a moderately significant variable.

[5 -12]

RESULTS

Strong Electrolytes

(a) Kinetics of Sodium Chloride Diffusion

(b) Kinetics of Calcium Sulfate Dihydrate Diffusion

Figure 5: Strong Electrolyte Diffusion Experiments The figures refer to the change in

conductivity in the reservoir for NaCl (a) and CaSO4 2H2O (b) in microSiemens per centimeter

vs. hours. The general erratic pattern can be attributed to chatter in the conductivity probe.

Generally, this data does not display a linear increase in conductivity that would have been

present in a zero-order release.

[5 -13]

Weak Electrolytes

(a) Kinetics of Ascorbic Acid Diffusion

(b) Kinetics of Acetysalisylic Acid Diffusion

[5 -14]

(c) Kinetics of Citric Acid Diffusion, Experiment 1

(d) Kinetics of Citric Acid Diffusion, Experiment 2

Figure 6: Diffusion of Weak Electrolytic Solutions The increase in hydronium ion

concentrations in the beaker for ascorbic acid (a), acetylsalicylic acid (b), and citric acid (c,d)

were calculated based on the change in pH in the general reservoir. Citric acid was conducted

twice for reproducibility.

[5 -15]

(a) Zero Order Representation of Citric Acid Diffusion (30% EVA)

(b) Zero Order Representation of Citric Acid Diffusion (40% EVA)

Figure 7: Citric Acid Trend Lines The change in hydronium ion concentrations for citric acid

diffusing through 30% EVA membranes (a) and 40% EVA membranes (b) were shown to be a

linear. This suggests that the release of citric acid into the beaker was pseudo-zero-order rate.

[5 -16]

Gravimetric Analysis of Alcohols

Figure 8: Gravimetric Analysis of Alcohols The change in mass of alcohols was shown to be

linear, thus correlating to a constant rate of release. Hence, all alcohols are shown to successfully

model pseudo-zero-order controlled release.

DISCUSSION

The readings of NaCl were on the low range of 0-14 μS/cm and erratic, and based on

Figure 5a, NaCl also did not appreciably diffuse through the membranes. Regarding CaSO4 2H2O, the conductivity readings were on the low range of 10-30 μS/cm and extremely erratic

(Figure 5b). These results can be considered as noise, because if CaSO4 2H2O did diffuse, the

conductivity readings should have consistently increased linearly and have been at least in the

hundreds (Figure 5b). Thus, CaSO4 2H2O did not appreciably diffuse through the membranes.

The testing of NaCl and CaSO4 2H2O with several membranes composed of up to 40% EVA

yielded essentially no change in the conductivity of the solution.

Looking at Figure 6a, the lack of an increase in the concentration of hydronium ion

indicates that ascorbic acid did not diffuse through any of the membranes. Thus, ascorbic acid

does not diffuse through membranes on the range of 30-40% EVA. Acetylsalicylic acid also did

not increase in hydronium ion concentration, so acetylsalicylic acid did not diffuse through any

membranes ranging from 30% to 40% EVA (Figure 6b). The decrease of H3O+ concentration

may be attributed to off-gassing of dissolved CO2 from the water, which increased the pH. This

off-gassing is noticed slightly in the citric acid 30% EVA trial (Figure 7c).

[5 -17]

The citric acid diffusion data (Figure 6c) exhibits pseudo-zero-order controlled release as

the graph shows linear data. The R2 value for the citric acid placed in the 30% and the 40%

membranes (Figure 7a and 7b) are both above .90 after 12 hours and 5 hours

respectively. Evidently, the diffusion was pseudo-zero order since the H3O+ concentration

increased at a constant rate. Figure 6d displays another trial of citric acid that did not diffuse

through the membrane. For this second set of trials, the thickness of the dry membrane may have

prevented the citric acid from diffusing at all, since thickness may vary after the film is cast.

The citric acid diffusion data (Figure 7) exhibits pseudo-zero-order controlled release as

the graph shows linear data after approximately 5 hours.

Figure 8 displays the gravimetric tests using alcohols. All of the tests with alcohols

exhibited diffusion at a pseudo-zero-order rate because the graph displays a linear relationship.

In addition, the rate of diffusion is generally correlated with the polarity of the membranes. The

higher polarity membranes yielded larger diffusion rates.

There is one possible explanation as to why ionic salts did not diffuse. As ionic salts

dissolve, they undergo solvation and develop ion-dipole attractions with water molecules. This

may prevent salts from diffusing through the membranes. Thus, membranes can only allow ionic

salt diffusion by exhibiting stronger intermolecular forces with the solutes than the water

molecules do.

