Investigation of the Binding of Single-Stranded DNA to Single-Walled Carbon. Nanotubes as Studied by Absorbance and Fluorescence Spectroscopy

Transcription

1 Investigation of the Binding of Single-Stranded DNA to Single-Walled Carbon Nanotubes as Studied by Absorbance and Fluorescence Spectroscopy A thesis presented to the faculty of the College of Arts and Sciences of Ohio University In partial fulfillment of the requirements for the degree Master of Science Maureen M. Heines August 2007

2 This thesis titled Investigation of the Binding of Single-Stranded DNA to Single-Walled Carbon Nanotubes as Studied by Absorbance and Fluorescence Spectroscopy by MAUREEN M. HEINES has been approved for the Department of Chemistry and Biochemistry and the College of Arts and Sciences by Liwei Chen Assistant Professor of Chemistry and Biochemistry Benjamin M. Ogles Dean, College of Arts and Sciences

3 Abstract HEINES, MAUREEN, M., M.S., August 2007, Chemistry Investigation of the Binding of Single-Stranded DNA to Single-Walled Carbon Nanotubes as Studied by Absorbance and Fluorescence Spectroscopy (98 pp.) Director of Thesis: Liwei Chen The effect of singled-stranded DNA concentration, single-walled carbon nanotube (SWNT) concentration, DNA base pair length, and ph on SWNT dispersion was investigated to gain a better understanding of how DNA binds to SWNTs. Changes in the near-infrared absorbance and fluorescence spectra were used to determine these effects. At low DNA concentration, very little DNA was bound to the SWNTs leading to poor SWNT dispersion. At higher DNA concentration, more DNA was bound to the SWNTs achieving better dispersion. Increasing the DNA concentration and base pair length resulted in better SWNT dispersion as observed from the increasing structure and intensity in the near-infrared spectra. The opposite effect was observed by increasing the SWNT concentration. An equilibrium binding constant was calculated using the Langmuir model but resulted in an implausible value. Varying the ph of DNA/SWNT samples showed DNA to act similar to SDS as a surfactant except under very basic conditions where the fluorescence intensity diminished above ph 11 for the DNA dispersed samples. Approved: Liwei Chen Assistant Professor of Chemistry and Biochemistry

4 For my mother.

5 Acknowledgments My foremost thanks goes to my thesis advisor, Dr. Liwei Chen. I wish to extend my sincere appreciation to Dr. Chen for his guidance, assistance, and insight that have helped shaped my research skills over the past two years. I would also like to thank Dr. Hugh Richardson for believing in me and pushing me to always try my hardest. To my thesis committee, Dr. Greg Van Patten and Dr. Jeff Rack, I would like to thank them for their time and careful consideration of this thesis. I would like to acknowledge the Ohio University Department of Chemistry and Biochemistry and their support of my research. I am grateful to my fellow chemistry graduate students who have provided me with a stimulating learning environment during my time at Ohio University. I would specifically like to thank Alyssa Thomas for her camaraderie and companionship both in and outside of lab and for always taking the time to help. Lastly, I would like to thank my family whose love and emotional support have helped me through the particularly difficult times. Without them, I would have never made it this far.

6 6 Table of Contents Page Abstract... 3 Dedication 4 Acknowledgments... 5 List of Tables... 8 List of Figures... 9 Chapter 1: History and Background Historical Background and Thesis Overview Theoretical Background on SWNT Spectroscopy Effects of ph and Surfactant Choice on Semiconducting SWNTs Chapter 2: Experimental Procedures, Instruments, and Materials Used Carbon Nanotube and Surfactant Preparation Apparatus, Instruments, and Parameters Materials Used Determination of DNA Concentration Determination of the Equilibrium Constant for DNA Adsorption/ Desorption Chapter 3: The Effect of Varying DNA or SWNT Concentration on ssdna/swnt Complexes Choice of SWNT Material Effect of DNA Concentration on Absorbance and Fluorescence Spectra Calculation of the Equilibrium Constant for DNA Adsorption/Desorption Effect of SWNT Concentration on Absorbance and Fluorescence Spectra Reduction of SWNTs as Observed by Changes in the NIR Spectra Discussion Chapter 4: The Effect of ssdna Base Pair Length on DNA Attachment to SWNTs Effect of Base Pair Length on Absorbance and Fluorescence Spectra Discussion... 75

7 7 Chapter 5: ph Effects of SWNTs Suspended with ssdna Effect of ph on Absorbance and Fluorescence Spectra Discussion Chapter 6: Conclusions References Appendix I... 98

8 8 List of Tables Tables Page Table 1. Samples #1-3 DNA concentrations from UV-vis measurements...43 Table 2. GT20 DNA concentrations from UV-vis measurements...49 Table 3. DNA concentrations from varying SWNT concentration experiments...58 Table 4. Samples #1-3 DNA concentrations re-measured after one month...63 Table 5. DNA concentrations from base pair effect tests...73

9 9 List of Figures Figure Page Figure 1.1. Diagram showing how a hexagonal sheet of graphite is rolled to form a tube...19 Figure 1.2. Diagram showing which chiral indices make up metallic or semiconducting SWNTs...20 Figure 1.3. Semiconducting SWNT Density of Electronic States...22 Figure 1.4. Mechanism for formation of protonated SWNT oxide...26 Figure 2.1. Sonicator apparatus...34 Figure 3.1. NIR absorbance spectra of SDBS and SDS with raw and processed SWNTs...38 Figure nm excitation fluorescence spectra of SDBS and SDS with raw and processed SWNTs...39 Figure nm excitation fluorescence spectra of SDBS and SDS with raw and processed SWNTs...39 Figure 3.4. Initial NIR absorbance spectra of samples #1-3 for the first GT30 concentration experiments before DNA concentration alterations...41 Figure 3.5. NIR absorbance after GT30 concentration alterations...42 Figure nm excitation fluorescence spectra after GT30 concentration alterations...42 Figure 3.7. NIR absorbance spectra of all samples for the second GT30 concentration experiments...44 Figure nm excitation fluorescence spectra for the second GT30 concentration experiments...45 Figure nm excitation fluorescence spectra for the second GT30 concentration experiments...45 Figure NIR absorbance of GT20 concentration experiments...46

