This work is concerned with creating predictive models for the optical properties of organic polymers, which will guide our experimentalist partners and allow them to target the most promising candidates. The RI model is developed based on a synergistic combination of first-principles electronic structure theory and machine learning techniques. The RI values predicted for common polymers using this model are in very good agreement with the experimental values.
Benchmarking methods for refractive index and polarizability calculation of organic polymers
This project concerns with benchmarking the accuracy of several DFT functionals and basis set for calculating the polarizability of polymers. In particular, we compare different model chemistries to determine an inexpensive and rational approach for the calculation of polarizability in HTPS framework. We compare the change in polarizability of non-conjugated and conjugated polymers with the increasing polymer chain length and subsequently present an extrapolation scheme for both types of polymers. This study gives guidelines to theoreticians in selecting methods for polymer studies.
Discovery of high refractive index polyimides
Polyimides have attracted much attention due to their exceptional thermal stability and ease of processability. Further, polyimides possess mechanical stability, good flexibility, flame resistance, radiation resistance and low dielectric constant, thus holding great promises for optoelectronic applications. However, these polymers have low RI values which limit the use in these applications. In this work, we identify novel polyimide candidates with high RI values using a high-throughput screening, materials informatics, and rational design framework.
Neural networks for the prediction of high refractive index organic molecules
High refractive index materials can be developed by tailoring the chemical structure. But, tailoring the structure using different functional groups could potentially lead to an infinite number of molecular candidates. It is impractical to empirically characterize a large number of candidates whereas computational analysis allows greater exploration at a mere fraction of the time and cost. In our approach, we create a large candidate library (~1.5 million) of small organic molecules using our in-house molecular library generator. However, applying first-principles and MD calculations on such a huge library is not a viable option. Therefore, we train a neural network algorithm using the data from 100,000 molecules, and use this model to predict the properties of 1.5 million candidates.
Polymers with environmental degradation
Degration model for the most favourable path for ester hydrolysis.
In this work, we investigate hydrolysis of polyesters using first-principles quantum calculations. In case of hydrolysis of linear esters, we observe the formation of a tetrahedral intermediate. The formation of this tetrahedral intermediate and subsequent cleavage of the ester bond is studied in the presence of different number of water molecules. The aim of the next phase of the project is to create a large number of potential candidates and determine the activation energies of these candidates using the protocol developed in the first project phase. To accelerate the process, we are in the process of casting this protocol into our virtual high-throughput screening (HTPS) framework.
ChemLG – A smart and massively parallel molecular library generator
ChemLG aims to extend and generalize library generation to identify molecular lead candidates and reaction networks in various materials applications such as photovoltaics, optoelectronics, and catalysis. This massively parallel generator offers a multitude of options to customize and restrict the scope of the enumerated chemical space and thus tailor it for the demands of specific applications. To streamline the non-combinatorial exploration of chemical space, we incorporate genetic algorithms into the framework. Genetic algorithms have shown to be efficient in optimizing chemical structures and generating useful compounds for different target applications. In addition to implementing smarter algorithms, we also focus on the ease of use, workflow, and code integration to make the this technology more accessible to the community.
ChemHTPS – A virtual high-throughput screening program suite for the chemical and materials sciences
The discovery of new compounds, materials, and chemical reactions with exceptional properties is the key to progress in chemistry. This process can be dramatically accelerated by means of the virtual high-throughput screening (HTPS) of large-scale candidate libraries. ChemHTPS aims to create a general-purpose, comprehensive, user-friendly, and black-box type suite, that will allow users to efficiently perform large in silico modeling studies and high-throughput analyses with low maintenance and without the need for expert knowledge. This suite provides a complete package for materials discovery, from generation of large molecule library to performing HTPS to easy storage/access of data with relatively low end-user maintenance.
GradeMaster is a python package to help faculty manage student roster, manage course grades, make projections and notify students. The graphical user interface (GUI) is built using PyQT library. The software suite features full class analysis (for all assignments and exams) and indivudual student stats and projections.
