Thin Film Solar Cells: State-Of-the-Art and the Future
Low temperature (cold) plasmas are indispensible in microelectronic device fabrication. Plasma etching, in particular, is used exclusively to transfer lithographically-defined patterns in polymer films into the underlying layer. Etching occurs when energetic ions impinge on the surface, promoting reactions with etchants to form volatile species. As thin film dimensions continue to shrink to the nm scale, etching selectivity becomes critically important. Selectivity may be achieved by using a nearly monoenergetic ion energy distribution (IED), with the peak ion energy placed between the threshold for etching one material vs. another. For example, atomic layer etching of Si may be achieved with an ion energy between the thresholds for chemical sputtering (Si with a chemisorbed chlorine layer) and physical sputtering (pure silicon).
A novel plasma reactor was developed to control the IEDs bombarding the substrate. A pulsed plasma was employed (square wave power modulation), and a DC bias was applied in the afterglow (during power off) on an electrode in contact with the plasma, resulting in a double-peaked IED. The mean energies of the two peaks, as well as the peak separation, could be varied at will by adjusting the applied DC bias and/or the discharge pressure. The full width at half maximum (FWHM) of the sharp peak corresponding to the DC bias diminished with decreasing electron temperature. The FWHM was adjusted by varying the time window in the afterglow during which DC bias was applied, and by changing the kind of the mostly noble gas feed to the plasma.
The ability to create monoenergetic IEDs enabled the study of Si etching near the threshold ion energy. Surprisingly, etching of p-type Si in Ar/Cl2 plasmas at ambient temperature was found to occur even for energies below the ion-assisted etching threshold. A series of well-defined experiments was employed to conclude that this ―sub-threshold‖ etching was due to photons from the plasma. The photon-assisted etching manifested itself as a constant background for ion energies below threshold. The importance of light from the plasma to etching was not realized thus far because it was masked by ion-assisted etching with high energy ions (100s of eV). The photonassisted etch rate becomes significant, compared to ion-assisted etching, for processes that require low (10s eV) ion energies to achieve high selectivity and low damage, such as atomic layer etching. Work supported by the DoE Plasma Science Center and NSF.
Department of Chemical and Biomolecular Engineering, University of Houston, Houston, TX
A one-dimensional two-phase model with global rate expressions for various reactions is used to study the performance of combined LNT-SCR catalyst systems at different temperatures. Pt/BaO/Al2O3 catalyst is used for NOx storage and reduction in LNT. Our model considers two types of NOx storage sites designated as ‗fast‘ and ‗slow‘ sites depending upon their proximity to Pt. LNT model includes reactions that describe NO oxidation, storage of NO2 in the form of nitrates during lean storage cycle and reduction to N2 in the subsequent rich cycle with H2 as a reductant. Cu-ZSM5 catalyst is used to selectively reduce effluent NOx in the SCR to form N2 and H2O. Model predicts the spatio-temporal concentration profiles of reactants and products inside the monolith channel, during periodic lean and rich cycling operation. It has been found that series arrangement of alternate LNT and SCR catalyst, by dividing them into equal halves, improves NOx conversion. For fixed total catalyst length as the number of LNT-SCR catalyst zones increase, NOx conversion increases and reaches a limit which is theoretically equivalent to that of mixed catalyst. Spatio-temporal profiles indicate complete storage of NH3 on SCR catalyst which can be efficiently used to reduce NOx slipping out of LNT. Various catalyst arrangements, the effect of different catalyst loading and cycle time are investigated in the present work.
Many bioanalytical technologies employ a label, typically a dye, enzyme or fluor, to signal the presence of analyte. While these labels are common and well-developed, such systems can suffer from low signal strength and label instability, and elaborate instrumentation is often required for detection. This work uses micron-sized magnetic sample-prep particles as light-blocking labels in optical, semi-homogeneous, diagnostic immunoassays based on microfabricated linear retroreflectors. Retroreflectors return light directly to its source and are readily detectable with inexpensive optics. When the target is present, magnetic sample-prep particles decorated with antitarget antibodies can assemble on the surface in front of the assay reflector and substantially reduce reflectivity. A difference imaging approach that can see single 1.0 um particles is used to detect and quantify signal intensity from each 1 mm x 1 mm array of retroreflectors. The magnetic properties of the particles are useful in sample preparation and concentration. The assay is implemented in a microfluidic format for Rickettsia conorii detection, using fluidic force discrimination to increase specificity.
1 Dept. of Chemical and Biomolecular Engineering, 2 Dept. of Mechanical Engineering, 3 Dept. of Chemistry, University of Houston, Houston, TX 77004, USA
Sickle cell anemia pathogenesis is evoked by polymerization of a mutant hemoglobin, HbS. To test the hypothesis of free heme effects on the kinetics of HbS polymerization, we design and assemble a microfluidics device with optional microheating. We will employ silicone oil to produce a train of droplets of HbS solution in a microchannel of size 5×40 μm2. The microheater system will locally heat the resulting picoliters droplets from 5 C to 25 C.
After tests with several polymers with different composition and chemical properties, we chose to build the microchannel from SU-8 – 3005. The protocol for microchannel preparation includes: starting with 3 ml of SU-8, we apply two-step spin coating with step 1, for 10 sec, with speed 500 rpm and acceleration 100 rpm/s, and step 2, for 30 sec, in which the speed is 2300 rpm and the acceleration—230 rpm/s; soft bake at 65 C for 1 min and 95 C for 10 min with ramp; cooling for 8 min; ultraviolet exposure time for 15 sec; post-exposure bake at 65 C for 1 min and 95 C for 8 min, with no ramp; cooling for 2 min; and, finally, development for 15 sec at ~5 C.
