-
Polysemanticity and Capacity in Neural Networks
Authors:
Adam Scherlis,
Kshitij Sachan,
Adam S. Jermyn,
Joe Benton,
Buck Shlegeris
Abstract:
Individual neurons in neural networks often represent a mixture of unrelated features. This phenomenon, called polysemanticity, can make interpreting neural networks more difficult and so we aim to understand its causes. We propose doing so through the lens of feature \emph{capacity}, which is the fractional dimension each feature consumes in the embedding space. We show that in a toy model the op…
▽ More
Individual neurons in neural networks often represent a mixture of unrelated features. This phenomenon, called polysemanticity, can make interpreting neural networks more difficult and so we aim to understand its causes. We propose doing so through the lens of feature \emph{capacity}, which is the fractional dimension each feature consumes in the embedding space. We show that in a toy model the optimal capacity allocation tends to monosemantically represent the most important features, polysemantically represent less important features (in proportion to their impact on the loss), and entirely ignore the least important features. Polysemanticity is more prevalent when the inputs have higher kurtosis or sparsity and more prevalent in some architectures than others. Given an optimal allocation of capacity, we go on to study the geometry of the embedding space. We find a block-semi-orthogonal structure, with differing block sizes in different models, highlighting the impact of model architecture on the interpretability of its neurons.
△ Less
Submitted 11 July, 2023; v1 submitted 4 October, 2022;
originally announced October 2022.
-
Adversarial Training for High-Stakes Reliability
Authors:
Daniel M. Ziegler,
Seraphina Nix,
Lawrence Chan,
Tim Bauman,
Peter Schmidt-Nielsen,
Tao Lin,
Adam Scherlis,
Noa Nabeshima,
Ben Weinstein-Raun,
Daniel de Haas,
Buck Shlegeris,
Nate Thomas
Abstract:
In the future, powerful AI systems may be deployed in high-stakes settings, where a single failure could be catastrophic. One technique for improving AI safety in high-stakes settings is adversarial training, which uses an adversary to generate examples to train on in order to achieve better worst-case performance.
In this work, we used a safe language generation task (``avoid injuries'') as a t…
▽ More
In the future, powerful AI systems may be deployed in high-stakes settings, where a single failure could be catastrophic. One technique for improving AI safety in high-stakes settings is adversarial training, which uses an adversary to generate examples to train on in order to achieve better worst-case performance.
In this work, we used a safe language generation task (``avoid injuries'') as a testbed for achieving high reliability through adversarial training. We created a series of adversarial training techniques -- including a tool that assists human adversaries -- to find and eliminate failures in a classifier that filters text completions suggested by a generator. In our task, we determined that we can set very conservative classifier thresholds without significantly impacting the quality of the filtered outputs. We found that adversarial training increased robustness to the adversarial attacks that we trained on -- doubling the time for our contractors to find adversarial examples both with our tool (from 13 to 26 minutes) and without (from 20 to 44 minutes) -- without affecting in-distribution performance.
We hope to see further work in the high-stakes reliability setting, including more powerful tools for enhancing human adversaries and better ways to measure high levels of reliability, until we can confidently rule out the possibility of catastrophic deployment-time failures of powerful models.
△ Less
Submitted 9 November, 2022; v1 submitted 3 May, 2022;
originally announced May 2022.
-
Water Self-Dissociation is Insensitive to Nanoscale Environments
Authors:
Solana Di Pino,
Yamila A. Perez Sirkin,
Uriel N. Morzan,
Verónica M. Sánchez,
Ali Hassanali,
Damian A. Scherlis
Abstract:
Nanoconfinement effects on water dissociation and reactivity remain controversial, despite their importance to understand the aqueous chemistry at interfaces, pores, or aerosols. The pKw in confined environments has been assessed from experiments and simulations in a few specific cases, leading to dissimilar conclusions. Here, with the use of carefully designed \textit{ab-initio} simulations, we d…
▽ More
Nanoconfinement effects on water dissociation and reactivity remain controversial, despite their importance to understand the aqueous chemistry at interfaces, pores, or aerosols. The pKw in confined environments has been assessed from experiments and simulations in a few specific cases, leading to dissimilar conclusions. Here, with the use of carefully designed \textit{ab-initio} simulations, we demonstrate that the energetics of bulk water dissociation is conserved intact to unexpectedly small length-scales, down to aggregates of only a dozen molecules or pores of widths below 2 nm. The reason is that most of the free-energy involved in water autoionization comes from breaking the O-H covalent bond, which has a comparable barrier in the bulk liquid, in a small droplet of nanometer size, or in a nanopore in the absence of strong interfacial interactions. Thus, dissociation free-energy profiles in nanoscopic aggregates or in 2D slabs of 1 nm width reproduce the behavior corresponding to the bulk liquid, regardless of whether the corresponding nanophase is delimited by a solid or a gas interface. The present work provides a definite and fundamental description of the mechanism and thermodynamics of water dissociation at different scales with broader implications on reactivity and self-ionization at the air-liquid interface.
