THE CHARACTERIZATION OF XANTHENE DYES ON A GLASS SUBSTRATE FROM DIFFERENT COATING TECHNIQUES

Chapter 1 of this dissertation is an investigation of rhodamine 560 (Rh560), a cationic dye similar to the well-studied rhodamine 6G (Rh6G). The spectral properties of Rh560 and Rh6G have similar changes as a function of thickness. At low surface coverage the spectra indicate monomers, at 1-2 monolayers dimers dominate, and thicker films show larger aggregates. The difference between Rh6G and Rh560 is that the transition from monomer to dimer occurs at different thickness, ~1.2 nm for Rh6G and ~0.5 nm for Rh560. This difference is accounted for by the molecular size. Chapters 2 and 3 describe photophysical investigations of the zwitterionic dyes, Sulforhodamine 640 (SRh640, Chapter 2) and sulforhodamine B (SRhB, chapter 3). The charge appears to have little effect as the thin film behavior in the zwitterions. As with Rh6G and Rh560, the films change from monomer to dimer to aggregate. The monomer to dimer transition occurs at ~1 nm for SRh640 and at ~1.5 nm for SRhB, consistent with their molecular sizes.


FIGURE PAGE
Chapter 1:      Over the years, there has been an interest in the properties of fluorescence dyes such as xanthene dyes, especially rhodamine 6G. Rhodamine 6G is a well-known cationic fluorophore used as a fluorescent tracer due to its high fluorescence quantum yield. [2][3][4] In high concentrations, rhodamine 6G has been known to form aggregates, which can significantly affect the photophysical properties. [5][6][7][8] Exciton theory has been used to predict the splitting of the two excited states based on dipole-dipole interaction in the dimer that causes spectral changes. For dimers there are two extremes labeled H-and J-dimers. An H-dimer has a sandwich structure and is identified by a blue shift in the absorption spectrum with respect to the monomer. Spectra of H-dimers are typically found at lower wavelength and are non-emissive. On the other hand, a Jdimer is a head-to-tail dimer and shows a red shift in the absorption spectrum with respect to a monomer and are emissive. [9][10] Results show that aggregation of the dye molecules can be controlled depending on certain preparation conditions. [11][12][13] For rhodamine 6G the aggregation is dependent on the thickness of the thin film. As the thickness of the thin film decreases we see that the emission is higher and have shorter wavelength maxima whereas when the thin films increase in thickness we see a quench in emission and longer wavelength maxima. The structure of the dye and thickness of the thin films can be detected from the absorption spectra. 2 By controlling the aggregation of the fluorescent dye we gained an understanding of its photophysical properties to improve our sensors efficiency. In this paper we are investigating the spectroscopic details of a xanthene dye similar to rhodamine 6G, rhodamine 560 also known as rhodamine 110. Similar to rhodamine 6G, rhodamine 560 is a cationic dye in the Xanthene family. [14][15][16][17] Scheme 1. Structures of Rhodamine 560 and Rhodamine 6G.

Experimental Methods:
Pre-cleaned Borosilicate glass microscope slides were used from Scientific Products. These glass slides were cut into a certain length and width, 3. spectrometer. The spectra were collected using an integration time ranging from 0.20 s to 10.0 s, depending on the concentration and from a wavelength of 300-800 nm. The fluorescence spectrum was measured using a Horiba Fluorolog-3 Fluorimeter. For each sample steady state fluorescence measurements were measured at an incidence angle of 60 degrees relative to the excitation beam with a slit width of 2 nm. The excitation light source was used at a wavelength of 450 nm. The emission wavelength ranged from 460 to 800 nm. Excitation spectra were also measured. The excitation range was done in four scans, one from 400 nm to 510 nm and the second one is from 525 nm to 700 nm using a detection wavelength of 520 nm. The third one is from 400 nm to 610 nm and the fourth one is from 625 nm to 700 nm using the a detection wavelength of 620 nm. A Filmetrics microscope was used to determine the thicknesses. Lifetime measurements were made using a Horiba Fluorohub timecorrelated single photon counting (TCSPC) system. A Horiba NanoLED N-460 pulsed diode laser was used as the light source with a wavelength of 464 nm, a repetition rate of 1.0 MHz, and 160 ps pulse duration time with a power output of about 7 pJ/pulse.