We arrived at this protocol after a lengthy optimization. We determine the dependence the thickness of polymer layer on a glass substrate on the rotation velocity and polymer volume. We optimize the conditions of baking, exposure and development, starting from parameters in the published literature. An ice bath is used to control the activity of developer solution.
We found that some salt solutions cause swelling of the SU-8 polymer and closure of the microchannel. Fortunately, tests revealed that hemoglobin solution as well as silicone oil do not act in this way.
We fabricated a thin film copper microheater by using an AZ 1500 photoresist mask followed by evaporation and deposition of metal. The quality of the photoresist mask was crucial for the quality of the microheater. The resistance of the microheater will be used to measure the temperature. For this, we found that this resistance depends linearly on temperature. Next, we used remote temperature sensing by an infrared camera to determine the dependence of the heater
temperature on the passing current.
To bond top and bottom glasses, which hold, respectively, the microchannel and microheater, we treat the SU-8 surface with oxygen plasma and bake in oven at 100C for 10 min.
A Diesel Particulate Filter (DPF) is used to remove particulate matter (PM) from the effluent of a diesel engine. A major technological challenge in the operation of this ceramic Cordierite filter is to prevent formation of local high temperatures that can melt the DPF or generate a thermal stress that may cause cracking. Most previous studies of the temperature rise during the DPF periodic regeneration (combustion of the deposited PM) considered cases in which the inlet velocity to all the parallel channels was uniform. A wide-angled cone (diffuser) is sometimes used to connect the diesel engine exhaust pipe to the DPF leading to a non-uniform velocity to the inlet channels, with the highest attained at the DPF center. We used a PM deposition and regeneration computational model to investigate the impact of the inlet cone on the DPF behavior under stationary feed conditions. The cone led to mal-distribution of the deposited PM, with the highest thickness in the DPF center. The highest regeneration temperature when using an inlet cone may be quite higher than when it is not used. Moreover, the inlet cone can lead to higher temperature gradients and the resulting thermal stresses may crack the ceramic support. The largest radial thermal gradients are encountered close to the wall in the downstream section of DPF, shortly after the temperature inside the filter reaches its peak. The cone leads to a slightly faster regeneration than when it is not used.
1Department of Chemical and Biomolecular Engineering, University of Houston, TX-77204
²Department of Mechanical Engineering, University of Houston, TX-77204
³Ford Motor Company, Dearborn, MI-48125
Traditionally, the three-way catalytic converter (TWC) is controlled via an inner-loop and outer-loop strategy using a downstream HEGO sensor and an upstream UEGO sensor. With this control structure, we rely on the HEGO voltage to indicate if the fractional oxygen storage level (FOS) is either unity (`bucket full’) or zero (`bucket empty’). However, if the FOS level of the catalyst could be measured or modeled, then a more pro-active TWC control strategy would be
In this work, we present a low-dimensional TWC model for control and diagnostics. The reduced order model retains the essential features of a TWC and provides high fidelity estimates of the fractional oxygen storage, yet is computationally efficient for real-time implementation. This simplified model was tested on multiple vehicles (P415 and S197) and with differently aged catalysts (full and partial volume). Finally, a novel outer loop control algorithm was developed around this model, where the fractional oxygen storage level (`bucket level’) is used for TWC controls and the total oxygen storage capacity (TOSC) (`bucket capacity’) is used for diagnostics. We also propose a catalyst aging model that can be used to update the oxygen storage capacity in real time so as to capture the change in the kinetic parameters with aging.
Department of Chemical & Biomolecular Engineering, University of Houston, Houston, TX, United States.
The selective catalytic reduction (SCR) of NOx with NH3 is considered to be a highly promising technique for the efficient reduction of highly detrimental NOx (to N2) emitted from diesel engine vehicles. Amongst the various catalysts available for SCR, Fe and Cu-based zeolite catalysts are considered to be highly stable and efficient towards NOx conversion over a broader temperature range. Cu-zeolite has been found to be a very good low temperatures (< 350 C) NOx conversion catalyst. However, Fe-zeolite has been found to be a better catalyst at higher temperatures (> 350 C).
This is a comprehensive experimental study coupled with the kinetic modeling of key SCR reactions. This study is focused on the determination of kinetics and mechanistic aspects of various SCR reactions. A detailed analysis of the extent of mass transfer limitations is provided. A systematic study of SCR reactions is carried out on a combined system of Fe and Cu-zeolite monolithic catalysts to determine if a high NOx conversion could be sustained over a wider temperature range than with individual Fe- and Cu-zeolite catalysts. Three configurations of combined Fe and Cu-zeolite catalysts were compared:
― “Sequential brick” catalyst comprising Fe-zeolite and Cu-zeolite monolith
― “Mixed washcoat” catalyst comprising a washcoat layer having equal mass fractions of Feand Cu-zeolites
― “Dual layer” catalyst comprising monolith coated with individual layers of Fe- and Cu zeolites of different thicknesses and mass fractions
The experiments included NO and NH3 oxidation, standard SCR (NO+NH3), fast SCR (NO+NO2+NH3), and NO2 SCR (NO2+NH3).All the reactions follow similar mechanism on both the Fe and Cu-zeolite catalysts. Monolith catalysts with different washcoat loadings, washcoat thicknesses and lengths indicate the presence of washcoat diffusion limitations at intermediate temperature range in all the SCR reactions. These diffusion limitations should be considered in any kinetic model and catalyst designs. A detailed analysis of the effect of temperature on the transitions between the controlling regimes (for various SCR reactions) is presented. Finally, amongst different combinations of Fe- and Cu-zeolite combined catalyst systems, a dual-layer catalyst system with a thin Fe-zeolite layer on top of a thick Cu-zeolite layer resulted in achieving high NOx removal efficiency over a wide temperature range.
