Synthesis of Metal-Coated Fluorescent Nanoparticles for Medical Applications
REU Student: Arly Parent
Graduate Mentor: Michael Klopfer
Faculty Mentor: Dr. Ravinder K. Jain
●Abstract
This paper is about our attempt to synthesize metal-coated nanoparticles with enhanced fluorescence. This enhancement can lead to better imaging capabilities in medical applications such as with deep-tissue imaging. Advancements in this area allow more and more medical options to be taken without the immediate use of surgery. More tests can be run “non-invasively,” diagnoses can be made more quickly, the length of time spent in medical facilities for tests and recovery can be lessened, and as a result, the cost of care can decrease. We use 2 different fluorophores as cores: Cadmium Selenide and Ruthenium.
●Purposes & Uses
We are creating these nanoparticles for use in medical scanners. They will serve to provide deeper scanning ability in a safe and controllable way. They will provide an increase in and allow us control of many factors—including penetration and signal filtering from the excitation source.
●Physical Phenomena and Background
In synthesizing our particle, we are taking advantage of a peculiar phenomenon. This is the phenomenon called local plasmonic resonance. Used in conjunction with the natural fluorescence of the fluorophores we are using, we endeavor to increase the fluorescence of the particle.
Core
The core of our nanoparticle is a fluorophore. In the case of the Cadmium Selenide quantum dot, we have a semiconductor core. When electrons in the valence band are given enough energy to jump to the conduction band, they leave behind an empty state that is termed a hole. This hole is the absence of an electron and quantum mechanically can be treated as a positive particle. When the electron and hole become bound, this state is called an exciton. The distance in this pairing is called the exciton Bohr radius and it is this that gives our quantum dot its properties. Quantum dots are semiconductors so small that they are confined quantum mechanically in all 3 dimensions—much like the archetypical example of the “particle-in-a-box.” The effects of this dominate when the size of the quantum dot is less than or equal to the exciton Bohr radius. The effect is that there is now exhibited the phenomenon of emission in energy levels again rather than in bands. Our nanoparticle uses this to give us tunable fluorescence.
Fig. 1 Exciton Bohr Radius-Band to Discrete Levels
Fluorescence occurs after an electron is excited and returns to the hole it left, falling to a ground state and thus emitting a photon. This emission is fluorescence. The allowed energy states are dependent on the size of the particle and are in this way evidenced in the wavelength of the fluorescence.
We take this idea a step further with two photon absorption fluorescence. Two proton absorption fluorescence (TPAF) is a nonlinear process in which two lower energy photons are absorbed nearly simultaneously and emit a single higher energy photon. TPAF intensity actually increases quadratically with excitation intensity[1].
Dielectric Spacer
The dielectric spacer is the buffer between the metal shell and the quantum dot. It is necessary because metals are known phosphorescence quenchers and it serves to extend the lifetime of and protect the core fluorophore from the shell. It is chosen carefully with respect to its bandgap so that it will not interfere with the absorption of the photons by the interior fluorophore. It is not desirable to have the spacer be absorbing the light meant for the quantum dot or dye. Furthermore, the spacer is not only a buffer. It facilitates the creation of the optimal dimensions for electric field enhancement with the shell[1]. Silicon dioxide, Zinc Sulfide, and Titanium Dioxide (Titania) are our current candidates for the spacer.
Shell
Certain metals such as gold, silver, platinum, and palladium have interesting optical properties. Upon irradiation of these metal surfaces with light of a resonant frequency, there is induced a coherent oscillation of the conduction electrons. This will maximize absorption and will enhance the electric field near the oscillation of free electron density, termed as the quasiparticle, the plasmon[Fig. 2].
Fig. 2 Local Surface Plasmonic Resonance
Project Implementation/Experiments
●Modeling
Synthesizing nanoparticles of this nature can be time-consuming and costly. There are too many variable parameters—spacer thickness, coating thickness, core/spacer/metal composition,etc. Checking each particular parameter is not viable. Thus, we must use computer modeling. My first task was modeling plasmonic quantum dots using JCMWave, a finite-element software suite with a multitude of capabilities in electromagnetic modeling. JCMSuite uses Finite-Element Method (FEM) simulation tools in order to solve Maxwell's equations for an almost arbitrary amount of shapes and for arbitrary refractive indices. In this case, I created a mesh to model a 2-dimensional cross-section of the particle I wished to model [Fig. 2]. If we take a 2D cross-section of a sphere (1/2 circle), we create a cylindrically symmetric layout that JCM can solve on a 2D computational domain. In essence, I modeled a tri-layered Gold-coated Cadmium Selenide plasmonic quantum dot in water being struck by an 800 nm plane wave.
Fig. 3 JCMWave Mesh of 3 Layer Spherical Nanoparticle
1Define the shape of the particle using the primitives supported by JCMWave. Currently, it supports rectangles, circles, and polygons of arbitrary shape.
2Define the “computational domain.” When we create a 3D solid from a 2D figure in JCM, we have to not only model the cross-section but a surrounding “computational” domain (Priority=-1) .