Results and Discussion
The glass substrates were deposited by either dip coating or spin coating in a series of concentration solutions of Rh560. The two different coating techniques were both set at a certain speed (1200 RPM) and at a certain pulling rate (~60 mm/min).
The dip coated samples were measured twice: immediately after the dip coat with both sides of the substrate and again after one side was polished clean. All absorption spectra of the two-side coated samples were twice the intensity of the sample with the one-side cleaned. This indicates the uniformity of the coating. Henceforth all reference to dip coated samples are single sided. As shown in Figure 1 there is a trend from the two coating techniques that shows similar spectral properties. As expected for the casting and in solution form, the absorbance increase as concentration increases.
However it does change between solution and thin films. In solutions, at lower concentrations, the absorption maximum is at 500 nm with a shoulder at 469 nm. As the concentration increases we noticed there is no shift and we can see no changes in the line shape throughout the concentrations. This suggests no aggregation occurs in solution.
However, thin films show different results from solutions. In spin coating we noticed at lowest concentration the maxium peak is at ~503 nm with a shoulder at ~468 nm. As the concentration increases the line shape broadens and there is a slight shift in the main peak to ~510 nm and the shoulder to ~479 nm. This broadened peak suggests aggregation is occuring. Similar spectral behaviors occurred in dip coating.
At the lowest concentration the maximum wavelength is at ~503 nm with a shoulder of ~468 nm. As the concentration increases the lineshape broadens and there is slight shift of the main peak to ~503 nm and the shoulder to ~468 nm. Between the two coating methods, the absorbance from spin coating is slightly greater than dip coating at the same concentration. With both spin coated and dip coated samples being broader at low and high energies this would imply an oblique dimer forms on the glass substrate.  techniques it can be seen there is a trend where at low concentrations the peak starts out with a narrow peak with the broad shoulder and as the concentration increases the peak shifts towards lower energy at ~500 nm with a prominent broad shoulder at ~470 nm. This peak has both broad low and high energies shoulders, which implies the presence of oblique dimers in thicker films, similar to what was found in Rh6G. 2 The spin coated samples show broader peaks than dip coated samples at high concentrations, which indicates there are more molecules on the surface when spin cast than dip cast. assigned to an oblique exciton dimer. As shown in the absorption spectra the increasing thickness affects a change in intensity, but it does not require any additional peaks to fit. It is evident in figure 7, which shows the intensity as a function of film thickness, both coating methods show a similar trend. As the thickness increases the fluorescence intensity reaches at maximum then drops at films that are more than 0.5 nm thick. This implies that we are seeing a transition from primarily monomers to primarily dimers. In previous work, it is shown that the transition for Rh6G is around ~1 nm. This would mean Rh560 is more susceptible in aggregation than it is with Rh6G.