10 Figure nm excitation fluorescence spectra of GT20 concentration experiments...47 Figure nm excitation fluorescence spectra of GT20 concentration experiments...47 Figure nm excitation fluorescence values plotted against GT20 concentration for designated wavelengths...48 Figure nm excitation fluorescence values plotted against GT20 concentration for designated wavelengths...49 Figure Plot of 1/bound DNA concentration against 1/unbound DNA concentration for the GT20 data...51 Figure Changes in the bound DNA concentration plotted against the unbound DNA concentration...52 Figure Unbound DNA concentration plotted against the total DNA concentration...53 Figure Plot of 1/surface coverage versus 1/unbound DNA concentration for the GT20 data...54 Figure NIR absorbance of spectrum GT30 with low SWNT concentration...55 Figure NIR absorbance spectrum of GT30 with mid SWNT concentration...55 Figure NIR absorbance spectrum of GT30 with high SWNT concentration...56 Figure NIR absorbance spectra of all three GT30 samples with varying SWNT concentration...56 Figure nm excitation fluorescence spectra of all three GT30 samples with varying SWNT concentration...57 Figure nm excitation fluorescence spectra of all three GT30 samples with varying SWNT concentration...57 Figure NIR absorbance spectra of sample 1 in February and in April...59 Figure nm excitation fluorescence spectra of sample 1 in February and April...60 Figure NIR absorbance of sample 2 in March and in April

11 11 Figure nm excitation fluorescence spectra of sample 2 in February and April...61 Figure nm excitation fluorescence spectra of sample 2 in February and April...61 Figure NIR absorbance spectra of sample 3 in March and in April...62 Figure nm excitation fluorescence spectra of sample 3 in February and April...62 Figure nm excitation fluorescence spectra of sample 3 in February and April...63 Figure 4.1. NIR absorbance spectra of GT10-30 with processed SWNTs...67 Figure nm excitation fluorescence spectra of GT10-30 with processed SWNTs...68 Figure nm excitation fluorescence spectra of GT10-30 with processed SWNTs...68 Figure 4.4. NIR absorbance spectra of GT10-30 with raw SWNTs...69 Figure nm excitation fluorescence spectra of GT10-30 with raw SWNTs...69 Figure nm excitation fluorescence spectra of GT10-30 with raw SWNTs...70 Figure 4.7. NIR absorbance spectra of GT10-30 with raw SWNTs...71 Figure nm excitation fluorescence spectra of GT10-30 with raw SWNTs...71 Figure nm excitation fluorescence spectra of GT10-30 with raw SWNTs...72 Figure NIR absorbance spectra for GT10-30 with raw SWNTs, second experiments...74 Figure nm excitation fluorescence spectra for GT10-30 with raw SWNTs, second experiments...74 Figure nm excitation fluorescence spectra for GT10-30 with raw SWNTs, second experiments...75

12 Figure 5.1. NIR absorbance spectra of GT30/SWNT samples with varying ph...78 Figure nm excitation fluorescence spectra of GT30/SWNT samples with varying ph...79 Figure nm excitation fluorescence spectra of GT30/SWNT samples with varying ph...79 Figure 5.4. NIR absorbance spectra of GT20/SWNT samples with varying ph...80 Figure nm excitation fluorescence spectra of GT20/SWNT samples with varying ph...81 Figure nm excitation fluorescence spectra of GT20/SWNT samples with varying ph...81 Figure 5.7. NIR absorbance spectra of SDS/SWNT samples with varying ph...82 Figure nm excitation fluorescence spectra of SDS/SWNT samples with varying ph...83 Figure nm excitation fluorescence spectra of SDS/SWNT samples with varying ph...83 Figure Percent relative intensity versus ph for SWNTs dispersed with GT Figure Percent relative intensity versus ph for SWNTs dispersed with GT Figure Percent relative intensity versus ph for SWNTs dispersed with SDS

13 13 CHAPTER 1: History and Background 1.1 Historical Background and Thesis Overview Since the initial discovery of carbon nanotubes, scientists have devoted much effort to elucidating their synthesis and properties. Significant advances in carbon nanotube research have shown them to possess unique attributes including exceptional strength, excellent thermal conductivity, and either metallic or semiconducting properties. Such unique properties make carbon nanotubes suitable for many different applications. While progression in carbon nanotube research has been impressive, there is still much to learn about these novel structures. One area of particular interest is the interaction between carbon nanotubes and biological molecules. A better understanding of how single-stranded DNA and single-walled carbon nanotubes interact will be vital for future carbon nanotube applications such as sensors for biological molecules. What follows is a review of the history of carbon nanotubes that encompasses the findings presented in this thesis. The first long needle-like tubes of carbon were observed by Iijima in 1991 using a transmission electron microscope (TEM). 1 Iijima facilitated an arc-discharge method by which multi-walled carbon nanotubes were grown from the cathode end of graphite electrodes. These nanotubes were found to be up to a micron in length, a few tens of nanometers in diameter, and made up of 2 to 50 multiple carbon shells. Each shell, or layer, was found to be similar in nature to a rolled up sheet of graphene. Two years later, the first single-walled carbon nanotubes (SWNTs) were synthesized by independent