1 Department of Chemical and Biomolecular Engingineering, University of Houston, Texas
2 Halliburton Energy Services, Duncan, Oklahoma
We present a novel semi-empirical rheological model to describe the effects of temperature, shear rate, pH and polymers/cross-linkers/additives on viscosity of gelling acids. We extend the twoscale continuum model (validated earlier for Newtonian fluids) to analyze the wormhole formation with gelling acids through single/dual carbonate cores. We also estimate the optimum injection rate for wormhole formation with gelling acids, develop scaling criteria for width and speed of propagation of gel and reaction fronts, and show that these strongly depends on rheological parameters. In addition, we identify the mechanism of flow diversion and show that the larger gel resistance in high-perm zone is the key to achieve uniform stimulation. Finally, we develop some guidelines for optimal uniform stimulation of carbonates with gelling acids.
Keywords: Rheology, Gelling Acids, Pattern Formation, Reactive Dissolution, Acidization, Carbonate Reservoir, Wormhole, Porous Media, Modeling and Simulations
Today‘s state-of-the art integrated circuits contain over 2 billion transistors over a very small area of few cm2 and the widths of the feature size are as small as 22 nm. This incredible innovation is made possible by plasma etching. Despite its success, plasma etching is not a well controlled process due to complex mix of neutral radicals, positive ions and UV photons interacting simultaneously to the substrate surface. One of the reason for process drifts in plasma etching is the loss of etchant (like Cl radicals in Cl2 plasma) to the chamber walls in many heterogeneous reactions
like atom recombination, formation of higher silicon-chlorides on SiOxCly deposited films during Si etching etc. Despite its importance, plasma surface interactions are very less understood.
Spinning wall method has been used to understand heterogeneous reactions on plasma rector chamber walls during and after silicon etching in Cl2 based inductively coupled plasma. In this method, a small cylindrical portion of the chamber wall is rotated and the surface is periodically exposed to the plasma and then to the differentially pumped diagnostic chamber housing either an Auger electron spectrometer for in-situ surface analysis or a quadruple mass spectrometer for line-ofsight desorption mass spectrometry. Reactants in the plasma impinge and stick on the rotating substrate, react and desorb over a time comparable to the substrate rotation period which can be varied from ~ 1 ms to 40 ms. Molecules desorbing from the spinning substrate are detected by lineof- sight mass spectrometry. Plasma reactor also houses a silicon electrode which can be etched in chlorine plasma by RF powering the electrode. During Si etching, etch products (SiClx, x = 1 – 3) deposit a dense amorphous silicon-chloride layer on the chamber walls which was characterized insitu by Auger electron spectroscopy. After stopping the etching of the silicon electrode, etch products continue to form on the walls due to the etching of SixCly deposited film resulting in a detection of mass spectrometer signals at m/e = 63 (SiCl+), m/e = 98 (SiCl2+), m/e = 133 (SiCl3+) and m/e = 170 (SiCl4+). Along with the etch products many heavy oxy-silicon-chlorides were also observed desorbing off the chamber walls. This interplay of chlorine plasma with SixCly coated chamber walls will be discussed. Also, effect of surface contaminants like ―Titanium‖ on delayed Langmuir-Hinshelwood (L-H) recombination probability of oxygen atoms in O2 ICP will be presented.
Hydraulic fracturing of horizontal wells is crucial for economic production of natural gas from unconventional reservoirs. For transverse fractures, existing approaches for making decisions on fracture spacing and length are based on sensitivity analysis using reservoir and/or fracture propagation simulators. The use of commercial simulators relies on assessment of various what-if scenarios, which are set up based on experience and data collected from completed wells in the same or similar reservoirs. Because of the large number of possible decision variables, the decision making process is complex, potentially leading to non-optimal. A computer-aided tool that could facilitate the process would allow engineers to efficiently converge to solutions that are at or close to the optimum.
The proposed approach relies on a novel efficient numerical optimization tool. The optimization objective function is Net Present Value (NPV) for horizontal wells drilled in the direction of the minimum horizontal stress, which allows multiple transverse fractures. Key decision variables are number of transverse fractures in a horizontal well; horizontal well length; proppant concentration; injection rate of fracturing fluid; injection time; fluid performance index and fluid consistency index. Key design constraints are fracture growth control, and geometric constraints. To set up the optimization, critical reservoir properties are assumed to be available from a variety of sources (e.g., seismic, prior production logs). The modified PKN model along with Carter‘s Equation for material balance is used for hydraulic fracturing propagation where fracture width, half length and net pressure are calculated. Productivity of multiple fractured horizontal wells is computed by using analytical equations. A robust optimization algorithm is used for finding the optimum parameters.
The capability and robustness of the proposed approach is demonstrated for a tight gas reservoir for which horizontal well geometry and its fracturing treatment is designed based on numerical simulations. A commercial simulator served as virtual reservoir with properties calibrated based on SPE case studies. The simulation show optimization can significantly increase NPV over other available solutions. Sensitivity analysis is done on various decision variables to show their significance in decision making for well stimulation. Also, significance of choosing optimum number of operating years is discussed in order to avoid missing optimum recovery. It is shown how important it is to find the optimum number of years for a particular formation which can give the maximum profit.
The proposed tool can provide a quantitative assessment of optimal horizontal well stimulations. It is envisioned to be used in an integrated decision support system that will help engineers to use a multitude of available data for development of unconventional gas resources.