A B
A B Table 2. Deconvolution from the fluorescence spectra from both spin-coated and dipcoated samples as shown for Peak position (λmax) and FWHM (Γ). A Gaussian function was used to describe each peak.
The fluorescence spectra from both coating techniques were deconvoluted and the results are shown in Table 2. Both coating methods required only three peaks to fit the spectra at ~525 nm, ~548 nm, and ~ 590 nm. The monomer was assigned at the peak of ~525 nm. The oblique exciton dimer is assigned to the peak of ~548 nm. Also, at higher concentration larger aggregation is assigned to ~590 nm that shifts to lower energy as the aggregates grow.
As shown in figure 8 thickness of the thin films have no effect on the line shape of the excitation spectra detected at 520 nm. However, when detecting at 620 nm maxima wavelength maxima shift to lower energy and the line shape broadens.
These results demonstrate that when detecting at the monomer wavelength, ~520 nm, all of the absorbed light is emitted from monomer. When detected at 620 nm, both monomers and aggregates contribute to the emission that is, some of the energy in the excited monomer is transferred to aggregates.  However, the spin-coated sample shows that the excitation spectra are slightly narrower than the absorbance. This behavior is accentuated in the thickest sample for both coating techniques. This implies that only absorption into monomers leads to emission. Although the oblique dimers absorb some light, these excited states do not contribute significantly to the emission.      thickness where Rh6G has a broad peak while Rh560 has a broad peak with shoulder which implies aggregation is much influential in Rh560 than Rh6G. For Rh560, the transition from monomers to aggregates is at the thickness of 0.5 nm. However, Rh6 G the transition is found to be at around interaction. An H-dimer has a face to face arrangement causing a blue shift with of the absorbance maximum respect to the monomer. These H-dimers are non-emissive. Jdimers are oriented in a head to tail arrangement and are emissive. This arrangement is causing a red shift with of the peak maximum respect to a monomer. Oblique dimers are oriented in an angle between two monomer transition moments. This arrangement results in both a lower and higher energy peaks with respect to the monomer. [8][9][10][11] Layering the neighboring dye molecules in certain geometry onto substrates can alter their optical properties.
In chapter one, we discussed the spectral properties of Rhodamine 560, another cationic fluorescent dye similar to Rhodamine 6G. Results show very similar spectral behavior of Rh6G and Rh560 in thin films. Controlling the concentration can prevent or induce aggregation. Higher concentration tends to form aggregates and at low concentration monomer like structure is layered across the glass substrates. In this work, we will investigate sulforhhodamine 640. Sulforhodamine 640 is an anionic dye in the Xanthene family unlike Rhodamine 6G and Rhodamine 560, which are a cationic dye.

Experimental Methods
Pre-cleaned Borosilicate glass microscope slides were used from Scientific Products. These glass slides were cut into a certain length and width

Results and Discussion
In this study, Sulforhodamine 640 thin films were produced by dip coating and spin coating onto glass substrates in a range of different concentrations from 1. As the concentration increases the line shape broadens and there is a gradual shift of the main peak to 586 nm and the shoulder to 546 nm. Between the two coating techniques, the absorbance from spin coating is four times greater than from dip coating at the same concentration. This is consistent with the spin coated sample being is broader at both high and low energies onto the glass substrate for a given concentration. Figure 2 shows the normalized absorbance of the two casting methods and in solution form. The solution spectrum shows that the same narrow absorption peak with a higher energy shoulder. There are no shifts. However, when the fluorescent dye is cast either by dip coating or spin coating the line shape of the spectra broadened.
For both coating methods, the absorbance spectra maximum shifts to lower energy as the film thickness increases. The high energy shoulder also increases. These observations imply the presence of oblique dimers in the thicker films, similar to what was found in Rh6G. 16  Table 1. Deconvolution from the absorbance spectra from both spin-coated and dipcoated samples as shown for Peak position (λmax) and FWHM (Γ). A Gaussian function was used to describe each peak.
Deconvolution of the absorbance spectra for both coating techniques were done, and all the spectra were fit to three peaks and the results are given in Table 1.
For spin coating, at low concentration where the thickness is in the submonolayer to monolayer range, there have two main peaks, ~590 nm with higher energy shoulder around ~ 550 nm and a hidden feature at 510 nm. The peak at ~550 nm is assigned as arising from a monomer while the peaks at ~590 nm and ~510 nm are assigned to an oblique exciton dimer. As the concentration increases no additional peaks are needed to fit the spectrum but the exciton peaks become more prominent. Dip coating shows similar behavior but with the peaks slightly shifted to ~530 nm for the monomer and ~590 nm and ~490 nm for the oblique exciton dimer.      Table 2. Deconvolution from the fluorescence spectra from both spin-coated and dipcoated samples as shown for Peak position (λmax) and FWHM (Γ). A Gaussian function was used to describe each peak.
Deconvolution of fluorescence spectra for both coating methods are done and most of the spectra were fit to three peaks as shown in Table 2. At lowest concentrations the monomer is assigned at ~530 nm and the dimer is assigned at ~580 nm. At higher concentrations, higher aggregates are forming and for that peak is assigned to ~600 nm.     When there are more neighboring dye molecules higher order aggregates can form.
This caused the fluorescence to be reduced even though it has a higher absorbance.
Therefore, the thickness on our samples affects the spectral behavior.

