46 research outputs found
Statistical mechanical theory of equilibrium structure and miscibility of polymer nanocomposites: effects of polymer chemical heterogeneity and architecture, and nanoparticle surface corrugation and softness
No full textMotivated by the persistent interest in different nanoparticles added to various polymer matrices, the Polymer Reference Interaction Site Model (PRISM) theory is extended and applied to study the thermodynamics, statistical structure, and miscibility of diverse polymer nanocomposites (PNCs). Under chemistry-matched conditions and in the absence of interfacial attractions between a spherically smooth nanoparticle and the matrix fluid, the polymer-induced depletion attraction is dominant and induces entropic phase separation. The depletion attraction can be potentially reduced by modifying the nanoparticle surface topography as recently observed in experiments. Two types of surface-modified nanoparticles have been considered in this thesis – (1) spheres with ordered roughness on the surface and (2) soft polymeric nanoparticles with surface fluctuations and fuzziness. Monte Carlo integration and other computational techniques have been developed to compute the effective interactions between such particles. The morphologically diverse particles introduce additional length scales, making the physics non-monotonic, subtle, and rich. The common advantage with using either of the particles is reduced contact aggregation and enhanced miscibility. Optimal surface corrugation and/or particle softness allow monomer penetration resulting in favourable (entropic) mixing. However, high enough degree of corrugation/softness can also result in destabilization by excluding the polymer from its interior.
Another route of developing new nanocomposites is by tuning the polymer-particle interfacial chemistry. Prior work has established three states of spatial organization, namely depletion, steric stabilization and bridging, depending upon the effective interfacial attraction strengths. Introducing polymer chemical heterogeneity via the use of AB copolymers offers additional control over the equilibrium structure. Specifically, two types of copolymers are considered – (1) random copolymers (RCP) of disordered sequence and (2) ordered, alternating multiblock copolymers (MBCP). Quantum chemical calculations are combined with the polymer liquid state theory to predict structure and miscibility. The chain connectivity, monomer sequence, copolymer composition and differential wettability results in unique frustration in the system leading to novel states of organization of the polymer around the nanoparticles. In the context of strongly attractive nanoscopic fullerenes, this results in improved miscibility relative to the corresponding homopolymers. For some of the systems studied, maximum dispersion is predicted at an intermediate copolymer composition due to packing correlations and differential wetting effects with favourable comparison to experiments.Submission published under a 24 month embargo labeled 'U of I Access', the embargo will last until 2017-12-01The student, Debapriya Banerjee, accepted the attached license on 2015-07-29 at 10:23.The student, Debapriya Banerjee, submitted this Dissertation for approval on 2015-07-29 at 10:24.This Dissertation was approved for publication on 2015-07-31 at 15:42.DSpace SAF Submission Ingestion Package generated from Vireo submission #8651 on 2016-03-08 at 11:04:52Made available in DSpace on 2016-03-08T17:21:18Z (GMT). No. of bitstreams: 2
BANERJEE-DISSERTATION-2015.pdf: 12084334 bytes, checksum: 2e5ace4141f110955bfbb7a00b222bb8 (MD5)
LICENSE.txt: 4215 bytes, checksum: 32c70542a37bb0dee64ad93fa9116069 (MD5)
Previous issue date: 2015-07-31Embargo set by: Seth Robbins for item 91477
Lift date: 2018-03-08T17:22:13Z
Reason: Author requested U of Illinois access only (OA after 2yrs) in Vireo ETD systemU of I Only Restriction Lifted for Item 91477 on 2018-03-09T10:15:29Z
Possible knowledge: forms of literary and scientific thought in early modern England
No full textThis dissertation argues that the emergence of a new intellectual paradigm I call “possible knowledge”—encompassing projective, probable, counterfactual, hypothetical, conjectural, and prophetic ways of thinking—shaped literary and scientific writing in Renaissance England. The project uncovers a prehistory of scientific probability, still perceived as an Enlightenment-era phenomenon, by focusing on a constellation of speculative modes of knowing that drew on the imagination in the face of epistemic uncertainty. Possible knowledge emerges from elements crucial to our understanding of the literary, including mimesis, utopian discourse, and dramatic enactment, and it crosses generic boundaries. The disruption of prophetic certainty, for instance, informs the action in William Shakespeare’s _Macbeth_, while the unrepeatable