Nanoporous zeolites are utilized in many industrial processes because of their thermal stability, tunable acidity, and shape-selectivity. Notably, MFI-type zeolite (termed ZSM-5) is active for various reactions, which include hydrocarbon isomerization, alkylation of hydrocarbons, and selective catalytic reduction (SCR) of NOx. The shape-selectivity of ZSM-5 is attributed to its 3-D network of interconnected 10-membered ring channels. Much attention has been given to the activity and identity of catalytic sites within the interior pores; however, active sites located on the external surfaces of zeolite catalysts may be equally, if not more important. Indeed, the synthesis of nanosized zeolite catalysts can dramatically increase the available external surface area for reactions. Differences between the inner and outer surface
activity of zeolite catalysts is not well understood, which necessitates increased research efforts to develop an improved fundamental knowledge of active sites and structure-function relationship. To this end, we are synthesizing core-shell MFI-type zeolites composed of an active (ZSM-5) core and a non-active shell of silicalite-1 (siliceous analogue of ZSM-5 with the same MFI structure). The silica shell ―passivates‖ the external active sites of the catalyst, which will allow us to perform systematic catalytic tests to distinguish the internal and external activity. The ZSM-5 catalysts synthesized in our lab are ~200 nm in size, which provides a large specific external surface area and relatively short internal diffusion path length. We have developed a technique for growing thin layers of silicalite-1 (5 to 10 nm) on ZSM-5 nanocrystals, which will
allow us to examine changes in reactivity without increasing mass transport limitations. Initiatives in catalyst synthesis and testing are combined with computational modeling using density functional theory (DFT) to calculate the stability of bulk and surface acidic sites and their activity for the NOx SCR reaction. Preliminary computational studies have focused on NH3 adsorption energies at two different ZSM-5 T-sites and future kinetic studies will identify the most active sites for NOx SCR using NH3 as the reducing agent. This unique combination of synthesis, testing and modeling represents a synergistic approach in rational catalyst design that allows us to develop structure-function relationships for the rational design of zeolite catalysts for NOx SCR.
Amphiphilic diblock copolymers, which form micelle structures in selective solvents, offer a great advantage of tunability in physical characteristics as compared to low molecular weight surfactants. Their micelles in aqueous solvents have been a subject of great interest in drug delivery applications for their high loading capacity and targeted drug delivery. The aim of this work is to understand the mechanisms governing the relaxation of diblock copolymer
micelles which have been perturbed from equilibrium, particularly the exchange of single chains that occurs as the system returns to the equilibrium state. The present work focuses on amphiphilic diblock copolymers containing poly(ethylene oxide) (hydrophilic) and polycaprolactone (hydrophobic) blocks, which spontaneously self-assemble into spherical micelles in water. Time-resolved small angle neutron scattering experiments are being used to probe the kinetics of exchange of single chains in selectively deuterated diblock copolymer micelles. The overall sizes of the micelles are characterized through dynamic light scattering experiments. Future experiments will focus on the effect of crystallization of the core block (polycaprolactone) on the micelle self-assembly kinetics.
Combinatorial design approaches to improve enzymatic or microbial production of a metabolite often rely on the throughput and/or selectivity of the screening system used. Regulatory proteins controlled by ―effector‖ molecules naturally couple molecular recognition to changes in gene expression, providing a platform for linking in vivo molecular synthesis to a readily detectable phenotype (e.g. GFP expression). We are developing customized molecular reporters by
engineering regulatory protein effector recognition. In addition to enabling high throughput screening, customized regulatory proteins are useful tools in metabolic engineering applications. We initially focused on the well-characterized regulatory protein AraC. Using fluorescence-activated cell sorting, libraries of >108 AraC variants are rapidly screened for
desired regulatory properties in the presence and absence of selected small molecules. This strategy led to the isolation of AraC variants that selectively report in vivo concentrations of the metabolites mevalonate and triacetic acid lactone (TAL). These reporters were subsequently used to screen for improved production of mevalonate and TAL by E. coli clones expressing
mutants of the respective heterologous biosynthesis pathways. The range of molecules accessible by variants of the AraC ligand binding pocket is now being explored. Other regulatory protein platforms are also being developed to further broaden our molecular reporting repertoire. For example the regulators TetR and ActR serve as platforms for detecting natural
and ―unnatural‖ products such as antibiotics. We aim to better understand which residue positions are most effective for altering ligand binding, resulting in a streamlined design process with optimized protein libraries.
Colloidal suspensions are ubiquitous in industrial applications, ranging from paints and coatings used in nanofabrication techniques to drilling fluids used in the upstream petroleum industry. Many of these applications involve flow of the suspensions in micro-scale geometries. While the flow properties of hard-sphere suspensions in micro-scale geometries have been studied extensively, the effects of inter-particle attraction on the confined flow behavior are not well understood. In this study, we are investigating the flow properties of colloidal suspensions with varying strengths of attraction in bulk and confined geometries. We used a slightly charged model system of poly(methylmethacrylate) spheres that are suspend in a refractive-index and density-matched solvent, and induce a controlled short-range inter-particle attraction by adding non-absorbing linear polystyrene. To characterize the bulk rheological properties of attractive
suspensions, we measure the viscosity as a function of shear rate using steady shear rheology. At low strengths of attraction, the suspensions exhibit shear thinning at low shear rates and shear thickening at high shear rates separated by a region where viscosity is independent of shear rate at intermediate shear rates. As the attraction strength is increased and the suspension approaches the gel state, the shear-thickening regime at high shear rates is not observed and the steady-shear viscosity was nearly constant even at the highest shear rates measured. To determine how confinement modifies the flow properties of attractive suspensions, we image the flow in microscale channels using confocal microscopy. We compute the flow velocity as a function of flow rate to determine the maximum shear rate across the micro-channel. We then observe the microscale structures that are formed during and after cessation of flow at different shear rates and compare to bulk rheological behavior. In future work, we correlate the micro-structures to flow rate, the range and strength of the inter-particle attractions, the geometry of the micro-channel, and the size of the particles.
