John Tumbleston
Ph.D.
In this thesis, the groundwork is established for a new type of bulk heterojunction (BHJ) organic solar cell geometry that has photonic crystal (PC) photoactive layers. This design is motivated by the need to improve light absorption without increasing active layer thickness, which for many BHJ systems, degrades electrical performance. It is demonstrated that with the right choice of materials and cell dimensions, quasiguided or resonant modes are excited near the band edge of a variety of BHJ blends to enhance absorption. Resonant modes are predicted by first developing a scattering matrix optical model and then observed in wavelength-, polarization-, and angular-dependent reflection and photocurrent measurements. PC cells are fabricated using a facile nanopatterning technique, where highly ordered arrays of submicron features are constructed over large areas in a single step. Optical and electrical function of this new cell architecture is fully explored in this thesis. Through optical measurements and modeling, PC devices show clear enhancements in light absorption. On the other hand, the impact of the nonplanar geometry on electrical performance is not as easily deduced due to the multitude of electrical processes that lead to photocurrent generation. First, the electrical properties of the electron transporting layer that interfaces with the BHJ nanopattern and provides optical contrast in the PC greatly affect parasitic resistances in the solar cell. By including resistance losses in a drift/diffusion numerical model that describes electrical performance, it is shown that these losses greatly influence fundamental steps leading to photocurrent generation. This is confirmed with experiment by comparing two BHJ material systems that have different affinities for exciton separation. Second, significant levels of free carrier recombination are predicted by the electro-optical model due to the relatively long transport paths in the nanopattern features. To test this prediction, an experimental technique is developed to measure the transport lengths of photogenerated electrons and holes in BHJ solar cells. It is found that transport lengths of positive and negative carriers are mismatched and helps explain both PC electrical performance and recent conflicting results of planar BHJ solar cells in the literature.
Dissertation
Download here (PDF)Rudresh Ghosh
Ph.D.
Thin films are widely used in various applications, including but not limited to simple reflective coatings for mirrors, electrodes for lithium batteries, conducting substrates for electronic circuits, gas sensors and solar cells. As the scope of their applications has widened over the years so has the need to obtain different structural motifs for thin films. A large variety of fabrication techniques are commonly employed to obtain these structures. Pulsed laser deposition (PLD) can be used to obtain films varying from extremely compact and only a few angstroms thick to micron thick porous structures. In this dissertation I introduce a model for predicting different structures as a function of laser parameters and deposition environments in a pulsed laser deposition system. This is followed by a comparison of simulated and experimentally obtained structures. I then use this model to obtain tailored structures suited for individual applications. One of the unique structures obtained using the PLD consists of vertically-aligned structures with nanoparticles as their building blocks. I investigate the superiority of this unique structure over random nanoparticle networks as photoanodes for titanium dioxide (TiO 2 )-based dye-sensitized solar cells (DSSC). UV-Vis studies show that there is a 1.4 x enhancement of surface area for PLD-TiO2 photoanodes compared to the best sol-gel films. PLD-TiO2 incident photon to current efficiency (IPCE) values are comparable to 3 x thicker sol-gel films and nearly 92% absorbed photon to current efficiency (APCE) values have been observed for optimized structures. I also examine the suitability of PLD-synthesized niobium oxide (Nb 2 O5 ) and tantalum-doped titanium oxide (Ta: TiO2 ) as photoanode materials. For optimized PLD-Nb2 O5 based DSSCs IPCE values up to 40%, APCE values around 90% and power conversion efficiency of 2.41% were obtained. DSSCs made of PLD-Ta:TiO2 show enhanced photocurrents as well photo efficiency over those based on PLD-TiO2 . I show that this improvement is due to the slower recombination rates because of the presence of Ta5+ . I also show the applicability of other structures obtained using the PLD. I use very compact thin films for obtaining the band edges of different potential candidates suitable for photoanodes of dye sensitized photoelectrochemical cells (DSPEC). This is done by a combination of x-ray and ultraviolet photoemission spectroscopy (XPS-UPS) and inverse photoemission spectroscopy (IPS). I also use porous tungsten oxide (WO 3 PLD films as H sensors. I determine the optical constants for WO3 in the colored and bleached states and use them to correlate the gasochromic and electrochromic behavior of WO3 . I also introduce polymer encapsulation of WO3 thin films to protect them from aging.
Dissertation
Download here (PDF)Emily Ray
Ph.D.
