Nanotechnology Research Clay Harris and Rob Dolan
Synchrotron IR Microspectrometry of Nanotubes and Nanowires       Applications of carbon nanotubes in modern science are both broad and far-reaching. Almost all fields of science stand to benefit from nanotechnology research. The purification of carbon nanotubes is of particular interest as a means to distribute chemically similar nanotubes. These structurally identical nanotubes can be bound to pharmaceuticals to be distributed in a targeted fashion in the body. As the nanotubes pass through tissues they transport the bound pharmaceutical as well, and are released from the nanotube into the body at specific locations and times. However, structurally different nanotubes have the potential to fail to release the pharmaceutical at the correct time, or to fail to target the proper body tissue. To ensure release and targeting, nanotubes must undergo a variety of purification processes to assure obtaining as many nearly identical nanotubes as possible.       While these purified nanotubes are chemically detectable, preliminary studies at the NSLS have suggested that these nanotubes may not be as pure as originally thought. Using the synchrotron light source, samples of both unpurified and purified nanotubes were examined in the infrared (IR) spectral region. Variations in spectral data clusters as the sample was translated spatially suggest the presence of many different components. This raises several questions about the purification process, and the reliability of chemical detection of varying degrees of purification. How can it be certain that one has made a desired nanotube? How does one located a desired nanotube in a complex mixture? And how can one determine the level of variation in a purified sample of nanotubes?	Image of nanotubes to be scanned Click image for more images
Data Analysis and Interpretation       By reducing sample data to their principal components, plotting those components on a quantile-quantile (QQ) plot, and fitting multiple lines to the QQ graph, it is possible to break the data set into its primary data clusters. Calculating distances from the center of these clusters to all other points in the data set yield chemical maps corresponding to the concentration of each unique cluster. A current goal is to analyze these clusters and locate physical sites where variations in the data occur. However, due to the volume of data obtained at the NSLS, and the small number of people available to interpret the data, analysis takes some time. The spatial resolution of conventional synchrotron IR microspectrometry (µm scale) in comparison to the size of the nanotubes (nm scale), complicates spectral data for nanotube analysis. Near-field techniques (see below) may soon help in this regard.
		Current Research and Solutions       Hyperspectral imaging tends to be data rich and information poor. Data volume has become a bottleneck in many areas of research, increasing the time it takes to both analyze and interpret data. Several methods are being developed that preprocess data and reduce the time it takes to interpret the data. These methods have been termed integrated computational imaging (ICI), hyperspectral integrated computational imaging (HICI), and integrated sensing and processing (ISP).       Using ICI, computers both capture and analyze an image during the detection process. ICI reduces the data load on the computer by downloading some of the computations to the image sensor itself (Lodder, 2004). In HICI, a spectral image is encoded at many different wavelengths simultaneously. ICI and HICI incorporate Felgett's advantage, or multiplex advantage, in which collecting data at all wavelengths at once reduces the collection time required for obtaining the same signal-to-noise ratio when compared to a dispersive technique that collects only one wavelength at a time. HICI methods for collecting data are advantageous because they compress a high volume of spectral information into a single image, which saves time and decreases the final volume of data collected.       Through the ISP process, the steps by which signals are detected and interpreted is being completely reorganized, reworked, and optimized. Originally began by Defense Advanced Research Projects Agency (DARPA), the project seeks to craft new sensor systems that treat the total structure as a single end-to-end process that can be optimized globally. ISP has created new mathematical algorithms and produced new statistical methods by which both the speed of sensing and amount of data are increasing, and the information gained is more meaningful. By developing and implementing new ICI and ISP methods and instruments, the bandwidth, volume of data to be interpreted, and amount of time taken to interpret data are each being significantly reduced, while maintaining a high level of meaningful information (Lodder, 2004).
