Optode-bead-based Functional Chemical Imaging of 2D Substrates

Document Sample
Optode-bead-based Functional Chemical Imaging of 2D Substrates Powered By Docstoc
					               OPTODE-BEAD-BASED FUNCTIONAL CHEMICAL IMAGING OF 2D
                                  SUBSTRATES




                                     Punkaj N. Ahuja

                       Submitted for the degree of Master of Science




                                  Advisor: Miklós Gratzl

                          Department of Biomedical Engineering

                            Case Western Reserve University

                                      August, 2011




       	
  
	
  
       1	
  
                           CASE WESTERN RESERVE UNIVERSITY

                               SCHOOL OF GRADUATE STUDIES


                             We hereby approve the thesis/dissertation of
                             Punkaj Ahuja________________________________

                                candidate for the ___________Master’s__________degree *.



                      (signed)____________Miklos Gratzl__________________
                                      (chair of the committee)

                      ____________________Andrew Rollins________________

                      _________________Shawn McCandless________________

                      ________________________________________________

                      ________________________________________________

                      ________________________________________________


                    (date) ______June 7, 2011________


       *We also certify that written approval has been obtained for any proprietary material
       contained therein.
       	
  




	
     	
  
       2	
  
Contents
1.      Motivation                                                       7
        1.1 Biological Sensing of Surfaces                               7
        1.2 Scanning Electrochemical Microscopy                          8
        1.3 Scanning Electrochemical Microscopy basics                   8
        1.4 Scanning Electrochemical Microscopy in Biological Settings   10
        1.5 Bulk Optode Technology                                       11
        1.6 Bulk Optodes and Potentiometric Ion Selective Electrodes     13
2.      Optode-bead-based Functional Chemical Imaging of 2D Substrates   15
        2.1 Materials and Methods                                        16
        2.2 Results                                                      22
               2.2.1 Optode bead calibration                             22
               2.2.2 Raw Data                                            23
                      2.2.2a Lateral Diffusion                           23
                      2.2.2b Spherical Diffusion                         24
                      2.2.2c Spread Distance Analysis                    25
               2.2.3 Linear Interpolation                                27
        2.3 Generation of Color Maps                                     27

               2.3.1 Thresholding                                        27
               2.3.2 Delaunay Triangulation                              28
               2.3.3 Color Map Generation with Circular Symmetry         26
               2.3.4 Normalized Color Maps                               29
               2.3.5 Effect of normalization of color map                30
        2.4 SECM Simulations                                             31
               2.4.1 SECM Simulation (1) Hit or Miss                     31
               2.4.2 SECM Simulation (2) Tiling                          32


	
  
3	
  
        2.5 Optode-membrane Based Function Imaging                           33
        2.6 Discussion                                                       34
               2.6.1 Calibration                                             34
               2.6.2 Raw Data and Spread Distance Analysis                   34
               2.6.3 Color Maps                                              36
               2.6.4 SECM Simulation (1) Hit or Miss                         36
               2.6.5 SECM Simulation (2) Tiling                              37
               2.6.6 Bead Clumping and Dispersion                            37
               2.6.7 Overall Comparison of Functional Optode-bead-based Imaging and
               SECM                                                           38
        2.7 Conclusions and Future Work                                      39
Acknowledgements                                                             39
References                                                                   55




	
  
4	
  
List	
  of	
  figures	
  
        1        Modes	
  of	
  SECM	
  imaging	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .8	
  

        2        Typical	
  Bio-­‐SECM	
  setup.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .10	
  	
  

        3        Optode	
  bead	
  calibration.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .20	
  

        4        Calibration	
  plots.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  ..	
  .	
  20	
  

        5        Raw	
  data	
  for	
  acidic	
  diffusion.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  21	
  

        6        Raw	
  data	
  for	
  basic	
  diffusion.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  22	
  	
  

        7        Spread	
  distance	
  for	
  lateral	
  diffusion.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  23	
  	
  

        8        Spread	
  distance	
  for	
  spherical	
  diffusion.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  ..	
  .	
  .24	
  	
  

        9        Thresholding	
  mask	
  images.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .25	
  	
  

        10       Delaunay	
  triangulation.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  ..	
  .	
  .26	
  	
  

        11       Color	
  map	
  with	
  circular	
  symmetry.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  27	
  	
  

        12       Normalized	
  color	
  maps	
  and	
  comparison	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  ..	
  .	
  28	
  	
  

        13       Spread	
  distance	
  color	
  intensities.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  29	
  	
  

        14       Hit	
  or	
  miss	
  simulation	
  (lateral	
  diffusion)	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .30	
  

        15       Hit	
  or	
  miss	
  simulation	
  (spherical	
  diffusion)	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  30	
  	
  

        16       Tiling	
  simulation.	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  .	
  ..	
  .	
  .	
  .	
  .31	
  	
  




	
  
5	
  
        Optode-bead-based Functional Chemical Imaging of 2D Substrates
                                         Abstract
                                            By
                                    PUNKAJ AHUJA
        We introduce a new technology for functional imaging of surfaces using

microscopic optode beads. The imaging technology produces a map of 2D chemical

distribution, obtained continuously and simultaneously at each pixel using a charged

coupled device (CCD) camera. Ionic distributions can be imaged with this technique

using ionophore-chromoionophores interactions in microscopic optode beads or

membranes. Response to pH diffusion from a 140 um channel were measured. Lateral

and spherical diffusion processes are compared. Data was resolved at each pixel using

linear interpolation of three closest beads with Delaunay triangulation and color-

concentration maps were created assuming circular symmetry. The approach is similar to

Scanning Electrochemical Microscopy, an imaging modality that relies on the scanning

of an electrochemical microsensor along a surface of interest. Potential or current

measures local concentrations of an analyte at each scanned position. Scanning of the

microelectrode tip is, however, time consuming and the tip itself may run into physical

obstruction at uneven surfaces. Functional optode-bead imaging provides an imaging

modality that can produce similar functional 2D maps as scanning electrochemical

microscopy does, but without the need for physical scanning of an electrode. The

functional optode bead technology has a variety of potential including measurements in

systems scanning electrochemical microscopy could not be applied. Some biomedical

applications include imaging of pH in epithelial cell monolayers and multicellular tumor

spheroids (MCTS), which are meant to represent tumor microenvironments.

	
  
6	
  
1. Motivation


1.1 Biological sensing of surfaces
        Imaging of ion distributions in biological samples could lead to a better

understanding of the effects of ionic diffusion on a wide variety of systems. While there

are common modalities which can electrochemically measure concentrations, there are

many processes which are difficult to measure or difficult to confirm. For example,

Potassium concentration and diffusion from channels has been linked to brain seizure

susceptibility [1], however the potassium concentrations across the specific areas of the

brain have not been imaged or adequately studied to date.. Multicellular tumor spheroids

(MCTS) are models which are used to mimic tumor microenvironments as studies to

measure mass transport, drug uptake, hypoxia and acidity are difficult to study in native

tumors [2, 3]. The viability of this model however has not been confirmed in specific

studies, since it is difficult to measure both pH and oxygen distributions throughout the

spheroid. We introduces an optode-bead based technology which could be used to

measure the ionic distributions and provide functional imaging of surfaces for a variety of

analyte, (i.e. H+, Na+, K+, lactate, Ca+, OH-, Cl-) particularly of interest in biological

sensing studies. [23, 18, 25]


1.2 Scanning Electrochemical Microscopy




	
  
7	
  
        Scanning electrochemical microscopy (SECM) is a technique developed in the

late 1980s for the characterization of surface reactivity at various interfaces including

liquid-liquid, solid-liquid, and liquid-gas [26]. It is used to monitor the electrochemical

activity of a surface [4]. SECM initially was used to study surfaces but has recently been

applied to biological samples and specifically the study of living organisms [27].