The incidence of light with momentum components outside the light cone on the surface of a negative permittivity material results in the excitation of a surface plasmon polariton and the enhancement of the incident signal when there is momentum and energy conservation. This process has an impact across many fields including imaging, optical computing, signaling, and photovoltaic devices, among others. I examine the role and tunability of light-surface plasmon interactions in several applications. I demonstrate a tuned metamaterial grating system that allows the signal from evanescent waves to be detected in the far field in the visible regime. I fabricate a metamaterial that is tuned to support surface plasmons that couple to visible light across a wide range of wavelengths. I characterize the plasmonic response through a simple technique wherein a the reflection from a subwavelength grating on a metamaterial indicates surface plasmon coupling when its intensity dips. With this I demonstrate that the reflection trends match well with simulation, indicating that coupling of light to surface plasmons occurs at the expected crossing points. The strength of coupling (denoted by the drop in reflection) however, is less than expected. Transmission measurements reveal a depolarizing effect that accounts for the decrease in evanescent light enhancement by the surface plasmons and is due to the surface roughness at the interfaces between the metal and dielectric. I also use a tuned metamaterial perforated with a subwavelength array of circular apertures to exhibit extraordinary transmission in the visible. I compare the transmission of the metamaterial to that of a thin film of Ag with equivalent thickness that has fewer plasmon modes and a resonance position in the UV to find that for 400 nm, both thin films exhibit a transmission minimum at 650 nm. Both film spectra have plasmon-aided extraordinary transmission peaks where there is momentum and energy conservation between the evanescent waves produced by the diffraction grating and the surface plasmons in the metamaterial at 570 nm and 700 nm. Here, more light is transmitted through the holes than is incident on them. Furthermore, I see that the surface plasmon generation by the holes themselves is negligible compared to those generated by the surface plasmon. I then explore the mechanism of increased external quantum efficiency with plasmonic structures in organic bulk heterojunction solar cells. I build an inverted bulk heterojunction solar cell with a Ag back cathode patterned with a diffraction grating to separate the possible mechanisms of enhanced current production. I-V curves from the patterned cell signify a total efficiency 3 times larger than a flat reference cell and the incident photon to electron conversion efficiency exhibits peaks where there is an increase in interaction path length of the incident light in the active layer due to scattering and none at the surface plasmon resonance position leading to the conclusion that the increase in performance is due to scattering and not plasmon generation.
Dissertation
Download here (PDF)Kristen Alexander
Ph.D.
Surface enhanced Raman spectroscopy (SERS) was originally discovered in the 1970s with the observation that organic molecules adsorbed onto a metal surface exhibit greatly enhanced Raman scattered light intensities when illuminated with a laser source. Enhancements of approximately 106 over regular Raman scattering have been commonly observed and proposed applications of SERS-active sensors exist over a wide range of fields, including chemical analysis, healthcare, food safety and national security, spurring an intense scientific interest in the area. More recently, observations of single- molecule SERS have demonstrated enhancement factors greater than 1013 at random ‘hot spots’, but so far, these enhancement factors are poorly understood due to lack of reproducibility and lack of methodical characterization of such spots. Theoretical calculations have shown that the dominant field enhancements are specifically localized in the crevices between metal nanoparticles and are strongly dependent on particle morphology, excitation wavelength and, perhaps above all, particle-particle coupling. The focus of this thesis is to address experimentally theoretical predictions by fabricating SERS configurations and to make definitive measurements of the SERS magnitude at interparticle hot spots. In this work, metal nanoparticles have been directed to form ordered arrays exclusively of metal nanoparticle dimers with control over orientation, size and interparticle spacing. In order to achieve unprecedented control of the material and geometric variables, elastomeric substrates were used to change particle-particle distance while holding all other physical parameters constant. This fundamental new approach to hot spot creation has opened doors to a new family of SERS substrates, where the turning on/off of a hot spot is as easy as flipping a switch. Most recently, I have demonstrated the feasibility of this approach with long nanorods that show an outstanding theoretical SERS match with the characteristic polarization dependence expected of such nanostructures. Additionally, this thesis demonstrates the feasibility of creating SERS-active dimers over a large area using a capillary force deposition technique which has further been used to compare the SERS enhancement factors derived from dimers to those of longer linear nanoparticle chains, ultimately demonstrating the practicality of the dimer configuration over more complex nanostructures.
