First principal calculations of optical and electrical properties of Sc

First principal calculations of optical and electrical properties of Sc, Ti
and V doped SnO2 used in photovoltaic applications
M. Boujnah1,*, A. Benyoussef 1, A. El Kenz1 and M. Loulidi1
1
Laboratory of Magnetism and Physics of High Energies, Department of Physics, B.P. 1014, Faculty of Sciences,
Mohammed V University, Rabat, Morocco
* Corresponding author: boujnah.mourad@gmail.com
Abstract:
The full-potential linearized augmented plane wave method (FP-LAPW) based on the density
functional theory (DFT) and Boltzmann's Transport theory, are employed to investigate
theoretically the electronic structure, optical and electrical properties of Sc, Ti and V doped rutile
SnO2. The FP-LAPW based on the new potential approximation known as the Tran–Blaha
modified Becke–Johnson exchange potential approximation (mBJ). The calculated band
structure and density of states (DOS) exhibit a band gap of pure SnO2 (3.3 eV) closer to the
experimental one. As well, our results indicate that the average transmittance in the 400 to 1000
nm wavelength region was 90%. The high transmittance and electrical conductivity indicate that
hexagonal doped SnO2 system is a potential as material for solar energy applications.
Key words: FP-LAPW, mBJ, transparent conductive oxide, electrical conductivity,
Transmittance.
A flexible implementation of the 2PT method for calculating thermodynamic properties from molecular dynamics simulations
Miguel A. Caro,1,2,⇤ Jouko Lehtomäki,1 Tomi Laurila,2 and Olga Lopez-Acevedo1
1
COMP Centre of Excellence in Computational Nanoscience, Department of Applied Physics,
Aalto University, Espoo, Finland
2
Department of Electrical Enegineering and Automation, Aalto University, Espoo, Finland
⇤
mcaroba@gmail.com; miguel.caro@aalto.fi
The 2PT method [1] is an approach that allows to calculate thermodynamic properties of
a molecular system (e.g., its entropy) based on its density of states (DoS) [2]. The DoS of
a system is accessible from molecular dynamics (MD) simulations by simply calculating the
velocity spectrum directly from the atomic trajectories. In some cases converged results can
be obtained from simulations as short as 20 ps. With current computer power this regime is
accessible to not only classical MD but also ab initio MD.
We will present our new implementation of the 2PT method, DoSPT [3], which is designed
to be highly flexible and suitably geared for obtaining free energies of solvated systems. We
will give a brief update on the state of the porting of our DoSPT code (written in Fortran) to
Python, which will allow native ASE integration (currently on an ASE development branch).
Finally, we will show an application of our code to computing accurate electrochemical reduction
potentials from first principles simulations [3].
Figure 1: DoS for the oxidized and reduced forms of ferrocene methanol solvated in water. Within the
2PT framework, thermodynamic properties can be obtained as functionals of the DoS.
[1]
[2]
[3]
S.-T. Lin, M. Blanco, and W. A. Goddard III, “The two-phase model for calculating thermodynamic properties of liquids from molecular dynamics: Validation for the phase diagram of
Lennard-Jones fluids”, J. Chem. Phys. 119, 11792 (2003).
P. H. Berens, D. H. J. Mackay, G. M. White, and K. R. Wilson, “Thermodynamics and quantum
corrections from molecular dynamics for liquid water”, J. Chem. Phys. 79, 2375 (1983).
M. A. Caro, T. Laurila, and O. Lopez-Acevedo, “in preparation”, , (2016).
1
Charge Transfer Excitons in Organic Semiconductors from
First Principle : The Challenges
Reyhaneh Ghassemizadeh, Michael Walter
Photocurrent generation in organic photovoltaics (OPVs) relies on the dissociation
of the excitons into free electrons and holes at donor/acceptor heterointerfaces.
The low dielectric constant of organic semiconductors leads to strong Coulomb
interactions between electron-hole pairs that should in principle oppose the
generation of free charges. The exact mechanism by which electrons and holes
overcome this Coulomb trapping is still unclear. Our goal is to gain a deeper
understanding of the nature of CT excitons in a donor/acceptor interface as present
in organic solar cells. Single donor/acceptor pairs may be trapped in cold
hydrogen or rare-gas matrix, where spectra can be recorded in unprecedented
precision. We aim to look precisely into the vibronic transitions important in CT
excitons as their fingerprints. The theoretical approach in this study is Density
Functional Theory (DFT) for the ground state and Time Dependent DFT
(TDDFT) for excited states. The accurate description of van der Waals (vdW)
interaction between molecules is needed also.
Success stories and some challenges with reactivity studies of metals and oxides
Karoliina Honkala
Department of Chemistry, Nanoscience Center
P.O. Box 35
40014 University of Jyväskylä
Finland
In my presentation, I will give an overview on our recent activity with metal and oxide surfaces and
bulk structures. In particular, I will discuss our efforts on understanding chemoselectivity in simple
organic transformations over closed-packed transition metal surfaces and how this can be
improved. The oxides systems to be addressed range from activity of Au clusters on metalsupported ultrathin oxide films to reducibility and oxygen mobility in transition metal oxides.
Powerful post-processing tool, advanced QM/MM interface and
self-interaction correction
Elvar Örn Jónsson
Department of Applied Physics, Aalto University,
FIN-00076 Aalto, Finland
eojons@gmail.com
We describe three implementations in GPAW: (1) improved post-processing tool to obtain
localized orbitals and atomic charges from calculated results, (2) coupling with a semiempirical force field in a self-consistent QM/MM scheme, and (3) improved accuracy of the
calculations by applying explicit self-interaction correction. The improved post-processing
tool is based on a generalized Pipek-Mezey charge localization scheme and is shown to have
several advantages over Bader analysis and Wannier functions [1]. The QM/MM
implementation involves at this stage the combination of DFT and the single center multipole
expansion (SCME) formalism for polarizable potential functions which is currently available
for water molecules [2]. The self-interaction correction has been implemented in a fully
variational and self-consistent way with complex optimal orbitals and has been applied to
both molecular [3] and solid state systems [4].