Next generation lithography requires highly sensitive, ultrathin resists that offer sub-22nm resolution with minimal line edge roughness. Chemically amplified (CA) resists can provide the required sensitivity to radiation, but these materials are approaching their intrinsic resolution limits. Extending the lifetime of CA resists require a comprehensive understanding of the physical and chemical variables that control image formation in thin films, as well as a robust methodology to construct accurate, quantitative models for the coupled reaction-diffusion mechanism.
Our objective is to develop a new method to study image formation in ultrathin films of CA resists. We use grazing incidence and transmission small angle x-ray diffraction to measure the spatial extent of deprotection in three dimensions with nanoscale resolution. Our short term aim is to collect diffraction data for latent image formation as a function of post exposure bake time and temperature. The model CA is poly(hydroxystyrene-co-tertbutylacrylate), PHOST-PtBA
from DuPont Electronic Materials, 40% PtBA, loaded with triphenyl perfluoro-1-butanosulfonate photoacid generator. The spatial distribution of photoacid is generated with electron beam lithography. Deprotection reactions are completed at temperatures in the range of 125oC – 140oC. X-ray Diffraction data are acquired at the Advanced Photon Source of Argonne National Laboratory. The latent image structure is calculated from diffraction data through an inverse solution method: First, we build a model for the structure of deprotected patterns that includes size, shape, periodicity, and width of the deprotection interface. Second, we simulate the diffraction data and refine each parameter through nonlinear regression to obtain agreement between simulated and experimental diffraction profiles. Over the long term, data for latent image evolution as a function of film thickness and polymer substrate interaction strength will be collected and analyzed.
In the last decade, immunotherapy using antibodies, vaccines, and adoptive cell therapy has emerged as a highly effective approach for the treatment of human cancer1. Immunotherapy works by harnessing the power of the immune system and its ability to recognize and eliminate cancer cells. All malignant tumors contain variable numbers of lymphocytes, predominantly Tcells, which are referred to as tumor-infiltrating lymphocytes (TIL). Initially believed to be a
byproduct of tumor derived inflammation, the role of TIL has since been clarified to be cancer fighting sentinels, with the number of TIL having a prognostic significance in diverse cancers. Since these TIL have been shown to be tumor-specific, the ability to harness these TIL for therapeutic purposes has been recognized as a powerful form of personalized medicine taking
advantage of the adaptive immune response highly-specific for each person‘s own cancer. Adoptive cell therapy (ACT), based on the adoptive transfer of TIL or chimeric antigen receptor T-cells (CAR), has shown considerable promise even in late stage tumors refractory to all other treatment methods with clinical response rates of ~50%. In addition to the advantages of standard immunotherapy listed above, ACT offers additional benefits (1) Ability to employ the full range of effector functions of T-cells including cytotoxicity and cytokine secretion (2) the ability to target even micrometastases, and (3) the potential to proliferate in vivo within the host thus increasing persistence, surveillance and destruction capabilities. In the last decade considerable progress has been made in identifying culture conditions, cytokine supplements and host conditioning treatments to maximize the potential of ACT. Regrettably however, ACT treatments result in a wide range of patient outcomes from no identifiable benefit to complete remission and this is hardly surprising given the complexity of the human T-cell responses even compared to similar responses in mice where complete remission is routinely reported. Singlecell functional heterogeneity of cells infused for ACT has been largely ignored and consequently their efficacy and persistence in vivo following ACT is unpredictable. Since tumor-specific antigens are not clearly defined bulk assays employing populations of TIL and targets to measure cytotoxicity and cytokine secretion although widely employed are largely insensitive to function correlated with antigen-specificity. In collaboration with Profs. Radvanyi and Cooper at M.D.
Anderson Cancer Center we have developed a high-throughput assay based on arrays of fabricated nanowells to interrogate the functionality of TILs/CAR T-cells at the single-cell level to identify both T-cell mediated cytotoxicity and cytokine secretion.
Oble, D.A., Loewe, R., Yu, P. & Mihm, M.C., Jr. Focus on TILs: prognostic significance of tumor infiltrating lymphocytes in human melanoma. Cancer Immun 9, 3 (2009).
Varadarajan, N.et al. A high-throughput single-cell analysis of human CD8+ T cell functions reveals discordance for cytokine secretion and cytolysis. J Clin Invest (2011)
Rheumatoid arthritis (RA) affects ~1% of the world‘s population and is the most common autoimmune disease afflicting humans. It is characterized by chronic inflammation that can progress to joint destruction and ultimately functional disability requiring lifelong treatment. Early diagnosis of RA is essential to prevent irreversible joint destruction. A review of the research efforts over the last decade reveals several important findings with regards to the
development of arthritis: (i) the prevalence of antibodies against citrullinated proteins (ACPA, ~60-75 % of RA patients, 96 % specificity), (ii) the appearance of ACPA precedes the onset of disease and can thus be used as an early diagnostic marker and (iii) immunity towards citrullinemodified proteins underlies arthritis (at least in rodents). Although the importance of ACPA, both as a causative agent and as a diagnostic marker, has been established, the reactivities of single ACPA and the eptiopes they recognize on citrullinated proteins/peptides, is unknown. Isolation and characterization of protein targets of ACPA would shed light on the underlying mechanism of RA and would offer (i) earlier diagnosis and (ii) routes for therapeutic intervention. In collaboration with Prof. Agarwal at UT Health Science Center we have employed a novel high-throughput methodology, microengraving that employs arrays of fabricated nanowells to isolate antibodies from stimulated memory B-cells from human blood. In conjunction with single-cell PCR cloning we will characterize the sequence and specificity of the ACPA antibodies.