        SECM utilizes a microelectrode (ME) that serves as a mobile probe and scans a

surface, recording measurements (changes) in the electrochemical potential. The SECM

electrode also has a positioning system that provides mobility to the probe in the

horizontal and vertical (x, y, and z) directions. As the probe scans, a steady-state current

is measured using a potentiostat. Reference and auxiliary electrodes are incorporated in

the setup outside of the area of scanning. There are a variety of applications for SECM

that include corrosion studies [5], surface modification studies [6] and uptake and release

from biological cells [7].


1.3 Scanning electrochemical microscopy basics


        SECM microelectrodes (ME) are defined by the dimensions and geometry.

Typically, they are smaller than the thickness of the diffusion layer in surrounding

solutions, typically ranging from tens of micrometers to nanometers. Their geometry

varies depending on intended application and can include disks, spheres and hemispheres


	
  
8	
  
[8]. For SECM measurements, it is a requirement for the ME to reach steady-state

conditions rapidly. For disk ME, the steady state current, with a large insulating sheet is

defined in the equation below


                                                             (1)


where	
  ne	
  is the number of electrons in the electrochemical process, F is the faraday

constant, D is the diffusion coefficient of the active species in solution, c is the

concentration of that species, and r is the radius of the electrode. Steady-state conditions

are defined when the sensing area of the electrode is localized within a constrained

volume at the electrode/solution interface. This requires that the electrode position be

held close in proximity to the surface.


        The current measured at the electrode interface, of an active, dissolved substance,

A, is proportional to the flux of A, JA . This can be described by Fick’s first law where x

and t represent space and time.




                                                                   (2)	
  

	
  
This indicates that the concentration profile of the active substance is dependent upon

both distance from the electrode and time. The SECM response is influenced by multiple

factors including tip geometry, tip to surface distance, and ratio of radii of insulating

material to metal electrode wire.




	
  
9	
  
         There are two modes of SECM imaging, constant-height and constant-distance.

In Constant height imaging, the ME is scanned over the surface in x- and y- directions

maintaining a fixed distance (z-) above the sample. In contrast, constant distance SECM

uses feedback to maintain a specific ME-to-sample distance. Due to higher resolution

and low chance of tip crash, the constant distance mode is most frequently applied to

SECM in biological scanning..



                                                         pH 7.5



                                                         Figure 1. The figure shows the two modes
                                                         of SECM imaging, constant-height and
                                                         constant distance.
1.4 Scanning electrochemical microscopy in biological settings


         SECM (Bio-SECM) application to biological samples has been relatively recent

[11]. It has been suggested that it could possibly quantify the flux of molecules entering

or leaving a cell, probe local electrochemical reactions occurring at or inside a cell, and

perform measurements on both a single cell and confluent assemblies or tissues.


         In order to use SECM for biological studies there are a variety of properties the

setup must ensure. These include, appropriate dimensions, stability for time scale of the

experiment, sensitivity and selectivity for analyte, and high signal-to-noise ratio [9].

These requirements however provide some practical limitations to the use of SECM for

biological settings.


         Bio-SECM has been used to perform extra and intracellular oxygen gas

measurements, and cell metabolite measurements [10, 7, 21].



	
  
10	
  
         Bio-SECM requires that the ME can be scanned over a surface with the guidance

of a high-resolution positioning system. As the electrode scans the surface, steady-state

current is monitored. Steady-state current is only seen as the active species is mass

transfer limited for potentials exceeding the redox potential of the species.


         The typical instrumental setup for Bio-SECM is bulky. It includes an optical

microscope, a x-, y- translation stage, an electrochemical cell, a ME, a constant distance

controller, and a z-axis positioning system, within a Faraday Cage, atop a vibration

isolation table [20] The setup requires significant room and stable conditions during the

SECM experiment. These experiments can be lengthy, upwards of hours, at a time. The

figure 2 shows a possible setup for a Bio-SECM experiment. We believe that optode

beads and optode membranes could potentially be much less bulky.


                                             Figure 2. Typical Bio-SECM setup (a)
                                             optical microscope (b) x,y translation
                                             stage (d) camera (e) electrochemical cell
                                             (f) ME (g) constant distance controller (h)
                                             z-axis positioner (i) faraday
                                             cage/vibrational isolation table




1.5 Bulk optode technology


         Ion selective bulk optodes were first developed in the late 1980s [4]. They are

similar in ways to ion selective electrodes in terms of components; however there are

some general differences. The key components of both types of sensors are ionophores,

	
  
11	
  
lipophilic complexing agents. These agents are capable of reversibly binding specific

ions. Ionophores can be either charged or neutral in their uncomplexed form. In addition

to the selective ionophore, a second ionophore is included which specifically interacts

with a reference ion. This ionophore undergoes a change in optical characteristics when

it undergoes complexation with analyte. These ionophores are typically referred to as

chromoionophores. The chromoionohore is pH sensitive, so as its degree of protonation

depend on the activity of competing ions and the target ions. The activity of target ions

can be calculated from the absorbance changes of the sensing film. Optical sensors of

this type are referred to as bulk optodes because the analyte is extracted not only on the

surface of the membrane but into the bulk.


         Polymer films, typically polyvinyl chloride (PVC) are used as water-immiscible

matrices for the ionophores.


         During measurements the bulk of the optode film is assumed to be in equilibrium

with the sample. The equilibrium process depends on ion activities and not directly on

concentration. The reason single ion activities can be measured is due to the fact that

either the pH or the reference ion in the sample solution is known. Thus, the optode

response is dictated by both the sample activity of the desired ion and the interfering ion

which co-participates in maintaining equilibrium. This is due to the fact that

electroneutrality must be maintained in the bulk phase. A phase transfer equilibrium of

two distinct ions must be established. Depending on the ion to be detected, a variety of

either neutral carrier or charged-carrier optodes can be designed. Either a competitive ion

exchange or carrier mediated co-extraction equilibrium provides optical response. The



	
  
12	
  
complexation reaction of the ions leads to optical response which can cause change in

absorbance, fluorescence, or phosphorescence depending upon the ionophores used.


         In developing bulk optodes, the most common approach is using

chromoionophores which are pH sensitivity along with second ionophores for specific

ions of interest. H+ sensitive ionophores are extremely selective and it can be assumed

that complexation with other ions will be negligible [12]. A potential drawback is the

cross-sensitivity, as a result of sensitivity to two competing ions in the membrane, to pH

and the other ion. This can be dealt with by either buffering the sample or measuring the

opposing ion simultaneously.


         Our bulk optode system was based pH sensitive optode beads that contained two

electrically neutral ionophores. The ionophores were the chromoionophores, sensitive to

H+ , a Sodium ionophore, and trapped lipophilic anionic sites which provides the

membrane the cation exchange properties necessary for measurements, and to maintain

bulk electroneutrality.