In our research group, one of the projects is to discover trends in data derived from different quantum chemical methods. The work focuses on the analysis and comparison of results of various molecular properties derived from different flavors of Density Functional Theory (DFT). These patterns in data have significant implications for the utility of the employed approximations and the design of new quantum chemical techniques. Electronic structure calculations of about 3 million molecular candidates has been evaluated in the Harvard Clean Energy Project (HCEP). The calculations were performed using different flavors of DFT. As there are many flavors, large number of correlations are required to understand trends between the flavors leading to a computationally expensive process. As part of HPC project, I have determined these correlations using gradient descent method and implemented parallel gradient descent to decrese the computation time. The code developed also include parallel data reading and determines multiple correlations simultaneously. The code is written in Python programming language and the parallel computing is implemented using MPI4Py package.
Enhanced electrical conductivity of suspended carbon nanofibers: Effect of hollow structure and improved graphitization
In this work we demonstrate that at the same pyrolysis temperature, higher graphitization can be attained in carbon nanofibers (CNFs) with hollow core compared to solid nanofibers, resulting in higher electrical conductivities in hollow CNFs. Through controlled electrospinning, single polymeric (PAN/PMMA) nanofibers are suspended across the walls of a photolithographically patterned SU8 microstructure, followed by pyrolysis to obtain single suspended solid, porous and hollow CNFs. It is found that hollow CNFs with a shell thickness of ~35 nm demonstrate the highest electrical conductivity (~105 S/m) which is almost an order of magnitude higher than those of solid CNFs (~104 S/m). Further, porous CNFs of the same diameter show lower electrical conductivity compared to the solid fibers (~103 S/m).
In this review report, a comprehensive review of the drug delivery applications of Pluronics is presented. Various structures of Pluronics/drug folumations i different therapies using these formulations is dicussed. The review includes various chemically modfied Pluronics that are studied in the past and their effects on the drug encapsulation and strcutre stability. Behaviour of in-vivio release of drugs on internal and external stimuli and different modifications in these encapsulated micelles for effective drug delivery have been reviewed. Additionally, an overview of the significance of Pluronic systems in overcoming the blood-brain barrier is also presented.
This review report includes an overview of the structure and dynamics of Pluronic micelles. The fundamentals of Pluronics synthesis, thermodynamics of micelle formation, molecular structure and solvent quality effects on the micelle structure and dynamics have been reviewed. Various applications of Pluronic micelles are also reported, with drug delivery being the main focus.
The problem of hydrocephalus has been noted since centuries, but only has the problem been understood in the late 19 century. Since then, constant efforts are being made to come up with effective diagnosis for hydrocephalus. Silicone shunts were developed in early 1950’s which would rescue the problem of hydrocephalus. But the problems related to mechanical and infectious complications, over drainage of the cerebrospinal fluid and loculations were observed related to silicone shunts. There have also been few reports of silicone allergy which were causing the shunt to fail. Up to date there has been numerous efforts and trials done with modified silicone shunts as well as other polymeric shunts to overcome these problems. The hunt for the novel material for shunts which has no allergies and that can overcome the noted deficiencies of silicone shunts is still going on. In this article, I present a review on hydrocephalus, current treatments and their limitations and a new material to be used as a shunt. I present a chemically modified polycarbonate-urethane (PCU) as a potential polymeric material to be used in the development of shunts for treating hydrocephalus.
Functionally graded hydroxyapatite-alumina-zirconia biocomposite: Synergy of toughness and biocompatibility
Functionally Gradient Materials (FGM) are considered as a novel concept to implement graded functionality that otherwise cannot be achieved by conventional homogeneous materials. For biomedical applications, an ideal combination of bioactivity on the material surface as well as good physical property (strength/toughness/hardness) of the bulk is required in a designed FGM structure. In this perspective, the work aims at providing a smooth gradation of functionality (enhanced toughening of the bulk, and retained biocompatibility of the surface) in a spark plasma processed hydroxyapatite-alumina-zirconia FGM bio-composite.
This project concerns the synthesis of hydroxyapatite-silver (HA-Ag) and carbon nanotube-silver (CNT-Ag) composites via spark plasma sintering (SPS) route. Ag reinforcement resulted in the enhancement of hardness (by ~15%) and elastic modulus (~5%) of HA samples, whereas Ag reinforcement in CNT, Ag addition does not have much effect on hardness (0.3 GPa) and elastic modulus (5 GPa). The antibacterial tests performed using Escherichia coli and Staphylococcus epidermidis showed significant decrease (by ~65-86%) in the number of adhered bacteria in HA/CNT composites reinforced with 5% Ag nanoparticles. Thus, Ag-reinforced HA/CNT can serve as potential antibacterial biocomposites.