Klareskog, L., Ronnelid, J., Lundberg, K., Padyukov, L. & Alfredsson, L. Immunity to citrullinated proteins in rheumatoid arthritis. Annu Rev Immunol 26, 651-675 (2008).
Bradshaw, E.M., et al. Concurrent detection of secreted products from human lymphocytes by microengraving: cytokines and antigen-reactive antibodies. Clin Immunol 129, 10-18 (2008)
1. Dept. of Chemical & Biomolecular Engineering, 2. Dept of Mechanical Engineering,
3. Dept of Chemistry, University of Houston, Houston, TX 77204-4004, USA
Numerous problems in material science and engineering, chemistry, biochemistry and molecular biology require exact control of temperature in miniscule volumes. Solutions tested for such studies are typically held in microfluidics channels and often monitored under high powered optical microscopes.
We set to design a method to control the temperature in a microfluidics channel of width 40 um and height 5 um. We chose to build a heater consisting of a metal band of approximately 100 nm thickness and 10 um width. The exact band dimensions were adjusted so that the total heater resistance would provide the desired heating power. We chose a heater geometry whereby two concentric rings are connected to one another and to the feeding electrodes by spokes positioned at 90o with respect to one another. The heater power was chosen to ensure that a temperature jump from 5 to 25 C could be completed within 100 milliseconds. The resistance of the heating element is used to measure the resulting temperature.
In the UH Nanofabrication facility, we tested the viability of different metals and metal alloys (Cu, Cr, Pt, Nichrome) as the structure material for the resistive heater. Best adhesion to the glass substrate and acceptable thermoelectric characteristics were achieved by using a 10 nm thick layer of chromium under a 100 nm layer of copper. The recipe and parameters for heater fabrication on glass substrate were modified from those in literature, which are for deposition on a silicon substrate. We optimized the procedure for deposition of the AZ-resist pattern defining the heater shape. Best results were obtained with the use of a UV filter.
The manufactured heater was tested and calibrated using a hot plate allowing temperature stabilization within 0.1 C. Additional calibration was carried out with an infrared camera. In this way, for each heater, we document the dependence of its power on the driving current and of its resistance—on temperature. We solved numerically the equation of thermal diffusion for the geometry of the heater to find the temperature distribution on the microfluidics device during local temperature control. With the data on the power and temperature-dependent resistance, we will design a Labview-based proportional-integral derivative algorithm to ensure temperature stability within 0.1C and fast egress to the desired temperature setting.
1. Dept. of Chemical & Biomolecular Engineering, 2. Dept of Mechanical Engineering,
3. Dept of Chemistry, University of Houston, Houston, TX 77204, USA
Sickle cell hemoglobin (HbS) polymerization is the primary pathogenic event of the deadly sickle cell anemia. Delaying or suppressing polymerization is potential treatment strategy. Recent work in our group has found that heme, which may be released in the red cell cytosol due to the instability of HbS strongly enhances the rate of polymerization even at relatively low concentration. Further tests of this finding require the accumulation of large volumes of kinetics data as a function of the concentrations of both heme and HbS. Towards this goal, we designed and assembled a microfluidics device.
For representative results, the probed volume needs to be comparable to the volume of a red blood cell. Hence, the channel in which the solution is held is 5×40 um2. The HbS solution is separated by silicone oil into droplets of 600 picoliters, whose length in this channel will be 30 um. Such droplets of deoxy-HbS solution (the form that undergoes polymerization) will be pumped into the observation area of the channel at 5oC. Using a microheater, temperature in the observation area will be raised to 25oC. The solution concentration will be chosen such that at this latter temperature the solution is supersaturated with respect to the HbS polymers. Polymerization will be monitored by DIC (Differential Interference Contrast) microscopy.
This poster presents the current progress for the manufacture of the microfluidics chip. In tests to date, the utility of different photoresist polymers (PDMS and SU-8 of different formulations) as structure material was tested. The recipe and parameters for proper channel fabrication on a
glass substrate (in contrast to silicon typically discussed in literature) have been fine-tuned through extensive experimentation in the UH Nanofabrication Facility. We tested the swelling of SU8 layers by different salt solutions. We showed that despite the small dimension of the channel, water and silicone oil freely flow through it. To ensure bonding with the glass closing the channel, the SU-8 polymer layer, after curing, is treated with ionizing plasma. The surface
modification after this treatment was characterized using atomic force microscopy. Several bonding procedures were tested. A schematic for the control of the droplet size through variation of the flow rates of silicone oil and HbS solution using Labview is currently being developed.
Colloidal suspensions are frequently encountered in a variety of industrial and technological settings, many of which require these materials to be spread into thin films, flowed through porous media, or otherwise forced into confined geometries. Confinement is known to change the structure, dynamics, and phase behavior of suspensions in which the colloidal particles behave as hard spheres. However, most technological suspensions, such as paints, drilling fluids, inks, and coatings, have more complicated inter-particle interactions that involve both repulsion and attraction of varying strength and range. Though it is well known that more complicated interactions lead to a rich variety of structures and phases (such as gels and cluster fluids) in bulk suspensions, the effect of confinement on these types of materials is almost entirely unknown. Here, we use confocal fluorescence microscopy to study how confinement influences the three dimensional structure, individual particle dynamics, and overall phase behavior of a model attractive colloidal suspension. We find that increasing confinement is qualitatively similar to increasing attraction strength in that both induce progressively more solid-like behavior in the samples. Depending upon the attraction strength, solidification occurs through formation of either a colloidal crystal or heterogeneous colloidal gel. Furthermore, the onset of the solid-like behavior occurs at larger sample thicknesses (less confined) as the attraction strength is increased. These conclusions have important implications for handling attractive colloidal suspensions. In particular, materials in which the particles are homogeneously distributed in bulk may have a heterogeneous particle distribution when confined, thus affecting the material
properties. In addition, material rheology can change dramatically during a fluid-to-gel phase transition, which implies that flow processing techniques suitable for bulk suspensions may need to be modified when flow occurs in confined geometries. Our future work will focus on exploring the physical mechanisms that drive solidification in attractive colloidal suspensions, which will help improve handling of these materials in their various applications.
