E LECTROCHEMISTRY : A DJUSTMENT OF M ONOLAYER CHARGE WITH pH by By MARK DE LA ROSA B.S. in Chemistry, Metropolitan State College, 2008 A thesis submitted to the Faculty of the Graduate School of the University of Colorado in partial fulfill ment of the requirements for the degree of Master in Chemistry 2012
! "" #$"%!&$' %"%!()*!&$'!+,%&'*!)(!-."'/.'! 0'1*''!23 +,*4!0'!5,!6)%, $,%!2''/!,77*)8'0!23 +,*4!9/0'*%)/ :!;$,"* <,**3!9/0'*%)/ =,"!<"/ 9>1>%&!??:!?@A? !
! """ D EDICATION I de dicate this work to Dr Anderson for pushing me to finish and for his guidance to insure the quality of the work. I also want to thank my son for the patience and for dealing with me in the library.
! "8 0'!5,!6)%,:!+,*4 Electrochemistry: A djustment of Monolayer charge with pH #$'%"%!0"*'.&'0!23! +,*4!9/0'*%)/ !"#$%&'$ ( ( B%"/1!%'5(! C ,%%'D25 '0 !D)/)5,3'*%! )(!D)5'.>5'%!&$,&!.)/&,"/! ")/"E,25' !(>/.&")/,5! 1*)>7%: !,0F>%&"/1!&$'!7= !)(!&$'!,0F,.'/&!%)5>&")/ !,55)G%!&$'!.)/&*)5!)(!&$'! "/&'*,.&")/%!)(!&$' D)0"("'0! '5'.&*)0'!,/0!&$'!%)5>&")/H I/!&$"%!G)*4 : !%'5( J ,%%'D253! )(! K J D'*.,7&) 7*) 7")/".!,."0 !L+M9N!)/!1)50!'5'.&*)0'%! ,55)G'0 !'8,5>,& ")/ )(!&$'! '5'.&*)0'!2'$,8")*! ,&!0"(('*'/&!7=O%!>%"/1!.3.5". J 8)5&,DD'&*3!,/0! .$*)/).)>5)D'&*3H!! ; 3.5".!8)5&,D'&*3! G,%!>%' 0! &)!0'&'*D"/'!&$'!%>*(,.' !.)8'*,1'! )/ !&$'!1)50!'5'.&*)0'!)(!&$'!K J D'*.,7&) 7*) 7")/".!,."0% !G,%! ?HK LP@HQN !RA@ J S D)5T.D ? H !!#$'!"/&'*.'7&!)(!. $*)/).)>5)D'&*3! 75)&%!()* !&$'!K J D'*.,7&) 7*) 7")/".! ,."0 !D)0"("'0!'5'.&*)0'% !,&!7=U!G,% !()>/0!&)!2' !UHQ R A@ J U V !,/0 :!G,% !()>/0!&)!2' ?HK R A@ J W !,&!7=!KH!!X$'/!& $'!7=!)(!&$'!,0F,.'/&!%)5>&")/!G,%!,0F>%&'0!&)!7=!U:!&$'! .,*2)R35".!, ."0!1*)>7!)(!&$'!+M9!0'7*)&)/,&'% !.,>%"/1!,/!'5'.&*)%&,&".!*'7>5%")/! G"&$!&$'!,/")/".!('**).3,/"0'!%7'."'% H!!#$"%!7*'8'/&%!&$'!,/")/".!*',.&,/&!(*)D! ,77*),.$ "/1 !&$'!%>*(,.' !,/0!.*',&'%!,!5).,5!.)/.'/&*,&")/!0'("."& H! !! X "&$!, !/ J 2>&,/'&$")5 !D)0"("'0!'5'.&*)0': !&$' !.$*)/).)>5)D'&*3 !"/&'*.'7& !G,%!()>/0!&)!2'! "/0'7'/0'/&!)(!7= H #$'!()*D!,/0!.)/&'/&!)(!&$"%!,2%&*,.&!,*'!,77*)8'0H!!I!*'.)DD'/0!"&%!7>25".,&" )/H 977*)8'0Y!! +,*4!9/0'*%)/
! 8 TABLE OF CONTENTS C HAPTER I. M OLECULAR S ELF A SSEMBLY M ONOLAYERS .1 Self Assembled Monolayers 1 Electrostatic assembled Monolayers ....7 II. E LECTROCHEMISTRY .10 Electrochemical Theory. 10 Electrochemical Measurements .. ..15 Cyclic Voltammetry .... 15 Chronocoulometry .. 18 III. E XPERIMENTAL M ETHODS ...20 Chemic als and their Source ...20 Elect rode/ Instrument Preparation.. ..20 Cyclic Voltamme try... ..21 Chronocoulometry... 22 Desorption using Cyclic Voltammetry.. ..22 IV. R E SULTS AND D ISCUSSION : C HARACTERISTIC OF M ODIFY AND U NMODIFIED E LECTRODE .24 Chronoamperomter y ..24 Cyclic Voltammetr y .. .25 Unmodified Electrodes..25 Modified MPA Electrode...27 Reductive Desorption of the SAM on the Electrode .28
! 8" Chronocoulometry.30 Comparison of Monolayer Behavior i n pH 3 and pH 7 ...30 Electrodes Modified with n alkanethiols at pH 7 and pH 3 33 V. C ONCLUSION .....35 B IBLIOGRAPHY ..3 7
! 8"" LIST OF TABLES )&"*+ ( 1. Surface Coverage determi ned by Cyclic Voltametry28 2. Chronocoulometry intercept calculate at the last 80%,70% and 60% and standard deviation for different runs .. 3 1 3. Comparison of the Alkan ethiols to 3 Mercaptopropionic Acid .. .... .34
! 8""" ,-.)(/0(0-1234. ( 0567%+ ( 1. Simple Generic Alkanethiol attach ed to the gold ...2 2. Shows an (!3 X!3) R 30Â¡ with the sulfur group(blue) 5 Â apart ....4 3. Typical cyclic voltammogram for reversible O + ne < => R .......16 4. Picture of the three electode system as used in these electrochemical experiments ..21 5. Comparison o f the blank electrodes .......26 6. Shows the Au electrode modified with MPA .. ..27 7. Sample of the desorption of 3 mercaptopropioni c acid using Cyclic voltammetry between 0 and 1.3 V ..29 8. Comparis ons of the blank electrode at pH 3 (blue) and pH 7 (blue) in chronocoulometry ... 32 9. Modified electrode in chronocoulometry with 3 mercaptopropionic acid at pH 7 (blue) and pH 3 (red) ...3 3
! "R LIST OF EQUATIONS Equation 1. Current ......11 2. Nernst Plank equation in one dimension...11 3 F ick 's first law.. .. ...12 4. F ick 's second law.. .13 5 F ick 's first law in spherical polar coordinates.. ...13 6. Cottrell equation.....14 7. Surface coverage of a monolayer... 17 8. Anson equation ..18 9. Cottrell equa tion.....24
! A C HAPTER I M OLECULAR S ELF A SSEMBLE M ONOLAYERS Self Assembled Monolayers Self assembled monolayers (SAMs) are highly ordered molecular structures that form on different surfaces. The study of these mon olayers is useful to many areas of science; including application to biosensors, biochemistry, the electroplating industry, materials science, and surface chemistry. 1, 2, 3, 4,5 The broad interest is in part because self assembly is a principle found in natu re that accounts for many of the interesting structures that occur naturally. For example, the lipid b i layer membrane, the double helix structure of DNA, and protein folding are all driven by through space interactions that characterize self assembling s ystems. The self assembly of organic monolayers is used to model complex systems because they are easy to prepare, they model the building blocks of more complex structures, and they allow for the creation of well defined nanostructures. 1 For that reaso n, this thesis investigates the role of through space interactions in determining the properties of the interfacial region. A significant amount of research has been performed with self assembled monolayers since in the 1980's. The molecules used to p repare monolayers are typically made up of a head group this is attached to the substrate surface, the alkyl backbone, and the end group that is exposed to the adjacent solution, as shown in F igure 1. The head group needs to have a high affinity for the s ubstrate material to insure the stability of the monolayer. The alkyl chain provides the opportunity for through space interactions (e.g. van der Waals forces) that allow the individual molecules that comprise
! ? the monolayer to assemble into a well defined structure. The end group introduces additional interactions that contribute to the ensemble structure of the monolayer, and also offers a reactive group for subsequent chemical modification of the interfacial assembly. Figure 1 Simple generic alkanethi ol attach ed to the gold. Yellow represents the sulfur group while black is a carbon group. There are many ways to prepare self assemble d monolayer s using organic molecules. A common method is molecular self assembly where a dilute solution of a molecule is exposed to a clean substrate for a period of time and the molecule spontaneously adsorbs to the substrate. Over time, the molecules assemble into a thermodynamically stable structure. The molecular self assembly technique is most commonly applied to either organic silane molecules adsorbed onto silicon substrates, long chain organic carboxylic acids adsorbed onto silver or aluminum substrates, or organic mercaptans adsorbed onto gold substrates. In the work described in this thesis, I will focus on o rganic mercaptans adsorbed to gold substrates.