1.5 Bulk optodes and potentiometric ion selective electrodes.


         While ion-selective electrodes and ion optodes utilize similar active components

and composition, but they use differing response mechanisms. ISE’s measure a potential

which is directly correlated to the activity of the analyte in the sample. Optical sensors

rely on changes in the sample composition to induce significant concentration changes in

the sensing film such that optical response can be detected (e.g. change in color). The

optode response is dependent on ratios of two ion activities one of which has to remain

fixed.

	
  
13	
  
         Differences in compositions exist between ISEs and ion optodes. For bulk

optodes, an excess of counter ion ionic sites is common. In many cases when the pH is

held constant in sample, the chromoionophore is in excess. This ensures that the counter-

ionophore or indicator can be fully protonated. The excess ionic sites are

counterbalanced by sample ions which are extracted and complexed by the lipophilic

carrier. In the case of measuring pH, it is important that the concentration of H+

chromoionophore is much smaller than the ion carrier, and ionic sites, in the following

order: L+T > R-t > C+t .


         Further, optode membranes must remain very thin in order to maintain the proper

response mechanism. If the PVC based membrane is greater than a few microns thick,

the membrane can deteriorate after contact with sample solutions. The introduction of

water to the membrane can cause turbidity which causes change in response. If the

membrane is only a few microns thick, however, this impact is not seen [13].


         The selectivity of optical sensing schemes versus potentiometric is similar as well,

and both can be effectively tuned for specific ranges by incorporating fractions of active

membrane components. The selectivity of optodes can be altered by adjusting polarity

of membrane material, often defined by the plasticizer (i.e. diocytl sebacate, a non-polar

plasticizer) [14, 15]. The analysis range of optodes is significantly smaller than that of

ISEs, with 2 to 4 orders of magnitude compared to 5 to 9 respectivly. However, as

mentioned, optodes can be tuned for specific ranges and multiple sensors can show

highly selective measuring ranges. The response time of ISEs is significantly faster that

of optodes, millicseconds to seconds, and seconds to minuts, respectively [15]. Optodes

require equilibrium between the sample and the bulk of the membrane, which requires

	
  
14	
  
diffusion within the polymeric film to be the rate-limiting step. Response times for

optodes can be in the order of seconds. Because the response time is limited by the

diffusion within the film, the use of microscopic optode beads provides relatively fast

response compared to bulk membranes. The bulk response of beads, because of size can

ensure response on the scale of seconds, or even milliseconds, for individual beads with

diameters ~2 um.


         The lifetime of both ISEs and ion-optodes is significantly impacted by the loss of

plasticizer, carrier, or ionic site caused by leaching into the sample [27]. If any of these

materials are leached into the sample, there is a shift in the equilibria for the optode

membrane response. Optical sensors, due to their minimal thickness can have increased

leaching rate in the sensing film, about 100 times faster than ISE membranes [15].

Further, photobleaching can occur in the chromoionophore with strong excitation

sources, particularly fluorescence [16]. High lifetime can be ensured by incorporating

active components onto the polymeric membrane for example [17]. Optodes have the

advantage compared to ISEs in that they can be much more miniaturized. Any hole in the

ISEs membrane causes an electrical short, which limits the size of ISEs. This is

advantageous for our applications to studying chemical distributions at surfaces. Optodes

are often considered ideal for continuous monitoring for environmental analysis where

rapid measurements are not critical [15].


2. Optode-bead-based functional chemical imaging of 2D substrates


         Here we describe a new, continuous method for mapping local proton and

hydroxyl ion concentrations across a surface. Diffusion of both hydroxyl ions and protons


	
  
15	
  
are examined using a 150 um source. Both cylindrical and spherical diffusion was

examined. Measurements were made using pH-sensitive optode beads dispersed at

random points on the surface. The information from the beads at each time point is then

used to develop a concentration topograph of the surface. We also include comparisons

of results achieved with our system and the results that would be expected with a

typical SECM experimental setup.


2.1 Materials and Methods


Reagents


         9-(diethylamino)-5-[(2-octyldecyl)imino]benzo[a]- phenoxazine (hydrogen ion

selective chromoionophore III), Bis[(12-crown-4)methyl]-2-dodecyl-2-methylmalonate

(bis(12- crown-4)), sodium tetrakis[3,5-bis(1,1,1,3,3,3-hexafluoro-2-methoxy-2-

propyl)phenyl]borate (NaHFPB), bis(2-ethylhexyl)sebacate (BEHS), poly(vinyl chloride)

(PVC), were purchased from Sigma-Aldrich and used without purification.


         Type 1 Agarose was used for all membranes (Sigma-Aldrich). Agar membranes

were prepared from 0.0001 M phosphate buffered saline with pH 7.4 at 25°C (Sigma-

Aldrich) containing 0.138M NaCl and.0027M KCl. Deionized water (specific resistance

>18.2 MΩ cm) by a Milli-Q water system (Millipore Corp., Bedford, MA, USA) was

used in all solutions.




Procedure


Preparation of microscopic optode beads.

	
  
16	
  
         PVC/BEHS microscopic beads were prepared to be used to produce color

changing optode beads. A spray dry method was used as explained below: a THF

solution containing 1 wt% PVC and 1 wt% BEHS was sprayed with a nebulizer under

heated air stream from a heat gun and PVC/BEHS particles (2.5 ± 1.0 um in diameter,

n=100) were collected in a cyclone chamber. To add the color changing component to

the beads, 300 mg of PVC/BEHS beads were mixed in 50 mg of BEHS solution

containing 0.5 mg of hydrogen ion selective chromoionophore III, 1.6 mg of NaHFB, and

10 mg of sodium ionophore, bis(12-crown-4) were added and mixed thoroughly [18].


Preparation of agar membrane with dispersed optode beads for calibration.


         pH-sensitive optode beads were manually dispersed on a glass slide (22 x 22 x 1.5

mm). A second glass slide was pressed against this to allow for further dispersion beads

by static charges. The slides were separated and two, single-folded pieces of aluminum

foil (14 um thick) was placed on two opposite sides of a slide. An identical, clean slide

was then clamped along the pieces of aluminum. The clamped structure was placed in 1

wt% agar and by capillary action a gel layer is wicked through between the slides with

pH sensitive optodes beads dispersed throughout.


Preparation of agar membrane with dispersed optode beads for cylindrical

diffusion.


         pH-sensitive optode beads were manually dispersed on a glass slide (22 x 22 x 1.5

mm). A second glass slide was pressed against this to allow for further dispersion beads

by static charges. The slides were separated and two pieces of aluminum foil (7 um

thick) was placed on two opposite sides of a slide. An identical, clean slide was then

	
  
17	
  
clamped along the pieces of aluminum. The clamped structure was placed in 1 wt% agar

and by capillary action a gel layer is wicked through between the slides with pH sensitive

optodes beads dispersed throughout.


Preparation of agar membrane with dispersed optode beads for spherical diffusion


         pH-sensitive optode beads were dispersed as described above. A 1 mm thick

spacer (Fisher Scientific microscope slide) was placed on opposite sides between two

glass slides, one with dispersed beads and one blank and clamped together. Agar

membrane was formed as described above.


Preparation of agar plug for diffusion in acidic direction.