NOTES
The authors declare no competing financial interest.

Introduction
There has been a significant amount of research in sensing. [1][2][3][4][5][6] There has been an interest in sensing explosives which can be a challenging task to achieve. 4,[7][8] In our research group we have been interested in sensing vapor gas explosives using fluorescence based sensors. Our sensor system is based on a three layer system, where we have a glass substrate coated with a transparent polymer, and then a fluorophore.
Our results showed a great deal in efficiency with both the sensitivity and selectivity of gas phase explosives. However, the interfacial effects are not fully understood yet. 9 Thus, it is important to look at the photophysics of the fluorophore in thin films.
Xanthene dyes are widely known. They have favorable photophysics such as high quantum yields and photostability. These have been used in many applications as a lasing medium in dye lasers and fluorescent markers in biological studies. 6,[10][11][12][13][14][15][16][17] The most studied on is Rhodamine 6G. Rhodamine 6G is a cationic fluorescent dye.
However, these xanthene dyes especially for rhodamine are known to aggregate, which can affect the photophysical properties. 18 Depending on the preparation conditions, aggregation of the dye molecules can be controlled. From our previous work showed that aggregation of rhodamine 6G is dependent on the thickness of the thin film. Emission is at its highest when the thin films are approximately one monolayer and when the films are thicker the emission intensity is quenched. Structures were identified using the absorption spectra. 17,[22][23][24][25][26][27][28] From Chapter 1, the results for rhodamine 560 showed similar results to rhodamine 6G. Though we were seeing aggregation occur regardless of which coating method was used, aggregation only occurs at higher concentration. At lower concentration we tend to see more monomer like structure in our thin films.
In this work we will investigate sulforhodamine B (SRhB) in thin films.
Sulforhodamine B is a zwitteronic dye in the xanthene family. Sulforhodamine B is different from rhodamine 6G due to its different functional groups where as sulforhodamine B has tertiary amines, a sulfate group and a sulfonic acid attached to a benzene ring. Rhodamine 6G has only two secondary amines, methyl groups, and an ester attached to the benzene ring.

Experimental Methods
Pre-cleaned Borosilicate glass microscope slides were used from Scientific Products. These glass slides were cut into a certain length and width

Results and Discussion
Sulforhodamine B thin films were produced by two different coating methods, dip coating and spin coating onto glass substrates. Different ranges of concentrations, from 1.0 x10 -3 M to 1.0x10 -7 M were deposited onto the glass substrates. Figure 1 shows  shows that the line shape is independent of the concentration. However, the thin films show different results. When the fluorescent dye is cast either by dip coating or spin coating the line shape of the spectra is broader than in solutions. For spin-coated samples the maxima absorbance peak shifts to lower energy as the film thickness increases. And the high energy shoulder increases as the thickness increases.
However, there is no gradual shift of the peak maximum for the dip-coated samples.
There is a change in the line shape where it broadens, and the higher energy shoulder increases as the concentration increases. This implies the presence of oblique dimers in the thicker films, similar to what was found in Rh6G. 17  of film thickness, Figure 4, the absorbance is nonlinear for both coating methods. This implies that dimerization occurs at the thicker region.  Table 1. Deconvolution parameter for absorbance spectra of spin coating and dip coating. Peak position (λ max ) and FWHM (Γ). A Gaussian function was used to describe each peak.
Deconvolution of the absorbance spectra for both coating techniques were done, and all the spectra were fit to three peaks and the results are given in Table 1 for both coating methods. Both coating methods give similar results. The peak at ~535 nm is assigned to absorption from a monomer. The pair of peaks at ~490 nm and ~575 nm are assigned to an oblique exciton dimer.  The line shape is such that the lower energy peak dominates over the higher energy shoulder. As the concentration further decreases the change in the line shape occurs where the once dominate lower energy shoulder starts to dissipate and the higher energy starts to dominate. Shown in Figure 6, the normalized emission spectra demonstrate the line shape and spectral shift changes as the concentration changes.
Two broader peaks grow at lowest concentration with low intensity and a lower energy shoulder. As the concentration increases, the two broad peaks shift gradually, and the higher energy feature starts to dissipate and the lower energy starts to dominate, reaching at a maximum intensity at a concentration of 1.0x10 -4 M. At higher concentrations the lower energy peak is nearly gone, the higher energy shoulder slowly decreases, and a newer lower shoulder start to appear. This implies that aggregation is occurring. Similar behavior is observed for the emission spectrum for dip coating.  Table 2. Deconvolution parameter for fluorescence spectra of spin coating and dip coating. Peak position (λ max ) and FWHM (Γ). A Gaussian function was used to describe each peak.
the monomer is assigned at ~520 nm and the excimer are assigned at ~592 nm. At higher concentrations, higher aggregates are forming and for that peak is assigned to ~620 nm.