! K An estimate of the bond enthalpy between the gold and sulfur is 40 45 kcal mol 1 indicating that adsorption forms a strong interaction between the molecule and the substrate; this value is comparable i n strength to a covalent bond. 11 Gold is generally an inert material that has a high surface energy. This can lead to contamination, suggesting that the Au surface must be carefully cleaned prior to formation of the monolayer. Bain et al. 12 compared hi gh energy surfaces (e.g. gold) to low energy surfaces (e.g. polyethylene and Teflon) and found that the lower energy surface was less susceptibly to contamination. This result suggests that gold substrates modified with n alkanethiols will be able to prot ect the substrate from oxidation and other chemical contaminations. According to past research, a fully saturated surface of n alkanethiols on gold exhibit an (! 3 x !3) R 30 Â¡ unit cell in which the hydrocarbons chains are tilted approximately 32Â¡ from the normal to the gold substrate. The saturated n alkanethiol molecules were also shown to be spaced approximately 5 Â apart within the monolayer, and to have an area per molecule of 21.6 Â 2 Lower coverage's exhibit different unit cells (e.g. a ( 5 !3 X !3 ) R 3 0Â¡ when the coverage is 40% of saturated) suggesting that the molecules that comprise the monolayer are spaced further apart and are, on average, tilted at greater angles from the surface normal. When forming the monolayer, the van der Waals (VDW) forces a mong the molecules that comprise the monolayer are the interactions that determine the ensemble structure of the monolayer. 1, 8 By adding different functional groups to the end of the alkanethiol this changes the extent of the van der Waals interactions a mong the hydrocarbon chains and results in greater steric effects that influence the structure of the monolayer. 1
! Q Figure 2 Shows an (!3 X!3) R 30Â¡ with the sulfur group(blue) 5 Â apart. The gold circles are the gold surface The formation of self asse mbled monolayers is influenced by many factors, including temperature, solvent, mercaptan concentration, mercaptan alkyl chain length, the cleanness of the gold substrate, and kinetics of adsorption. 12 Within the monolayer, the alkyl chains assemble into a n all trans conformation; and, it is found that as the alkyl chain gets longer, the monolayers form a more densely packed structure. Ulman shows that there are differences between monolayers formed with short chain thiols (n<9) and long chain thiols (n>9 ). 11 Porter et al. systematically studied different n alkyl thiol chain lengths (n= 1, 3, 5, 7, 9, 11, 15, 17, 19, and 21) and find that (1) the shorter groups develop a monolayer structure with more alkyl chain tilt than is found with longer alkyl
! Z thiols, (2) shorter alkanethiols form less dense monol ayers, and (3) monolayers formed from shorter alkyl chains have more disorder than the longer chains( n"9). This is because the van der Waals interactions with the longer chains are stronger. This result suggests that through space interactions drive the assembly process. 12 Ulman compared monolayers prepared from n octadecyl mercaption and tert butyl mercaption. They find that the n octadecyl mercaptant forms monolayers on gold that are a factor of 290 710 stronger than those formed with the t butyl me rcaptan. 11 This is due to both the steric hindrance of the tert butyl group and the shape of the tert butyl group preventing formation of strong van der Waals interactions between the adjacent mercaptans. This result again shows that self assembling syste ms prefer to form structures that use long chain linear molecules to maximize the driving for ce of the through space interactions Several other studies show that n alkanethiols are stable at room temperature, but over 70Â¡C the monolayer tend s to desorb f rom the gold surface. This behavior is dependent on the length of the alkyl chain. 1,12 Nuzzo et al. showed that for temperatures between 170 230 Â¡C there was also sulfur loss from hexadecanethiolate. 20 The higher temperature disrupts the van der Waals interactions, effectively "melting" the monolayer. With the weakened lateral interactions, only the gold substrate interaction must be overcome for the molecules that comprise the monolayer to desorb. Bain et al. show that the kinetics of n alkanethiol ad sorption onto Au substrates is a two step process when conducted with moderate concentration (e.g. 1mM) of n alkanethiols. 12 The first is a fast step that takes a few minutes and during which the contact angle of the modified s ubstrate reaches 80 90% of th e final value. The second
! W step requires several hours before the monolayer reaches its final condition. Bain et al. concluded from this work that, to form a good monolayer, the adsorption solution should to have dilutions up to 10 5 mole/L and that the a dsorption should take place over several hours. The end group or the terminal functional gro up plays a key role in the self assemble d monolayer. The end group can have an impact on the tilt angle and effect the addition of other groups due to steric cons traint s or the types of interaction s with the end group 1 Nuzzo et al. shows that SH(CH 2 ) 16 OH had a tilt angle of 28Â¡ and the alkyl chains are in a trans configuration that interferes with the maximum packing on the gold surface. 8 Smaller groups with SH( CH 2 ) 6 OH had behaviors similar to an alogous alkanethiols having a packing density of 21.5 Â 2 ; but they are found to be sensitive to the humidity. Alkyl chains with terminal carboxyl group s showed similar tilt angles as found with alkyl chains having a h ydroxyl group ; however, the alkyl chains exhibit a high er density gauche d effect. 1 Tarasevich et al. used bovine a o r tic endothelial cell (BAEC) growth to functionalize and to see which terminal group provides the best substrate (CH 2 OH, CO 2, CH 3 and CO 2 H ). The y concluded that for the ir experiment the carboxyl group work s the best. 32 The terminal functional group plays a significant role in the self assemble monolayer depending on what the researcher wants.