         A polystyrene weigh dish (Fisher Scientific, 41 x 32 x 8 mm) was used as the

substrate. A source was formed by placing a hot soldering iron near a needle tip (Talon

American) and placing the hot needle tip on the substrate. A small source was formed,

~150 um diameter. 1 wt% agar was made in 0.0001M PBS, adjusted to pH 3. The pH of

the solution was adjusted with HCl. The substrate was placed upside down on ~10 ml

slightly boiling DI water. A drop was placed on the substrate above the source and was

allowed to sit above the boiling DI for ten minutes.


Preparation of agar plug for diffusion in basic direction.


         Identical steps were taken to produce the 150 um source in the substrate. 1 wt%

agar was made in 0.0001M PBS and adjusted to pH 8. The pH of the solution was

adjusted with KOH. The plug was formed using the same procedure described above


Optode bead calibration.

	
  
18	
  
         A 14 um membrane with optode beads dispersed throughout was soaked in PBS

for thirty minutes. The beads were then exposed to pH solutions between 5.0 to 8.0 in

0.5 pH increments. Solutions were made by adjusting 0.1M PBS with HCl and KOH

stock solutions. A pH electrode was used to ensure accurate pH values (Fisher

scientific). The calibration was done first going in the basic direction of the pH scale

from PBS (pH ~7.36) to 7.5 and 8.0, and then in the acidic direction in 0.5 pH units to pH

5.0. The calibration was then repeated from the acidic side back to the basic end of the

scale. The membrane was exposed to solution for 30 minutes. Images were taken

approximately every 5 minutes with a CCD camera (Kodak Easyshare DX4900).


Experimental setup for diffusion in acidic direction.


A 0.0001M PBS solution was adjusted to pH 3.5 for lower reservoir and pH 7.5 for the

upper reservoir. The optode bead membrane was placed flush with the surface of the

upper reservoir covering the source. A glass slide was used above the membrane during

cylindrical diffusion experiments. In both cylindrical and spherical diffusion experiments

a 1 g weight was placed on top of the membranes to hold them in place.


Experimental setup for diffusion in basic direction.


A 0.0001M PBS solution was adjusted to pH 8.5 for lower reservoir and pH 4.5 for the

upper reservoir. The membranes were treated as described in acidic direction

experiments.


Image Acquisition.




	
  
19	
  
A scion microscope camera was used to take pictures of the dispersed beads. ImageJ

software was used for image acquisition and storage. Images were taken every 30

seconds throughout the length of each experiment.


Image Analysis for circular symmetry concentration maps.


Matlab was used for image analysis. A linear interpolation was performed on each image

to obtain data for 10 second intervals. A mask was created for the images which included

beads (data) and excluded the substrate (include figure of mask). A Delaunay

triangulation was then performed on each image using the same mask with the function

GRIDDATA. A function was written for averaging overlapping circular rings (5 pixels in

diameter) of the GRIDDATA images to obtain a concentration map shown in figures

(11). Rings were made from the center and created outward in 5 pixel increments. Each

ring included data from the two adjacent rings, one outer and one inner ring to produce a

5 pixel ring with 15 pixel information.


Scanning electrochemical microscopy simulation (1), hit or miss.


A simulation was done to compare SECM measurements with measurements obtained

with optode beads. Typical parameters were assumed based on source size. An electrode

tip size of 80um was chosen for the source size ~150um. A step size of 80um was used

with an acquisition time of ~4s and scan rate of 40 um/s. (Citation)


Scanning electrochemical microscopy simulation (2), tiling.


A simulation was done to further compare SECM measurements with color map obtained

with optode beads. The same parameters were used as simulation (1), although the


	
  
20	
  
electrode tip was assumed to be square. The average color at each measurement point

was plotted continuously over time using the color maps obtained at each time point.


2.2 Results


2.2.1 Optode bead calibration


         ImageJ was used to obtain RGB values at 10 random points with bead

information in each image. To adjust for ambient light variations, each RGB value was

normalized using Pythagoras normalization shown below, where R is red value, G is

green value, and B is blue value, nX represents the normalized value. This calibration

also provides with a general color indication to relate local measurements actual

concentrations.




                                                     	
  (3)	
  



         Images for pH units 5.5 to 8.0 are shown in the panels below:

	
  
	
  




	
  
21	
  
                        a                           b                                   c

                                                          Figure 3. Raw images of optode-bead
                                                          pH calibration at t=30 minutes. (a) pH
                                                          5.5 (b) pH 6.0 (c) pH 6.5 (d) pH 7.0 (e)




                        d
                        d                            e



The calibration plots of Red, Green, and Blue intensities from a single experiment are

shown below.




                                          a                                                   b




                                                         Figure 4. normalized
                                                         calibrations for pH-
                                                         sensitive optode beads.
                                                         (a) red intensity, (b) green
                                                         intensity, (c) blue
                                                         intensity
                                          c




2.2.2 Raw Data:


	
  
22	
  
2.2.2a Lateral diffusion experiments
Acidic direction:


         Images were taken every 30 seconds for the acidic direction experiments. Raw

data at multiple time points is shown in the figure below. Images are shown in four

minute increments starting with time=2min. The images show the spread of protons from

the source across the surface in the x- and y- directions. Diffusion is limited from z-

direction by a glass side placed directly over the optode-bead membrane.




                           a                               b                                 c




                            d                              e                                     f

                                                                  Figure 5. each panel
                                                                  contains an image of
                                                                  dispersed beads in response
                                                                  to lateral H+ diffusion. (a)
                                                                  t=2 min (b) 6 min (c) 10
                                                                  min (d) 14 min (e) 18 min
                                                                  (f) 22 min (g) 26 min (h)
                                                                  30 min
                           g                               h

Basic direction


	
  
23	
  
         Images were taken every 30 seconds for the acidic direction experiments. Raw

data at multiple time points is shown in the figure below. Images are shown in four

minute increments starting with time=2min. The images show the spread of hydroxyl

ions from the source across the surface in the x- and y- directions. Diffusion is limited

from z-direction by a glass side placed directly over the optode-bead membrane.




                              a                              b                                 c




                              d                              e                                  f

                                                                       Figure 6. each panel
                                                                       contains an image of
                                                                       dispersed beads in response
                                                                       to lateral OH- diffusion.
                                                                       (a) t=2 min (b) 6 min (c) 10
                                                                       min (d) 14 min (e) 18 min
                                                                       (f) 22 min (g) 26 min (h)
                                                                       30 min
                              g                              h


2.2.2b Spherical diffusion
Acidic and basic directions
         For spherical diffusion, beads were dispersed on the surface of a 100 um thick

agar membrane. To ensure that the thickness of the agar membrane would not limit

	
  
24	
  
diffusion in the z-direction, an estimate was made using the diffusion coefficient of

protons in agar [19]. The equation below was used to estimate diffusion time, where

D=2.226 *10-9 m2s-1 , z is thickness (100 um), and t is time. Raw data images of the

diffusion of protons and hydroxyl ions across the surface can be found in Appendix 1.


                                                =         (4)

                                               = 2        (5)
2.2.2c Spread distance analysis
Lateral diffusion
         The plots below show the distance of diffusion of either protons or hydroxyl ions

at various time points for all experiments done. The diffusion distance was defined as the

transition color found at the boundary of diffusion. Distance was measured in pixels,

where 1 pixel ~ 6 um.




Figure 7. The left panel shows spread distance of protons in lateral diffusion experiments (n=6).
The right panel shows spread distance of hydroxyl ions in lateral diffusion experiments (n=4)

Spherical diffusion
The plots below show the distance of diffusion of either protons or hydroxyl ions at

various time points for experiments. The diffusion distance was defined as the transition




	
  
25	
  
color found at the boundary of diffusion. Distance was measured in pixels, where 1 pixel

~ 6 um.