epic events in John Milton’s _Paradise Lost_ reveal why contemporary experimental methods—which could produce only probable knowledge about the natural world—were insufficient to explicate prelapsarian states of being. I engage with the history and philosophy of science to show how the techniques of writing associated with possible knowledge are visible across modern disciplinary divides: the error and the endlessness that govern Edmund Spenser’s epic-romance, _The Faerie Queene_, are at the heart of the modern scientific epistemology laid out in Francis Bacon’s inductive method. And as Margaret Cavendish’s utopian experiment with cognitive realms in _The Blazing World_ underscores, possibility could allow authors intellectual freedom and creativity in their engagement with the material world. By focusing on hypothetical and suppositional modes of thinking, I map the contours of the humanities and the sciences as these began to assume their modern disciplinary forms.Ph. D.Includes bibliographical referencesby Debapriya Sarka
Binding Characteristics of Anticancer Drug Doxorubicin with Two-Dimensional Graphene and Graphene Oxide : Insights from Density Functional Theory Calculations and Fluorescence Spectroscopy
No full textThere has been a perpetual interest in identifying suitable nano-carriers for drug delivery. In this regard, graphene-based two-dimensional materials have been proposed and demonstrated as drug carriers. In this paper, we have investigated the adsorption characteristics of a widely used anticancer drug, doxorubicin (DOX), on graphene (G) and graphene oxide (GO) by density functional theory calculations and fluorescence and X-ray photoelectron spectroscopies. From the calculated structural and electronic properties, we have concluded that G is a better binder of DOX compared to GO, which is also supported by our fluorescence measurements. The binding of DOX to G is mainly based on strong pi-pi stacking interactions. Consistent with this result, we also found that the sp(2) regions of GO interact with DOX stronger than the sp(3) regions attached with the functional groups; the binding is characterized by pi-pi and hydrogen-bonding interactions, respectively.</p
Semi-supervised Learning using Triple-Siamese Network
No full textMissing data problem is inevitable in mostly all research areas including Artificial Intelligence, Machine Learning and Computer Vision where we have modicum knowledge about the complete dataset. One of the key reasons of missing data in AI is insufficiency of accurately labeled data. To solve a classification problem using ML or training a Deep Neural Network model, we need a huge amount of labeled data. It is difficult to get labeled data but unlabeled data is inexpensive and available easily. It is usual that we get no more than a single element per class to train our models due to unavailability of enough labeled training data. Strict privacy control or accidental loss may also cause missing data problem. One of the ways of getting training data labeled is using human-in-the-loop, but budget constraints can prevent that option. The objective of this research is to recover the complete signal or missing labels of the dataset using state-of-the-art Machine Learning and Computer Vision techniques. We propose a novel network trained with a few instances of a class to perform Metric Learning. We then convert our dataset to a graph signal and recover the graph completely using Recovery algorithm in Graph Fourier Transform. Our approach performs significantly better than Graph Neural Network and other state-of-the-art techniques
Rakhaldas Banerjee\u27s \u27Shashank\u27: History vs. Narrative/ রাখালদাস বন্দ্যোপধ্যায়ের ‘শশাঙ্ক’ : ইতিহাস বনাম আখ্যায়িকা
No full textIn his narrative Shashanka, Rakhaldas Bandyopadhyay sought to rekindle the spirit of a humiliated and oppressed Indian populace by portraying the independent Bengal of the 7th century against the backdrop of 19th and 20th-century colonial India. At that time, the extraordinary valor of the Sikhs, Marathas, and Rajputs was well known throughout the country. Yet, instead of choosing any of these non-Bengali figures as his hero, Rakhaldas presented Shashanka, the ancient sovereign ruler of Bengal, who had liberated Gauda - present-day Bangladesh - from the shackles of subjugation.
Bandyopadhyay’s narrative constructs a powerful dichotomy between freedom and subjugation, strength and weakness, through which he explores the conflict between Hinduism and Buddhism in 7th-century India. One of the most contentious questions in history is: was Shashanka truly anti-Buddhist? History has often portrayed Shashanka through a one-sided lens, resulting in a simplistic evaluation of his character. Rakhaldas Bandyopadhyay, however, diverged from this conventional historical narrative in his Shashanka, attempting instead to offer a more nuanced view.