! U Electrostatic Assembled Monolayers Electr ostatic, layer by layer assembly was introduced by Decher in 1991. 7, This technique takes advantage of through space interactions to alternately deposit polycations and polyanions along a charge d substrate. In this manner, interfaces can be modified in a way that introduces many different chemical and physical properties. In this thesis, I will us e electrochemical methods to probe the interfacial charge state of a modified surface. The charge state of the interface is central in establishing conditions t hat allow subsequent electrostatic deposition of polyelectrolytes.. 5,7,16 The technique of layer by layer self assembly uses primarily electrostatic interactions between polyelectrolytes to create the modified interface. There are, however, secondary in teractions that can also be leveraged to create interfacial structures. These secondary interactions include hydrophobic, hydrogen bonding and dispersion forces that exist between molecules. Like molecular self assembly, the through space interactions be tween the individual polymers drives the electrostatic assembly process. Frequently a molecularly assembled monolayer with an ionizable terminal group is deposited first on the substrate to serve as a template for the electrostatic assembly of the polyelec trolytes. The dissociation of the confined functional group depends on several factors. These include the strength of the lateral interactions among the molecules that comprise the monolayer, the pH of the adjacent solution, and the ionic strength of the adjacent solution. The pKa of polyacrylic acid (PAA) is around 6.79 with no additional salts, but drops to 4.68 in high ionic strength solutions. 20 Clark et al. observe the interactions of the functional groups at different pH values and how they intera ct with PAA, and PMAA ( poly (methacrylic acid)). At low pH, the interaction between the
! [ PAA and the terminal carboxylic acid group of the monolayer are primary by hydrogen bonding. 18,20 When the pH is increased to 4.8, the hydrogen bonding contribution decreases and the carboxylic acid and the amine group s would ionize causing the electrostatic interactions to start to dominate. At this pH, hydrophobic interactions also play a key role in the assembly process. For example, Clark showed that as the nu mber of secondary amines increased, the polarity of the polyamine backbone also increases. Linear polyethyleneimine (LPEI) will not absorb to the oligoethylene glycol (EG) SAMs between a pH of 2.5 7 because of the steric repulsion of their hydrate layer. LPEI, however, will absorb to the carboxylic acid surface due to the strong electrostatic interactions. At pH values greater than 7, the PAA and the polyamine are ionized. When the solution pH is equal to 7, the carboxylic acid group is highly ionized, c ausing the electrostatic interactions between the SAMs and the polyelectrolyte being deposited to increase. The hydrogen bonding and hydrophobic forces are much less important than the electrostatic forces under these conditions. 21 The influence of molecul ar weight of the polymer on the polyelectrolyte deposition depends on the ionic strength of the solution. 22,23 Baur shows that at low ionic strength (I# 0.1 NaCl), the bilayer thickness is independent of the molecular weight of the polyelectrolyte; but, a t higher ionic strengths (I "0.1 NaCl), then the thickness o f the polyelectrolyte bilayer depends on the molecular weight of the polymer. 22 Clark et al. also compare the thickness of the polyelectrolyte bilayer with and without salt present in the solution. With salt, they found that the polycation/polyanio n bilayer has a thickness of 50 100 Â. Without the salt, a thickness of 10 25 Â is found. The salt lowers the effective charge density of the polyion and reduces the electros tatic driving force to ass embly 22
! S Self assembled monolayer's are very useful in many fields of study and have a variety of applications. Electrostatic interaction is a common way to create interfacial assemblies. When using this technique, there are man y factors that need to be considered. For example, the pH of the solution must be within a specific range to maintain appropriate ionization of both the polycation and the polyanion. Depending on the desired thickness of the assembly, the addition of the salt to the solution to control the ionic strength and the molecular weight of the polymers must also be considered. Other interactions, like hydrophobic, hydrogen bonding and dispersion forces, can also play a key role in determining how the polymers w ill assembly along the interface. In this thesis research, I will explore how the experimental conditions impact the ionization o f a monolayer of 3 mercaptoprop ionic acid, and how we can use electrochemical techniques to qualitatively and quantitatively estimate the charge density of the interfacial monolayer.
! A@ C HAPTER II E LECTROCHEMISTRY Electrochemical Theory In electrochemical methods current flows when potential is applied to the electrode. The act of applying potential generates two types of processes: non Faradic and Faradic. The non Faradic processes occur when charge does not cross the interface, but the applied potential induces the movement of charge in the adjacent solution and current is measured. This is due to the charging of th e interface. When the potential of an electrode is altered, that results in a change in the excess charge present at the interface of the conductor with the adjacent electrolyte solution. This charge on the electrode surface induces the accumulation of i ons of opposite charge on the solution side of the interface. It is the redistribution of ionic charge on the solution side of the interface that results in the charging current. This process is rapid, and once the interface is fully charged (assuming no further change in the potential), the charging current decays exponentially in time to zero. 10,37 Faradic processes occur when charge is transferred between the electrode and oxidizable/reducible species in the adjacent electrolyte solution. When the e lectrons (charge) cross the electrode/solution interface in a heterogeneous redox process, then faradaic current is generated. The faradaic current is proportional to the rate at which the reactant species is transported from the bulk solution to the elec trode interface. 10,37,38 The transport of reactant through the solution is the flux, and the current measured is proportional to this flux.