Figure 8. The left panel shows spread distance of protons in spherical diffusion experiments
(n=4). The right panel shows spread distance of hydroxyl ions in spherical diffusion experiments
(n=2)
2.2.3 Linear interpolation
         Images were taken every 30 seconds during experiments. In order to retrieve a

more precise map at multiple time points and compare with SECM experiments, a linear

interpolation was performed on each image to obtain experiment images every ten

seconds. The equation used for the interpolation is below. Matlab was used to perform

the interpolation and the code can be found in Appendix 2.


                             1=( 2− 1)( 3− 1)( 3− 1)+ 1                   (6)
2.3 Generation of color map
         For each image taken, the end goal is to generate a concentration map using the

information of the color of the beads, RGB values. This requires a multi-step process

which includes extraction of data, a Delaunay triangulation, color averaging using

circular symmetry, and pixel normalization. The following are described and results are

shown in the following section.


	
  
26	
  
2.3.1 Thresholding
         In order to extract data from each image a thresholding scheme was developed to

obtain a mask of the image. For each experiment, an image was normalized using the

Pythagorus normalization (eqn 3). 10 random pixels where beads were dispersed were

chosen and the red and blue values were obtained. From these values a threshold range of

red intensity and blue intensity was used and applied over the entire image. The result

was a mask of the image where pixels which represented beads were displayed as white

(R, G, & B = 1) and background/substrate was displayed as black (R, G, & B=0). This

map was used for Delaunay triangulation for each image in the same experiment. Below

two sets of dispersed beads with corresponding thresholding masks are shown.	
  



                                                               Figure 10. the figure
                                                               shows (a, c) sample raw
                                                               data images used to
                                                               generate (b, d)
                                                               thresholded masks used
                                                               for Delaunay triangulation


                        a                              b




                            c                              d
2.3.2 Delaunay triangulation

         Using the mask generated by thresholding, each image was processed in matlab

using the GridData function. This function allowed a Delaunay triangulation to be done

	
  
27	
  
on the entire image using only bead data in corresponding to the mask. Each image was

first decomposed to three stacks, a Red, Green and Blue stack, respectively. The

triangulation was calculated for each of these stacks. The stacks were then compiled once

again to produce a color image. The code for this function can be found in Appendix 3 .

For analysis, each image took approximately nine-seconds. Below an example of the

Delaunay triangulation and raw data are shown in the figure. If there was not enough

surrounding information provided from the mask to generate a color using the

triangulation, a value of 0 was given for that pixel, which explains the black edges.




   Figure	
  11.	
  (a)	
  	
  Delaunay	
  triangulation	
  and	
  	
  	
  	
  	
  	
  (b)corresponding	
  raw	
  data	
  image.	
  	
  	
  



2.3.3 Color map generation with circular symmetry
The Delaunay triangulation provides sufficient information at most pixels based on the

corresponding masks. However, the result is also biased at the vertices of the triangles

made by the function, as described above. In order to smooth the image, circular

symmetry was assumed. Color-averaged rings were then calculated growing outward

from the source. Each ring included an average color of its own diameter (5 pixels) and

also the outer 5-pixel ring and inner 5-pixel ring, thus creating a 5 pixel ring with 15-

pixel averaged information. These rings were grown from the source across the entire


	
  
28	
  
image. The function can is attached in Appendix 4 . For each image, the function takes

approximately 25 seconds to generate a color map. The ring radius or distance from the

center was created using the equation for a circle, shown below, and then applied to the

adjacent rings. In the equation below the cX and cY represent center x and y position of

source. Various parameters can be set to adjust size of ring.


                                 =( −     )2+ −     2    (6)
Example images of color maps and corresponding raw images are shown in the figure

below. We assume circular symmetry to be valid because of the source, a microscopic

cylinder, which would provide diffusion in all directions in the 2D field to be described

as a function of distance form only a central point. Thus we are assuming that diffusion

all points at a specific distance from the source take the same concentration value.


                                                                Figure 12. (a) raw bead data of diffusion
                                                                of protons (b) circular symmetry color
                                                                map generated from raw data (c) raw bead
                                                                data for diffusion of OH- ions (d) circular
                                                                symmetry color map generated from raw
                                                                data



                     a                             b




                                                        2.3.4 Normalized color map
                                                                 In order extend the range

                     c                             d    of colors represented in the color

map between 0 – 255, and provide more uniform ring colors with growth from center, a

Pythagorean normalization was done on each pixel as explained in Eq 3 .


	
  
29	
  
         Below is a figure showing a comparison of a color map that is normalized

compared to a non-normalized image. This is followed by color maps at various time

points for lateral acidic and basic direction experiments. More color maps can be seen in

Appendix 5 .


                                                   Figure 13. (a) shows a color map
                                                   without normalization (b) shows the
                                                   same color map after normalization (c-
                                                   f) show the normalized color map at 4
                                                   minute intervals for diffusion of
                                                   protons (g-j) show the normalized color
                                                   map at 4 minute intervals for diffusion
                    a                       b      of hydroxyl ions.




                    c                       d                      e                     f




                    g                       h                          i                     j




2.3.5 Effect of normalization of color map



	
  
30	
  
             The plots below show the red color intensity of each 5 pixel ring at various time

points. There are plots for both pre-normalized color maps and normalized color maps to

show the effect that normalization has on the images. The red green and blue color

intensity plots can be found in Appendix 6 in full size with time labels for each plot.




                                                         a                                                   b




                                                  c
                                                           d
         Figure 14 : (a) pre-normalized red color intensity for the diffusion of protons at various time
         points (b) normalized red color intensity for diffusion of protons at various time points (time
         increases in downward direction) (c) pre-normalized red color intensity for the diffusion of
         hydroxyl ions at various time points (d) normalized red color intensity for diffusion of hydroxyl
         at various time points (time increases in downward direction)



2.4 SECM simulations
2.4.1 SECM simulation (1) Hit or miss:
             Final images of the hit or miss simulation for lateral and spherical diffusion in the

basic direction are shown below. Movies of the experiments have been made with

	
  
31	
  
growing black circular rings with diameters defined by transition color are available. The

final image shown includes the black ring for that time point. The simulation shows the

raw data obtained by the optode-bead imaging at a single time point. It also includes all

SECM measurements which would have been made over the course of the entire

experiment. Simulations are shown with the starting point as a uniform grid around the

source, and also with the starting point at the source directly.




            Figure 15: the left panel shows the hit or miss SECM simulation for lateral diffusion
            with the source centered around the simulation region; the right panel shows the hit or
            miss SECM simulation with the simulation starting at the source




          Figure 16. The panel above shows the hit or miss SECM simulation for spherical
          diffusion with the source centered around the simulation region




2.4.2 SECM simulation (2) tiling:

	
  
32	
  
         Final images of the tiling SECM simulation are shown in the figures below. They

were obtained using the normalized color maps for each experiment. Simulations for

acidic and basic lateral diffusion experiments are shown. Movies of compiled images are

available. Further, images of the simulation with the source as the starting point of the

simulation are also shown for the diffusion of hydroxyl ions.