△ Less
Submitted 6 July, 2023; v1 submitted 26 April, 2021;
originally announced April 2021.
-
On Preheating in Higgs Inflation
Authors:
Yuta Hamada,
Kiyoharu Kawana,
Adam Scherlis
Abstract:
Recently, the problem of unitarity violation during the preheating stage of Higgs inflation with a large non-minimal coupling has been much discussed in the literature. We point out that this problem can be translated into a strong coupling problem for the dimensionless effective coupling, and that the existence of these problems is highly dependent on the choice of higher-dimensional operators be…
▽ More
Recently, the problem of unitarity violation during the preheating stage of Higgs inflation with a large non-minimal coupling has been much discussed in the literature. We point out that this problem can be translated into a strong coupling problem for the dimensionless effective coupling, and that the existence of these problems is highly dependent on the choice of higher-dimensional operators because they can significantly change the background dynamics and the canonical normalization of the fluctuations around it. Correspondingly, the typical energy of particles produced during the first stage of preheating can remain comparable to or below the cutoff scale of the theory. As an example, we numerically calculate the particle production in the presence of a specific four-derivative operator of the Higgs field, and confirm the statement above. Our argument also applies to multi-field inflation with non-minimal couplings.
△ Less
Submitted 8 February, 2021; v1 submitted 9 July, 2020;
originally announced July 2020.
-
Driven Liouville-von Neumann Equation for Quantum Transport and Multiple-Probe Green's Functions
Authors:
Francisco Ramírez,
Daniel Dundas,
Cristián G. Sánchez,
Damian A. Scherlis,
Tchavdar N. Todorov
Abstract:
The so called Driven Liouville-von Neumann equation is a dynamical formulation to simulate a voltage bias across a molecular system and to model a time-dependent current in a grand-canonical framework. This approach introduces a damping term in the equation of motion that drives the charge to a reference, out of equilibrium density. Originally proposed by Horsfield and co-workers, further work on…
▽ More
The so called Driven Liouville-von Neumann equation is a dynamical formulation to simulate a voltage bias across a molecular system and to model a time-dependent current in a grand-canonical framework. This approach introduces a damping term in the equation of motion that drives the charge to a reference, out of equilibrium density. Originally proposed by Horsfield and co-workers, further work on this scheme has led to different coexisting versions of this equation. On the other hand, the multiple-probe scheme devised by Todorov and collaborators, known as the hairy-probes method, is a formal treatment based on Green's functions that allows to fix the electrochemical potentials in two regions of an open quantum system. In this article, the equations of motion of the hairy probes formalism are rewritten to show that, under certain conditions, they can assume the same algebraic structure as the Driven Liouville-von Neumann equation in the form proposed by Morzan et al. [J. Chem. Phys. 2017, 146, 044110]. In this way, a new formal ground is provided for the latter, identifying the origin of every term. The performance of the different methods are explored using tight-binding time-dependent simulations in three trial structures, designated as ballistic, disordered, and resonant models. In the context of first-principles Hamiltonians the Driven Liouville-von Neumann approach is of special interest, because it does not require the calculation of Green's functions. Hence, the effects of replacing the reference density based on the Green's function by one obtained from an applied field are investigated, to gain a deeper understanding of the limitations and the range of applicability of the Driven Liouville-von Neumann equation.
△ Less
Submitted 10 May, 2019;
originally announced May 2019.