Nanopantography, a massively parallel nanofabrication technique, uses a large area collimated ion beam, directed at an array of electrostatic microlenses (fabricated using conventional microelectronics manufacturing methods) on a substrate. By applying appropriate voltages to the microlenses and “rocking” the substrate in a controlled fashion, ion “beamlets” entering the lenses focus to spots that can be rastered across the substrate, “writing” virtually any nanopattern
with dimensions down to 10 nm. The goals of this research are (a) to improve the resolution of Nanopantography to 3nm, and (b) to increase the writing speed, making Nanopantography a manufacturing-worthy process.
To achieve these goals, an advanced ultra high density plasma source was designed and built so that the system can operate with a smaller ion beam extraction aperture to improve resolution and, at the same time, have a high ion beam flux to improve throughput. Unique features of the new plasma source are: (a) a compact plasma volume combined with a 1.5 kW RF pulse generator, to yield 10X higher plasma density, compared to the original system. (b) a directcoupled
automatic matching network to reduce parasitic power dissipation and improve plasma stability. (c) a magnetic field to confine the plasma, and further enhance plasma density by at least 3X. (d) space charge neutralization to prevent Coulomb explosion of the high flux ion beam. For this purpose, electron emitting Yttria Iridium filaments were used. The new plasma
source was characterized with a Langmuir probe. Plasma density was measured for a range of powers, pressures and magnetic fields. The plasma density was in excess of 1.0×1012 cm-3 at 20 mTorr and 200 W with 50 Gauss magnetic field intensity. There was an optimum magnetic field intensity (which shifted up as pressure increased), that maximized plasma density.
Advances in density functional theory (DFT) have enabled theoretical calculations to describe catalytic reactions with great detail and an accuracy that compares favorably to experimental data. Such calculations allow for reasonable predictions of catalytic activity and lead to the possibility of computer-based catalyst designs. Using DFT and microkinetic modeling as main computational tools, our preliminary results for the CO oxidation reaction indicate that a highly active mixed catalyst with two different types of sites can be created from much less active components. Hence, we want to explore the feasibility of bifunctional catalyst design and suggest types of reactions where multiple functionalities can improve the overall reaction rate.
Department of Chemical & Biomolecular Engineering University of Houston 77204-4004
A crystallite-scale model is developed and incorporated into a reactor scale model to study the effects of Pt dispersion during the periodic lean-rich operation on a Pt/BaO/Al2O3 NOx storage and reduction (NSR) catalyst. The model is an extension of a previously developed regeneration model [D. Bhatia, M.P. Harold and V. Balakotaiah, Catal. Today 151 (2010) 314] and accounts for crystallite-scale diffusion limitations in the storage phase. The storage model is based on the concept of NOx spillover from Pt to BaO and diffusion of stored NOx in the barium phase. The model predicts the main features of NOx storage, such as the increase in NOx breakthrough time for increasing Pt dispersion at fixed Pt loading. The predicted NOx storage increase with Pt dispersion is a result of the increase in exposed Pt area which increases the specific NO oxidation activity, and of the increased interfacial perimeter between Pt and BaO, which promotes the rate of spillover. A sensitivity analysis of the stored NOx diffusivity reveals the importance of this process. The combined storage and regeneration model is used to simulate the entire lean rich cycles to study various cycle-averaged variables such as NOx conversion and NH3 selectivity. Simulations show the effectiveness of the catalyst to utilize the storage sites. The results reveal that the low dispersion catalyst (8%) is not able to utilize all of the available barium sites because NOx diffusion is too slow to access barium sites far from the Pt crystallites. The dispersion also affects the extent of axial uniformity of stored NOx during a typical 1-2 minute lean phase. The axial distribution is much more uniform for the low dispersion catalyst whereas a rather sharp storage front is evident for the high dispersion catalyst. Comparisons of the model predictions to experimental data reveal good agreement in the cycle-averaged trends. The model is used to study various storage and regeneration timing protocol which shows that shorter storage time is required for lower dispersion catalysts to achieve higher cycle-averaged NOx conversion. The model is useful in elucidating the complex transient phenomena occurring in the lean NOx trap.
Keywords: NOx, Hydrogen, Platinum, Barium, NOx storage and reduction, Lean NOx trap, Dispersion, Spillover
Department of Chemical and Biomolecular Engineering, University of Houston, Houston, TX
Atomistic models of silica nanoparticles, fullerenes and polyethylene are employed to study the curvature-induced effects on the polymer bound layer. Using large systems with connectivity-altering Monte Carlo simulations we sample the structure of melt in the proximity of nanoparticles. Density profiles reveal bulk like polymer density beyond 2 nm from the surface. Two important quantities used in characterizing adsorption; absolute adsorbance and surface concentration show strong dependence upon particle curvature. A detailed analysis of polymer adsorbed on the nanoparticle surface by defining tail, train and loop segments suggests that this phenomenon is not only a result of geometrical arguments but also of chain stiffness that leads to formation of shorter train segments upon increasing curvature. Our study shows that curvature effects become extremely profound for particles with sizes close to the polymer Kuhn length where interactions are governed by entropic contributions. Therefore, apart from enthalpic interactions, polymer stiffness needs to be accounted for quantitative prediction of adsorption behavior of polymer melt at nanoparticle surfaces.