[1] Pipek–Mezey orbital localization using various partial charge estimates, S. Lehtola
and H. Jónsson, Journal of Chemical Theory and Computation 10, 642 (2014).
[2] A Transferable H2O Interaction Potential Based on a Single Center Multipole
Expansion: SCME, K. T. Wikfeldt, E. R. Batista, F. D. Vila and H. Jónsson, Physical
Chemistry Chemical Physics 15, 16542 (2013).
[3] Charge Localization in a diamine cation: A rigorous test of energy functionals for
electronic systems, X. Cheng, Y. Zhang, E.Ö. Jónsson, H. Jónsson and P.M. Weber,
Nature Communications 7, 11013 (2016).
[4] Calculations of Al dopant in alpha-quartz using a variational implementation of the
Perdew-Zunger self-interaction correction, H. Gudmundsdóttir, E. Ö. Jónsson and H.
Jónsson, New Journal of Physics 17, 083006 (2015)
Analyzing genetic algorithm behavior using clustering
Mathias Jørgensen1*, Bjørk Hammer1
1
iNANO, Aarhus University
* mj@inano.au.dk
A genetic algorithm (GA) often produces a large amount of intermediate data before
converging to the global minimum. Hidden within this intermediate data lie trends
that can reveal important information about GA performance and behavior. In the
case of optimizing atomic structures, the intermediate data represents a vast
configuration space of atomic positions of local minima structures. We are using
hierarchical clustering to develop a tool for analyzing these structures. Looking at
C9H7N data from a large number of convergence attempts, we find that the structures
have a significant tendency to form clusters where the similarity of two structures is
based on interatomic carbon-carbon and carbon-nitrogen distances. With this new
knowledge of geometric characteristics in different regions of configuration space, we
are able to track which regions the GA visits over the course of an entire convergence
attempt. This includes analyzing behavior in converged versus non-converged
attempts where we find that certain clusters are restraining the GA search. In the
future, we will utilize this type of clustering analysis to rationally improve the GA
design.
Figure: Progression of the genetic algorithm searching for the global minimum
(quinoline).
Gold-doped Ag29-xAuxnanoclusters
Rosalba Juarez-Mosqueda*‡, Sami Malola‡ and Hannu Häkkinen‡
‡
Department of Physics and Department of Chemistry, Nanoscience Center, University of
Jyväskylä, FI-40014 Jyväskylä, Finland
* rosalba.r.juarez-mosqueda@jyu.fi
ABSTRACT
Pure and Ag-doped Ag29-xAux nanoparticles protected by 1,3-benzenedithiol (BDT) and
triphenylphosphine (TPP) have been recently synthesized and proven to be stable under
ambient conditions. Presumably, the incorporation of 1-5 Au atoms in the [Ag29xAux(BDT)12(TPP)4]
-3
nanoclusters
(NCs)
enhance
noticeably
their
stability
and
photoluminescence (PL) quantum yield (QY). Herein, we used density functional theory
(DFT) to investigate the stability and optical properties of [Ag29-xAux(BDT)12(TPP)4]-3NCs
with x=0-5. Our results confirm that at room temperature, all the possibleAu-doped isomers
can co-exist in the synthetic mixture, since the maximum energy difference between the
diverse isomers is only ~1.6 meV. We found that the increase in the stability of the Audoped NCs is due to the marked tendency of Au to attract electronic charge density from
the neighboring atoms that, as consequence, renders the Au-P bonds ~0.5 eV stronger than
the analogous Ag-P bonding. Furthermore, the large positive charge around the P atoms
that are bonded to Au, indicates the enhancement of the ligand-to-metal charge transfer
(LMCT), suggesting that the interactions between the doping Au and the TTP ligands could
play an important role in the enhancement of the PL QY.
Pyridine adsorption and diffusion on Pt(111)
investigated with Density functional theory
Esben Leonhard Kolsbjerg
Interdisciplinary Nanoscience Center (iNANO)
Gustav Wieds Vej 14
Aarhus University
DK-8000 Aarhus C
ABSTRACT
On late transition metal surfaces, the binding of aromatic molecules is a key step in many
industrial processes, including the sensitive chemicals production in the pharmaceutical and
petroleum refining by catalysis1. The classical aromatic molecule benzene on Pt(111) is one
such system that has been studied intensively. This system now serves as a benchmark for
the assessment of theoretical methods2. To supplement this established literature this
project, take on the slightly more involved aromatic molecule pyridine, C5H5N. A replacement
of a CH by a secondary amine group make pyridine less symmetric than benzene.
Interesting aspects comes with the lone pair on N and adds changes to the proven energy
balance upon adsorption, diffusion and reaction. Both the aromatic π-system and the lone
pair can now be the key interaction in reactions.
Pyridine by itself has many interesting applications, this include both acting as solvent for
chemical reactions3, as well as test subject for the development of Surface-Enhanced Raman
Scattering4, and its chemical variants are used as a precursors in the synthesis of many
different pharmaceuticals5.
With van der Waals-corrected Density Functional Theory (vdW-DFT) the adsorption, diffusion
and dissociation of pyridine, C5H5N, on Pt(111) is investigated. A systematic and intensive
search in the adsorption potential energy landscape for local minima reveals adsorption
parallel to the surface for the intact pyridine. With a network of interconnected local minima,
the most favorable diffusion path for pyridine is found to have a 0.53 eV barrier. The
preferred path is composed of small single rotational steps with a carbon-carbon double
bond hinged above a single Pt atom while the pyridine remains parallel to the surface. The
source of the preferred pathway is argued to follow from the π-bond for C2-Pt being stronger
than the equivalent CN-Pt. Our calculations support previous experimental observations from
the literature for pyridine on Pt(111).
REFERENCES
1.
2.
3.
4.
5.