-
The Goldilocks zone: Towards better understanding of neural network loss landscapes
Authors:
Stanislav Fort,
Adam Scherlis
Abstract:
We explore the loss landscape of fully-connected and convolutional neural networks using random, low-dimensional hyperplanes and hyperspheres. Evaluating the Hessian, $H$, of the loss function on these hypersurfaces, we observe 1) an unusual excess of the number of positive eigenvalues of $H$, and 2) a large value of $\mathrm{Tr}(H) / ||H||$ at a well defined range of configuration space radii, co…
▽ More
We explore the loss landscape of fully-connected and convolutional neural networks using random, low-dimensional hyperplanes and hyperspheres. Evaluating the Hessian, $H$, of the loss function on these hypersurfaces, we observe 1) an unusual excess of the number of positive eigenvalues of $H$, and 2) a large value of $\mathrm{Tr}(H) / ||H||$ at a well defined range of configuration space radii, corresponding to a thick, hollow, spherical shell we refer to as the \textit{Goldilocks zone}. We observe this effect for fully-connected neural networks over a range of network widths and depths on MNIST and CIFAR-10 datasets with the $\mathrm{ReLU}$ and $\tanh$ non-linearities, and a similar effect for convolutional networks. Using our observations, we demonstrate a close connection between the Goldilocks zone, measures of local convexity/prevalence of positive curvature, and the suitability of a network initialization. We show that the high and stable accuracy reached when optimizing on random, low-dimensional hypersurfaces is directly related to the overlap between the hypersurface and the Goldilocks zone, and as a corollary demonstrate that the notion of intrinsic dimension is initialization-dependent. We note that common initialization techniques initialize neural networks in this particular region of unusually high convexity/prevalence of positive curvature, and offer a geometric intuition for their success. Furthermore, we demonstrate that initializing a neural network at a number of points and selecting for high measures of local convexity such as $\mathrm{Tr}(H) / ||H||$, number of positive eigenvalues of $H$, or low initial loss, leads to statistically significantly faster training on MNIST. Based on our observations, we hypothesize that the Goldilocks zone contains an unusually high density of suitable initialization configurations.
△ Less
Submitted 12 November, 2018; v1 submitted 6 July, 2018;
originally announced July 2018.
-
The Stochastic Axion Scenario
Authors:
Peter W. Graham,
Adam Scherlis
Abstract:
For the minimal QCD axion model it is generally believed that overproduction of dark matter constrains the axion mass to be above a certain threshold, or at least that the initial misalignment angle must be tuned if the mass is below that threshold. We demonstrate that this is incorrect. During inflation, if the Hubble scale is low, the axion tends toward an equilibrium. This means the minimal QCD…
▽ More
For the minimal QCD axion model it is generally believed that overproduction of dark matter constrains the axion mass to be above a certain threshold, or at least that the initial misalignment angle must be tuned if the mass is below that threshold. We demonstrate that this is incorrect. During inflation, if the Hubble scale is low, the axion tends toward an equilibrium. This means the minimal QCD axion can naturally give the observed dark matter abundance in the entire lower part of the mass range, down to masses $\sim 10^{-12}$ eV (or $f_a$ up to almost the Planck scale). The axion abundance is generated by quantum fluctuations of the field during inflation. This mechanism generates cold dark matter with negligible isocurvature perturbations. In addition to the QCD axion, this mechanism can also generate a cosmological abundance of axion-like particles and other light fields.
△ Less
Submitted 15 August, 2018; v1 submitted 18 May, 2018;
originally announced May 2018.
-
A Quantum-Mechanics Molecular-Mechanics scheme for extended systems
Authors:
Diego Hunt,
Veronica M. Sanchez,
Damian A. Scherlis
Abstract:
We introduce and discuss a hybrid quantum-mechanics molecular-mechanics (QM-MM) approach for Car-Parrinello DFT simulations with pseudopotentials and planewaves basis, designed for the treatment of periodic systems. In this implementation the MM atoms are considered as additional QM ions having fractional charges of either sign, which provides conceptual and computational simplicity by exploiting…
▽ More
We introduce and discuss a hybrid quantum-mechanics molecular-mechanics (QM-MM) approach for Car-Parrinello DFT simulations with pseudopotentials and planewaves basis, designed for the treatment of periodic systems. In this implementation the MM atoms are considered as additional QM ions having fractional charges of either sign, which provides conceptual and computational simplicity by exploiting the machinery already existing in planewave codes to deal with electrostatics in periodic boundary conditions. With this strategy, both the QM and MM regions are contained in the same supercell, which determines the periodicity for the whole system. Thus, while this method is not meant to compete with non-periodic QM-MM schemes able to handle extremely large but finite MM regions, it is shown that for periodic systems of a few hundred atoms, our approach provides substantial savings in computational times by treating classically a fraction of the particles. The performance and accuracy of the method is assessed through the study of energetic, structural, and dynamical aspects of the water dimer and of the aqueous bulk phase. Finally, the QM-MM scheme is applied to the computation of the vibrational spectra of water layers adsorbed at the TiO$_2$ anatase (101) solid-liquid interface. This investigation suggests that the inclusion of a second monolayer of H$_2$O molecules is sufficient to induce on the first adsorbed layer, a vibrational dynamics similar to that taking place in the presence of an aqueous environment. The present QM-MM scheme appears as a very interesting tool to efficiently perform molecular dynamics simulations of complex condensed matter systems, from solutions to nanoconfined fluids to different kind of interfaces.