                                                                                Figure 17. The left panel
                                                                                shows the final tiling image
                                                                                for lateral diffusion in acidic
                                                                                direction with a grid
                                                                                surrounding the source. The
                                                                                right panel shows the final
                                                                                tiling image for lateral
                                                                                diffusion in basic direction
                                                                                with a grid surrounding the
                                                                                source. The bottom panel
                                                                                shows the tiling for diffusion
                                                                                in basic direction with starting
                                                                                point at source




2.5 Optode-membrane based functional imaging


         In addition to using optode-beads, we have looked into the use of optode

membranes for functional imaging of 2D surfaces. The experiments done were identical

to lateral diffusion experiments described above, however rather than an optode-

	
  
33	
  
bead/agar membrane, a thin, ~4 um thick optode membrane (PVC/DOS) was placed 10

um above two diffusion sources of varying sizes. This technology allows continuous and

simultaneous imaging at each pixel with a CCD camera, and without the need of

interpolation and development of concentration maps post processing. Optode-

membrane based imaging has the potential to be implemented in studies where ionic

distributions will not be limited by diffusion through the membrane, maintaining a

variety of applications suitable only for optode-bead based functional imaging. Appendix

7 shows images of lateral diffusion experiments done with optode-membranes.


2.6 Discussion


2.6.1 Calibration:


         The calibrations for the optode beads provide information to correspond color

response with specific pH values. It shows that the range of the red color intensity to

change the most with a change in the pH. Blue intensity also has a significant dynamic

range to be used to indicate color change and correspond with pH values. The green color

intensity change with pH is limited indicating that it could be used as an internal

reference color. The experiments also showed that the beads when dispersed and clumps

are avoided; response time to changes in pH


can be in the order of seconds. This is necessary when trying to incorporate the beads to

rapid changes including diffusion.


2.6.2 Raw data and spread distance analysis




	
  
34	
  
         The raw data images provide data in pixels containing beads which can then be

extracted by thresholding. Once thresholded, these values can be used to generate color

values at each pixel corresponding to concentrations. The raw data can also be used to

determine the approximate spread distance of either protons or hydroxyl ions from the

source at any given time point. This transition region is defined as a color which

represents an intermediate color between the orange and blue extreme levels. The spread

distance and speed of the diffusion can be influenced by a wide variety of parameters.

Comparing lateral diffusion of protons and hydroxyl (ions, we see that spread distance

increases faster and that overall spread is faster. We believe that this is a result of a

variety of factors. Firstly, protons are significantly smaller and lighter than hydroxyl ions,

providing them with larger diffusion coefficient in the agar membrane. Further, the pH

level of the acidic solution is a very high level, pH 3.5 indicating a very large number of

free protons for diffusion. In comparison, the pH of the basic solution that is diffusion,

pH 8.5 is only slightly basic. This limits the speed and total distance of spread from the

source for the basic direction.


         The data also shows that for spherical diffusion the diffusion of protons is initially

faster than that of hydroxyl ions which can again be due to the reasons mentioned above.

The spherical diffusion experiments when compared with lateral diffusion show that

there is a significantly slower diffusion or spread in the x- and y- direction for spherical

diffusion. This is due to the fact that we have no longer limited diffusion in the z-

direction by removal of the glass slide. With diffusion allowed in x- y- and z- direction

for the experiments, we would expect slower spread in the x- and y- direction as

observed.


	
  
35	
  
2.6.3 Color maps


         The color maps provide color intensities at each pixel for Red-Green-Blue

variables. These three values can be used to correspond with calibrations to obtain

concentrations at each pixel. The plots shown in figure 12, 13 comparing the pre-

normalized and normalized images show that the normalization eliminates many

variations throughout the image due to ambient light variation, thresholding issues and

other factors that impact the acquisition of the images. This can particularly be seen in the

red color region for the acidic direction experiments and in the blue color region for the

basic direction experiments. This is due to the fact that the original color, or main color

shown in corresponding images is initially red or blue, respectively.


2.6.4 SECM simulation (1) hit or miss


         The hit or miss simulation shows a comparison of the potential of optode bead

imaging compared to the current technology of SECM. The hit or miss simulation

indicates how many measurements would be made in the transition region of the

diffusion profile, oftentimes the region of interest. This is important because SECM

assumes that it obtains enough information from scanning the surface to reconstruct a

concentration profile. Further an SECM scan of the region, 1.77 x 1.77 mm would take

approximately one hour, during the course of this experiment it is clear that there is a lot

which could change in the regions of scan prior or after scanning. This can lead to

incorrect or inconsistent data. We show through the simulations that with the diffusion of

protons and hydroxyl ions from a single source, SECM would often miss many transition

points and provide an inaccurate concentration profile for both lateral and spherical


	
  
36	
  
diffusion. In comparing the two, it is clear that the spherical diffusion experiment would

provide better data than lateral diffusion, with more ‘hits’, represented by colored in

black circles. The number of hits for spherical diffusion was twelve compared to six for

lateral diffusion. When the source was used as the starting point for the experiment, there

were significantly more ‘hits’ for lateral diffusion, however reconstruction from SECM

would still result in inaccurate results.


2.6.5 SECM simulation (2) tiling


         The tiling simulation shows the actual data that would be collected by the SECM

if, instead of collecting current, it would collect colors represented by the pH-sensitive

optode beads. The simulations show that there is a significant region which SECM

would generate a map which does not correctly show the diffusion profile of protons or

hydroxyl ions when the source is centered in the area of scanning. Further, if there were

any changes in the region which was already scanned there will be even more inaccurate

measurements recorded. With the scan beginning at the source, the SECM still provides

inaccurate representation of the diffusion profiles.


2.6.6 Bead Clumping/dispersion


         The size of individual optode beads, ~2 um is advantageous to the beads in that it

provides a very small electrode which can be placed practically anywhere to make

measurements. However in order to obtain an even dispersion of individual optode beads

is very difficult due to various interactions between individual the PVC/DOS beads

which can lead to clumping. In order to reduce clumping in the beads a variety of

solvents, including ethanol and ether were tried, and also sonication in DI water and the

	
  
37	
  
solvents. Sonication showed some improved dispersion but it was not a dependable

method due to inconsistency in dispersion. In order to reduce clumping in the next

generation of beads, the inclusion of a nonionic solvent could be added during particle

formation, for example Tween 20. This has been shown to reduce clumping in

PVC/DOS microbeads [22, 24]. It is important to prevent clumps to ensure fast response

time.


2.6.7 Overall comparison of optode-bead imaging and SECM


         Optode bead imaging allows several advantages to the SECM measurements. The

dispersion of beads allows for an ‘electrode’ to be placed at any desired region for

measurements to be made by the user. With interpolation of data a profile of the entire

region can be obtained at each time point. SECM allows a measurement of only a single

point at each time point. Further, without any scanning of the region, we eliminate any

potential convection effects which could interfere with concentration profiles. The

optode beads also provide continuous measurements, responding to changes in

concentration in order of seconds. This provides continuous results at each point

throughout the entire experiment, where again SECM provides only a single

measurement at each point at a specific time. The optode bead measurements do not

have time limitations since, measurements are made continuously. SECM experiments

can take multiple hours which can result in a variety of discrepancies in data which often

go overlooked. The setup for optode beads is also less invasive than an SECM

experiment which requires a very large setup as described earlier. Finally, functional

optode imaging has the potential to make measurements where the SECM ME would be

unable to.