14 14 methods from two separate groups. 2,3 Both groups used a method of vaporizing carbon with a metal catalyst to produce single-layer nanotubes approximately 1 nm in diameter and several microns long. Iijima s group used an iron catalyst in an argon/methane atmosphere and found the SWNTs deposited in the soot. 2 Bethune s group from IBM used cobalt in a helium atmosphere and found the tubes in both the soot and attached to the chamber walls. 3 Along with these electric arc vaporization methods, laser ablation and chemical vapor deposition have been among the more popular methods of SWNT synthesis. Arcdischarge and laser ablation methods require solid-state carbon precursors for nanotube growth and high temperatures for the vaporization of carbon. 4 Though these two methods produce large quantities of byproducts, recent progress in these methods have produced higher yields of high quality nanotubes with nearly perfect structures. 5, 6 Chemical vapor deposition (CVD) requires lower temperatures than the arc-discharge or laser ablation methods. CVD also uses a metal catalyst and hydrocarbon gases instead of solid-state carbon for its carbon atom source. The quality of the CVD method of SWNT synthesis has improved over the years such that the quality of nanotubes produced by this method are comparable to those of arc-discharge and laser ablation Early calculations predicted that graphene s unusual electronic structure would allow SWNTs to exhibit either metallic or semiconducting properties Depending on the manner in which a graphene sheet was rolled to form the nanotube, the SWNT could be metallic, semiconducting with a small band gap, or semiconducting with a band gap inversely proportional to its tube diameter. 16 In 1992, these calculations were supported

15 15 when I-V plots produced from scanning tunneling microscopy and spectroscopy (STM/STS) provided evidence for both metallic and semiconducting nanotubes. 17 Further STM studies confirmed that the electronic properties of SWNTs are related to changes in their chirality and tube diameter Conductivity of metallic and semiconducting nanotubes have been measured with one study showing a resistance range of kω for metallic tubes (at temperatures ranging from K), kω for large diameter semiconducting tubes (at room temperatures), and resistances in the MΩ range or higher for small diameter semiconducting tubes (diameters less than 1.5 nm at room temperatures). 22 Other experimental results have shown current densities for metallic SWNTs in the range of 10 9 A/cm 2 which is much larger than the maximum of 10 5 A/cm 2 for metals. 23, 24 Further electronic studies on SWNTs showed that due to their one-dimensional electronic structure, very high electron mobility in metallic tubes was possible where electronic transport occurred ballistically and without heat production over lengths of micrometers. 23, 25, 26 Superconductivity has even been observed at low temperatures in metallic nanotubes. 27 There have also been extensive theoretical studies on the mechanical properties of SWNTs. The flexibility of nanotubes was examined by Iijima et al. using molecular dynamic simulations. The results showed that bending of the nanotubes was completely reversible up to Using an empirical lattice dynamics model, Lu was able to calculate a Young s modulus of ~ 1 TPa, a shear modulus of ~ 0.45 TPa, and a bulk modulus of ~ 0.74 TPa. 29 These values are comparable to those of diamond. Hernandez and his group found similar values for the Young s modulus at ~ 1.24 TPa. 30 There have

16 16 been a variety of experimental studies which have examined the elastic modulus of SWNTs with methods such as proximal probe tip indentation and TEM analysis of vibrations. The Young s moduli for these experiments range anywhere from TPa up to 5 TPa. 31 Despite these large differences in Young s modulus values, SWNTs still exhibit an exceptionally high elastic modulus, fracture strain sustaining capability, and a large elastic strain Furthermore, compared to steel wire, the strength and densitynormalized modulus of typical SWNTs are ~ 56 and ~ 19 times greater, respectively. 33 The exceptional electrical, mechanical, and physical properties of SWNTs along with their low density make them highly desirable for many potential applications. SWNTs are being considered for field emission electron sources for devices such as flat panel displays due to their small radius and length of the nanotube Their tiny size, high surface area, hollow center, and chemical inertness make SWNTs appealing for applications as chemical and biological sensors. Carbon nanotubes have promise as probe tips for atomic force microscopy (AFM) and STM instruments; due to their high elasticity and again to their small size, carbon nanotubes offer the benefit of highresolution imaging. 31 SWNTs may also play a vital role in the miniaturization of electronic circuitry and devices. However, in order to utilize SWNTs in any of these potential applications, they must first be separated into individual tubes. Carbon nanotubes have a tendency to aggregate in bundles with binding energies of ~ 500 ev per micrometer of tube-tube contact due to van der Waals forces. 39 This bundling leads to poor solubility, no observable emission spectra, and rather diffuse absorption spectra. This issue has recently been resolved by dispersing carbon nanotubes

17 17 in water with surfactants. The first attempts to separate SWNTs involved solubilizing them in various organic solvents. 40, 41 Later studies facilitated the use of polymer 42, 43 surfactants such as polyvinyl pyrrolidone and gum Arabic for SWNT separation. Though these methods produced SWNTs with some separation, the samples were still dominated by aggregated nanotubes. In 2000, Vigolo et al. attempted to separate SWNTs with the surfactant sodium dodecyl sulfate (SDS) which freed the tubes from their bundles. 43 O Connell et al. perfected this technique in In these experiments, the SDS served as a coating in the form of a close-packed columnar micelle that prevented the tubes from re-aggregating. Any remaining bundled tubes in solution were removed by centrifugation. In contrast to earlier studies, the semiconducting tubes in the SDS samples from O Connell s studies were found to fluoresce in the near-infrared region of nm. A new surfactant, single-stranded (ss) DNA, was auditioned by Zheng et al. one year later. 45 ssdna was tested as a surfactant because molecular modeling suggested it would bind to the carbon nanotubes through π-stacking interactions of DNA s aromatic nucleotide bases with the carbon nanotube sidewall. The DNA s binding free energy to the nanotube is suggested to rival that of nanotube-nanotube interactions. They found that poly T (thymine) yielded the highest dispersion efficiency among the four different bases (thymine, guanine, adenine, and cytosine). They also concluded that ssdna appeared to be more efficient in dispersing carbon nanotubes compared to previous polymers when they found that 1 mg DNA could disperse an equal amount of carbon nanotubes in 1 ml volume. During this experiment, Zheng et al. also found DNA to be an efficient means of