By presenting the actions and motivations of both Shashanka and the Buddhist monks, he created space for a re-evaluation of Shashanka’s character. Many of the allegations against Shashanka appear, in Rakhaldas\u27s portrayal, to be exaggerated. Hence, he provides counter-arguments to defend Shashanka and absolve him of such accusations. Conversely, he casts doubt on the conduct and ideals of the Buddhist monks by portraying them in a tainted, morally ambiguous light.
In those of Rakhaldas Bandyopadhyay’s novels that depict the socio-cultural fabric of the Hindu-Buddhist era, the Buddhist monks are often shown in a degenerate and decayed form. As a novelist, he fails to maintain an impartial perspective in this regard. His portrayal of Buddhist monks reflects a nationalist sentiment, rooted in his ideological convictions. Consequently, Shashanka is presented as the epitome of masculine prowess. Through the heroic tale of this Hindu patriarch and his immense strength, there is a discernible and conscious attempt to assert Hindu national pride
Protein Folding Activity of the Ribosome (PFAR) –– A Target for Antiprion Compounds
No full textPrion diseases are fatal neurodegenerative diseases affecting mammals. Prions are misfolded amyloid aggregates of the prion protein (PrP), which form when the alpha helical, soluble form of PrP converts to an aggregation-prone, beta sheet form. Thus, prions originate as protein folding problems. The discovery of yeast prion(s) and the development of a red-/white-colony based assay facilitated safe and high-throughput screening of antiprion compounds. With this assay three antiprion compounds; 6-aminophenanthridine (6AP), guanabenz acetate (GA), and imiquimod (IQ) have been identified. Biochemical and genetic studies reveal that these compounds target ribosomal RNA (rRNA) and inhibit specifically the protein folding activity of the ribosome (PFAR). The domain V of the 23S/25S/28S rRNA of the large ribosomal subunit constitutes the active site for PFAR. 6AP and GA inhibit PFAR by competition with the protein substrates for the common binding sites on the domain V rRNA. PFAR inhibition by these antiprion compounds opens up new possibilities for understanding prion formation, propagation and the role of the ribosome therein. In this review, we summarize and analyze the correlation between PFAR and prion processes using the antiprion compounds as tools
Molecular Recognition in Partially Folded States of a Transporter Protein: Temperature‐dependent Specificity of Bovine Serum Albumin
No full textControlling effective interactions and spatial dispersion of nanoparticles in multiblock copolymer melts
No full textFinding Limit Cycles in self-excited oscillators with infinite-series damping functions
No full textIn this paper we present a simple method for finding the location of limit cycles of self
excited oscillators whose damping functions can be represented by some infinite convergent
series. We have used standard results of first-order perturbation theory to arrive at
amplitude equations. The approach has been kept pedagogic by first working out the cases
of finite polynomials using elementary algebra. Then the method has been extended to
various infinite polynomials, where the fixed points of the corresponding amplitude
equations cannot be found out. Hopf bifurcations for systems with nonlinear powers in
velocities have also been discussed
Image-Label Recovery on Fashion Data Using Image Similarity from Triple Siamese Network
No full textWeakly labeled data are inevitable in various research areas in artificial intelligence (AI) where one has a modicum of knowledge about the complete dataset. One of the reasons for weakly labeled data in AI is insufficient accurately labeled data. Strict privacy control or accidental loss may also cause missing-data problems. However, supervised machine learning (ML) requires accurately labeled data in order to successfully solve a problem. Data labeling is difficult and time-consuming as it requires manual work, perfect results, and sometimes human experts to be involved (e.g., medical labeled data). In contrast, unlabeled data are inexpensive and easily available. Due to there not being enough labeled training data, researchers sometimes only obtain one or few data points per category or label. Training a supervised ML model from the small set of labeled data is a challenging task. The objective of this research is to recover missing labels from the dataset using state-of-the-art ML techniques using a semisupervised ML approach. In this work, a novel convolutional neural network-based framework is trained with a few instances of a class to perform metric learning. The dataset is then converted into a graph signal, which is recovered using a recover algorithm (RA) in graph Fourier transform. The proposed approach was evaluated on a Fashion dataset for accuracy and precision and performed significantly better than graph neural networks and other state-of-the-art methods