R. A. Sheldon and H. V. Bekkum, John Wiley & Sons (2008).
M. Saeys, M.-F. Reyniers, G. B. Marin, and M. Neurock, J. Phys. Chem. B, 106, 7489 (2002)
A. P. Tanberg, J. Am. Chem. Soc., 36, 335 (1914).
M. Moskovits, Rev. Mod. Phys., 57, 783 (1985).
J. Magano and R. D. Joshua, Chem. Rev., 111, 2177 (2011).
Analyzing (mostly Plasmonic) Excitations with Real-time TDDFT
Mikael Kuismaa), Tuomas Rossib), Paul Erharta)
a) Department of Physics, Chalmers University of Technology, SE-412 96 Gothenburg,
Sweden
b) COMP Centre of Excellence, Department of Applied Physics, Aalto University School
of Science, FI-00076 AALTO, Finland
We have recently implemented a very efficient time-dependent density functional theory code
based on localized basis functions [1]. In combination with GLLB-SC potential [2], the
method yields very promising results for plasmons of noble metal clusters. [1] In this talk, I
will summarize the main results and principles for efficient simulation of plasmons of noble
metal particles. Although the concept of plasmon can be fairly simply understood in terms of
charge density fluctuations, the details of build-up of a plasmon resonance are easily hidden if
only absorption spectrum or even time-dependent induced potential is available. A Kohn–
Sham decomposition of transition density matrix technique is presented in context of real time
TD-DFT, which allows equivalent assignment of the electron-hole origin of excitations than
quantum chemistry field is accustomed to with Casida-type response equations. To test the
robustness of the method, we will show the equivalence of these approaches to the ’orbital
assignment problem’ by analyzing excitations and comparing to recently computed and
synthetized molecular solar thermal energy storage systems [3]. At the last part of the talk,
transition density matrices are used to analyze and characterize plasmonic excitations on
various systems.
Figure 1: The most dominant hole- and electron orbital and their transition density for
plasmon peak of a small silver cluster.
[1] Localized surface plasmon resonance in silver nanoparticles: Atomistic first-principles
time-dependent density-functional theory calculations – M. Kuisma, A. Sakko, T. P. Rossi, A.
H. Larsen, J. Enkovaara, L. Lehtovaara, and T. T. Rantala – Phys. Rev. B 91, 115431 (2015)
[2] Kohn-Sham potential with discontinuity for band gap materials – M. Kuisma, J. Ojanen, J.
Enkovaara, and T. T. Rantala — Phys. Rev. B 82, 115106 (2010)
[3] Comparative Ab-Initio Study of Substituted Norbornadiene-Quadricyclane Compounds
for Solar Thermal Storage — M. Kuisma, A. Lundin, K. Moth-Poulsen, P. Hyldgaard, and P.
Erhart — J. Phys. Chem. C, 120, 3635 (2016)
Efficient van der Waals functionals and other GPAW
developments
Ask Hjorth Larsen1,2 , Mikael Kuisma3
Nano-Bio Spectroscopy Group and European Theoretical Spectroscopy Facility (ETSF), Universidad del Paı́s Vasco, CFM CSIC–UPV/EHU–MPC & DIPC,
20018 San Sebastián, Spain
2
Departament de Quı́mica Fı́sica & Institut de Quı́mica Teòrica i Computacional
(IQTCUB), Universitat de Barcelona, c/Martı́ i Franquès 1, 08028 Barcelona,
Spain
3
Department of Physics, Chalmers University of Technology, SE-412 96 Gothenburg, Sweden
1
The Chalmers–Rutgers van der Waals density functionals (vdW-DF) have been
available and in use for a long time in GPAW. Typically, vdW-DFs are useful for
describing weak bonds between components that are relatively stable by themselves, such as intermolecular bonds. vdW-DFs have an explicitly non-local correlation energy which is considerably more complicated to evaluate than a semilocal functional. With large bond lengths and multicomponent systems, calculations
tend to become quite large, and performance becomes an issue. The old vdW-DF
implementation in GPAW parallelizes only up to 20 cores and therefore cannot
be applied to large systems. We thus present libvdwxc, a scalable library of van
der Waals functionals that aims not to be notably more painful than any semilocal
functional in general.
Furthermore I will briefly present a number of recent GPAW features which
include improved parallelization of exchange–correlation functionals and Poisson
solvers by parallelization over all available cores, as well as the norm-conserving
sg15 pseudopotentials.
1
Designing Excitons in van der Waals Heterostructures
S. Latini⇤ , K. T. Winther, T. Olsen and K.S. Thygesen
Center for Nanostructured Graphene (CNG) and Center for Atomic-scale Materials Design (CAMD),
Department of Physics, Technical University of Denmark, 2800 Kgs. Lyngby, Denmark
*E-mail: simola@fysik.dtu.dk
The possibility of tailoring excitonic properties by vertically stacking two-dimensional layers in
van der Waals heterostructures (vdWHs) represents a new paradigm in material science. Ab-initio
calculations should in principle provide a powerful tool for guiding the design of vdWHs, but in
their traditional form they are limited to commensurable structures with a few layers. Here, we
overcome these limitations rephrasing the exciton many-body problem in terms of an e↵ective 2D
hydrogenic Hamiltonian (Mott-Wannier model) where the interaction between the electron and the
hole is screened by the characteristic spatially non-local dielectric function of vdWHs. We calculate
the dielectric properties using a multiscale approach where the dielectric functions of the individual
layers (the dielectric building blocks) are computed ab-initio and coupled together via the long-range
Coulomb interaction1,2 . The dielectric functions of more than 50 di↵erent 2D materials are available
in an open database together with the software for solving the coupled electrostatic equations3 .
We demonstrate that while for monolayers and few-layers vdWHs the spatial non-locality of the
dielectric function is well described by its linear approximation in reciprocal space1,4,5 , for vdWHs
where the excitonic radius is comparable or smaller than their thickness, it is crucial to account for
non-linearities. With a proper description of the dielectric screening, we are then able to calculate
accurate exciton binding energies in agreement with experiments. As an illustration, we show in
fig. 1, how the peculiar non-Rydberg exciton series in supported WS2 , recently observed experimentally, is well reproduced by our calculations. Our method can even be applied to the design
of more complex many-body excitations where the electron and the hole are localized in distinct
layers, namely inter-layer excitons (see fig. 2 (a)). Beside being a↵ected by the dielectric environment, inter-layer excitons properties can be further tuned engineering the electron-hole separation.