	
  
38	
  
For example, in cancer drug delivery studies it is assumed that a multicellular tumor

spheroids (MCTS) accurately depicts cell behavior in cancerous tissue. This however has

not been functionally tested. Using optode beads dispersed at the bottom of the half

spheroid imaging of concentration profiles for a variety of analytes can be measured. This

would provide significant insight into a region that cannot be done with current

technologies.


2.7 Conclusions and future work


         The work discussed introduces the potential of optode-bead based technologies

potential applications and advantages to current imaging techniques. Compared to

SECM, optode-bead or optode-membrane functional imaging can provide a continuous

and simultaneous 2D ionic distribution without the need of a physical scanning electrode.

Using interpolation, an entire concentration map can be obtained without disturbing a

biological system. Further, if used in simultaneously with other technologies, including

SECM, optode-based imaging can provide insight into ionic distributions which, as of

yet, have not been studied due to limitations in technology. We plan to use both optode-

beads and optode membranes to study a variety of biological systems, including single-

cell epithelial layers, specifically ion channel diffusion mechanisms, and MCTS models

to study mass transport, drug uptake and their interactions with hypoxia and acidity

simultaneously while keeping the spheroid model intact.


Acknowledgments




	
  
39	
  
The author would like to acknowledge the committee members Dr. Miklos Gratzl, Dr.

Shawn McCandless, Dr. Mihailo Rebec, and Dr. Andrew Rollins, the assistance of Dr.

Sumitha Nair, Sreenath Narayan and Dr. Maria Peshkova.




	
  
40	
  
Appendix 1: raw data




Left: acidic
cylindrical raw data,
with mask;
Right: basic
cylindrical raw data
with mask




	
  
41	
  
Appendix 2
%this function is used to generate interpolated images in and provide
output images at specified times, based on two input images at
specified times.

function [interp_image] = inter(input_image1, input_image2, t1, t2,
interp_t)

  close all

input_image1=double(input_image1);
input_image2=double(input_image2);

input_image1r=input_image1(:,:,1);
input_image1g=input_image1(:,:,2);
input_image1b=input_image1(:,:,3);

input_image2r=input_image2(:,:,1);
input_image2g=input_image2(:,:,2);
input_image2b=input_image2(:,:,3);


interp_imager=input_image1r+((input_image2r-input_image1r)/(t2-
t1)).*(interp_t-t1);

interp_imageg=input_image1g+((input_image2g-input_image1g)/(t2-
t1)).*(interp_t-t1);

interp_imageb=input_image1b+((input_image2b-input_image1b)/(t2-
t1)).*(interp_t-t1);


interp_image=[interp_imager, interp_imageg, interp_imageb];


interp_image=reshape(interp_image, size(input_image1));
interp_image=uint8(interp_image);




	
  
42	
  
Appendix 3
function [output_image] = ColorInGridfinal(input_image, content_mask)
% This function takes a color diffusion image, input_image, and finds a
% guess for the data at every location where the beads do not exist.
The
% meat of the matter uses Delauney triangulation, as implemented by the
% MATLAB function 'griddata'
% Written by: Sreenath, spn5@case.edu
% Input Variables:
    % input_image = color image with a single diffusion source
% Output Variables:
    % output_image = color image with the colors "interpolated" into
the
        % blank space in input_image - "interpolated" is in quotes
because
        % this is not an interpolation; the algorithm fills in the bead
        % data using Delauney triangulation

%% Set up Defaults
input_image = double(input_image);

%% Decompose colors of input_image into separate variables
% this is done, though it is bulky, so that logical indexing can be
used
input_image1 = input_image(:,:,1);
input_image2 = input_image(:,:,2);
input_image3 = input_image(:,:,3);

% normalize
normfactor = sqrt(double(input_image1.^2 + input_image2.^2 +
input_image3.^2));
input_image1 = input_image1 ./ normfactor;
input_image2 = input_image2 ./ normfactor;
input_image3 = input_image3 ./ normfactor;

%% Generate content mask, so that we know where the beads are
% if needed, this thresholding can be improved to get rid of the junk
% around the edges - it would nice to have graphical user interface to
make
% this step better
%content_mask = input_image3<.44 | input_image1<.47; %acidic direction
oct14

content_mask = input_image3<.51 | input_image1<.56; %basic direction
oct 14

content_loc = find(content_mask);
[content_locY, content_locX] = ind2sub(size(input_image), content_loc);

%% Interpolation Using griddata
% rescale original input images (before normalization)
input_image1 = input_image(:,:,1)./256;

	
  
43	
  
input_image2 = input_image(:,:,2)./256;
input_image3 = input_image(:,:,3)./256;

% generate a grid of every x and y point:
[X,Y] = meshgrid(1:size(input_image, 2), 1:size(input_image, 1));

% grid the data onto each of those x and y points, using the previously
% computed locations of data, for each color separately:
output_image1 = griddata(content_locX, content_locY,
input_image1(content_loc), X, Y, 'cubic'); %#ok<*FPARK>
output_image2 = griddata(content_locX, content_locY,
input_image2(content_loc), X, Y, 'cubic');
output_image3 = griddata(content_locX, content_locY,
input_image3(content_loc), X, Y, 'cubic');

%% Remake a color image for output
output_image = [output_image1, output_image2, output_image3];
output_image = reshape(output_image, size(input_image));

% save test_run.mat




	
  
44	
  
Appendix 4
function [out] = ringstest(im)
im = im2double(im);
imR = im(:,:,1);
imG = im(:,:,2);
imB = im(:,:,3);
outR = ringsGray(imR);
outG = ringsGray(imG);
outB = ringsGray(imB);

out = zeros(size(im));
out(:,:,1)=outR;
out(:,:,2)=outG;
out(:,:,3)=outB;

function [out] = ringsGray(im)
startRad = 0;
ringRad = 5;
ringDiff = 5;
ringNum = 400;
% im is a grayscale image
[sx sy] = size(im);
%cx = ceil(sx/2);
%cy = ceil(sy/2);
cx=275;
cy=500;

[Y,X] = meshgrid(1:sy,1:sx);
rad = ((X-cx).^2+(Y-cy).^2).^0.5;
out = zeros(sx,sy);
m = out;
for i=0:(ringNum-1)
    lowRad = startRad+i*ringDiff;
    highRad = startRad+i*ringDiff+ringRad;
    outRad = startRad+i*ringDiff+2*ringRad;
    %ADD high rad2....and include in mask
    mask = (rad>=lowRad) & (rad<highRad) & (rad<=outRad);
    if sum(mask(:))==0
         break;
    end
    ringCol = sum(sum(mask.*im))./sum(mask(:));
    out = out + mask.*ringCol;
end




	
  
45	
  
Appendix 5 Color maps
                         t=0 min




                          t=2




                        t=6




	
  
46	
  
         t=12




         t=18




           t=36 min




	
  
47	
  
Diffusion of hydroxyl ions
                             t=1 min




                             t=2 min




                             t=6 min




t=12 min




	
  
48	
  
         t=18 min




              t=36 min




	
  
49	
  
Appendix 6
Diffusion of hydroxide ions:

    154	
  
    152	
  
    150	
  
    148	
  
    146	
  
    144	
  
    142	
  
    140	
  
    138	
  
    136	
  
              0	
       50	
     100	
     150	
     200	
  


Green normalized intensity

    160	
  

    150	
  

    140	
  

    130	
  

    120	
  

    110	
  

    100	
  
              0	
       50	
     100	
     150	
     200	
  


Blue normalized intensity
Diffusion of protons:




	
  
50	
  
    156	
  
    154	
  
    152	
  
    150	
  
    148	
  
    146	
  
    144	
  
    142	
  
    140	
  
    138	
  
    136	
  
              0	
       50	
     100	
     150	
      200	
  


Green normalized intensity



    180	
  

    170	
  

    160	
  

    150	
  

    140	
  

    130	
  

    120	
  

    110	
  

    100	
  
              0	
       50	
     100	
     150	
     200	
  


Blue normalized intensity
Diffusion of protons:




	
  
51	
  
    152	
  

    150	
  

    148	
  

    146	
  

    144	
  

    142	
  

    140	
  

    138	
  

    136	
  

    134	
  
              0	
                50	
                100	
                   150	
                     200	
                     250	
  


Pre-normalized green intensity

    180	
  

    170	
  

    160	
  

    150	
  

    140	
  

    130	
  

    120	
  

    110	
  

    100	
  
              0	
     20	
     40	
       60	
     80	
        100	
     120	
         140	
     160	
       180	
     200	
  


Pre-normalized blue intensity
Diffusion of hydroxide ions:




	
  
52	
  
    170	
  

    165	
  

    160	
  

    155	
  

    150	
  

    145	
  

    140	
  

    135	
  

    130	
  

    125	
  

    120	
  
              0	
     20	
      40	
       60	
        80	
        100	
         120	
         140	
          160	
       180	
          200	
  


Pre-normalized green intensity

    155	
  

    150	
  

    145	
  

    140	
  

    135	
  

    130	
  

    125	
  

    120	
  

    115	
  

    110	
  
              0	
     20	
     40	
      60	
       80	
        100	
        120	
         140	
         160	
      180	
      200	
  


Pre-normalized blue intensity




	
  
53	
  
Appendix 7: Membrane experiments


                                       t=1 min




                                   t=6 min




	
  
54	
  
References
[1] Katsuya Yamada, et al. Protective Role of ATP-Sensitive Potassium Channels in
Hypoxia-Induced Generalized Seizure. Science, 2001, 1543.
[2] Howell et al. Multicellular spheroids: a new model target for in vitro studies of
immunity to solid tumor allografts. J Natl Cancer Inst 1977, 1849-53
[3] Howell et al. The multicellular spheroid as a model tumor allograft. I. Quantitative
assessment of spheroid destruction in alloimmune mice. Transplantation, 1978, 3, 136-
40.
[4] Bard, A.J. et al. Scanning electrochemical microscopy. Theory of the feedback mode.
1989, 61, 1221-27
[5] Pahler M, Santan JJ, Schuhmann W, Souto R. Application of AC-SECM in corrosion
science: local visualization of inhibitor films on active metals for corrosion protection.
Chemistry, 2010, 905-11.
[6] Zhang, M. et al. SECM for imaging and detection of latent fingerprints. Analyst.
2009, 134, 25-30.
[7] Sun P, Carpino J, Mirkin, Nanoelectrochemistry of mammalian cells. PNAS, 2008,
105, 443-48.
[8] Zoski, C. G., Handbook of Electrochemistry. Elsevier Science, Amsterdam, 2007.
[9] Stulik et. al. Microelectrodes. Definitions characterization and applications. Pure
Appl. Chem. 2000, 72, 1483-1492.
[10] Mauzeroll, J, Bard, A.J., Monks, T. Menadione metabolism to thiodione in
hepatoblastoma by scanning electrochemical microscopy. PNAS, 2004, 101, 175282-
587.
[11] Beaulieu I, Kuss S, Mauzeroll J. Biological Scanning Electrochemical Microscopy
and Its Application to Live Cell Studies. Anal chem. 2011, 83, 1485-1492.
[12] Bakker E, Lerchi M, Simon, W. Synthesis and characterization of neutral hydrogen
ion-selective chromoionophores for use in bulk optodes. Analytica Chimica Acta, 1993,
278, 211-25.
[13] Li Z, Li X, Harrison J. Dual-sorption model of water uptake in Poly(vinylchloride)-
based ion-selective membranes: experimental water concentration and transport
parameters. Anal Chem, 1996, 68, 1717-1725.
[14] Bakker E, Buhlmann P, Pretsch E. Carrier-based ion-selective electrodes and bulk
optodes. 1. General characteristics. 1997, ACS, 97, 3083-3132.
[15] Bakker E, Buhlmann P, Pretsch E. Carrier-based ion-selective electrodes and bulk
optodes. 2. Ionophores for potentiometric and optical sensors. 1997, ACS, 97, 3083-3132.


	
  
55	
  
[16] Thomas Rosatzin, Petr Holy, Kurt Seiler, Bruno Rusterholz, Wilhelm Simon,
Immobilization of components in polymer membrane-based calcium-selective bulk
optodes. Anal Chem 2002, 64, 2029-2035.
[17] Bakker, E.; Lerchi, M.; Rosatzin, T.; Rusterholz, B.; Simon, W. Synthesis and
characterization of neutral hydrogen ion-selective chromoionophores for use in bulk
optodes. Anal. Chim. Acta 1993, 278, 211.
[18] Tohda K, Gratzl M. Micro-miniature autonomous opoticla sensor array for
monitoring ions and metabolites 1: design, fabrication and data analysis.
[19] Davies E, Huang Y, Hook, J, Lillford P. Dynamics of water in agar gels studied
using low and high resolution H NMR spectroscopy.
[20] Schulte A, Schuhmann W. Scanning Electrochemical Microscopy as a Tool in
Neuroscience.
[21] Yasukawa et. al. Characterization and imaging of single cells with Scanning
electrochemical microscopy. Electroanalysis, 2000, 12, 653-59.
[22] Retter R, Peper S, Bell M, Bakker E. Flow cytometric ion detection with plasticized
Poly(vinyl Chloride) microspheres containing selective ionophores.
[23] Tan S, Hauser P, Seiler K, Suter G, Simon W. Reversibble optical sensing
membrane for the determination of chloride in Serum. Analytica chimica acta, 1991, 35-
44.
[24] Peper S, Retter R, Bell M, Bakker E. Monodisperse Plasticized Poly(vinyl chloride)
fluorescent micrsospheres for selective ionophore-based sensing and extraction. Anal
Chem, 2001, 73, 6083-87
[25] Bychkova V, Shvarev A. Surface area effects on the response mechanism of ion
optodes: a preliminary study. Anal Chem, 2009, 7416-19.
[26] Sanders W, Vargas R , Mark R. Anderson. Characterization of Carboxylic Acid-
Terminated Self-Assembled Monolayers by Electrochemical Impedance Spectroscopy
and Scanning Electrochemical Microscopy. Langmuir 2008 24 (12), 6133-6139

[27] Scanning electrochemical microscopy coupled with intracellular standard addition
method for quantification of enzyme activity in single intact cells.Gao N, Jin W. Analyst,
2007., 1139-46
[28] Polymerized Nile Blue derivatives for plasticizer-free fluorescent ion optode
microsphere sensors. Ngeontae W, Chao X, Bakker E. Analytica Chic Acta. 2007, 124-
33.




	
  
56	
  

				
DOCUMENT INFO
Categories:
Tags: biology
Stats:
views:3
posted:11/19/2013
language:
pages:56