Zeta Potential Measurements of Various Substrates and Amino Acid Adsorption via Rotating Disk
Maureen Tang and Derek Eguae-Obazee
Carnegie Mellon University
Received: May 16, 2006
Abstract
Present techniques for measuring the zeta potential of mineral surfaces use a pressure gradient to force fluid through a small slit creating a streaming current, but a new instrument by Sides and Hoggard1-3 measures zeta potential using flow across a rotating disk. The first objective of this semester’s work was to confirm the utility of this instrument for measuring zeta potential of planar surfaces. The second objective was to explore the utility of using the instrument for measuring adsorption of amino acid onto mineral surfaces. The first objective was completed; zeta potential as a function of pH was measured with excellent comparison to literature values for mica, sapphire, polycarbonate, teflon, and gold. More work must be done in order to complete the second objective. The addition of lysine caused a 10 mV drop in the zeta potential of sapphire, but subsequent trials detected aging effects in the surface and were slow and unresponsive to lysine addition. Furthermore, the surface took much longer to reach equilibrium. Repeating the experiment with a new sapphire disk brought equilibration times back to normal but demonstrated no response to lysine addition. Future work should attempt to reproduce the initial results and develop a better understanding of amino acid adsorption. The effect of different cleaning procedures should be especially investigated.
Background and Theory
Functional groups on a surface in solution naturally dissociate to become positively or negatively charged4. Counterions from the solution then electrostatically bind to the surface at the inner Helmholtz plane (IHP). Directly beyond these ions is the surface of shear, or outer Helmholtz plane (OHP). Counterions outside the OHP are in the diffuse double layer; they follow a Boltzmann distribution and decay exponentially in concentration. The unequal distribution of charge generates an electrostatic potential. The zeta potential is the electric potential at the surface of shear. Poisson’s equation relates potential to charge density by
[1]
where Φ is potential and ρecharge density. Surface charge q2 can be back-calculated from zeta potential by
[2]
given a bulk ion concentration Cinf, charge per ion z, and zeta potential ζ. Experimental methods for measuring zeta potential of particles differ, but the traditional method of measuring surface zeta potential is the parallel plate streaming potential measurement. A pressure drop ΔP across a channel of height H causes flow, which generates a convective current of charged ions. Because this surface current js is dependent upon the charge density, and the charge density depends on the potential distribution, the surface current is related to the zeta potential by
[3]
Charge must be conserved through the entire system, so ions will ohmically conduct a current back through the bulk solution. This conductive current generates a streaming potential, which can be directly measured and related to the surface current. The zeta potential is then
[4]
The parallel-plate method of measuring zeta potential is labor-intensive and slow. The new instrument by Sides and Hoggard measures zeta potential when the surface is a rotating disk1-3. Flow across a rotating disk is significantly more complex than that across a surface. For the velocity profile over a rotating disk, the streaming current jsr is
where[5]
and ω is angular rotation rate. The streaming current and streaming potential vary with axial distance from the rotating disk. Directly at the surface,
[6]
but at 1 mm away from a 3 cm disk the above quantity equals 0.854. Thus,
[7]
when Φ is measured between a point 1 mm from the disk’s center and a point in the bulk solution.
Amino acids are zwitterions, meaning that they have both acidic and basic functional groups5. In water, the carboxylic acid group, basic amino group, or both will dissociate to become positively or negatively charged. Some amino acids have additional acidic or basic groups on their sidechains. Lysine, a basic amino acid, is shown below.
Carboxlyic and amino groups are weak acids and bases. They will donate and accept protons in order to maintain a characteristic equilibrium constant called the pKa. Amino acids thus act as buffers, keeping the pH of solution at a constant level. Also, the overall charge of an amino acid depends on the pH of the solution. Above a characteristic value called the pI, amino acids will be negatively charged, and below the pI amino acids are positively charged. The higher the pI, the more basic (positively charged) the amino acid.
Surfaces have a similar characteristic pH value called the pHzc.above which the surface is negative and below which it is positive. The pHzc is also called the isoelectric point (IEP). Churchill et al found through AFM measurements that maximum adsorption of amino acid on quartz (SiO2) occurred when the pI and pHzc were at maximum difference from each other because the surface and amino acid were oppositely charged6. Similarly, Wieland et al found that the basic amino acids lysine (pI = 9.74) and arginine strongly adsorbed while acidic and neutral amino acids exhibited limited adsorption onto sapphire (Al2O3)7. The pHzc of sapphire depends on the orientation of the crystal structure; Kershner et al measured it for three different crystal orientations at 4.5, 5.5, and 6.5, and Franks and Meagher measured it between 5.0 and 6.08,9.
Experimental
A polycarbonate disk was used to hold the samples, bounded by double sided tape. The disk was attached to the Zetasizer that could control its speed of rotation. The disk was then placed in electrolyte solution in a teflon container. In the container, there were also two electrodes one placed close to the disk and one faraway. The electrodes were than attached to a voltammeter that measured the potential difference between the electrodes. A conductivity meter was also placed in the solution and was calibrated each day to make sure that it was reading accurately. Finally, a pH meter was placed in solution to measure changes in zeta with pH. It was calibrated at a pH of 4, 7, and 10.