Nanoparticle dynamics can be investigated by dynamic light scattering on traditional way, however, this technique can not be applied to concentrated dispersions or a more complicated system. In this study, a new technique, fluorescent differential dynamic microscopy(DDM) along with bright field DDM was used to examine the dynamics of two different size particles (Fluorescent Microsphere Suspensions: polystyrene beads 400 nm and 100 nm) at different concentrations. The results obtained from both of these techniques are quite identical to DLS and suggest that DDM could be applied to characterize the particle dynamics in a more extensive way.
Recently several efforts have been underway to use nanoparticle dispersions to improve the exploration and production of sub-surface hydrocarbons, specifically as rheology modifiers, profile-modifiers or interfacial agents. The dynamics for those nanoparticle dispersions is distinctly different from those of chemical tracers (small molecules) and those colloidal (Micronsized) particles, because of the comparable magnitudes of dispersions and transport coefficient. In this study, we will study the dynamics of nanoparticle dispersions in micropatterned channels with well-defined obstacles (such as cylindrical posts) whose surface chemistry can be tailored to simulate a model system to understand the diffusion of such dispersions in porous media. The differential dynamic microscopy (DDM) technique is used in this complicated system to characterize the particle dynamics.
Kidney stone disease (urolithiasis) affects 10–15% of the population worldwide where over 80% of renal stones are formed from the growth and aggregation of calcium oxalate monolydrate (COM) crystals. Despite increasing urolithiasis incidence rates, there has been no advancement in treatment or prevention in the last 30 years. We are developing a bottom-up approach to prevent COM crystallization through the design of crystal growth modifiers, termed inhibitors, that exhibit an affinity for binding to COM crystal surfaces and disrupting the addition of growth units through specific modifier-crystal binding to crystal surface sites (i.e. kink or step sites). In pathological biomineralization, COM crystallization is believed to be regulated in vivo by the interaction of proteins with crystal interfaces. Urinary proteins, such as osteopontin (OPN), Tamm‐Horsfall protein (THP), transferrin (Tf), and human serum albumin (HSA), are suggested to be inhibitors of COM nucleation, growth, and/or aggregation. We will present kinetic studies of COM growth in the presence of urinary proteins and biomolecules most commonly observed in proteomic assays of the organic matrix in human stones. These studies reveal that several constituents are potent inhibitors of COM. These results will be the basis for the design peptide growth modifiers as drug targets for COM stone disease. To this end, we are collaborating with scientists at Rensselaer Polytechnic University to pursue synergistic studies of high-throughput peptide synthesis, crystal growth characterization, and molecular modeling to identify peptide- COM molecular recognition and use this information as a general heuristic for the design of potent, high efficacy modifiers. Peptides are ideal candidates for COM growth inhibition because they are generally nontoxic and are ―programmable,‖ meaning the peptide sequence order, functional composition, and structure can be designed with high efficiency for inhibiting COM growth. Since the surfaces of COM crystals are calcium‐rich, we have chosen to study amino acids with acidic moieties (i.e. Asp, D, or Glu, E), which bind to COM surfaces via a Ca2+bridge. We have studied a first generation library of peptides with different binder sequences. These studies consist of the following: (i) Bulk crystallization experiments to assess changes incrystal habit (i.e. size, morphology, and orientation) through a combination of microscopy measurements (optical, SEM, and AFM); (ii) Kinetic studies using ion-selective electrode (ISE)and atomic absorption spectroscopy (AAS) measurements to assess the peptide efficacy and potency; and (iii) Interfacial studies using AFM to monitor the dynamics of surface growth and the influence of peptide-COM nteractions on the crystal topography. Preliminary results from these studies reveal that several urinary proteins and peptide analogues are effective COM growth inhibitors. These results will help guide the design of future peptide libraries to eventually find a set of potent inhibitors that can be used in long-term in vivo testing as possible drug candidates for COM stone prevention.
Reversible addition-fragmentation chain transfer (RAFT) polymerization has been utilized for the synthesis of triblock copolymers that will be evaluated for their potential as thermoplastic elastomers. Stearyl acrylate, derived from fatty acids such as oleic acid, has been used for the preparation of sustainable thermoplastic elastomers. Poly(stearyl acrylate) was successfully synthesized using two chain transfer agents (CTAs): S-1-dodecyl-S‘-(α,α‘-dimethyl-α-acetic acid)trithiocarbonate (CTA1) and 1,4-Bis(thiobenzoylthiomethyl)benzene (CTA2). The resulting polymers had well-controlled molecular weight distributions and low polydispersity indices, and were used for the subsequent polymerization of block copolymers. First, poly(stearyl acrylate) synthesized with CTA1 was further used as a macro-CTA in the RAFT polymerization of styrene, resulting in the diblock copolymer poly(stearyl acrylate-b-styrene) with a narrow molecular weight distribution. This synthesis will be extended to prepare poly(styrene-b-stearyl acrylate-b-styrene) triblock copolymers. Second, the end-groups of poly(stearyl acrylate) synthesized with CTA2 will be converted to hydroxyl groups and the polymer will be used as a macroinitiator for the ring-opening polymerization of lactide, resulting in poly(lactide-b-stearyl acrylate-b-lactide) triblock copolymers. The tensile properties of both classes of triblock copolymers will be evaluated for applications as thermoplastic elastomers.