! AA When considering a heterogeneous reaction, the current response of the redox reaction has two primary contributio ns: the mass transport of the reactant species to the electrode surface and the rate at which the electron crosses the heterogeneous boundary. In most redox reactions either the mass transport or the heterogeneous electron transfer limits the reaction r ate, and experimental conditions can be adjusted to favor one or the other of these. 10,37 By definition, the flux (J) is a measure of the mass transport through a unit area in space. The current (i) in an electrochemical measurement that is proportional to the flux. i= nFAJ where i is the current, J if the flux, F is Faraday's constant, n is the number of electrons and A is the area (in my experiments the Area is the Area of the electrode surface). Mass transfer, or flux, represents the movement of m ass from one location in space to another. There are three different contributions to solute mass transport; diffusion, electrical migration, and convection. Mass transfer to an electrode can be described mathematically by the Nernst Plank equation. In one dimension, the Nernst Plank equation is given by: In the Nernst Plank equation the first term is the contribution from diffusion, the second term is the contribution from electrical migration, and the final term is the contribution
! A? from convectio n. These contributions are independent of each other and experimental conditions can be established that isolate the individual contributions to the mass transfer from each other This allows simplification of the understanding of the origin of the role o f each of these contributions to the flux in an individual measurement. 10,37,38 For example, adding excess inert electrolyte suppresses the electrical migration of the reactant and effectively eliminates the contribution to the total flux of our reactant by electrical migration. Further, convection can be suppressed by not stirring the solution. Under these conditions, the mass transport is limited to diffusion only. These conditions (e.g. excess inert electrolyte and not stirring the solution) are gene rally the conditions that we use in our experiments. With mass transport only by diffusion, describing the mass transfer as a function of time can be accomplished by solving the Nernst Plank equation. Assuming the contributions from electrical migration a nd convection are zero, the Nernst Plank equation simplifies to Fick's first law of diffusion. Fick's law describes the flux due to the rate of diffusion. Specifically, Fick's laws describe the movement of species through solution under the influence of a concentration gradient. Fick's first law states that the flux is proportional to the concentration gradient. 10,37
! AK Fick's second law describes the change in the concentration gradient as a function of time. These are one dimensional representations of Fick's laws. The one dimensional Fick's laws adequately describe experimental behavior if conditions of semi infinite linear diffusion (e.g. diffusion to an infinite plane) are established. These conditions occur in el ectrochemical systems when the distance that the reactant diffuses during the time frame of the experiment is small relative to the dimensions of the electrode. These conditions are satisfied when either the experiment ha s a very short duration, or the el ectrode dimensions are very large. If either of these conditions are not met, then Fick's laws mu st consider radial diffusion (e.g. diffusion to a point) 10,37 Under radial diffusion conditions, Fick's laws can be expressed in spherical pola r coordinates (only Fick's law is given below): where r is the distance from the point. In experiments that use a large electrode, when experimental conditions are set for the complete oxidation or reduction of the reactant at the electrode surface, a concentrati on gradient is established in the volume of solution adjacent to the electrode. This volume of solution is referred to as the diffusion layer. As the diffusion layer grows
! AQ with time the concentration gradient decreases. As the gradient decreases, so too does the flux. This is experimentally manifested by a decrease in the observed current. 37,38 Solving Fick's laws for these experimental conditions results in the following relationship between current (i) and time (t) for an experiment in which the pot ential is stepped from a value when no heterogeneous reaction occurs to a value where the beyond the standard potential of the redox reaction. This relationship is known as the Cottrell equation, where D 0 is the diffusion coefficient, C is the conce ntration of the reacting species and t is the time duration of the experiment following the potential step. This equation describes the current behavior as a function of time in an experiment in which the concentration of the reactant is maintained at 0 a djacent to the electrode surface 10 This is accomplished by holding the electrode potential well past (50 100 mV) the standard potential of the redox reaction. By controlling the experimental parameters, we determine the major mode of mass transfer that the redox couple is subject to. Under the conditions of this experiment, the contributions to the flux, as described by the Nernst Plank equation, by electrical migration and convection are minimized; and, the vast majority of the flux of the reactant is determined by diffusion of the reactant under the influence of the concentration gradient. The current response in any given experiment will be determined by the specific potential function that is applied to the electrode. In this
! AZ work, we will rely on two primary experimental methods: cyclic voltammetry and chronocoulometry. Electrochemical Measurements There are a variety of analytical electrochemical techniques that can be used to determine the properties of self assembl ed monolayers confined to gold electrode surfaces. These techniques are used to study the structures, adsorption and characteristics of the monolayers. Many different techniques were employed to measure SAMs based on the type of experim ental conditions, type of SAMs and the permeabilit y of the monolayer to the adjacent solution. Wang et al. demonstrated the adsorption of L cystenie SAMs on gold electrode us ing scanning electron microscope(SCM) and attenuated total reflection Fourier transformed (ATR FTIR) to characterize the structu re. Cyclic voltammetry(CV), chronocoulometry(CC), and a.c. impedance were also utilized to determine the electrochemical properties. 4 Cyclic Voltammetry Cyclic voltammetry (CV) is a common technique for measuring qualitative information about a heterogen eous electrochemical reaction. CV is a rapid way to measure the thermodynamics and kinetics of redox processes, and is often the first experiment that is conducted with characterizing a redox system. In the Cyclic voltammetry measurement, a voltage i s applied to the "working" electrode using a linear potential ramp. The voltage is scanned linearly in time, and the faradic current is measured as a function of the applied potential. The response has a characteristic shape, marked by a rising current t hat peaks before it decays. The peak
! AW and decay is characteristic of the competition between mass transport and heterogeneous kinetics that characterizes the CV experiment. The current vs. potential response can be described as either being reversible (e. g. diffusion controlled where the heterogeneous kinetics are fast and have little impact on the I vs. E profile) or irreversible (e.g. where the charge transfer is slow and kinetics dominate the I vs. E profile). 38 The cyclic voltammetry I vs. E profile is divided into a forward portion and a reverse portion. The reduct ion reaction is designated the cathodic response, and the oxidation reaction is designated as the anodic response. Figure 3 shows the forward reaction for the oxida tion of Fe 2 + to Fe 3 + and in the reverse reaction Fe 3 + is reduced to Fe 2 + Figure 3 Typical cyclic voltammogram for reversible O + ne < => R