Nanotube data clusters (click for larger image)
Molecular Factor Computing Near-Field Scanning Optical Microscopy (MFC-NSOM)       Conventional NIR microscopes limit spatial resolution to about lambda/2, making the best NIR spatial resolution 0.5-1 µm (Lodder, 2005). Details become lost after light travels more than a wavelength from a feature smaller than lambda/2, but with smaller wavelengths higher resolutions can be gained. To overcome this limitation, our research is looking into the development of near-field HICI devices.       The near-field effect is achieved when a subwavelength-sized optical element is placed at a subwavelength distance between the optical element and the sample. Because the light travels only over a short distance, diffraction and far-field characteristics are eliminated. Near-field techniques can enable a resolution of 80-120 nm for a light source of 1200 nm (Lodder, 2005).
		MFC-NSOM using the NSLS as a light source (click for larger image)
By developing instruments capable of realizing sub-micron spatial resolutions, the boundaries of NIR microscopy can be expanded. NSOM makes possible the in vivo imaging of cell membranes and other biological systems, the study of single molecules, and the collection of new information on the orientation and behavior of molecules. Building on this principle by the use of MFC, in which molecular filters are used to encode specific wavelengths of light corresponding to the analyte of interest, speeds near-field spectrometric imaging sufficiently to make it practical in biological systems..       Using an MFC NIR-NSOM instrument, samples of both purified and unpurified carbon nanotubes will be analyzed to create an optical method to quickly determine if the desired structures are being produced, and where. Materials are selected for use as molecular filters by comparing their transmission spectra to the factor loadings correlating to the analyte of interest. This produces images whose colors correlate to the concentration of the analyte. For example, by searching for key differences between unpurified and purified nanotube spectra, it is possible to find a quick and easy means of identifying unpurified elements of a purified sample. Because the light passes through a limited number of nondispersive optics, the Jacquinot advantage (throughput advantage) also applies. This means that a higher signal-to-noise ratio is maintained when compared to more traditional NIR Methods.
Calculating MFC Loadings (click for larger image)
Binding an Anti-CRP molecule to a Nanowire       Recent studies have shown that binding a receptor to a nanotube, and then that nanotube/receptor to a protein, can affect nanotube conductivity. Such protein-bound nanowires can be used as sensors to search for specific proteins in vitro or even in vivo. By bonding an anti-CRP (C-reactive protein) to a nanowire, we hope to show that a CRP "sniffing" nanowire sensor can be constructed. CRP (which binds to anti-CRP) indicates inflammation that can lead to matrix protein degradation and the rupture of vulnerable atherosclerotic plaque. If CRP is present, it will bind to the anti-CRP on the sensor, causing an electrical current of the nanowire to change. By searching for CRP above and below vulnerable atherosclerotic plaque in a flowing blood stream, it may be possible to locate trouble areas in which plaques are likely to rupture even in the presence of arthritis (which dramatically elevates CRP) in the patient.       Attaching anti-CRP to the nanowire can be complicated. The streptavidin process, which results in the formation of a protein layer over the surface of the nanotubes, offers a potential means for building the sensor. Biotinylated anti-CRP is allowed to bind to the streptavidin layer, which anchors the sensing protein to the nanotube surface (Lenihan et al., 2004). Simple electronics monitoring current in the nanowire permit the detection of CRP.
Nanotechnology at ASRG       By developing an MFC-NSOM, and combining it with the concept of sensing arrays utilizing ISP methodology capable of ICI and HICI, sensors will be created that not only obtain large quantities of data, but also analyze that data immediately. The development of a nanowire "sniffer" capable of measuring CRP in vulnerable plaque will provide a method for the identification of patients susceptible to sudden cardiac death. Further research and development of these analytical tools and methods may reduce the number of deaths due to vulnerable plaque.
References: Lenihan, J. S.; Gavalas, V. G.; Wang, J.; Andrews, R.; Bachas, L. G. Protein Immobilization on Carbon Nanotubes Through a Molecular Adapter. J. Nanosci. Nanotech. 2004. Vol. 4, pp 600-604(5). Lodder, R. A.; Harris, J. C. NIR-NSOM Using Hyperspectral Integrated Computational Imaging. 2005. Cassis, L. A.; Urbas, A.; Lodder, R. A.; Hyperspectral integrated computational imaging. Anal. Bioana. Chem. 2004.