18 18 separating metallic and semiconducting nanotubes; when running these samples on an ion-exchange column, they found metallic tubes would elute first off of the column followed by the semiconducting tubes. Further studies by Zheng s group found that a combination of alternating guanine and thymine bases with a total length of 20 to 90 bases resulted in the best nanotube separation. 46 DNA itself has many interesting properties that make it useful for potential applications in constructing objects on the nanometer scale and assembling/connecting nanoparticles, molecules, or materials to/on a substrate The π-stacking interaction between DNA s bases may have electronic properties appropriate for use in molecular electronics. 50 Along with carbon nanotubes, DNA has potential applications in bio and optical sensors Despite all of these possibilities for device applications, the study of ssdna as a surfactant for SWNT separation is relatively new and there is still much to be learned. For this thesis, the effects of DNA and SWNT concentration, DNA base pair length, and ph on the binding of DNA to the SWNT was investigated by NIR optical spectroscopy. A theoretical background on the unique structure of SWNTs which leads to their absorbance and fluorescence in the near-infrared region is examined in the second half of chapter 1. Details of experimental procedures such as sample preparation, materials, and instrumentation are described in chapter 2. Chapter 3 explores the effects of DNA and SWNT concentration on the NIR spectra where chapter 4 examines the effect of varying DNA base pair length on these spectra. Finally, chapter 5 shows the

19 19 results of varying the ph of DNA/SWNT complexes compared to varying the ph of a SDS/SWNT complex. 1.2 Theoretical Background on SWNT Spectroscopy SWNTs can be thought of as cylinders of rolled up graphene sheets. The nanotube chirality is defined by the chiral vector, C r, and the chiral angle, θ. The chiral vector is h perpendicular to the long nanotube axis. The chiral angle determines the amount of rotation, or twist, in the nanotube. The structure of SWNTs is described by the chiral vector in the following equation: where a r 1 and a r 2 Figure ). r C h r r = na + m 1 a 2 are unit vectors and n, m are integers known as the chiral indices (see (1) Figure 1.1: Diagram showing how a hexagonal sheet of graphite is rolled to form a tube 54

20 20 The chiral angle can range between 0 to 30. In the case of a 0 angle, the nanotube is called a zig-zag type. A chiral angle of 30 is called an armchair type. The general rules for determining whether the SWNT is metallic or not can be determined from the chiral indices. When n = m or when n-m = 3i, where i is a nonzero integer, the SWNTs are metals. When n-m 3i, where again i is a nonzero integer, then the SWNTs are semiconductors. Figure shows which pairs of chiral indices represent metallic and which represent semiconducting nanotubes with the red dots representing metallic tubes and the black circles representing semiconducting. Figure 1.2: Diagram showing which chiral indices make up metallic or semiconducting SWNTs 55

21 21 SWNTs possess a quasi-one-dimensionality in their π-electron structure which results in sets of sharp maxima in their electronic densities of states known as van Hove singularities. SWNT have strong transitions for light polarized along the tube axis. However, allowed dipole absorption and emission transitions will only occur between van Hove singularities in matching valence and conduction sub-bands. These transitions and their energies are denoted by E ii, with i = 1, 2, 3, etc. Carbon nanotubes consist of covalently bonded carbon atoms where each carbon is bound to three others through sigma bonds. The remaining p-electron is part of a delocalized π-electron system. SWNTs can be formed with different diameters and chiralities which results in a variety of π-electron band structures and excitations among the various nanotube types. The SWNTs that are semiconducting in nature possess a band gap which allows for these tubes to fluoresce in the near-infrared (NIR) region. On the other hand, metallic nanotubes, which have no band gap, will not emit in the NIR. The optical spectroscopy of semiconducting SWNTs is dominated by E 11 and E 22 transitions between the conduction and valence bands. The values of E 11 and E 22 vary with tube structure. According to Kasha s rule for molecular photophysics 56, the semiconducting tubes will only emit measurably in the E 11 transition. Figure represents the density of electronic states for a semiconducting SWNT. The valence and conduction bands are labeled v and c respectively for the van Hove singularities. The valence and conduction bands are also labeled with subscripts depicting the sub-band index. The values for energy and density of states are arbitrary.

22 22 Figure 1.3: Semiconducting SWNT Density of Electronic States 57 In order to observe E 11 and E 22 transitions in SWNTs, they must first be separated from one another. As mentioned previously, a van der Waals force of ~ 500 ev binds the tubes to each other which quenches their fluorescence due to either non-radiative energy transfer to neighboring bundled SWNTs or transfer to nearby metallic tubes which do not fluoresce in the NIR. These large van der Waals energies can be overcome by dispersing SWNTs in an aqueous surfactant. The surfactant prevents the tubes from aggregating thus allowing for their separation from bundles. The better the SWNTs are separated, the greater the fluorescence intensity and structure. Bundled tubes can be discriminated from dispersed tubes by broadened peaks with no defined structure in the absorbance and fluorescence spectra. Therefore, the effects of increasing the surfactant concentration

