Research Theme T3


(T3)  Development of Catalytic Nanopumps Utilizing Carbon Nanotubes and Layer-by-Layer Processing

CSULA Faculty Participants:  Frank Gomez, Guo-meng Zhao

Penn State Faculty Participants:  Ayusman Sen, Tom Mallouk

There is a need to develop non-mechanical nano- and microscale pumps in various microfluidic platforms that are able to function without an external power source, providing precise control over the flow rate in response to specific signals [167—183]. The development of such pumps, where the flow rate is specifically controlled in response to external stimuli is the first requirement in the design of the next generation of smart devices. Concomitantly, layer-by-layer (LBL) processing and related techniques provide an interesting opportunity for microfluidic platforms [184—188]. The wide-ranging functionality of LBL techniques makes it possible to tailor, at the nanoscale, physical properties including the mechanical, electrical, and optical properties of the coatings as well as the chemical functionality of the coatings. In the latter case, functionalization of reactive groups within and at the surface of the multilayer coatings can be used to immobilize bioactive molecules including adhesion tripeptides, antibodies, enzymes, and DNA. The ability to tailor interfaces has led to many of the technological advances in materials in recent decades across all domains, and is particularly relevant when considering nanoscale effects [189—191]. CNTs are interesting molecular wires with unique electronic properties that are highly sensitive to its environment and vary significantly with changes in electrostatic charges and surface adsorption of various molecules [192—194]. In addition, CNTs have better biocompatibility, as shown by selfpropelling of the glucose oxidase- and catalase-modified CNT bundles in glucose [195]. They are promising materials for the design of many functional thin films including those for catalytic membranes [196, 197], actuation [198], and mechanical thin film applications [199]. The ability to control the architecture of CNT thin films at nanometer and micrometer-sized scale is integral in tailoring the film properties and their functionality [200]. For example, vertically-aligned oriented geometries vis-à-vis vertically-aligned CNTs (VACNTs) have been used as an element in bulk composite materials [201]. With this in mind, it would be interesting to examine catalytic nanopumps on VACNTs functionalized by LBL techniques.

In this project we propose to: 1) examine the ability of enzymes to act as surface pumps that are attached via electrostatic interaction onto VACNTs modified by LBL polyelectrolytes; 2) study the chemotactic behavior of small particles on these surfaces, and; 3) utilize chemometric approaches to optimize the design of the VACNT and LBL structures and experimental conditions, thereof. Through solution processing, the LBL technique will provide molecular assembly in 3D on high-permeability nanoscale (and spaced) VACNT scaffolds. The nanoporous elements have ultra-high permeability allowing for the surface to be easily accessed through solution processing while the LBL will maintain the novel ultra-high permeability via multiple layer depositions. This attribute has two major implications for the proposed work: (i) LBL assembly can occur into the VACNT forests and the process can be repeated many times while maintaining the needed ultra-high permeability, and; (ii) the resulting nanoporous features with LBL-functionality can be utilized directly in microfluidic platforms (fluids and gases will still permeate) presenting a controlled way of studying chemotactic behavior of materials by catalytic motors.

Figure 2. Micro-patterned VACNT features in microfluidic channels: schematics of fabrication and functionalization. Not to scale.This work will examine an extension of the chemotactic effect using microfluidics, leading to a better understanding of the collective behavior of catalytic pumps. In addition, it is proposed that modifications in the type of LBL polyelectrolyte used will elicit variances in the ability of enzymes to act as surface pumps. This project brings together four excellent investigators from CSULA and Penn State, synergistically pulling together key new contributions from each: Gomez has developed a myriad of analytical techniques incorporating microfluidics, materials, and devices/instruments for probing a myriad of chemical and biochemical interactions and has utilized chemometrics in optimizing microfluidic platforms and experimental design [202— 219]. Zhao has studied the magnetic and electrical properties of carbon nanotubes for a variety of applications [220—223]. Sen [167, 169—174, 176—178] and Mallouk [179—183] have been at the forefront of studying catalytically driven nanomotors and pumps creating a new paradigm for molecular-level engineering of functional materials. The VACNT fabrication process utilized in this work is schematically presented in Figure 2. Catalyst patterns will be defined by standard photolithography on substrate (e.g. SiO2 or Au) wafers. A 10nm Al2O3 layer and a 1nm Fe layer will be subsequently deposited by electron-beam evaporation. After lift-off, the patterned substrates will be inserted into a CVD furnace for growing VACNTs. Standard CNT growth techniques will result in forests of multi-walled (2-3 concentric walls), VACNT, with an average tube diameter of 8nm and an average inter-CNT spacing of ~80nm, giving a 1% volume fraction. Incorporation of the patterned CNT structures into devices is achieved using standard soft lithography techniques. VACNT materials are a unique architecture that should allow them to be used to examine chemotactic behavior in microfluidic platforms. Their use will allow novel pumping assemblies and, hence, delivery of materials. The ability to functionalize can potentially allow for the coupling of pumping to delivery of materials and the construction of hydrogels. Furthermore, the unique qualities of CNTs, allow for their study by electrochemical means.