! AU In a reversible cyclic voltammetry there are several parameters that are well defined and characteristic of this behavior The rati o of the cathodic peak current to the anodic peak current is equal to one, and the separation between the cathodic peak potential and the anodic peak potential is equal to 59/n mV at 25Â¡C for reversible redox processes. 10,37 In addition, the forward and re verse responses are independent of the scan rate. When the gold electrode is coated with a monolayer, then the shape of the cyclic voltammetry I vs. E response changes from that observed for reversible behavior. Specifically, the separation between the c athodic and anodic peak potentials increases. This is because the monolayer limits the kinetics of the electrode reaction by creating a "barrier" through which the electronic charge must traverse before participating in the redox reaction. Cyclic voltam metry can also be used to quantify the amount of material adsorbed to the electrode surface. Walczak et al. coated a gold electrode with a monolayer, then desorbed the monolayer in the presence of a 0.05 M KOH solution during a cathodic potential sweep. This caused the reduction of the Au S interaction and the subsequent desorption of the mercaptan molecule. 30 Reductive desorption of the monolayer provides a method for measuring the surface coverage of the monolayer on the gold. The surface coverage o f the monolayer ( $ ) is calculated by: where Q is determined by integrating the area under the I vs. E curve. 10,37,38
! A[ Chronocoulometry Chronocoulometry(CC) is a technique used to study the adsorption of electroactive species at the electrode. 31,35 Chronocoulometry me asures the charge passed during a potential step experiment. In this experiment, the potential of the working electrode is stepped from a value where no faradaic electrochemistry occurs to one where diffusion controlled mass transport occurs. The charge measured as a function of time is due to a combination of faradaic and non faradaic processes. CC allows the measurement of electrode surface area, diffusion coeffi cients, analyte concentration, and electrode kinetics in addition to the quantization of th e adsorbed species. 37 The Cottrell equation is used for describing the chronoamperometry response, which is measure of current vs. time. The Anson equation is calculated by integrating the Cottre ll equation to get the following where Q is the amount of charge in coulombs at a certain time (t) and Q dl is the charge used by the capacitate of the electrode and electrolyte double layer. With the adsorption of electroactive species, then the term nFA% is added to the Anson equation to quantify the charge associated with the redox chemistry of the adsorbed species. 10,31 The CC data is analy zed by plotting the charge (Q) vs. the square root of time. The intercept of the Q vs. t 1/2 plot is quantitatively representative of Q dl the charge required to charge the double layer, and nFA%, the charge required to reduce/oxidize the surface adsorbed species If Q dl can be independently determined, then the amount of material adsorbed to the surface can be determined. Although simple in principle, a problem arises when
! AS measuring Q dl from charging the electrode, as it is difficult to determin e indepen dently. One way that this has been approached in the past is to run a double potential step chronocoulometry experiment where the charge on the electrode is returned to its original state in the second potential step 28,31,36
! ?@ C HAPTE R III E XPERIMENTAL M ETHODS Chemicals and their Sources A pH 7 solution was prepared from NaH 2 PO 4 monobasic (Sigma reagent Grade), Na 2 HPO 4 dibasic anhydrous (Sigma reagent grade) and KC l (Sigma reagent grade). The pH 3 buffer was made by addin g 1.0 mole/L HC l (Mallinckrodt 37% Assay) to the pH 7 buffer and adding an appropriate amount of KCl to the solution to make the ionic strengths of the two bu ff er solutions equal The final ionic strength of the buffers was 0 .22 M. Solutions of K 4 Fe(CN) 6 *3H 2 0 (98.5 % ACS grade) from ACROS Organics (0 .005 M ) were prepared with 0.5 M of KC l supporting electrolyte and the different buffer s olutions A dsorption solutions were prepare using 0.005 M of 3 mercaptopropioni c acid (99%)(MPA), n butanethiol (99% ), n octanethiol (97%) and n dodecanethiol (99%) in 95% ethanol (Aldrich). Electrode/ Instrument Preparation Cyclic voltammetry (CV), chronoamperometry, and chronocoulometry were performed with a 2 mm diameter (area of 3.142 m m 2 ) gold electrode from CH Instrument, Inc. The Au working electrodes was polished with 0.05 &m alumina micropolish fro m Buehler, and subsequently rinse d with deionized water All measurements were performed using a 604B electrochemical analyzer from CH Instruments (Austin, TX). A standard three electrode system was used, with a platinum wire as the secondary electrode and saturated calomel reference electrode. (Figure 4 ) All samples were placed in a CH Faraday cage to reduce the ambient electronic noise.
! ?A Samples were purged with N 2 for 10 minutes to remove O 2 in the solution prior to the electrochemical reaction. Figure 4 Picture of t he three electode system as used in these electrochemical experiments Cyclic Voltammetry Cyclic voltammetry was used initially to clean the A u working electrode. This was conducted by cycling the applied potential between 1.6 V to 0.3V for 21 s w eep segments in a solution of 0.5M sulfuric acid. Analytical C V scans w ere performed in 0.5M KCl solutions containing 0 .005 M K 4 Fe(CN) 6 *H 2 O in order to determine the standard potential of the redox reaction and to determine how mo dification of the Au electrode influences the interfacial interact ions with the redox co uple in the adjacent
! ?? solutions The CV experiment was conducted by cycling the a pplie d the potential from 0V to 0.6V and then back to 0V. Chronocoulometry Chronocoulometry measurements for the unmodified and modified electrode s were performed in pH7 and pH3 buffer solutions. Experiments with t he unmodified working electrode w ere first ru n in pH7 then pH3 solutions. The working electrode was then exposed to the mercaptan solution for 20 min utes ( 3 mercaptoprop ionic acid, n butanethiol, n octanethiol, and n dodecanethiol were used in this thesis research); and, when removed from the mercap tan solution, the electrode was rinsed with ethanol. Chronocoulometry measurements were then conducted with the modified electrode in the same pH7 and then pH3 buffer solutions. Desorption using Cyclic Voltammetry Cyclic voltammetry was also u sed for redu cti ve desorption to remove the self assembled monolayer from the working electrode. Desorption of the mercaptan occurs when a negative potentials is applied to the electrode in 0.5M KOH solution s The negative applied potential reduces the Au S interacti on causing the mercaptan to desorbs from the substrate surface. The current from this process can be integrated to quantitatively determine the amount of mercaptan originally adsorbed to the electrode. Prior to all measurements, t he unmodified Au working electrode was calibrated by finding the surface area by using chronoamperometry measurements in 0 0 5M KC l solutions containing 0 .005 M K 4 Fe(CN) 6 *H 2 O. The results were plotted according to Current vs. the inverse square root of Time and from the slope of t his plot the area of the
! ?K electrode can be determined u sing the Cottrell equation. The area of the first electro de was 3.40 ( 0 .12) mm 2 a nd for the second electrode 3.64 ( 0 .11) mm 2
! ?Q C HAPTER IV R ESULTS AND D ISCUSSION : C HARACTERISTIC OF M O DIF IED AND U NMODIFIED E LECTRODE S Chronoamperometry Chronoamperometry was used to determine the surface area of the electrodes that were used throughout this research. This technique provides an accurate measurement of the microscopic electrode surface area provided that the concentration and diffusion coefficient of the analyte species is accurately known. For the surface area determination, an average of fifteen individual experiments were conducted for each of the electrodes used. The current respons e is given by the Cottrell equation (below), and the area of the electrode could be determined by plotting the current versus the inverse square root of time: The area of the electrodes is found from the slope of this plot. 37 Of the two electrodes use d in the course of these measurements, the first electrode had an average area of 3.4 ( 0.1) x 10 2 cm 2 and the second electrode had an average area of 3.6 (0.1 )x 10 cm 2 The geometric area of the gold disk electrode is 3.14 x 10 2 cm 2 The difference in t he measured area versus the geometric area is due to microscopic roughness of the gold substrate. The roughness factor (the experimental area divided by the geometric area) for these electrodes is ~1 .1, a value consistent with reports for similarly prepar ed gold disk electrodes.