23 23 while maintaining a constant SWNT concentration can be investigated. If increasing the surfactant concentration results in better surface coverage, and thus better separation of the SWNTs, one would expect to find an increase in structure and fluorescence intensity as surfactant concentration is increased. In addition to this investigation, the effects of altering the SWNT concentration while maintaining a constant surfactant concentration could be studied. Increasing the SWNT concentration should result in less surfactant per tube to disperse the SWNTs and a decrease in structure and fluorescence intensity would be expected. 1.3 Effects of ph and Surfactant Choice on Semiconducting SWNTs The effect of ph on semiconducting nanotubes has been studied using surfactants such as SDS, but not with using ssdna as a surfactant. One study suggested the changes observed in the NIR spectra from varying the ph is related to the reduction potential of the SWNTs. 58 Semiconducting SWNTs have a band gap that is inversely proportional to their tube diameter. As the chemical reactivity of a valence electron is dictated by its energy level, the changes in the nanotube valence energy level as a function of their band gaps could have an effect on chemical reactions. For example, recent studies have found that the reduction potential (the tendency of a chemical species to acquire electrons) of semiconducting SWNTs depends on the size of their band gap. 58, 59 These studies showed that peaks at longer wavelengths in the NIR absorbance spectrum (from smaller band gap tubes) were more easily diminished when oxidized than peaks at shorter wavelengths

24 24 from larger band gap tubes. Therefore, they concluded tubes with larger band gaps/smaller diameters have a higher reduction potential with lower v1 energy bands. They further explained that oxygen dissolved in aqueous solution can act as an oxidizing agent for the SWNTs where the presence of oxygen in the system can diminish the E 11 transitions beginning with small band gap tubes as the ph of the system is lowered. The effect of ph arises from the dissolved oxygen and is represented by: 4SWNT (reduced) + O 2 + 4H + 4SWNT + (oxidized) + 2H 2 O (2) They claim the oxygen can oxidize SWNTs upon lowering the ph, thus diminishing fluorescence intensity. As the ph of the system is lowered, the reduction potential of O 2 /H 2 O couple increases making it easier to oxidize the SWNTs. Since the smaller band gap tubes have a lower reduction potential, they will be oxidized before the larger band gap tubes and there should be greater changes in the NIR spectra for the small band gap tubes. A positive charge is generated on the SWNT for each electron withdrawn from the tube during oxidation. This effect was reversible upon increasing the ph and fluorescence intensity would return to its original intensity and was able to be repeated several times. This indicates that the reaction in equation (2) must be in some sort of equilibrium. If the reaction is in equilibrium, it could be forced to shift to the reactants side forming more reactants. By adding a strong base to deplete the proton concentration on the reactants side, the reaction could be forced back to the left.

25 25 Another study suggests that the dissolved oxygen actually forms a 1, 4- endoperoxide across an aromatic ring in the carbon nanotube honeycomb structure. 60 When the ph is lowered to acidic conditions, the 1, 4-endoperoxide can become protonated resulting in the endoperoxide ring opening up and a carbocation structure forming (see Figure 1.4). They claim that the positive delocalized hole formed from this process is responsible for luminescence quenching through a non-radiative Auger recombination process. The carbocation is stabilized from the delocalization that occurs in the nanotube. Increasing the ph above 3 leads to the removal of protons from the tube which results in a peroxide anion that can snap back into the 1, 4-endoperoxide. This is just the reverse of the mechanism illustrated in Figure 1.4. This again suggests some sort of equilibrium. Both this study and the study mentioned previously result in the electron depletion of SWNTs which leads to fluorescence quenching.

26 26 Figure 1.4: Mechanism for formation of protonated SWNT oxide 60 The choice of surfactant can enhance or diminish the ph effect. If an anionic surfactant such as SDS is used, ph effects are far more noticeable than if a non-ionic surfactant is used. It is suggested that a negatively charged surfactant decreases the ability of the SWNT to be reduced by stabilizing the positively charged oxidized form of the SWNT. 57 Lowering the reduction potential would make it easier to oxidize the SWNTs, thus changes in the NIR spectra upon oxidizing the SWNTs would be more noticeable. As the reduction potential of the large band gap tubes is higher than for the small band gap tubes, the small band gap tubes would be oxidized before the large as ph is lowered. Therefore, it is expected that using DNA (with its negatively charged backbone) as

27 27 surfactant would show a noticeable diminished fluorescence for smaller band gap nanotubes before larger band gap nanotubes upon lowering the ph.

28 28 CHAPTER 2: Experimental Procedures, Instruments, and Materials Used 2.1 Carbon Nanotube and Surfactant Sample Preparation General Procedure. Three different surfactants were used to suspend the SWNTs in solution; sodium dodecyl sulfate (SDS), sodium dodecyl benzene sulfonate (SDBS), or single-stranded (ss) DNA. Solid ssdna samples containing 10, 20, or 30 base pairs of guanine (G) and thymine (T) in a GT pattern were diluted with deionized (D.I.) water to form stock solutions of 5 or 10 mg/ml. The GT pattern was chosen as previous groups have found it to be the most efficient at separating the SWNTs. 46 The desired concentration of ssdna solution was prepared by diluting the stock solution with D.I. water to the appropriate volume. Solutions of SDS and SDBS were prepared at 1% by weight of the surfactant. Single-walled carbon nanotubes (SWNTs) in the range of mg were weighed out. The ssdna, SDS, or SDBS solution was added to the SWNT in a test tube. Some samples were left to soak for ~ 30 minutes while other samples went straight to the next step. After adding the surfactant solution to the SWNTs, the samples were then sonicated with a sonicator probe with the amplitude setting placed between 50-75%. The times of sonication varied from 20 minutes up to 1 hour. The sample was then transferred to a centrifuge to remove any remaining insoluble/bundled SWNTs from the sample. The samples were centrifuged at 12,000-13,000 rpm for 1 hour. The dispersed SWNTs in solution were then separated from any solid SWNTs at the bottom of the centrifuge tube by using a micropipette to remove the solution.