Figure 3. (A) VACNT filled in microchannel. (B) Example of microfluidic chip to be used for SPRi to detect particles. VACNT are shown not to scale.LBL assembly of alternating layers of oppositely charged polyelectrolytes will be constructed on VACNTs providing the formation of 5 –500 nm thick films with monolayers of various substances growing in a pre-set sequence. A cleaned substrate will be immersed into a dilute solution of a cationic polyelectrolyte (poly(ethylenimine) [PEI], poly(dimethyldiallylammonium chloride) [PDDA], poly(allylamine) [PAH], polylysine,) for a time optimized for the adsorption of a monolayer, then rinsed and dried. The next step will be the immersion of the polycation covered substrate into a dilute dispersion of polyanion (poly(styrenesulfonate) [PSS],poly(vinylsulfate), and poly(acrylic acid).) or negatively charged nanoparticles also for a time optimized for the adsorption of a monolayer, then rinsed and dried. The self-assembly of a polyelectrolyte monolayer and any monoparticulate layer sandwich unit onto the substrate is then completed. Subsequent sandwich units are self-assembled in a similar fashion. Different enzymes (catalase, glucose oxidase, and peroxidase) and polyions will be assembled in the pre-planned order in one film. Forces between nanoparticles and binder layers govern the spontaneous LBL self-assembly of the films which are primarily electrostatic and covalent in nature but also involve hydrogen bonding, hydrophobic and other types of interactions. The LBL assembly of varied polyelectrolytes will allow for the facile tuning of thicknesses, composition ratios, and porosities by controlling the degree of ionization of said polyelectrolytes with the pH of the solution that will affect the confirmation of the file and its morphology. Figure 3 shows a VACNT filled in a microchannel and an example of a microfluidic chip constructed to detect small tracer particles. The surface chemistry of the LBL functionalized- VACNTs will be analyzed using X-ray photoelectron spectroscopy (XPS), X-ray diffraction (XRD), atomic force microscopy (AFM), and scanning electron microscopy (SEM). We will determine the electrochemical properties of the materials by examining the sheet resistance and running cyclic voltammetry experiments on the materials. Lastly, we will examine the swelling behavior and mechanical integrity of the LBL functionalized-VACNTs.

Figure 4. Three-dimensional response surface for channel width (B)-sample volume (C).Chemometrics and computational methods can be useful in improving the performance of microfluidic-based processes and devices. Due to the large quantity of variables affecting the performance of these devices, the procedure to develop them cannot be considered a simple task. In this context, the term “optimization” refers to improving the performance by determining the best conditions at which the best response is obtained. Optimization and modeling can shorten the method development, which can often be a time-consuming process. Unfortunately, it is still a common practice to use the classical experimental approach based on the single variable approach (SVA) or one-variable-at-a-time strategy [224]. SVA might work in specific cases but usually requires excessive experimental work. Besides, it has a serious problem when variables interact with each other. In this case, the optimization of parameters by varying a factor while keeping all other parameters constant is misleading. In this context, the multivariate methods, whereby, all factors are varied simultaneously, are important issues. In addition, these approaches require less time, effort, and resources compared to univariate procedures. These methods facilitate optimizing all factors affecting the process while minimizing the number of experiments [225]. The approach to process the optimization is called response surface methodology (RSM). RSM is a collection of mathematical and statistical techniques that can be used for modeling and analysis of a process in which a response of interest is influenced by several parameters and where the objective is to optimize this response. These techniques have been extensively used in bioprocesses [226—229], industrial applications [230—232], and analysis of very complex mixtures [224, 233—235]. Although the application of RSM to microfluidic processes with different unpredictable variables would seem logical, there is a dearth of work in this area. Hence, we propose the use of chemometric approaches to optimize the design and experimental conditions in the development of catalytic nanopumps, research that is both transformative and novel. The use of RSM will allow us to optimize the spatial geometry of the VACNTs within the microfluidic platform, the LBL protocol, and experimental conditions. This additional integration of RSM adds to the transformative nature of the proposed work and will bring new insights into nanomotor studies and chemotaxis. Figure 4 is the 3D response surface showing the effect of channel width (B) and sample volume (C) on the color intensity (response) for a paper microfluidic device at a fixed value of channel length using RSM. These figures clearly show a positive effect for the channel width and a negative effect for the sample volume. It can be seen that when channel width was increased, the color intensity increased.

In a typical experiment using catalase as the model enzyme, previously assembled as part of the LBL assembly, microspheres will serve as tracer particles. In the presence of substrate (H2O2 in this case) the flow of particles will be visually examined. Furthermore, kinetic reaction data (kcat and KM for the immobilized enzyme) will also be determined from the fluid pumping experiments. Temporal and spatial variations in pumping as well as the pumping mechanism will also be examined. We further propose to use surface plasmon resonance imaging (SPRi) to visualize the particles (see Figure 3B). SPRi is a sensitive, label-free, and low-light optical method that eliminates the requirement for modified biological molecules and allows for real time observation. Here, SPRi contrast will allow for sensitive and accurate quantification of the tracer particles while in the microfluidic channel. Gomez has experience in SPR in previous microfluidic work [236—239]. For the hydrogel experiments, solutions of fluorescein dye (as model cargo) will be preloaded into the microchannels and will be absorbed onto the LBL-functionalized VACNT assemblies (using urease as the model enzyme). Varying the amount of substrate for the specific enzyme should liberate dye that can be easily monitored by fluorescence means. Hypothetically, if enzyme pumps can be activated by specific molecules, it is a logical goal to design enzyme-based microfluidic devices that can act both as sensors and pumps. This work will collectively design, develop, model, and study autonomous motors and pumps in microfluidicbased platforms and presents a controllable way of studying chemotactic behavior of these artificial catalytic micromotors. Our long-term goal is to develop and create smart materials that can autonomously and collectively interact with each other and their surroundings. While studies on active matter have focused on examining biological active swimmers or particles, we prefer a foundational approach (chemical) to understanding motility thereby allowing for specificity in material control that are novel and not possible from biological species. Hence, the five-year goal of this project will be to develop the capabilities within Cal State L.A., in collaboration with Penn State, to examine mechanistic questions associated with nanomotors and to enable the engineering of novel systems.