! ?Z Cyclic Voltammetry Unmodified E lectrodes Cyclic voltammetry is an experimental technique that can be used to characterize the electrode before and after a monolayer of the mercaptan is adsorbed on the electrode. In this method, a potential ramp is applied to the electrode and the current is monitored as a function of the applied potential. For the redox probe used (ferrocyanide), the current response has fast kinetics, is described by the Nernst equation, and has a very well de fined behavior. Any deviation from this behavior is diagnostic of the impact that the modification layer has on the heterogeneous reaction. In this work, the role of solution pH on the interfacial properties is being explored, especially with regard to t he impact that ionization of a carboxylic acid group at the terminus of the monolayer has on the electrode reaction Wang et al. showed using quartz crystal microbalance that the pKa of confined carboxylic acid groups is around 4. 29 Wang et al. also sh owed that the protonation of the carboxylic acid group is reversible and that there was no difference between the behaviors for a monolayer with an ionizable carboxylic acid group when the solution pH is 3 and the behavior with an unmodified gold electrode 4 This result suggests that the neutral monolayer has no impact on the interfacial behavior. They also show that the frequency of the quartz crystal oscillator changes at basic pH, consistent with deprotonation of the carboxylic acid group. They interpr et the frequency change to viscoelastic changes at the interface due to the electrostatic repulsion between neighboring anionic groups. Solutions with pH of 3 and a pH of 7 were chosen for this
! ?W work to create conditions wher e a monolayer of 3 mercaptopro p ionic acid is fully protonated (pH 3) and fully ionized (pH 7). Unmodified electrodes were polished and cyclic voltammetry experiments conducted in the pH 3 and pH 7 solutions containing the same concentrations of ferrocyanide. The peak current ratio o f the forward to reverse potential scans for the pH 3 solution is 0.99, and that for the pH7 solution is 0.90. These values compare favorably to the theoretical value of 1. The peak ratio can be affected by many factors, including chemical reaction with the redox process. 37 Figure 5 shows the similarity of the I vs. E behavior of unmodified electrodes at different pH values. Figure 5 C omparison of the blank electrodes. The red line shows CV at pH3 and the blue at pH7
! ?U Modified MPA E lectrode Whe n the electrode w as modified with 3 mercaptoprop ionic acid, the behavior is significantly different. The ratio of the forward to reverse peak current ratio of pH 3 is 0 .96 and pH7 is 0 .87. The smaller peak current ratio at pH 7 is characteristic of slow er heterogeneous kinetics, and is characteristic of a more difficult reaction. At pH 7, I t is expect ed the that MPA m onolayer will be fully ionized, and have electrostatic repulsion with the negatively charged ferrocyanide reactant ion. This electrostati c interaction will result in differences in the peak current response for the reactant and the product ion (ferricyanide) Figure 6 illustrates the difference s between the modified electrodes at different pHs Figure 6 Shows the Au electrode modified w ith MPA. Red is at pH3 and the blue is at pH7
! ?[ At pH 7, The carboxylic acid group deprotontates causing electrostatic interactions to increase and slowing down the kinetics of the heterogeneous reaction.. 5,29,30 Reductive Desorption of the SAM on the ele ctrode Surface coverage s of the molecules n butanethiol, n dodecanethiol, n octanethiol and 3 mercaptopropionic acid on the gold electrodes were determine d by cyclic volt ammetry in 0.5 M KOH. Table 1 summarize s and compares the surface coverage of each o f these monolayer s on gold as determined by the reductive desorption measurement. Reductive desorption is simply a cyclic voltammetry experiment in which the Au S interaction is reduced and in the process the molecules that comprise the monolayer desor bs from the electrode surface. By integrating the area under the reductive desorption peak(s), the charge associated with the desorption can be quantitatively determined. From this value and the stoichiometry of the reduction reaction, a quantitative det ermination of the surface coverage is obtained. Table 1 Surface coverage determined by cyclic voltametry Literature value didn't report error. ** Not enough samples to take a good standard deviation The surface coverage s of the monolayer are in agr eement with t he literature values except for dodecanethiol Figure 7 shows typical current vs. potential reductive desorption curve
! ?S f or one of the monolayers. The 3 mercaptopropionic acid has a short chain, which causes slow desorption of the sulfur group. Schreiber also sh owed that the shorter chain can not compete with the thiol gold interaction. 1 Longer chains like octanethiol and dodecanthiol form a densely packed monolayer ; and due to the chains being in a n all trans conf ormation have fewer gauche def ects. 1,11 The less defective monolayers are more difficult to desorb electrochemically. Figure 7 Sample of the desorption of 3 mercaptopropionic acid usin g Cyclic voltammetry between 0 and 1.3 V.
! K@ Chronocoulom e try Comparison of Monolayer B ehavior in pH 3 and pH 7 Chronocoulomtry measures the initial charge amount required to charge the interface (e.g. the double layer) and the charge amount required to oxidize/reduce the redox species immediately adjacent to the electrode, in this case Fe(CN) 6 3/ 4 in a phosphate buffer. The ability of chronocoulometry to quantitatively determine the surface charge makes this an ideal method for determining the influence that the electrostatic interaction of the monolayer has on the distribution of ions adjacent t o the modified electrode. The quantitative values presented will be given for a single electrode preparation. This is because the quantitative values determine the surface condition o f reactants present at the electrode T his is dependent on the preparat ion of the monolayer, and the day to day preparation of the monolayer is difficult to exactly reproduce. For this reason the experiment s will compare the qualitative trends in the data of the data within a series of experiments conducted on the same day with the same electrode preparation Chronocoulomtry was initially performed using an unmodified gold electrode for the oxidation of ferrocyanide in solutions buffered at pH 7 and pH 3. Multiple runs were performed with the same electrode. The ave rage i ntercept at pH 3 was 1.1 x 10 6 C and at pH 7 the intercept was 2.1 x 10 6 C. This result shows that the behaviors are similar un der these conditions, (F igure 8 ) This is consistent with our expectation that the solution pH should not have a significant impact on the interfacial behavior of the ferroc y anide. Chronocoulometry is calculated using the last 80% of the data, from which the slope and
! KA intercept is calculated. The first 20% of the data is discarded due to the charging process and potential not b eing fully established. Table 2 illustrates different runs in pH7 on a blank electrode calculating the last 80%, 70% and 60% of intercept using chronocoulometry. Data show there is little difference from 80% to 60% Relative standard deviation was calcu late approximately at 8% for pH7 for one electrode in table 2. This percentage remains consisted using different pH's and electrodes. Table 2 Chronocoulometry intercept calculate at the last 80%,70% and 60% and standard deviation for different runs. R elative standard deviation calculated (RSD)at pH 7 for 80%. Quantitatively this result represents the charge required to establish the double layer, and can be used as a reference value for experiments that make use of the ionizable monolayers.