Statement of the Problem
We have developed a fluorescence-based sensor using Xanthene derivatives. Due to their high quantum yield and strong signal in emissions Xanthene dyes are widely used. These dyes interact with analytes such as trinitrotoluene (TNT) and Trinitritrobenzene (TNB) and have shown high sensitivity in detection from fluorescent emission. 1 We also have done intense investigation on Rhodamine 6G and its photophysics properties. This led to our further understanding of the aggregation of the dye molecules and how the effect of thin film thickness plays a key role in aggregation. [2][3][4][5] Now we had extended our results to other xanthene dyes including Rhodamine 560, Sulforhodamine B and Sulforhodamine 640. We had compared the spectroscopic results from a widely used fluorophore such as Rhodamine 6G with other Xanthene dyes and investigate these fluorophores to optimize their photophysics properties and enhance our fluorescence-based sensor.

Justification for and Significance of the Study
Fluorescence-based sensors have been reported and are continuing to improve the methods in enhancing the fluorescent signal. A three layer structure which consists of a glass substrate, a polymer, and a fluorophore has been reported and shown as a sensor in detection of analytes. Results show that with a little amount of fluorophore present in their sensor, the signal. Rhodamine 6G is a common cationic fluorophore used as a fluorescent tracer due to its high fluorescence quantum yield. 1 In high concentrations, Rhodamine 6G has been known to form aggregates and higher order aggregation which can significantly affect the photophysic properties. 2 Results show aggregation of the dye molecules can be controlled depending on certain preparation settings. [6][7][8] For Rhodamine 6G the aggregation is dependent on the thickness of the thin film. As the thickness of thin films decreases we see that the emission is higher and have shorter wavelength maxima whereas when the thin films increases we see a quench in emission and longer wavelength maxima. The structure of the dye and thickness of the thin films can be found in the absorption spectra. 2 From controlling the aggregation of the fluorescent dye we gained an understanding of its photophysical properties to improve our sensors efficiently.
In this work we had expanded our investigation on the Xanthene dyes coated on glass substrates. The xanthene dyes we had studied are Rhodamine 6G, Rhodamine 560, Sulforhodamine B, and Sulforhodamine 640. Each xanthene dye is structurally different: the Rhodamine 6G and Rh560 are cationic dyes, Sulforhodamine B is zwitteronic dye and Sulforhodamine 640 is an anonic dye. We had investigated the absorption, emission, and excitation spectra of Rh6G, Rhodamine 560, Sulforhodamine B, and Sulforhodamine 640. 1 We will also investigate the preparation of the thin films from coating techniques spin cast and dip coating. [9][10][11][12] We will use our spectroscopic spectra as a function of thickness to determine the structure of these dyes which will give us understanding to improve our sensing methods.

Methodology or Procedures
The experiments were conducted by the start with a substrate preparation. Pre-cleaned Borosilicate glass microscope slides was used from Scientific Products. These glass slides were cut into a certain length and width. These slides were washed twice in a Branson 3510 sonicator. First wash contained only 95% ethanol and was sonicated for fifteen minutes. It was then followed by a second wash of distilled water for another fifteen minutes. Then the slides were dried by using a nitrogen gas tank.