Before each trial, a sample was prepared/cleaned depending on sample needs. Mica could be prepared by using tape to cleave the old surface and expose a new surface before each trial. One attempt to clean sapphire used acetone, but most trials rotated the disk at 700 rpm while in contact with a solution of 1 um aluminum oxide particles. This polished amino acid off the surface, which was then rinsed thoroughly in DI water. The use of Piranha etch was later investigated with the assistance of Travis Crites at the PTC. Dirty samples were submerged in 3:2 sulfuric acid: hydrogen peroxide mixture for approximately 20 minutes. The samples were then rinsed with DI water and methanol, then dried with an air hose. At this time, the cell-cleaning procedure was also modified. The cell was filled with KOH instead of water during sonication. Other samples tested included teflon, polycarbonate, and gold which were rinsed thoroughly with DI water before each trial.
Two types of experiment were done zeta dependence on pH and zeta dependence on conductivity. To change pH in a pH experiment hydrochloric acid and potassium hydroxide were used. In a conductivity experiment a 10 or 15 thousand micro Siemen potassium chloride solution was used. Our main focus was on the zeta dependence on pH but conductivity was closely watched and reported since zeta depends on conductivity of the solution.
Results and Discussion
1. Measurement of Zeta Potential
Mica
Mica was our most studied substrate and we used data obtained by Scales10 et al to test our setup’s robustness. Scales studied the zeta potential of mica solutions as a function of pH and conductivity. In his experiments, the slit configuration was utilized. He cleaned his samples with fine slurry, then soaked with a hot alkaline detergent, rinsed
Figure 3: The zeta potential of mica as a function of pH. Experimental results are compared to data Scales at concentration 10-3 M KCl.
in water, cleaned with ammonical peroxide solution, rinsed with water again, and then dried. Figure 3 shows the experimental data obtained for zeta potential as a function of pH plotted with data from Scales. In the high (7.5<pH<9.0) and low (3.0<pH<4.5) pH range our data followed Scales’ well. In the middle range (4.5<pH<7.5) our data was off by about 20%. This may have to do with the sensitivity of water to changes in pH at this range. Small amounts of acid or base can cause large changes in pH at this range.
Sapphire
Sapphire was actually previously studied by two groups one lead by Kershner8 et al and the other by Franks and Meagher9. Sapphire (aluminum oxide) was studied in a similar manner to mica but Kershner showed that zeta depends on the orientation of the sapphire substrate. Franks and Meagher did not see the dependence but did show the zeta versus pH relationship.
Franks and Meagher went through exhaustive cleaning to ensure that the sapphire surfaces were not contaminated. They had two cleaning procedures but there data was taken using cleaning procedure two which included an acetone, ethanol, potassium hydroxide, and nitric acid soncation each followed by a DI water rinse and a 16 hour soak in DI water. Franks and Meagher examined four different sapphire orientations the R, A, C, and M planes which vary in the orientation of the aluminum and oxygen atoms in space. Figure 4 shows three of the four different sapphire orientations that are terminated with hydroxyl ions.
Figure 4: Three different hydroxyl terminated sapphire surfaces where a) is the C-plane b) is the A-plane and c) is the R-plane crystallographic orientation8.
Kershner used a simpler but more dangerous Piranha etch to clean his samples followed by a deionized water rinse. His data includes the zeta vs. pH plots for three different sapphire orientations in 10 mM KNO3 solution. They tested the R, A, and C planes of sapphire. Kershner saw that there was about a 10 mV difference between the zeta potentials of R and C planes and that the A plane fell in between the two. Figure 5 shows the graph of zeta vs. pH for the different orientations.
Figure 5: Zeta potential vs. pH for the R, A, and C plane sapphire substrates obtained from the Kershner8 group.
The last figure for sapphire shows the experimentally measured zeta potentials plotted with the zeta potentials adapted from Franks and Meagher and Kershner for the R-plane sapphire substrates. Both the Franks’ research group and Kershner’s group saw that sapphire had an isoelectric point in between 4 and 5. Our sample had an isoelectric point in the same range.
Figure 6: Experimentally determined zeta potential of sapphire compared to Franks’ and Kershner’s R-plane sapphire substrates.
Teflon
Testing the zeta potential of teflon was important for our study since we were using a Teflon container to run our experiments. Teflon was chosen because of its resistance to chemical change and rigidity. The work of Zimmermann11 et al was done with a different type of teflon then we had studied but is presented for comparison purposes. Zimmermann prepared Teflon AF (Poly(tetrafluoroethylene-co-2,2-bis(trifluoromethyl)-4,5-difluoro-1,3-dioxole)) samples by fusing them from a polymer solution onto glass or silica wafers. Figure 7 shows the work of Zimmermann compared to our experimental measurements of the zeta potential as a function of pH. These measurements were taken at an ionic strength of 10-3 M KCl.
Figure 7: Zeta potential of Teflon (AF) as a function of pH.