As for the weak electrolytes, citric acid, ascorbic acid, and acetylsalicylic acid mostly

differ in polarity and solubility (Figure 5).

Citric Acid Ascorbic Acid Acetylsalicylic Acid

Solubility: 147.76 g/100 mL 33.0 g/100 mL 0.30 g/100 mL

Molar mass: 192.21 g/mol 176.12 g/mol 180.16 g/mol

pKa: 2.79 4.17 3.49

Figure 9: Comparison between Citric Acid, Ascorbic Acid, and Acetylsalicylic Acid based

on structure, solubility, and acidity This figure shows the differences in structure, solubility,

and acidity between the three weak electrolytes which may explain differences in diffusion

through the membrane.

[5 -18]

Our results show that citric acid was the only weak electrolyte that diffused through the

membranes. This can be explained by the fact that citric acid is the most polar of the three

because it has three carboxylic acid groups. Its polarity may enable its dissolution and diffusion

through the membrane. In addition, citric acid is much more soluble than ascorbic acid and

acetylsalicylic acid (Figure 5), so more citric acid is present in solution than ascorbic acid or

acetylsalicylic acid. Therefore, its concentration gradient is greater and more detectable than that

of the other two acids. The concentration gradients of both ascorbic acid and acetylsalicylic acid

are so small that if they diffused, it may take weeks, or even months, for a discernible change to

be detected.

The gravimetric analysis of several organic alcohols was unique compared to both the

weak and strong electrolytes, since each of the four alcohols tested produced a zero-order rate of

diffusion. Methanol, ethanol, and propanol are significantly smaller than the electrolytes in the

proceeding experiments; this could provide an explanation as to why this group of molecules was

the only one to consistently diffuse through the EVA membrane.

CONCLUSION

Impermeability of Ionic Salts

Multiple experiments with membranes of varying polarity have provided strong evidence

against the hypothesis that ionic salts can pass across moderately polar membranes. The most

polar membrane tested, which was 40mils in thickness and 40% EVA by composition, did not

allow ionic salts to diffuse into deionized water any better than the less polar membranes tested.

Hence, these extremely polar molecules cannot diffuse through a polymer membrane regardless

of the increased membrane polarity. It is speculated that these ions were too polar in comparison

to the moderately polar membrane.

The Polarity of the Solute Influences Diffusion

Although many weak, moderately polar organic acids were tested, only citric acid could

successfully diffuse across the membrane. Citric acid, acetylsalicylic acid, and ascorbic acid had

similar molar masses but varying solubility in water. Furthermore, citric acid has a polarity

closer to that of EVA. The innate intermolecular compatibility between EVA and citric acid

allows for citric acid molecules to dissolve into the membrane, and therefore diffuse through it.

The Polarity of Alcohols is Comparable to that of EVA

Alcohols were able to diffuse through membranes of varying polarity consistently. This

can be attributed to the fact that the alcohol molecules are smaller and less polar than the weak

organic electrolyte solutions tested previously. The hydroxyl functional group is less polar than

the carboxyl group, thus the polarity of alcohols is more comparable to EVA, allowing it to

diffuse.

Zero-Order Controlled Release Can Be Achieved Through Diffusion

Methanol, ethanol, propanol, and citric acid were able to successfully diffuse through the

EVA membrane. Regardless of the percent composition of EVA, the flux of the sampled

[5 -19]

substance was kept at a constant rate because the concentration gradient of the selected substance

was kept constant across the membrane using saturated aqueous solutions or vapor pressure

constants. Hence, the apparatuses modeled a pseudo-zero-order controlled release process.

FUTURE STUDIES

The team found that ionic salts do not diffuse through membranes composed of up to

40% EVA, but was unable to conclude about greater polarities due to limitations of lab

equipment. It may be possible that increasing the polarity of the membrane, or even changing the

membrane composition itself, may yield positive results. Creating an ionic membrane, perhaps

through cross-linking ionic substances, may be the best method to dissolve ionic substances. It

may be important to determine if intermolecular forces between the solute and the membrane can

overcome those between the solutes and the water molecules. Finding a membrane’s polarity or

material that diffuses ionic salts would have strong applications in the medical field, especially

for the diffusion of electrolytic solutions into the body.

Moreover, the weak electrolytes used were relatively large in size, which may have

affected the diffusion rate. Future teams should consider choosing solutes that vary in molecular

size. Decreasing molecular size may increase diffusion through interstitial spaces. Using

molecules like acetic acid, which is smaller in size, may be better. Molecular size and membrane

polarity may be two variables that can be tested to determine a relationship for rate of diffusion.

Another variable to consider is membrane thickness. Thinner membranes may allow solutes to

diffuse faster. However, this variable may be difficult to test because it is not easily controllable

due to the limitations of lab equipment.