Figure 2 shows, for example, how the inter-layer exciton binding energy in a MoS2 -WSe2 based
heterostructure can be varied over a broad range by simply intercalating a di↵erent number of h-BN
layers.
In summary, our approach is accurate, yet highly efficient and has the merit of providing a seamless
connection between the limit of isolated monolayer materials and the more complex case of van der
Waals heterostructures.
1
2
3
4
5
6
S. Latini, T. Olsen and K.S. Thygesen, Physical Review B, vol. 92, p. 245123, 2015.
K. Andersen, S. Latini, K.S. Thygesen, Nano Letters, vol. 15, p. 4616, 2015.
The dielectric building blocks and the QEH software can be downloaded from https://cmr.fysik.dtu.dk/vdwh/vdwh.html
P. Cudazzo, C. Attaccalite, I. V. Tokaltly and A. Rubio, Physical Review Letters, vol. 104, p. 226804, 2010.
O. Pulci, P. Gori, M. Marsili, V. Garbuio, R. Del Sole and F. Bechstedt,Europhysics Letters, vol. 98, p. 37004, 2012.
A. Chernikov, T.C. Berkelbach, H.M. Hill, A. Rigosi, Y. Li, O.B. Aslan, D.R. Reichman, M.S. Hybertsen, T.F. Heinz, Phys.
Rev. Lett., vol. 113, p. 076802, 2014.
0.0
0.0
(a)
0.1
0.1
n=5
W (r)(eV)
0.3
0.3
0.4
2D Hydrogen
1
2
3
n
4
0.5
0.7
WS2
0.6
n=4
0.4
5
0.8
200
WSe2
MoS2 @WSe2
0.2
0.3
n=3
0.6
Exp.
WS2 @hBN
0.5
Eb(eV)
0.2
Eb(eV)
MoS2
0.1
0.2
(a)
n=2
WS2 @hBN
100
(b)
0.5
0 1 2 3 4 5 6 7 8 9
n=1
1/r
0
100
0.4
(b)
200
r(Å)
FIG. 1: Excitons in supported WS2 . (a) Binding energies
of the lowest five excitons in freestanding WS2 and WS2
on hBN (experimental data from 6). (b) The screened
electron-hole interaction in WS2 monolayer adsorbed on
hBN along with the radial probability distribution of the
first five excitons.
no. hBN
FIG. 2: Inter-layer Excitons in MoS2 -WSe2 based heterostructures. (a) Schematic of the structure. (b)Binding
energy of the lowest intra and inter-layer (green) excitons
as function of the number of sandwiched hBN layers.
Orbital-free density functional theory in GPAW and self-consistent
calculations with kinetic energy functionals
Jouko Lehtomäki1⇤ , Alexander Karpenko1 , Leonardo A. Espinosa Leal1 ,
Miguel A. Caro1 , Olga Lopez-Acevedo1
1
⇤
Aalto University, Department of Applied Physics
jouko.lehtomaki@aalto.fi
Orbital-free density functional theory is an electronic structure method which in theory gives
much better scaling with respect to system size than the usual Kohn-Sham scheme by approximating kinetic energy as an explicit functional of density. We will give a small theoretical introduction
to the modern orbital-free density functional theory (OFDFT) and how GPAW allows us to tackle
problems in the field of OFDFT[1]. GPAW is one of the few calculators that allow us to solve allelectron OFDFT problem self-consistently with linear scaling with respect to the system size. We
discuss the differences between assesing kinetic energy functionals in all electron setting versus
with pseudopotentials.
We present few simple models for the kinetic energy functional. Especially we investigate a
simple functional, which is a linear parametrized combination of Thomas-Fermi and von Weizsäcker
kinetic energy functionals. We present results on the parametrization of this model for atoms[2]
and on the transferability of these parameters to small molecules[3].
[1] Lehtomäki J., Makkonen I., Caro, M., Harju, A., Lopez-Acevedo O., J. Chem. Phys. 141,
234102 (2014).
[2] Espinosa Leal L., Karpenko A., Caro M., Lopez-Acevedo O., Phys. Chem. Chem. Phys., DOI:
10.1039/C5CP01211B (2015).
[3] Karpenko A., Espinosa Leal L., Caro M., Lehtomäki J., Lopez-Acevedo O. (in preparation).
Strong 1D localization and highly electron-hole anisotropic charge
carrier masses in heavy halogen functionalized graphene derivatives
Lukas Eugen Marsoner Steinkasserer1*, Alessandra Zarantonello1 , and Beate Paulus 1
Institut für Chemie und Biochemie, Freie Universität Berlin, Germany
* marsoner@zedat.fu-berlin.de
Building upon the work of Karlický et al. [1], we explore the thermodynamic stability, electronic as
well as optical properties of selected heavy halogen functionalized graphene derivatives. [2] While
hydrogen-containing derivatives are predicted to display stabilities between those of fully
hydrogenated and fully chlorinated graphene, the substitution of hydrogen by fluorine significantly
increases their stability, bringing it close to that of fully hydrogenated graphene.
Analysis of the systems’ band structure relieves the presence of strong 1D-localization of the
charge-carriers. This localization further shows a marked electron-hole asymmetry, which might be
exploited to separate excitions created from optical excitations in photovoltaic applications.
Including both electron-electron as well as electron-hole interactions at the G0W0 and BSE [3] level
respectively, we find the systems to display optical adsorption peaks close to the Shockley-Queisser
limit, rendering them excellent candidates for applications in future solar-cell technology.
[1] F. Karlický et al., J. Chem. Phys., 137, 034709, 2012
[2] L. E. Marsoner Steinkasserer et al., in preparation
[3] J. J. Mortensen et al., Phys. Rev. B, 71, 035109, 2005; J. Enkovaara et al., J. Phys. Condens.
Matter 22, 253202 (2010); F. Hüser et al., Phys. Rev. B 87, 235132, 2013, T. Olsen et al., Phys. Rev.