Polycarbonate
Kirby12 et al studied various polymers using the slit configuration for microfluid polymer substrates. We are particularly concerned with polycarbonate because of its use in our disk configuration. To normalize the zeta potential Kirby proposed a term called the pC, which is just the negative log of the concentration in moles per liter. For consistency purposes this term was multiplied out and compared to the data we obtained. Kirby took data at various pCs but the data shown in Figure 8 is taken at a pC of 2.9, which corresponds to an approximant ionic strength of 10-3 M KCl.
Figure 8: The zeta potential for polycarbonate as a function pH. Kirby’s data was obtained factoring out the normalization term pC.
Gold
The next substrate studied was gold. This proved to have interesting trends due to the electrochemistry of the gold surface. Gold, unlike the other substrate appeared to have an inflection point at around a pH of 6 not seen by any other substrate. It also had an isoelectric point around 4.5. Giesbers13 et al studied gold at concentration 10-3 M by deposition of a thin layer titanium (5 nm) and a layer of gold (15 nm) onto to silica wafers in a slit configuration. The titanium was used to keep the gold on the silica surface. The gold surface was cleaned with a Piranha etch for two minutes. Kirby also found an inflection point of a different shape that started at a pH of 6. Figure 9 shows the data obtained by Kirby and experimentally.
Figure 9: Zeta potential of gold as a function of pH.
Silica
Silica is one substrate that has yet to be studied but is important to the study of adsorption of amino acids. Silica is a mineral that is found abundantly as quartz, which is known to be a chiral mineral. The specific adsorption of certain chiral amino acids onto quartz surfaces can be useful in the manufacture of many bioactive materials including pharmaceuticals14. In pharmaceuticals, most chiral amino acids are synthesized as racemic mixtures in which the two enantiomers are indistinguishable. The need to produce enantiospecific materials is important but due to the lack of quick and efficient ways to analyze racemic mixtures, producing enantiospecific materials can be exhaustive. The study of the adsorption of chiral amino acids onto quartz by way of the zeta potential is important for this reason. Theodoly15 et al has examined the zeta potential of silica surfaces by way of the slit configuration. The silica was cleaned in a Piranha solution and rinsed with DI water before being used. Figure 10 shows how the zeta changes with pH for a silica substrate at an ionic strength of 10-3 M KCl.
Figure 10: The zeta potential of silica as a function pH examined by Theolody et al and Wiese et al.
2. Effect of Lysine on Zeta Potential of Sapphire
The first attempt to measure the adsorption of lysine on sapphire demonstrated a significant (10 mV) drop with the addition of 0.1 mL of 0.01 M lysine solution (Figure 2). Two other trials measured a drop of approximately 5 mV.
Initial addition of the lysine, at 16 minutes, caused pH to decrease by 0.01 and conductivity to rise by 0.8 uS. These changes could not explain the change in zeta because the change in pH was not large enough and an increase in conductivity would cause zeta to increase, not decrease. Because these factors could not have caused the change in zeta, the lysine must have affected the surface by adsorption or some other mechanism. Subsequent attempts to repeat the results while using the polishing method of cleaning demonstrated an upward drift in zeta potential with and without addition of lysine (Figure 3). The surface also took much longer to reach initial baseline and to respond to lysine additions (Figure 4).
After the Piranha etch was used to clean the surface, the sapphire disk’s reponse time dropped to its initial time. However, the surface exhibited no response to lysine additions. The experiment was repeated, using a new sapphire disk. A curve reproducing the initial 10 mV drop and short response time was expected, but the actual results again had no response to lysine addition (Figure 4).
We have no current explanation for this behavior. The surface could be contaminated from previous trials, but a new surface should have been clean. A possible explanation is that contaminant from the cell, electrodes, or pH and conductivity meters adsorbed on to the disk before lysine addition. Sonication with KOH should have removed contaminant lysine, but the process might not have been effective.
Conclusion and Recommendations
Extensive study of the rotating disk method with various mineral surfaces shows that this method can be utilized in the study of zeta potential. Further study of minerals of interest will show that this can be useful with any mineral surface. Finding sufficient cleaning procedures for the various substrates can prove to be beneficial in the reproducibility of results. The ease of the rotating disk method makes it a preferable option in measuring zeta potentials and in adsorption studies.
Three trials, from March 30 and AM/PM April 4, appeared to show a reproducible drop with the addition of lysine. Subsequent trials gave increased equilibration times and decreased responses to amino acid addition. While these could be the results of aging effects, neither Piranha etch nor using a new disk demonstrated a return to the initial response although both elimanted the upwards drift in zeta. In order to reproduce the initial results, cleaning procedures must be more thoroughly investigated. Ozone cleaning, chromerge, or even a simple soak in concentrated base may be sufficient to remove amino acid without affecting the sapphire. The Schneider research group at the PTC may know more about cleaning surfaces. Sources of contamination also need to be determined. Ninhydrin might indicate whether or not lysine is entering trials before addition, and if so, where it is coming from.
After a cleaning procedure is developed, zeta potential data must be converted to adsorption curves. Surface complexation models like that presented by Scales for mica may have insight on how substrates bind to the sapphire surface. Collaboration with the Walker group may use known adsorption curves of well-characterized surfactants to determine the relationship between adsorption and zeta potential.