△ Less
Submitted 29 April, 2016;
originally announced April 2016.
-
Electron transport in real time from first-principles
Authors:
Uriel N. Morzan,
Francisco F. Ramirez,
Mariano C. Gonzalez Lebrero,
Damian A. Scherlis
Abstract:
While the vast majority of calculations reported on molecular conductance have been based on the static non-equilibrium Green's function formalism combined with density functional theory, in recent years a few time-depedent approaches to transport have started to emerge. Among these, the driven Liouville-von Neumann equation (J. Chem. Phys. 124, 214708 (2006)) is a simple and appealing route relyi…
▽ More
While the vast majority of calculations reported on molecular conductance have been based on the static non-equilibrium Green's function formalism combined with density functional theory, in recent years a few time-depedent approaches to transport have started to emerge. Among these, the driven Liouville-von Neumann equation (J. Chem. Phys. 124, 214708 (2006)) is a simple and appealing route relying on a tunable rate parameter, which has been explored in the context of semi-empirical methods. In the present study, we adapt this formulation to a density functional theory framework and analyze its performance. In particular, it is implemented in an efficient all-electron DFT code with Gaussian basis functions, suitable for quantum-dynamics simulations of large molecular systems. At variance with the case of the tight-binding calculations reported in the literature, we find that now the initial perturbation to drive the system out of equilibrium plays a fundamental role in the stability of the electron dynamics, and that the equation of motion used in previous tight-binding implementations has to be modified to conserve the total number of particles during time propagation. Moreover, we propose a procedure to get rid of the dependence of the current-voltage curves on the rate parameter. This method is employed to obtain the current-voltage characteristic of saturated and unsaturated hydrocarbons of different lenghts, with very promising prospects.
△ Less
Submitted 23 April, 2016;
originally announced April 2016.
-
Cluster Algebras and Symmetries of Regular Tilings
Authors:
Adam Scherlis
Abstract:
The classification of Grassmannian cluster algebras resembles that of regular polygonal tilings. We conjecture that this resemblance may indicate a deeper connection between these seemingly unrelated structures.
The classification of Grassmannian cluster algebras resembles that of regular polygonal tilings. We conjecture that this resemblance may indicate a deeper connection between these seemingly unrelated structures.
△ Less
Submitted 27 October, 2015;
originally announced October 2015.
-
Hedgehog Bases for A_n Cluster Polylogarithms and An Application to Six-Point Amplitudes
Authors:
Daniel E. Parker,
Adam Scherlis,
Marcus Spradlin,
Anastasia Volovich
Abstract:
Multi-loop scattering amplitudes in N=4 Yang-Mills theory possess cluster algebra structure. In order to develop a computational framework which exploits this connection, we show how to construct bases of Goncharov polylogarithm functions, at any weight, whose symbol alphabet consists of cluster coordinates on the $A_n$ cluster algebra. Using such a basis we present a new expression for the 2-loop…
▽ More
Multi-loop scattering amplitudes in N=4 Yang-Mills theory possess cluster algebra structure. In order to develop a computational framework which exploits this connection, we show how to construct bases of Goncharov polylogarithm functions, at any weight, whose symbol alphabet consists of cluster coordinates on the $A_n$ cluster algebra. Using such a basis we present a new expression for the 2-loop 6-particle NMHV amplitude which makes some of its cluster structure manifest.
△ Less
Submitted 27 October, 2015; v1 submitted 7 July, 2015;
originally announced July 2015.