Lett. 116, 056401 (2016), F. Rasmussen and K. S. Thygesen, arXiv:1511.00129 (2015)
Abstract for GPAW 2016
A General DFT Method for Computing Polaronic Conductivity in Battery Materials
Marko Melander1*, Elvar Ö. Jónsson2, Tejs Vegge1, Juan-Maria García-Lastra1
1
Department of Energy Conversion and Storage, Technical University of Denmark, DK-4000 Roskilde, Denmark,
2
COMP, Applied Physics Department, Aalto University, FI-00076 Aalto, Espoo, Finland
* Presenting Author: marmela@dtu.dk
ABSTRACT
Combining constrained density functional theory (cDFT) [4] with Marcus theory for electron transfer is an
efficient and accurate way for computation of charge transfer rates. Here, we present an implementation of
cDFT within the projector augmented wave -method (PAW) in the GPAW code. We also present how to
extract the electronic coupling element and reorganisation energy from the cDFT-PAW wave functions for
parametrisation of Marcus theory. Apart from previous cDFT implementations, our implementation utilises
modern non-norm conserving pseudopotentials, can use several wave function presentations, namely
numerical grid or using an LCAO and plane wave basis sets and has flexible periodic boundary conditions
ranging from 0D to 3D.
The method is applied for studying the polaronic transport in LiFePO4, a common cathode material in Liion batteries. Fast charge transfer in battery electrodes is paramount for quick and efficient charging and
decharging of a battery; the slow charge transfer at cathodes limits battery performance and improving
charge transfer is crucial [1]. As typical cathodes in Li-ion and Li-air batteries are semiconductors or
insulators, their electric conductance is due to consequent charge hopping known as polaronic conduction.
We present how cDFT is combined with Marcus theory to yield transfer kinetics from which conductivity
can be obtained after solving the master equation.
Fig. 1. Electron polaron (blue) on iron (brown) in LiFePO4 and its possible hopping directions (arrows).
REFERENCES
[1] J. Wang, X. Sun, Olivine LiFePO4: the remaining challenges for future energy storage, Energy
Environ. Sci, 8 (2015) 1110-1138
[2] P. Bai, M.Z. Bazant, Charge transfer kinetics at the solid-solid interface in porous electrodes, Nature
Comm., 5 (2014) 3585
[3] M. Melander, E.Ö. Jónsson, T. Vegge, J.M. García-Lastra, in preparation
[4] B. Kaduk, T. Kowalczyk, T. van Voorhis, Constrained Density Functional Theory, Chem. Rev., 112
(2012) 321-370
A DFT study of the effect of SO4 groups on the properties of TiO2 nanoparticles
O. Miroshnichenko1,2 , S. Posysaev1,2 and M. Alatalo1
University of Oulu, Theoretical Physics Research Unit, PL 3000, FI–90014 Oulu, Finland
2
Peter the Great St. Petersburg Polytechnic University, Department of Theoretical Mechanics, Polytechnicheskaya 29, 195251 St. Petersburg, Russian Federation
email: Olga.Miroshnichenko@oulu.fi
1
Titanium dioxide is one of the most investigated and most widely used semiconductor
metal oxides. Such popularity comes from its countless applications in an enormous
number of industry and technology fields with an increasing interest in nanosized TiO2
clusters. The properties of small particles are different from the bulk, which is very important in, for example, light scattering measurements, where the refractive index is needed.
The optical and electronic properties of titanium dioxide nanoparticles have shown to
be strongly dependent on the structure and size of the particle[1, 2]. Besides these size
and shape dependent changes, in the case of small particles, the effects of adsorbates become increasingly important, because in the applications, the nanoparticles often reside
in water or other, more complicated solutions. Therefore, it is essential to understand the
properties of TiO2 nanoparticles in more detail and to extend the earlier results to realistic
conditions, where adsorbates such as hydroxyls [3] and SO4 groups are present on the
surface of the clusters.
We have performed density functional theory (DFT) and time-dependent DFT calculations for titanium dioxide nanoparticles covered with varying number (1-4) of SO4 groups.
We find that SO4 groups significantly affect the structure of nanoparticles and also change
the photoabsorption characteristics. Moreover SO4 groups influence the structure during
the particle growth at the early stages of sulfate manufacturing method, allowing the particle to form in anatase structure instead of rutile. All calculations were performed using
the GPAW software package.[4]-[6]
[1] S. Auvinen, M. Alatalo, H. Haario, J.-P. Jalava, R.-J. Lamminmäki, J. Phys. Chem. C
115, 8484 (2011).
[2] S. Auvinen, M. Alatalo, H. Haario, E. Vartiainen, J.-P. Jalava, R.-J. Lamminmäki, J.
Phys. Chem. C, 117, 3503 (2013).
[3] O. Miroshnichenko, S. Auvinen, M. Alatalo, Phys. Chem. Chem. Phys., 17, 53215327 (2015).
[4] J. J. Mortensen, L.B. Hansen, K. W. Jacobsen, Phys. Rev. B 71, 035109 (2005).
[5] J. Enkovaara, C. Rostgaard, J. J. Mortensen et al., J. Phys.: Condens. Matter 22,
253202 (2010).
[6] S. R. Bahn, K. W. Jacobsen, Comput. Sci. Eng., Vol. 4, 56-66 (2002).
What
'
sne
wi
nGPAW andASE
Jens Jørgen Mortensen
Department of Physics, Technical University of Denmark
The talk will give an overview of GPAW and ASE and the latests new features. We will
also take a look at the whole development process from writing code to testing it and
documenting it.
Designing quantum spin Hall insulator in-plane heterostructures from
first principles: 1T'-MoS2 with adsorbates
Thomas Olsen*
Center for Atomic-Scale Materials Design, Department of Physics, Technical University of Denmark
* tolsen@fysik.dtu.dk
Interfaces between normal and topological insulators are bound to host metallic states that are protected
by time-reversal symmetry and therefore robust to disorder and interface reconstruction. Twodimensional topological insulators (quantum spin Hall insulators) offer a unique opportunity to change
the local topology by adsorption of atoms or molecules. Here, we apply first principles calculations to
show that the quantum spin Hall insulator 1T'-MoS2 exhibits a phase transition to a trivial insulator
upon adsorption of various atoms. It is then demonstrated that one-dimensional metallic boundary
states indeed arise in a “ribbon” geometry of alternating regions with and without adsorbed oxygen.