! K? Figure 8 Comparisons of the blank electrode at pH 3 (blue) and pH 7 (red ) in chronocoulometry When 3 mercaptopropionic acid was absorbed onto the electrode, however, very different chronocoulom e try results were obtained. The modified electrode exhi bited a s maller intercept (7.4 x 10 7 C) at pH 7; while, at pH 3, the same electrode h ad an average intercept of 2.3 x 10 6 C similar to the value found with the unmodified electrode When the results for an unmodified electrode and an electro de modified with 3 mercaptoprop ionic acid were compared, the chronocoulometry intercepts at pH 3 were similar but a difference between these two electrode preparations at pH 7 was found This result is interpreted in terms of the electrostatic repulsion between the negat ively charged interface at pH 7 and the anionic redox probe (ferrocyanide). The electrostatic repulsion creates a concentration deficit of the ferrocyanide in the volume of solution
! KK immediately adjacent to the electrode; consequently, the amount of initia l charge needed in the chronocoulometry experiment is smaller than if the ferrocyanide concentration were homogeneous throughout the solution up to the electrode surface. Figure 9 Modified electrode in chrono coulometry with 3 mercaptopropionic a cid at p H 7 (red) and pH 3 (blue ). Electrodes Modified with n alkanethiols at pH 7 and pH 3: comparison with results from 3 mercaptopropionic acid Three different n alkanethiols that were also used to modify the gold electrode, and chronocoulo metry was performed using these electrodes for the oxidation of ferrocyanide al so at pH 7 and pH 3. Table 3 summarizes the results of the initial charge found for these electrodes, and compares these results to those of 3 mercaptopropionic acid. All samples were prepare d under the same conditions for the chronocoulometry measurements and for the electrode modification.
! KQ Table 3 Comparison of the alkanethiols to 3 mercaptopropionic acid in chronocoulometry under similar conditions N butanethiol had almost no change in different pH while the 3 mercaptopropionic acid decreased in pH7 compare to pH3. Results from n butanethiol indicate that there is no difference in the chronocoulometry initial intercept at the different pH values. The 3 mercaptopropionic acid and n butanethiol have similar number of substituent T herefore are comparable due to the way that they are orientated and coverage on the gold surface The octanethiol and dodecanethiol have longer chains and are orientated different by on the gold. They a lso have stronger van der Waals interactions and therefore the comparison to 3 mercaptopropionc acid is invalid. These electrodes modified with monolayers of different n alkanethiols have behaviors that are independent of pH because the pH of the adjacent so lution has little to no effect on the charge of these monolayers. The monolayer prepared with 3 mercaptopropionic acid, on the other hand, is deprotanted at pH 7 creating an anionic interface. This interfacial charge causes the intercept to be smaller tha n what is found at pH 3, where the interface is neutral. This is due to the monolayer being negatively charge d causing a surface deficit, which will make the intercepts smaller.
! KZ C HAPTER IV C ONCLUSION In this research, we demonstrate that the self assem bl y of a monolayer of 3 mercaptopropionic acid can influence the distribution of ionic redox couples in the volume of solution immediately adjacent to the electrode surface. This distribution of ionic species is influenced by th e pH of the adjacent solution because that pH determines the ionization of the confined 3 mercaptoprop ionic acid monolayer. The surface cover age of 3 mercaptopropionic acid on gold was found to be 2.3(0.4)x10 9 mol/cm 2 which is consistent with literature values. 5 Monolayers p repared with other n alkanethiols were consistent with each other, and with literature results, in regards to the surface courage and the lack of influence of the pH of the adjacent solution on the monolayer behavior. The characteristics of 3 mercaptoprop ionic acid monoalyers that were deposited on gold electrodes were evaluated at differe nt pH's using cyclic volt a mmetry and chronocoulometry. As the pH of the adjacent solution increased to 7 the carboxylic acid group on the 3 mercaptopropionic acid depro to nates creating an electrostatic repulsion with the anionic ferrocyanide redox active species. This effectively creates a concentration gradient that leads to a concentration deficit of the redox active species in the volume of solution adjacent to the e lectrode. Chronocoulometry was utilized to com pare monolayers prepared from 3 mercaptopropionic acid to monolayers prepared from different n alkanethiols to demonstrate the influence of the electrostatic interactions at the electrode interface. The n alka nethiols used were n butanethiol, n octanethiol and n dodecanethiol. As the