3Define the relative permittivity and permeability of each layer of our multi-shelled particle and the computational domain.
Material / Rel. Permittivity / Rel. Permeability / Radius from CenterCdSe / 9.128865610908527 / 1.0 / 3 nm
Spacer (SiO2 or ZnS) / 1.205404496424333 / 1.0 / 20 nm
Metal (Au) / -24.236565042428122+1.458651718229687i / 1.0 / 25 nm
Domain (H2O) / √ 1.396 / 1.0 / —
4We must then model the incident light source that is striking our particle.
●We can specify a time-harmonic plane wave
●
●Where:
●And:
● = vacuum wavelength (m)
●= refractive index
●= Amplitude (V/m)
●In this case:
/ 800 nm/ [0.0, 10964158.361028377, 0.0]
/ [0.0, 0.0, 10000]
After running all of the calculations, JCMWave outputs the electric field strength for our model in the form of vectors.
Fig 4: Electric Field Strength XZ Plane
If we consolidate the magnitudes of the rays in the XZ plane, we can see the electric field enhancement occur in more than one area around the circle [Fig. 5]. The optimum size of each layer was determined by my mentors, Michael Klopfer and Ravinder Jain in their previous paper [1].
Fig 6: XZ Plane Electric Field Enhancement
of Gold Plasmonic Quantum Dot
A comparable palladium quantum dot was modelled as well. It too showed enhancement effects, providing evidence that palladium is a viable alternative to gold as a coating. Because our immediate application for this project is in the field of medical imaging, we are exploring lower cost metals with plasmonic properties. Though gold is one of the best, some other metals can be used for lower costs.
Fig. 7 XZ Plane Electric Field Strength of Pd Plasmonic Quantum Dot
●Synthesis
Synthesizing these nanoparticles is a complex task. It presents many problems that must be addressed. One of these is quenching. Because we must not only controllably coat a quantum dot with silica but coat that silica with a metal. Throughout this multi-coating process, there are many chances for quenching. To test our methods, we decided to separate the project into two different aspects:coating the fluorophore with silica and quoting silica with metals. I was tasked with the silica coating aspect and the metal coating was assigned to the group headed by Abhaya Datye.
Over this Spring semester, my focus was on coating the less toxic Ruthenium fluorophores. In order to do this, I used the Stober process. This is a sol-gel process used in many applications to create spherically-shaped silica particles. A two-part reaction, the Stober process consists of a hydrolysis and a
condensation.[2] In our experiments, we use TEOS(Tetraethyl Orthoslicate), Ammonium Hydroxide, Water, Ethanol, and Ruthenium. Hydroxyl anions react with TEOS to start the hydrolysis reaction[2].
The condensation reaction occurs immediately after, forming Si-O-Si bridges [2].
This overall reaction is:
We get our silica spheres. During this normal Stober process, we introduce Ruthenium dye. The dye is used to dope the silica spheres that are forming. Depending on the size of the sphere formed, a large large dye number can be obtained per bead[3].
Fig. 8 Size Control and Dye Amount Table
Reaction 1
In our first attempt, we used the procedure from Zhang et al. (2006) as our template for synthesizing our doped nanoparticles.
Procedure
147.5 mL ethanol
22.5 mL of 2*10-5 M Ruthenium/Ethanol Solution
30.1563 mL TEOS
4Begin magnetic stirring
5Add 1 mL of ammomium hydroxide dropwise while stirring
6Stir overnight. The solution should become cloudy.
7Centrifuge
8Redisperse in ethanol
Observations
Our first reaction using the Zhang et al. (2006) procedure failed to produce any fluorescence or any 'cloudiness' in the solution. No centrifuging was done on this sample.
Reaction 2
Procedure
150 mL of 2*10-5 M Ruthenium/Ethanol Solution
22.937 mL of TEOS
3Begin magnetic stirring
4Add 1 mL of ammomium hydroxide dropwise while stirring
5Stir overnight. The solution should become cloudy.
6Centrifuge
Observations
This reaction showed slight 'cloudiness' in the UV and even less in the visual band. However, fluorescent particles were centrifuged out of the solution. These particles were not uniform or well-defined. Further, they were fused together. They cannot be called Stober spheres. Though this wasn't a perfect success, we begin to get close to the sizes we are striving to create as well as the morphology we need for coating.
Fig. 9 Scanning Electron Microscope Images of Reaction 2(sonicated)
Reaction 3
This batch was created using a slight modification of Kah et al.(2008)[4].
Procedure
125 mL of 2*10-5 M Ruthenium/Ethanol Solution
21.19569 mL of TEOS
3Begin magnetic stirring
4Add 0.43478 mL of 14.8 M Ammonium Hydroxide while stirring vigorously
51.9565 mL of water
6Stir vigorously overnight
7White turbid suspension should be seen
8Centrifuge at 10,000 rpm for 30 minutes
9Wash with ethanol and water
10Repeat >5 times
Observations
This reaction is our most promising. The morphology of the silica spheres is much closer to the ideal spherical shape we are endeavoring to create. Though we have more spherically-shaped particles, the problem of them being fused together still remains. Even after sonication, we see them still fused. This is a problem for our next step—coating. In fact, it makes coating inadvisable for our purposes. This is an issue that must be resolved. In all, however, this is a large step toward viable, uniform,spherically-shaped Ruthenium-doped silica spheres.