The construction allows one to construct networks of one-dimensional metallic channels in sheets of
1T'-MoS2 by selectively adsorbing oxygen to specific regions. Spin-orbit coupling plays a crucial role
in these calculations and we will briefly discuss the implementation in GPAW.
Theoretical Study of MoS2 Contacts Modes with Metals
S. Posysaev1,2*, O. Miroshnichenko1,2, X. Shi1, M. Huttula1, W. Cao1 and M. Alatalo1
1
University of Oulu, Center for Molecular Materials, Theoretical Physics, PL 3000,
FI—90014 Oulu, Finland
2
Peter the Great St. Petersburg Polytechnic University, Department of Theoretical Mechanics, Polytechnicheskaya 29, 195251 St. Petersburg, Russian Federation
* sergei.posysaev@oulu.fi
In the near future, the manufacturers of logic transistors will face the inevitable failure
of the present technology to fulfill the Moore’s law requirement below 20 nm line
width. The solution is to go down in scale until the fundamental limit of a single atom
layer. Recently, a large variety of 2D materials, including graphene and single atom layers of rhodium and palladium, has been synthesized. However, these materials do not
have a bandgap.
Molybdenum disulfide, a widely studied semiconducting layered transition metal
dichalcogenide, possesses an indirect band gap of 1.3 eV in bulk and a direct band gap
of 1.9 eV in the monolayer [1]. Several studies [2, 3] have reported large resistance between contacts and semiconducting layer. It shows us the importance of more thorough
research of metal-mTMD (monolayer transition-metal dichalcogenide) edge contacts.
The experimental comparison has been presented between the basal (top) and edge
planes of MoS2 by using macroscopic molybdenite crystals [4]. On selective exposure of
only the basal or edge plane, a comparison of their electrochemical performances was
made. The edge plane of MoS2 crystal showed a significantly higher electrochemical
activity than the basal plane.
Our calculations are performed by means of density functional theory (DFT) implemented in program package GPAW. These calculations yield detailed information on
the electronic and atomic structure, properties of MoS2 surface cuts, properties in the
bulk and its complexes with transition metals. We have determined the most energetically preferable surface cuts of MoS2. Structure optimization between the surface of
MoS2 and Au with varied thickness was carried out. The tendency of gold to attach to
the ideal edge surface of MoS2 layers and forming a contact was observed.
[1]
A. Splendiani, L. Sun, Y. Zhang, T. Li, J. Kim, C.-Y. Chim, G. Galli, F. Wang,
Nano Lett. 2010, 10, 1271–1275.
[2]
H. Liu, A. T. Neal, and P. D. Ye, ACS Nano 6, 8563 (2012).
[3]
J. Kang, W. Liu, D. Sarkar, D. Jena, K. Banerjee, Phys. Rev. X 2014, 4, 031005.
[4]
S. M. Tan, A. Ambrosi, Z. Sofer, Š. Huber, D. Sedmidubský and M. Pumera,
Chem. – Eur. J., 2015, 21, 7170–7178.
DFT-Study of the hydrogenation of CO2 to Methanol
Thomas Reichenbach1,2,*, Krishnakanta Mondal1,2, Daniel Himmel3,6, Albert Bruix5, Bjørk
Hammer5, Ingo Krossing3,6, Michael Walter1,2,3, , Michael Moseler1,3,4
1
Fraunhofer Institut für Werkstoffmechanik, Wöhlerstraße 11, 79108 Freiburg
Freiburger Institut für Interaktive Materialien und Bioinspirierte Technologien, Georges-KöhlerAllee 105, 79110 Freiburg
3
Freiburger Materialforschungszentrum, Stefan-Meier-Straße 21, 79104 Freiburg
4
Institut für Physik, Universität Freiburg, Hermann-Herder-Str. 3, 79104 Freiburg
5
Interdisciplinary Nanoscience Center (iNANO), Department of Physics and Astronomy, Aarhus University, DK-8000 Aarhus C, Denmark
6
Institut für Anorganische und Analytische Chemie, Universität Freiburg, Albertstr. 21, 79104 Freiburg
2
* thomas.reichenbach@iwm.fraunhofer.de
As a result of the increasing amount of renewable fluctuating energies, an efficient possibility
to store the latter is needed. Additionally the CO2-Emission and consequently the global
warming is becoming the biggest problem humanity is facing. Within the project HyCO2 of
the Sustainability Center Freiburg the hydrogenation of CO2 using sustainable H2 is investigated to tackle both issues. Methanol is a key intermediate that can be further processed to
fuels with improved combustion properties. Currently the CO2-hydrogentation is already realised in an industrial scale using Cu/ZnO-catalysts, whereby the achieved process conditions
are not economically yet. To optimize the catalytic processes during methanol synthesis, the
reaction mechanisms have to be better understood.
The fundamental chemical reactions are studied using density functional theory, the catalyst is
modelled by ZnxOy/Cu(111). This oxide/metal-configuration showed a superior catalytic activity compared to its reverse counterpart as well as pure Cu(111) in experiments [1,2].
A representative model is found using the recently in ASE implemented genetic algorithm by
global optimising ZnXOY/Cu(111): A Zn3O-pattern is the characteristic interface between Cu
and ZnO. As a result the simplest model ensuring transferability to bigger models
Zn3OH/Cu(111) can be used to study the methanol synthesis mechanisms in a hydrogen rich
environment. The calculations indicate a hydrogenation of CO2 via a formate-intermediate,
which is a current debate in literature. In addition this small model is transferred to a molecular system, which is evaluated using gold standard coupled cluster to check the quality of the
DFT-results. A very systematic, order-maintaining underestimation of barriers is revealed.