29 29 DNA Concentration Experiments. The concentration of ssdna was varied while keeping the SWNT concentration the same. The ssdna used for the first experiments was 30 base pairs long consisting of guanine (G) and thymine (T), or GT30 (30 G and 30 T). All dilutions used D.I. water. 60 μl of stock GT30 (5 mg/ml) was diluted to 1500 μl total to form a 0.2 mg/ml solution. This solution was added to ~ 3 mg unprocessed, or raw, SWNTs. The sample was sonicated for 1 hour at 75% amplitude and then centrifuged for 1 hour at 13,000 rpm. The sample was then divided into three 500 μl portions. To the first of these portions, sample #1, 500 μl of D.I. water was added. Sample #1 was then sonicated for 30 minutes at 50% amplitude and centrifuged for 20 minutes at 12,000 rpm. There were no solid SWNTs at the bottom of the centrifuge tube after the second centrifugation. For the second 500 μl portion, sample #2, 500 μl of a 0.2 mg/ml solution of GT30 was added. The 0.2 mg/ml GT30 solution was prepared by diluting 20 μl of the 5 mg/ml stock GT30 to 500 μl. After this solution was added to sample #2, it was then sonicated for 30 minutes at 50% amplitude and centrifuged for 20 minutes at 12,000 rpm. This sample as well showed no deposited solid SWNTs after centrifugation. The third and final 500 μl portion, sample #3, had 500 μl of a 0.8 mg/ml GT30 solution added to it. This GT30 solution was prepared by diluting 80 μl of the stock 5 mg/ml GT30 to 500 μl. Sample #3 was also then sonicated for 30 minutes at 50% amplitude and centrifuged for 20 minutes at 12,000 rpm. Again, this sample yielded no solid SWNTs after the centrifuge step. The DNA concentration tests were repeated a second time but using more samples. D.I. water was used to prepare all GT30 solutions and for dilutions. The original

30 30 sample was prepared by diluting 45 μl of a stock solution GT30 (5 mg/ml) to 900 μl total. The solution was now 0.25 mg/ml in GT30. This solution was added to ~ 1 mg of raw SWNTs. The sample was left to sit 30 minutes before sonicating for 1 hour at 75 % amplitude. The sample was then centrifuged at 12,000 rpm for 1 hour. The sample was then divided into five 150 μl portions. All of these portions had various concentrations of GT30 added to form a total volume of 250 μl, followed by sonication for 30 minutes at 50% amplitude. Based on the previous concentration tests, these samples were not centrifuged after their second sonication step. To the first portion, sample A, was added 19 μl of stock GT30 diluted to 100 μl. Sample A was two times as concentrated in DNA as the original (the 0.25 mg/ml sample). For the second portion, sample B, 46 μl of stock GT 30 diluted to 100 μl was added to create a sample four times as concentrated as the original. To the third sample, sample C, was added 73 μl of stock GT30 diluted to 100 μl. Sample C was six times as concentrated as the original. The fourth portion, sample D, was to be eight times as concentrated as the original. To this sample was added 100 μl of stock GT30 with no dilution. Lastly, the fifth portion was just 150 μl of the original sample diluted to a total volume of 250 μl. One last set of tests was performed this time using GT20 instead of GT30. The original sample was prepared from a 10 mg/ml GT20 stock solution. 30 µl of the stock solution was diluted to 1500 µl to form a 0.2 mg/ml solution. To this solution was added ~ 0.5 mg of raw SWNTs. The sample was sonicated 1 hour at 75% amplitude and centrifuged at 13,000 rpm for 45 minutes. The sample was then divided into six 250 µl portions. The GT20 concentration was then varied from 0.2 mg/ml up to 2 mg/ml. The

31 31 first portion was prepared as 0.4 mg/ml by adding 250 µl of a 0.6 mg/ml GT20 solution (prepared from the stock solution). To the second portion was added 250 µl of a 1.4 mg/ml GT20 solution to prepare a 0.8 mg/ml sample. The third portion was prepared as 1.2 mg/ml by adding 250 µl of a 2.2 mg/ml GT20 solution. The fourth solution was prepared as a 1.6 mg/ml solution by adding 250 µl of a 3 mg/ml GT20 solution. To the fifth portion was added 250 µl of a 3.8 mg/ml GT20 solution to prepare a 2 mg/ml sample. After the addition of new DNA solutions to each portion, they were then sonicated at 50% amplitude for 30 minutes. The sixth and last portion was unaltered to be measured as the original sample concentration of 0.2 mg/ml SWNT Concentration Experiments. These samples were prepared in a similar manner to those from the DNA concentration tests except that the SWNT concentration was varied instead of the DNA concentration. For the first SWNT concentration tests, three different types of DNA were used (GT10, GT20, and GT30). These three DNA solutions were prepared by diluting 100 μl of a 5 mg/ml stock solution to 1000 μl. This resulted in a 0.5 mg/ml solution. Each of these solutions, GT10, GT20, and GT30, were added to ~ 1 mg of unprocessed, or raw SWNTs. The raw SWNTs are lightweight and tend to stick to the weighing paper therefore making it difficult to accurately weigh them. The result was that even though each of these three samples was supposed to contain 1 mg of SWNTs, the SWNT concentration varied in each sample. After adding each DNA solution to the SWNTs, the samples were sonicated for 1 hour at 75% amplitude and then centrifuged for 1 hour at 12,000 rpm.