Fig. 10 Scanning Electron Microscope Image of Reaction 3(sonicated)
Fig. 11 Scanning Electron Microscope Image of Stober Spheres
Reaction 4
Procedure
120 mL of 2*10-5 M Ruthenium/Ethanol Solution
21.1748 mL of TEOS
3Begin magnetic stirring
4Add 0.4 mL of 14.8 M ammonium hydroxide dropwise while stirring
5Let stir overnight
Observations
Visual inspection of the sample showed promise. A cloudy, white,turgid solution was seen and centrifuging produced a relatively large quantity of solid. This was expected. However, this reaction failed to produce any silica spheres. Instead, SEM (Scanning Electron Microscopy) showed large agglomerates of SiO2 even after we sonicated the sample. This will clearly not be useful in our project. It is unclear exactly what caused this. However, it is possible the centrifuging played a role in creating this agglomerate.
Fig. 12 Scanning Electron Microscope Image of Reaction 4(sonicated)
Issues
From the results of these four reactions, a number of issues arise. Our first issue is the agglomeration and fusion of the silica spheres. Even after sonication, they remain fused. They cannot be termed truly spherical if this is the case and cannot be easily coated using the methods of the Datye group. Possible reasons for this fusion could be hydrolyzed TEOS, contaminated/dilute ammonium hydroxide, and/or the centrifuging process. The hydrolyzed TEOS and the centrifuging process are the two most likely problem sources. To test this, in the future, we will have to try using newly opened TEOS and other methods of removing the particles from solution such as filtering.
Another problem that must be addressed as a result of the SEM imaging is the issue of not having a Ruthenium signature in our EDS results. What we see across all of the samples is that we get evidence of the other expected elements but even with multiple EDS scans at different voltages, we do not get a response for Ruthenium.
Fig. 13 Reaction 3 EDS Results
Initially, this was a puzzling result because the particles fluoresce clearly and brightly in the UV. It was unknown what was causing this phenomenon. Possible causes were believed to be either EDS malfunctions or improper washing of the particles before Scanning Electron Microscopy began. It was possible that EDS was not reading the Ruthenium that was expected to be and visually seemed to be in the sample. It was also believed possible that the sample was not adequately washed in order to remove all excess solution and thus excess ruthenium from the exterior of the silica particles which would cause visible fluorescence but likely not at concentrations high enough to provide enough counts for the EDS to make its determination.
Another possible cause thought viable was that the total amount of ruthenium in the solution was too minute to be read. This hypothesis was confirmed by a calculation of the weight percentage of ruthenium in the sample being less than 1%, a figure well under EDS' threshold.
Future Plans
The future of this endeavor lies in the coating of these spheres. Once the doped Stober synthesis process is perfected, the coating is the next and final step in the synthesis of this three-layer particle. From then these particles must be subjected to rigorous testing in various applications. Optimal spatial configurations of a large number of these particles must be discovered, tunings must be decided upon, as well as the brightness, intensity, and duration of the source.
Furthermore, for all of these particles, with respect to the Cadmium Selenide quantum dot particularly, a cytotoxicity assay needs to be performed in order to determine whether the dielectric spacer and metal shell will have the expected protective effect toward the human body.
Figure References
[Fig. 1] Webb, Brittany. "Quantum dots." (2010).
[Fig. 2] Sudip Nath, Subhra Jana, Mukul Pradhan, Tarasankar Pal, Ligand-stabilized metal nanoparticles in organic solvent, Journal of Colloid and Interface Science, Volume 341, Issue 2, 15 January 2010, Pages 333-352, ISSN 0021-9797, 10.1016/j.jcis.2009.09.049.
(
Keywords: Nanoparticles; Organosol; Catalysis; SERS
[Fig. 7] Zhang, Jian, et al. "Dye-labeled silver nanoshell-bright particle."The Journal of Physical Chemistry B110.18 (2006): 8986-8991.
●References
[1] M. Klopfer and R. Jain, "Plasmonic quantum dots for nonlinear optical applications [Invited]," Opt. Mater. Express 1, 1353-1366 (2011).
[2] Ibrahim, Ismail AM, A. A. F. Zikry, and Mohamed A. Sharaf. "Preparation of spherical silica nanoparticles: Stober silica."J American Sci6 (2010): 985-989.
[3] Zhang, Jian, et al. "Dye-labeled silver nanoshell-bright particle."The Journal of Physical Chemistry B110.18 (2006): 8986-8991.
[4] Kah, James CY, et al. "Synthesis of gold nanoshells based on the depositionprecipitation process."Gold Bulletin41.1 (2008): 23-36.