[1] S. D. Senanayake et al. The Journal of Physical Chemistry C 2016 120 (3), 1778-1784
[2] T. Fujitani, I. Nakamura, T. Uchijima, J. Nakamura, Surf. Sci. 1997, 383, 285-298.
Nanoplasmonics within time-dependent density-functional theory
Tuomas Rossi*, Martti Puska, Risto Nieminen
Department of Applied Physics, Aalto University, Finland
* tuomas.rossi@aalto.fi
Localized surface plasmons are collective electron excitations that dominate the optical (or near-optical)
response of metallic particles. These excitations are characterized by electron density oscillations
extending over the whole particle. This makes plasmonic excitations sensitive to the composition
and the shape of the particle, enabling their experimental tunability. Additionally, the response at the
plasmon resonance is strong and the associated charge density oscillation creates an enhanced electric
near-field around the particle. These properties make plasmons suitable and promising for a number of
applications from spectroscopy to nanomedicine. [1]
In recent years, along with the improving precision of experimental techniques, it has become increasingly important to understand the plasmonic response of nanoscale systems and structures with
nanoscale features. In such systems, the response is strongly affected by quantum phenomena such as
spill-out of surface charge and electron tunneling. The quantum effects originating from the electronic
structure are most commonly modeled computationally by using the time-dependent density-functional
theory (TDDFT) due to its feasible computational speed and decent accuracy. [2]
In this presentation, I describe the TDDFT approach for modeling the plasmonic response of finite
nanoscale systems. The focus of the presentation is on the methods and tools that are available
in GPAW. [3-5] These tools include the recent implementation of time-propagation TDDFT within
localized basis sets [5]. This highly-efficient approach enables first-principles modeling of systems up
to size scales that have not been easily accessible before. [5] However, when using localized basis
sets, the basis set incompleteness can be a problem because it is not straightforward to refine the basis
sets. Thus, a special attention needs to be paid on choosing the appropriate basis set carefully. [5-6]
In the presentation, I discuss this and other practical aspects of numerical calculations. Additionally,
I describe different ways of analyzing the plasmonic response and discuss some recent case studies
where we have applied the presented methodologies for nanoplasmonic systems. [7-8]
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
V. Giannini, A. I. Fernández-Domínguez, S. C. Heck, and S. A. Maier, Chem. Rev. 111, 3888 (2011).
N. J. Halas, S. Lal, W.-S. Chang, S. Link, and P. Nordlander, Chem. Rev. 111, 3913 (2011).
J. Enkovaara et al., J. Phys. Condens. Matter 22, 253202 (2010).
M. Walter, H. Häkkinen, L. Lehtovaara, M. Puska, J. Enkovaara, C. Rostgaard, and J. J. Mortensen,
J. Chem. Phys. 128, 244101 (2008).
M. Kuisma, A. Sakko, T. P. Rossi, A. H. Larsen, J. Enkovaara, L. Lehtovaara, and T. T. Rantala,
Phys. Rev. B 91, 115431 (2015).
T. P. Rossi, S. Lehtola, A. Sakko, M. J. Puska, and R. M. Nieminen, J. Chem. Phys. 142, 094114 (2015).
T. P. Rossi, A. Zugarramurdi, M. J. Puska, and R. M. Nieminen, Phys. Rev. Lett. 115, 236804 (2015).
T. P. Rossi et al. (unpublished).
E↵ect of vertex corrections in Hedin’s equations for predicting bandgaps in solids
Per S. Schmidt and Kristian S. Thygensen
Center for Atomic-scale Materials Design (CAMD)
Technical University of Denmark, Department of Physics
Fysikvej 1 2800 Kgs. Lyngby
The first-principles application of many-body perturbation theory at the
level of Hedin’s GW approximation[1] to the calculation of energy levels in
semiconductors and insulators has proved to yield values in good agreement
with experiments. In spite of this, the justification of neglecting vertex corrections, , in the self-energy is not obvious. In this work we improve upon
the widely used G0 W0 approximation for the self-energy by including these
vertex corrections in both the screened Coulomb interaction, W , and the
self-energy. The vertex is approximated through two di↵erent kernels and
compared, a renormalized adiabatic LDA kernel (rALDA[2]) and a jelliumwith-gap kernel (JGMsx[3]). Standard G0 W0 calcualtions systematically underestimate bandgaps and we show that the vertex corrections results in
increased bandgaps of solids, bringing them closer to experimental values. In
addition to being more accurate, the calculations are also shown to converge
faster with respect to basis set size.
The e↵ect of including vertex corrections only in W has also been investigated. The computational cost is similar but the resulting bandgaps decreased in size, and are thus further away from experimental values.
Per Schmidt
March 9, 2016
[1] L. Hedin, Phys. Rev. 139, A796 (1965)
[2] T. Olsen and K. S. Thygesen, Phys. Rev. B 88, 115131 (2013)
[3] C. E. Patrick and K. S. Thygesen, J. Chem. Phys. 143, 102802 (2015)
1
Singlet Fission and Repulsive Van der Waals forces
Oliver Stauffert
Department of Physics, Universität Freiburg oliver.stauffert@physik.uni-freiburg.de
We describe electronic structures for organic molecules, which are interesting for current research
on organic solar cells. By doing so, we presents additional insight on the processes called singlet
fission, which is a phenomenon that occurs in this field of work. In the process of singlet fission, a
single singlet excitation is split among two molecules and dissociates into two triplet excitations.
In the experiment only the singlet excitation and an additional, non radiating decay can be
observed. With DFT and it is possible to estimate the possibility of triplet fission on an energetic
basis. Also the the triplet states are feasible and characterizations of this triplet states are
conducted. By doing so, possibilities are given to experimentally show triplet population and proof
the occurrence of singlet fission.
Further we aim to investigate repulsive Van der Waals forces. Hereby the primary characteristics of
the Van der Waals forces are given by the London formula
UVdW ~ C6 / R⁶. The main dependency of the C6 coefficient is given by the possible excitations of
the molecule and their transition energies. For ground state molecules, the transition energies are
always positive and the Van der Waals interaction is generally attractive. However, for excited
molecules, negative transition energies occur, which can lead to repulsive Van der Waals forces.