32 32 For the second set of SWNT concentration tests, only GT30 was used. A solution was prepared by diluting 151 μl of stock 5 mg/ml GT30 to 1510 μl to form a 0.5 mg/ml solution. This sample was divided into three 500 μl portions and added to different concentrations of raw SWNTs. The first portion was added to ~ 0.5 mg of SWNTs, the second portion to ~ 1.5 mg SWNTs, and the third portion to ~ 2.5 mg of SWNTs. The samples were left to soak for 30 minutes before sonicating. Each sample was then sonicated for 1 hour at 75% amplitude and centrifuged at 12,000 rpm for 1 hour. Effect of DNA Base Pair Length Experiments. The effect of ssdna base pair length was examined by preparing DNA solutions with the same concentration (in mass/volume) and adding them to the same mass of SWNTs. Pure, or processed SWNTs were used for this experiment because they do not stick to weighing paper like raw SWNTs and are not as fluffy, thus making them easier to weigh out accurately. 0.3 mg/ml solutions of GT10, GT20, and GT30 were prepared. The stock solutions for GT10 and GT30 were 5 mg/ml, so 36 µl of each stock solution was diluted to 600 µl total with D.I. water. The GT20 stock solution was 10 mg/ml, so 18 µl of this stock solution were diluted to 600 µl total. The DNA solutions were then measured using an ultraviolet-visible spectrometer to determine their exact concentrations of the ssdna strands (in molarity) using Beer s Law. The ratios of the concentrations for GT10, GT20, and GT30 were 3:2:1 respectively. Next, 1 mg of processed SWNTs was weighed out and placed into a test tube for each of the three samples. Each DNA solution was added to the SWNTs and left to soak for 30 minutes. The samples were then sonicated at 75% amplitude for 1 hour followed by centrifugation at 12,500 rpm for 1 hour.

33 33 SDS and SDBS Samples. SDS and SDBS were also used as surfactants to disperse the SWNTs. The SDS and SDBS solutions were prepared by creating a 1% by weight surfactant solution. To make the 1% by weight solution, 20 mg of the surfactant (either SDS or SDBS) was added to 2 ml of D.I. water. This solution was then added to ~ 1 mg of either raw or processed SWNTs. The samples were sonicated for 1 hour at 75% amplitude and then centrifuged at 12,000 rpm for 1 hour. 2.2 Apparatus, Instruments, and Parameters Sonicator. For all sonication steps in SWNT suspended sample preparations, a Vibra Cell sonicator model VCX130 (net power output: 130 W, frequency: 20 khz) from Sonics and Materials, Inc. was used. Samples were placed in glass test tubes and submerged into an ice water bath. The sonicator probe was placed into the solution in the test tube being careful not to touch the probe to the test tube sidewalls or bottom. All sonication was performed between 50-75% amplitude. The apparatus set up is shown in Figure 2.1 below.

34 34 Figure 2.1: Sonicator apparatus Near-Infrared Spectrometer. A model NS1 NanoSpectralyzer TM from Applied NanoFluorescence was used to measure the absorbance and fluorescence spectra in the NIR of the ssdna/swnt samples. The fluorescence spectra were measured at excitation wavelengths of 658 nm and 785 nm. The quartz cuvettes for this instrument had a rectangular inner chamber. The path length was 10 mm in one direction and 4 mm in the perpendicular direction. The 10 mm path length was used for absorption measurements while the 4 mm path length was used for fluorescence. A minimum volume of 200 μl of solution was required to obtain an accurate measurement. The parameters for the NIR spectrometer were as follows: NIR fluorescence and absorption averaging set at 10 for all measurements, NIR fluorescence and absorption integrations set at 150 for all measurements.

35 35 Ultraviolet-Visible Spectrometer. An ultraviolet-visible (UV-vis) spectrometer from Agilent Technologies model HP8453 was used to measure the absorbance spectra of ssdna from the ssdna/swnt samples. The wavelength range for the spectrometer was nm. Cuvettes for this instrument were made of quartz with a 1 mm (0.1 cm) path length. Samples to be measured in the spectrometer were separated using a Microcon YM-100 Centrifugal Filter Unit from Millipore. The filter material was regenerated cellulose with a filtration area of 0.32 cm 2 and a diameter of 12.3 mm. ph Meter. To test the ph of various ssdna/swnt samples, a VWR symphony ph meter model SP70P was used. It contained a Ag/AgCl internal reference electrode. 2.3 Materials Used ssdna. All three types of ssdna (GT10, GT20, and GT30) were purchased from Integrated DNA Technologies, Inc. The GT30 DNA was modified using an ionexchange HPLC purification method, while the GT10 and GT 20 were prepared using a standard desalting method. The ssdna samples were purchased in solid form and were later diluted to prepare stock solutions. SWNTs. Both the unprocessed/raw and processed/pure HiPCo SWNTs were purchased from Carbon Nanotechnologies Incorporated. The raw tubes had not undergone any post processing and contained <35wt% ash content. Processed tubes were purified to remove large catalyst particles and contained <15wt% ash content. SDS and SDBS. Both SDS and SDBS solids were purchased from Sigma Aldrich and used without further purification.

36 Water. For all dilutions in these experiments, deionized (18.2 MΩ cm) water was 36 used. 2.4 Determination of DNA Concentration The molar concentration of DNA suspended SWNT samples was determined using data collected from an ultraviolet-visible (UV-vis) spectrometer. The absorbance spectrum of ssdna can be observed in this region. Three types of ssdna were used for this thesis: GT10, GT20, and GT30. The extinction coefficients at 260 nm for GT10, GT20, and GT30 are 189,200 L/(mole cm), 377,200 L/(mole cm), and 565,200 L/(mole cm) respectively. After measuring the absorbance of the ssdna solution, the Beer-Lambert Law 61 was used to calculate the molar concentration by: A = εbc (3) where A is the measured absorbance at 260 nm, ε is the extinction coefficient, b is a path length of 0.1 cm, and c is the concentration. 2.5 Determination of the Equilibrium Constant for DNA Adsorption/Desorption An equilibrium constant for the adsorption/desorption of DNA can be calculated using the Langmuir model where three approximations are employed: 1. Adsorption is complete once monolayer coverage has been reached, 2. All adsorption sites are equivalent and the surface is uniform, 3. The occupancy state of the adsorption site will not affect the probability of adsorption of desorption for adjacent sites. 62 In the Langmuir model, adsorption is described by:

37 37 R + M RM (4) ka kd In equation (4), R is the reagent DNA, M is unoccupied adsorption sites on the SWNT, RM is occupied adsorption sites on the SWNT, k a is the rate constant for adsorption, and k d is the rate constant for desorption. An adsorption site on the SWNT is defined as the area covered by the adsorption of a single GT strand. The binding constant could thus be written as: [ SWNTo / DNAb ] K = (5) [ SWNT ][ DNA ] Here, [SWNT o /DNA b ] is the concentration of the occupied sites, [SWNT u ] is the concentration of unoccupied sites, and [DNA u ] is the unbound or desorbed DNA concentration. At equilibrium, the change in fractional coverage with time is equal to zero so one could write: u u KP θ = (6) KP +1 where K is the equilibrium constant defined as k a /k d, P is reagent pressure, and θ is the surface coverage defined as sites occupied/total possible sites, or n/n m. However, in the experiments with DNA and SWNTs, the factor of pressure, P, is replaced by desorbed, or unbound, DNA concentration defined as C. Rearrangement of equation (5) into its linear form leads to: 1 n = + Kn (7) m C n m By plotting 1/n versus 1/C, the calculated slope will be equal to 1/Kn m. Using the value of 1/n m determined from the y-intercept, K can be calculated.

38 38 CHAPTER 3: The Effect of Varying DNA or SWNT Concentration on ssdna/swnt Complexes 3.1 Choice of SWNT Material SWNTs suspended with SDS and SDBS were examined for comparison to the DNA/SWNT samples. The samples were made with both processed and raw SWNTs. The absorbance and fluorescence spectra for these samples are shown in Figures Figure 3.1: NIR absorbance spectra of SDBS and SDS with raw and processed SWNTs

39 39 Figure 3.2: 658 nm excitation fluorescence spectra of SDBS and SDS with raw and processed SWNTs Figure 3.3: 785 nm excitation fluorescence spectra of SDBS and SDS with raw and processed SWNTs

40 40 Based on the SDS/SDBS results, raw SWNTs showed greater fluorescence intensity. This is likely to be related to the fact that the processed material was purified by using an acid to remove catalyst particles. Submitting the SWNTs to a purification treatment with strong oxidizing acids would partially oxidize the nanotubes. Oxidation of the SWNTs leads to a decrease in fluorescence intensity which is what was observed when comparing the fluorescence spectra of the processed and raw SWNTs. Therefore, raw SWNTs were chosen to prepare the DNA samples for the concentration experiments. 3.2 Effect of DNA Concentration on Absorbance and Fluorescence Spectra The NIR absorbance and fluorescence spectra for the first DNA concentration experiments are shown below. GT30 was chosen to prepare the samples due to other results (see chapter 4) which showed GT30 to have the greatest fluorescence intensity. The absorbance spectra of the three initial unaltered ssdna/swnt samples (all with starting concentration of 0.2 mg/ml GT30) are shown in Figure 3.4. The spectra show that the initial SWNT concentration for the three samples is the same.

41 41 Figure 3.4: Initial NIR absorbance spectra of samples #1-3 for the first GT30 concentration experiments before DNA concentration alterations After alterations in DNA concentration, the absorbance spectrum for each sample was remeasured and is shown in Figure 3.5. The spectrum shows the initial concentration of 0.2 mg/ml GT30 before alterations, sample #1 which was diluted by two, sample #2 which is two times the GT30 concentration of the diluted sample, and sample #3 which is five times the GT30 concentration of the diluted sample.

42 42 Figure 3.5: NIR absorbance after GT30 concentration alterations The fluorescence spectrum measured at 658 nm excitation for samples #1-3 and the original sample is shown in Figure 3.6. Figure 3.6: 658 nm excitation fluorescence spectra after GT30 concentration alterations

43 43 These samples were measured in a UV-vis spectrometer to determine the concentration of DNA. The total DNA concentration of each sample was measured first. The samples were then placed into a microfilter and centrifuged to separate the unbound DNA from the dispersed SWNTs. The microfilter s pores were small enough to allow free DNA to pass through, but not the SWNTs. The bound DNA concentration was calculated by taking the difference between the total DNA concentration and the unbound DNA concentration. The DNA concentration results are listed in Table 1. Table 1. Samples #1-3 DNA concentrations from UV-vis measurements Total µm Unbound µm Bound µm Sample Sample Sample These experiments revealed that by increasing the DNA concentration, the fluorescence intensity increased. The absorbance spectra of the original samples (before varying the DNA concentration) were all similar thus the samples all contained a similar concentration of SWNTs. Figure 3.5 showed that the absorbance peaks for each sample after varying the DNA concentration were similar. The measurement of the DNA molar concentration using UV-vis spectroscopy revealed that as the total DNA concentration increased, the DNA concentration bound to the SWNTs also increased. This agreed with

44 44 the increases observed in the fluorescence spectra for these samples; as there is an increase in DNA bound to the SWNT, there is an increase in fluorescence intensity. The second DNA concentration experiments were also prepared with GT30 but with more samples (see chapter 2). The NIR absorbance spectra for all samples are shown in Figure 3.7. Figure 3.7: NIR absorbance spectra of all samples for the second GT30 concentration experiments The 658 nm and 785 nm excitation fluorescence spectra for these samples are shown in Figures 3.8 and 3.9.

45 45 Figure 3.8: 658 nm excitation fluorescence spectra for the second GT30 concentration experiments Figure 3.9: 785 nm excitation fluorescence spectra for the second GT30 concentration experiments