-
First-principles molecular dynamics simulations at solid-liquid interfaces with a continuum solvent
Authors:
Veronica M. Sanchez,
Mariela Sued,
Damian A. Scherlis
Abstract:
Continuum solvent models have become a standard technique in the context of electronic structure calculations, yet, no implementations have been reported capable to perform molecular dynamics at solid-liquid interfaces. We propose here such a continuum approach in a DFT framework, using plane-waves basis sets and periodic boundary conditions. Our work stems from a recent model designed for Car-P…
▽ More
Continuum solvent models have become a standard technique in the context of electronic structure calculations, yet, no implementations have been reported capable to perform molecular dynamics at solid-liquid interfaces. We propose here such a continuum approach in a DFT framework, using plane-waves basis sets and periodic boundary conditions. Our work stems from a recent model designed for Car-Parrinello simulations of quantum solutes in a dielectric medium [J. Chem. Phys. 124, 74103 (2006)], for which the permittivity of the solvent is defined as a function of the electronic density of the solute. This strategy turns out to be inadequate for systems extended in two dimensions, by introducing new term in the Kohn-Sham potential which becomes unphysically large at the interfacial region, seriously affecting the convergence. If the dielectric medium is properly redefined as a function of the atomic coordinates, a good convergence is obtained and the constant of motion is conserved during the molecular dynamics simulations. Moreover, a significant gain in efficiency can be achieved if the simulation box is partitioned in two, solving the Poisson problem separately for the "dry" region using fast Fourier transforms, and for the solvated or "wet" region using a multigrid method. Eventually both solutions are combined in a self-consistent procedure, and in this way Car-Parrinello molecular dynamics simulations of solid-liquid interfaces can be performed at a very moderate computational cost. This scheme is employed to investigate the acid-base equilibrium at the TiO2-water interface.
△ Less
Submitted 30 April, 2009;
originally announced April 2009.
-
Simulation of Heme using DFT+U: a step toward accurate spin-state energetics
Authors:
Damian A. Scherlis,
Matteo Cococcioni,
Patrick Sit,
Nicola Marzari
Abstract:
We investigate the DFT+U approach as a viable solution to describe the low-lying states of ligated and unligated iron heme complexes. Besides their central role in organometallic chemistry, these compounds represent a paradigmatic case where LDA, GGA, and common hybrid functionals fail to reproduce the experimental magnetic splittings. In particular, the imidazole pentacoordinated heme is incorr…
▽ More
We investigate the DFT+U approach as a viable solution to describe the low-lying states of ligated and unligated iron heme complexes. Besides their central role in organometallic chemistry, these compounds represent a paradigmatic case where LDA, GGA, and common hybrid functionals fail to reproduce the experimental magnetic splittings. In particular, the imidazole pentacoordinated heme is incorrectly described as a triplet by all usual DFT flavors. In this study we show that a U parameter close to 4 eV leads to spin transitions and molecular geometries in quantitative agreement with experiments, and that DFT+U represents an appealing tool in the description of iron porphyrin complexes, at a much reduced cost compared to correlated quantum-chemistry methods. The possibility of obtaining the U parameter from first-principles is explored through a self-consistent linear-response formulation. We find that this approach, which proved to be successful in other iron systems, produces in this case some overestimation with respect to the optimal values of U.
△ Less
Submitted 18 May, 2007;
originally announced May 2007.
-
The Effect of Counterions on the Interactions of Charged Oligothiophenes
Authors:
Nicholas E. Singh-Miller,
Damian A. Scherlis,
Nicola Marzari
Abstract:
The functionality of conjugated polymer systems often relies on oxidations or reductions, in most cases mediated by the presence of counterions. The effect that the common counterion hexafluorophosphate (PF6-) has on the intermolecular interactions between charged oligothiophenes is investigated here using ab initio quantum chemistry methods. Counterions are explicitly included in the simulation…
▽ More
The functionality of conjugated polymer systems often relies on oxidations or reductions, in most cases mediated by the presence of counterions. The effect that the common counterion hexafluorophosphate (PF6-) has on the intermolecular interactions between charged oligothiophenes is investigated here using ab initio quantum chemistry methods. Counterions are explicitly included in the simulations of oxidized oligothiophenes and in the dimerization process. Our calculations provide quantitative and qualitative insight into the intermolecular interactions in oligothiophene-counterion systems and show that the intermolecular pi-stacking of oligothiophenes is not adversely affected by the presence of counterions, and that in fact oligothiophene dimerization is further stabilized by their presence.
△ Less
Submitted 28 September, 2006;
originally announced September 2006.