! KW conditions of the so lution became more basic, the 3 mercaptopropionic acid deprotonates which inhibits the ability of the anionic ferrocyanide reactant to approach the electrode s urface. The electrodes modified with n alkanethiols were constant in that they showed no change in behavior as the solution pH was altered. The data shows that electrodes modified with 3 mercapto pro pionic acid influences the ability of anionic species to approach the electrode surface. Modifying the electrode with ionizable molecules that make up the monolayer allows the control of interfacial interactions by adjusting the pH of the solution adjacent to the electrode surface. This suggests that one must carefully consider the molecular structure of the electrode when interpreting electrochemical results. It also suggests that solution pH can be a valuable parameter when modifying electrodes to be used in sensing applications. ( ( ( ( ( ( ( (
! KU BIBL I OGRAPHY ( ( AH \H!-.$*'"2'*:! !"#$"%&&'()'*+",-.%'*.(%).%' WZ!L?@@@N!AZA J ?ZWH ?H !]H!X,/1:!^H!_'/1:!;!\,/1:!`H!_$)>: '/#+")-0'#,'10%.2"#)-)-032(.-0'45%6(&2"3' Q[Q! L?@@@N![[ J S?H KH -H!a)/1:!]H!<": '7(#%0%.2"#.5%6(&2"3'-)8'7(#%)%"$%2(.&' Q?!LASSUN!U J AKH QH -H!X,/1:!aH !aB:!`H!`B: '/#+") -0'#,'9::0(%8'10%.2"#.5%6(&2"3' KQ!L?@@QN!QSZ J Z@@H ZH XH!-,/0'*%:!+H 6H!9/0'*%)/:! ;-)$6+(" ?Q!L?@@[N!A?UWW J A?UU@H WH ;HaH!_,/1D'"%&'*:!BH!^'*&)..":!;H6H!^',>.$,D7:!bH6H!-&,(()*0:! '10%.2"#.5(6(.-' 9.2!ZK!L?@@[N!WUU[ J WU[WH UH bH!a'.$'*:!]HaH!=)/1:!]H!-.$D"&&:! '<5()'*#0 (8'=(06& !?A@T?AA!LASS?N![KA J [KZH [H 6H!bH!c>EE):!2)"%:!aH!'!53& (.-0'45%6(&2"3 SU!LASSKN!SQZAH A@H !9H]H!^,*0:!54/'*: '10%.2"#.5%6(.-0'?%25#8&@'=+)8-6%)2-0&' -)8' 9::0(.-2(#)&A !? /0! '0 > ])$/!,/0!X"5'3!,/0!-)/%!I/.H :!c'G!d)*4!e?@@Af AAH !9H!B5D,/:!! 45%6 (&2"3 >'B%C %(D >' !SW!LASSWN!AZKK J AZZQH A?H ;H!aH!^,"/:!gH!^H!#*)>1$&)/:!dH!#,):!]H!g8,55:!bH+H!X$"&'%"0'%:!6HbH!c>EE): /#+")-0'#,'96%"(.-)'45%6(&2"3'*#.(%23 AAA:!LAS[SN K?A J K KZH AKH MH!#H!=,DD)/0:! 4+""%)2'#:()(#)'()'4#00#(8'E'F)2%",-.%'*.(%).% !Q!L?@@@N!QK@ J QQ?H AQH #H!9H!])/'%:!bH!MH!M'*'E:!^H!]H!])$/%)/:!6H!+H!;*))4%:! ';-)$6+(" !AA!LASSZN!AKA[ J AK?[H
! K[ AZH -H!-H!-$"*,&)*":!+H!\H!6>2/'*: '?-".#6#0%.+0%&' KK!L?@@@N!Q?AK J !Q?ASH AWH ;HgH!])*0,/:!^H!< H!\*'3:!-H!h)*/1>&$:!6H!+H!;)*/:! ';-)$6+("' A@!LASSQN!KWQ? J KWQ[H AUH !cH9H!h)&)8:! 'G-)#*2"+.2+"%8'?-2%"(-0& !A?!LASSSN!U[S J USWH A[H !XH^H!-&).4&)/:!+H\H!6>2/'* : '?-".#6#0%.+0%&' K@!LASSUN!?UAU J ?U?ZH ASH 6H!/,G,3:!6H!]H!X"55".>&:! ;-)$6+("' S!LASSKN!?UUZ J ?UUUH ?@H -H!gH!^>*4':!;H!]H!^,**'&&:! ;-)$6+(" AS!L?@@KN!K?SU J KK@KH ?AH !-H!':!MH!#H!=,DD)/0: ''*+:"-6#0%.+0-"'*.(%).%A !Q!LASSUN! AQA J AQWH ?KH ]HXH!^,>*H!LASSUN '=-""(.-2(#)'-)8'*2"+.2+"-0'&2 +8(%&'#,'*%H+%)2(-003'98"I%8' !#03%0%.2"#032% '?+0(20-3%"&A' M$a !#$'%"% :!+I#H ?QH #H!h)/0):!]H!+)*"&,:!!+H!i4,D>*,:!#H!-,"&):!hH!B)%,4":! /#+")-0'#,' %0%.2"#-)-32(.-0'45%6(&2"3' ZK?!L?@@?N!?@A J ?@ZH ?ZH aH!=H!g8,/%:!+H!]H!h'553:! 9)-0 32(.-0 '45%6 (&2"3 ZQ!LAS[?N!AU?U J AU ?SH ?WH !]H!=H!;$*"%&"':! /#+")-0'#,'10%.2"#-)-032(.-0'45%6(&2"3 !AK!LASWUN!US J [SH ?UH ]HMH!^>.$'*:!.+""%)2&%:-"-2(#)&>.#6J(&&+%&JKL M NJKL M N%>:8, A !;$*)/).)>5)D'&*3:!,..'%%'0!]>53!?A :!?@A?H ?SH ]!X,/1!:!'!53& (.-0 '45%6 (&2"3 !SW!LASS?N!Z??Q J Z??[H K@H ;H!9H!X"0*"1:!;H!;$>/1:!+H!aH!M)*&'*:! /#+")-0'#,'10%.2"#-)-032(.-0'45%6(&2"3' KA@! LASSAN!KKZ J KZSH
! KS KAH \H!;H!9/%)/:! '9)-032(.-0'45%6(&23 !K[!LASWWN!ZQ J ZUH K?H ;H!aH!#"0G'55:!-H !IH!g*&'5:!^H!aH!6,&/'*: ';-)$6+("' AK!LASSUN!KK@Q J KQAKH KKH +H!d>,/:!-H!_$,/:!`H!_$)>:!dH!<">:!:! ;-)$6+("' ?Q! L?@@[N![U@U J [UA@H KQH !+H!X,5.E,4:!!aHaH!M)7'/)':!6H!-H!a'"/$,DD'*:!^H!aH!<,D7:!;H ;$>/1:!+HaH! M)*&'*:! ;-)$6+(" !U!LASSAN!?W[ U J ?WSKH KZH !]H!=H!;$*"%&"':!6H!9H!i%&'*3)>/1:!\H!;H!9/%)/:! /#+")-0'#,'10%.2"#-)-032(.-0' 45%6(&2"3' AK!LASWUN!?KW J ?QQH KWH XH # H !d,7:!6H!#H!^>*4':!gH!9H!^5>2,>1$:!!6H!9H!a>*%&:! /#+")-0'#,'10%.2"#-)-032(.-0' 45%6(&2"3' AZS!LAS[KN!?[U J ?SKH KUH ]H !X,/1: '9)-032(.-0'10%.2"#.5% 6(&2"3O' j=!M>25"%$'*:!I/.H :!c'G!d)*4!LASSQNH K[H aH!M5'&.$'*: '9'=("&2'.#+"&%'()'10%.2"#8%'!"#.%&&%& :!?! /0 '0:!#$'!6)3,5!-)."'&3!)(! ;$'D"%&*3:!;,D2*"01'!e?@@SfH