Towards error bar estimates and hybrid functionals
Mikkel Strange*, Jens J. Mortensen, Karsten W. Jacobsen and Kristian S. Thygesen
CAMD, Department of Physics, Technical University of Denmark, Denmark
*mikst@fysik.dtu.dk
About 200.000 materials are “known” to exist, but even the most basic properties, such as band gap,
elasticity constants, etc. have only been determined for a relatively small number of them.
There is large interest and effort in creating large databases using high-throughput calculations, for
example in the NOMAD project [1] as well as big data analytics tools to look for structures and
correlations. The NOMAD project is briefly introduced; it aims at building a large database for
calculations based on the 50 most cited electronics structure codes as well as making big data
analytics tools.
One of the challenges in comparing results from different DFT calculations is that properties
converge differently with computational settings such as basis set, k-point sampling etc.
Here we analyze errors from computational settings and take a first step towards providing error bar
estimates for various DFT calculated properties obtained using GPAW. In addition a comparison is
made to results from other codes such as AIMS and VASP.
If time allows, some of the challenges of using functionals that include some fraction of exact
exchange, e.g. hybrid functionals, such as HSE and PBE0, in GPAW are discussed.
[1]AEuropeanCenterofExcellence:https://nomad-coe.eu
Efficient approach for calculating band gaps of van der Waals
heterostructures
Kirsten T. Winther* , Kristian S. Thygesen
CAMD, Department of Physics, Technical University of Denmark, Building 307, Kgs. Lyngby,
Denmark
* kiran@fysik.dtu.dk
The discovery of two-dimensional (2D) crystals such as graphene and MoS2 has lead to a new
class of materials known as van der Waals heterostructures (vdWHs), where stacking different
2D crystals gives rise to highly tunable electronic properties [1]. The GW approximation is the
state of the art approach for calculating quasiparticle band structures, and could provide a
powerful tool for modeling and guiding the design of vdWHs. But in the traditional form, such
calculations are only feasible for commensurable structures with a few layers, due to high
computational cost.
Here, we have extended a recently developed multiscale approach, the Quantum electrostatic
heterostructure (QEH) model [2], to calculate the band gap of van der Waals heterostructures
on the G0W0 level. In the QEH model the response functions of the individual layers are
calculated from ab-initio, and coupled classically via their long-range Coulomb interaction (see
Fig. 1). This significantly reduces computational cost, and the dielectric function of realistic,
incommensurable vdWHs comprising hundreds of layers can be calculated. Within the G0W0 +
QEH framework the quasiparticle bandstructure is obtained from a standard G0W0 calculation
for the isolated 2D layer [3] together with an added QEH correction to the quasiparticle
energies. This new approach enables the theoretical prediction of band gaps for a large range of
materials at a feasible computational cost.
Figure 1: Schematic of the QEH model.
References
[1] Geim, A. K. and Grigorieva, I. V. Van der Waals heterostructures, Nature 499 (2013) 419-425.
[2] Andersen, K., Latini, S., Thygesen, K. S. Nano Letters, 15 (2015) 4616–4621.
[3] Rasmussen, F. A. and Thygesen, K. S. J. Phys. Chem. C, 119 (2015) 13169–13183.
Bayesian error estimation functionals and further method
development at SUNCAT
Johannes Voss
SUNCAT Center for Interface Science and Catalysis
SLAC National Accelerator Laboratory, USA
vossj@slac.stanford.edu
The Bayesian error estimation (BEE) functionals developed in collaboration between SUNCAT and
CAMd have recently been extended to meta functionals [1,2] with improved description of both
surface reaction energetics and cohesive energies of solids [3]. The next generation of BEE
functionals will be hybrids including exact exchange contributions. Besides aiming at a better
description of transition metal oxide catalysts, the main motivation at SUNCAT to include exact
exchange contributions is to address challenges in the modeling and understanding of the
electrochemical interface.
In addition to an outlook for the functional development efforts at SUNCAT, an overview of further
method development projects at SUNCAT will be given.
Figure: BEE functional calculations of N 2 bond strength on transition metal surfaces showing a high
likelihood for ruthenium being the best catalyst for ammonia synthesis of the systems considered
(from [4]).
[1] Wellendorff et al., J. Chem. Phys. 140, 144107 (2014).
[2] Lundgaard et al., submitted.
[3] Pandey and Jacobsen, Phys. Rev. B 91, 235201 (2015).
[4] https://www6.slac.stanford.edu/news/2014-07-10-uncertainty-gives-scientists-new-confidencesearch-novel-materials.aspx – Medford et al., Science 345, 197 (2014).
Accurate spectroscopy with GPAW: absolute XPS and ResonantRaman spectra
Michael Walter1,2*
1
Center for Interactive Materials, University of Freiburg, Germany, 2Fraunhofer IWM, Freiburg, Germany
*Michael.Walter@fmf.uni-freiburg,de
𝐹,exp
Figure: a) Difference between experimental 𝐸𝐵
and calculated 𝐸𝐵𝐹,calc for various solids.
The numbers quoted and the shaded area give the PBE band gaps of the respective material.
b) Differences of relative shifts between pairs of elements for a certain compound. Multiple
points indicate individual experimental results.
In the first part of the presentation I show that absolute binding energies of core electrons in
molecules and bulk materials can be efficiently calculated by spin paired density-function
theory employing a Δ Kohn-Sham (ΔKS) scheme corrected by offsets that are highly transferable. These offsets depend on core level and atomic species and can be determined by comparing ΔKS energies to experimental molecular X-ray photoelectron spectra. The correct prediction of absolute and relative binding energies on a wide range of molecules, metals and
insulators is demonstrated [1].
The second part of the presentation is devoted to the calculation of resonant Raman spectra in
molecules and solids. As Raman scattering involves both absorption as well as emission of
photons it is a second order process that is rather involved. Therefore there are different approximations in the literature that partly exclude each other. I will give a comparison of these
approximations and their range of validity in organic molecules and carbon based material.
[1] Michael Walter, Michael Moseler, Lars Pastewka http://arxiv.org/abs/1511.06610