-
Density functional theory in transition-metal chemistry: a self-consistent Hubbard U approach
Authors:
Heather J. Kulik,
Matteo Cococcioni,
Damian A. Scherlis,
Nicola Marzari
Abstract:
Transition-metal centers are the active sites for many biological and inorganic chemical reactions. Notwithstanding this central importance, density-functional theory calculations based on generalized-gradient approximations often fail to describe energetics, multiplet structures, reaction barriers, and geometries around the active sites. We suggest here an alternative approach, derived from the…
▽ More
Transition-metal centers are the active sites for many biological and inorganic chemical reactions. Notwithstanding this central importance, density-functional theory calculations based on generalized-gradient approximations often fail to describe energetics, multiplet structures, reaction barriers, and geometries around the active sites. We suggest here an alternative approach, derived from the Hubbard U correction to solid-state problems, that provides an excellent agreement with correlated-electron quantum chemistry calculations in test cases that range from the ground state of Fe$_2$ and Fe$_2^-$ to the addition-elimination of molecular hydrogen on FeO$^+$. The Hubbard U is determined with a novel self-consistent procedure based on a linear-response approach.
△ Less
Submitted 12 August, 2006;
originally announced August 2006.
-
Stacking of oligo and polythiophenes cations in solution: surface tension and dielectric saturation
Authors:
Damian A. Scherlis,
Jean-Luc Fattebert,
Nicola Marzari
Abstract:
The stacking of positively charged (or doped) terthiophene oligomers and quaterthiophene polymers in solution is investigated applying a recently developed unified electrostatic and cavitation model for first-principles calculations in a continuum solvent. The thermodynamic and structural patterns of the dimerization are explored in different solvents, and the distinctive roles of polarity and s…
▽ More
The stacking of positively charged (or doped) terthiophene oligomers and quaterthiophene polymers in solution is investigated applying a recently developed unified electrostatic and cavitation model for first-principles calculations in a continuum solvent. The thermodynamic and structural patterns of the dimerization are explored in different solvents, and the distinctive roles of polarity and surface tension are characterized and analyzed. Interestingly, we discover a saturation in the stabilization effect of the dielectric screening that takes place at rather small values of $ε_0$. Moreover, we address the interactions in trimers of terthiophene cations, with the aim of generalizing the results obtained for the dimers to the case of higher-order stacks and nanoaggregates.
△ Less
Submitted 9 December, 2005;
originally announced December 2005.
-
A unified electrostatic and cavitation model for first-principles molecular dynamics in solution
Authors:
Damian A. Scherlis,
Jean-Luc Fattebert,
Francois Gygi,
Matteo Cococcioni,
Nicola Marzari
Abstract:
The electrostatic continuum solvent model developed by Fattebert and Gygi is combined with a first-principles formulation of the cavitation energy based on a natural quantum-mechanical definition for the surface of a solute. Despite its simplicity, the cavitation contribution calculated by this approach is found to be in remarkable agreement with that obtained by more complex algorithms relying…
▽ More
The electrostatic continuum solvent model developed by Fattebert and Gygi is combined with a first-principles formulation of the cavitation energy based on a natural quantum-mechanical definition for the surface of a solute. Despite its simplicity, the cavitation contribution calculated by this approach is found to be in remarkable agreement with that obtained by more complex algorithms relying on a large set of parameters. Our model allows for very efficient Car-Parrinello simulations of finite or extended systems in solution, and demonstrates a level of accuracy as good as that of established quantum-chemistry continuum solvent methods. We apply this approach to the study of tetracyanoethylene dimers in dichloromethane, providing valuable structural and dynamical insights on the dimerization phenomenon.
△ Less
Submitted 6 October, 2005;
originally announced October 2005.
-
Pi-stacking in charged thiophene oligomers
Authors:
Damian A. Scherlis,
Nicola Marzari
Abstract:
The pi-stacking of oxidized thiophene oligomers is investigated using ab-initio quantum chemistry methods. Dimers of singly-charged oligothiophenes are found to be unstable in the gas phase, but can be stabilized as bound dications in the singlet state by a polarizable solvent such as acetonitrile. Our calculations provide a detailed description of the mechanisms and the energetics involved in t…
▽ More
The pi-stacking of oxidized thiophene oligomers is investigated using ab-initio quantum chemistry methods. Dimers of singly-charged oligothiophenes are found to be unstable in the gas phase, but can be stabilized as bound dications in the singlet state by a polarizable solvent such as acetonitrile. Our calculations provide a detailed description of the mechanisms and the energetics involved in the dimerization phenomenon, and highlight the role and importance of the environment in the stabilization of the stacks. The need for accurate treatments of electronic correlations and of solvation effects for a realistic description of these materials is underscored.
△ Less
Submitted 13 September, 2004;
originally announced September 2004.