Currently, the membranes are limited to a range of 30-40% EVA because the films

cannot stay intact outside of these ranges. This issue limits the extent of our research because it is

impossible to accurately predict the behavior of solutes through polarities outside these ranges.

For example, citric acid diffuses through 12% EVA membranes and 30% EVA membranes, but

not through 40% EVA membranes. This potentially suggests that citric acid only diffuses in a

range of approximately 12-30% EVA. Thus, the high and low extremes of membrane polarity

prevent citric acid diffusion. The rate of diffusion may follow a bell curve shape with respect to

membrane polarity. Perhaps future groups should mathematically model diffusion of substances

through membranes before performing experiments. Testing this theory would require

experimenting with one solute and a wide range of membrane polarities, above 40% EVA.

It is also important that future groups can study the wide-ranging applications of zero-

order kinetics. For example, the zero-order principle can be applied in the pharmaceutical

industry to diffuse drugs at a rate that prevents underdose or overdose by supplying only a

therapeutic amount. Moreover, implementing zero-order diffusion in the agricultural industry

can allow nutrients to diffuse at a constant rate, eliminating the need to actively give nutrients to

the plants. Similarly, in the cosmetic industry, zero-order diffusion can be used to send fragrance

uniformly through a space.

Nevertheless, prior years of research have provided a myriad of results and conclusions

about zero-order rates and the diffusion of various substances through different membranes. The

team encourages future scientists to enhance this field by further learning how to diffuse ionic

[5 -20]

salts, small or large weak electrolytes, or multicomponent solutes, and seek to find an even

stronger relation between membrane polarity and diffusion rates.

REFERENCES

1. Ward C.J., Dewitt, M, and Davis, E.W. Halloysite Nanoclay for Controlled Release

Applications. Nanomaterials for Biomedicine, 2012. [Internet] [cited 2014 Jul 24] 209-

238. Available from: http://pubs.acs.org/doi/abs/10.1021/bk-2012-1119.ch010

2. Injection vs. Infusion. [Internet] [cited 2014 Jul 27] Available from:

http://www.alzet.com/products/ALZET_Pumps/injectionvsinfusion.html

3. Andrew J.P. Handbook of polyethylene: structures, properties, and applications.

[Internet] [cited 2014 Jul 27] Available from:

http://books.google.com/books?hl=en&lr=&id=OPuWyxwJwJwC&oi=fnd&pg=PR5&dq

=structure+of+polyethlyene&ots=VstmHfCXeu&sig=I7ZRFIih_qNNfVSVmtRLn7CztF

Q#v=onepage&q=structure%20of%20polyethlyene&f=false

4. Polymeric membranes. [Internet] [cited 2014 Jul 27] Available from:

http://www.hyfluxmembranes.com/polymeric-membranes.html

5. Dormier A et al. Relative influences of polarity and crystallinity on zero-order kinetic

release through a polymer membrane. New Jersey Governor's School in the Sciences,

2013. [Print] [cited 2014 Jul 27].

6. Wijmans J., and Baker, R. W. The solutions-diffusion model: A review. Journal of

Membrane Science, 1995. [Internet] [cited 2014 Jul 30] 107: 1-21. Available from:

http://www.scribd.com/doc/115865737/J-Wijmans-R-Baker-The-Solution-Diffusion-

Model-A-Review

7. Aquino J et al. Pseudo-zero-order kinetics across a polymer membrane using a saturated

solution reservoir system. New Jersey Governor’s School in the Sciences, 2012.

[Internet] [cited 2014 Jul 27]. Available from: http://www.drew.edu/wp-

content/uploads/sites/99/Team5Cincotta12.pdf

8. Hansen C. Hansen’s Solubility Parameters: A User’s Handbook. [Print] New York: CRC

Press, 2007. [cited 2014 Jul 27].

9. Harper CA, Edward MP. Plastics materials and processes: a concise encyclopedia. [Print]

New York: Wiley-Interscience, 2003. [cited 2014 Jul 27].

10. Ascorbic Acid. [Internet]. [cited 2014 Jul 30] Available from:

http://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=54670067#itabs-3d

11. Citric Acid. [Internet]. [cited 2014 Jul 30] Available from:

http://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=311&loc=ec_rcs

12. Aspirin. [Internet]. [cited 2014 Jul 30] Available from:

http://pubchem.ncbi.nlm.nih.gov/summary/summary.cgi?cid=2244&loc=ec_rcs

13. The solution process. [Internet]. [cited 2014 Jul 31] Available from:

http://users.humboldt.edu/rpaselk/ChemSupp/Images/Hyd_Na+Wiki.jpg