Methods Development

Density-functional theory (DFT) methods have been successfully applied to simulations of molecules, clusters, surfaces, and solids. However, as the simulated systems grow in size, it becomes necessary to employ simplified methods, such as the density-functional-based tight-binding (DFTB). DFTB uses parameterized two-center approximations to make the computations much faster and scalable. As a result, systems containing several thousands of atoms can be simulated with good accuracy at a small fraction of DFT's computational cost. We have been working together with SCM (http://www.scm.com) in the development of DFTB algorithms for the calculation of structural, optical, and electronic properties (among others) of systems with vastly different chemical natures and sizes, such as those shown here.

MOFs and SURMOFs

MOFs and SURMOFs are crystalline porous materials composed of two components: metal or metal oxide moieties and organic molecules known as linkers. The MOFs can be synthesized via reticular chemistry. The choice of metal and linker has significant effects on the structure and properties of the particular MOF. MOFs have broad applications because of two key attributes: their extremely large surface area and the large variety of structure and functionality of their constituting building blocks. MOF applications include gas storage, gas separation, quantum sieving, catalysis, sensors, proton and electric conductivity etc. Using the available arsenal of computational methods (electronic structure methods and classical FF methods) our research on MOFs and SURMOFs is focused on:

	Prediction of MOFs for applications (screening) The reticular synthesis of MOFs leads to many millions of possible frameworks. It is impossible to synthesize so many structures and so we use computational methods to predict which structures should be synthesized.
Quantum sieving (H2/D2 separation) The strongly attractive open metal sites present in MOF frameworks of the CPO-27 family can be used for hydrogen isotope separation, giving very high separation temperatures at temperatures above the N2 liquid boiling point.
	Electronic and magnetic properties of SURMOFs The unusual coupling of the spin ½-ions within the resulting one-dimensional chains in SURMOF-2 is found to stabilize a low-temperature ferromagnetic phase with a Curie temperature above 20 K.

Ultrathin Materials

Ultrathin materials are extremely interesting as building blocks of next-generation nanoelectronic devices because it is much easier to fabricate circuits and other complex structures by tailoring 2D layers into desired forms. Due to their atomic scale thickness and extraordinary electronic properties, 2D materials are the ultimate limit of miniaturization in the vertical dimension and allow shorter transistors before short channel effects set in.

Exploitation of strong quantum confinement effects in a layered noble metal dichalcogenideleads to single-material logical junction. The proposed device, made of PdS2 , with channel lengths of about 2.5 nm, shows negligible tunnel loss and thus marks the limit of integration density for a conventional transistor.

Proton-transfer reactions in soft matter

Proton-transfer reactions and mechanisms are essential in technological processes, e.g. in proton-exchange membrane fuel cells (PEMFCs). The fuel cell membrane is responsible for the transport of protons between the electrodes. Different membrane materials are used to date, of which Nafion is the most common one. However, the proton transport in Nafion membranes critically depends on the water content, which limits the operating temperature to the boiling point of water.
State of the art research on water-free proton transport favors protogenic groups other than the sulfonic-acid group that is essential in the Nafion membrane. Possible candidates are phosphonic acid and imidazole. We study these materials with respect to their proton-transporting properties using quantum and classical molecular dynamics as well as Monte-Carlo methods.

Fig. 1: Alignment of Imidazole guest molecules in a MIL-53(Al) MOF channel.		Fig 2: Sampled NN and NH distances (upper panels) and partial N and H charges (lower panels) during proton transfer in liquid imidazole (left column) and a MIL-53(Al) MOF channel (right column). The statistical errors are highlighted as shaded areas.

References

Hydrogen storage in graphene-based nanostructures

Hydrogen storage by physisorption in graphene-based nanostructures is certainly comparable to the performance of state-of-art "reference" materials such as metal-organic frameworks (MOFs) [their uptake normally does not exceed ~7 wt% at 77 K]. As a result, graphene-based materials represent a practical alternative to MOFs which are (in many cases) humidity-sensitive (i.e., subject to hydrolysis) and very expensive. Our recent focus has been on computational studies (classical Grand canonical Monte Carlo) in multilayered graphene systems with defect-induced additional porosity. We considered model systems which consist of infinite parallel graphene layers with "holes" of different sizes (~6-30 Å). We found that the incorporation of defects (preferably small "holes", up to 10 Å in diameter, cf. Figure 1) results in the increase of theoretically possible surface areas beyond the limits of ideal defect-free graphene (~2700 m²/g) with the values approaching ~5000 m²/g. This leads to promising hydrogen storage capacities of up to 6.5 wt% at 77 K (15 bar). Modelling was supported by an experimental work performed in the group of Dr. Talyzin (Umeå University, Sweden).

Figure 1. An example of a 3D stacking of perforated graphene sheets (a perspective view).

References

Metal-organic frameworks

In recent years the group of Prof. Seifert has gained a wide experience in simulating the properties of porous solids, mainly metal-organic and covalent organic frameworks as well as porous carbons. This experience involves the calculation of gas adsorption isotherms by classical GCMC methods, identification of adsorption sites by molecular dynamics simulations at the DFTB (density-functional based tight binding) level of theory and the flexibility ("breathing") of metal-organic frameworks based on ab initio molecular dynamics. Furthermore, there is a strong expertise in topological modeling of not-yet-synthesized structures applied to zeolitic imidazolate-based frameworks (ZIFs, cf. Fig. 1).

Figure 1. Most promising LiB(imid)₄ frameworks for hydrogen storage applications.

References

DFT investigations of magnetic properties in metal-organic frameworks (MOFs)

Metal organic frameworks (MOFs) gained very high interest over the last decade due to their high porosity, low density and absorption behaviour. Some MOFs exhibit transition metal centers, which can lead to a certain magnetic ground state within the system. In order to investigate the magnetic ground state of e.g. the exible MOF DUT-8(Ni) [1,2], ab initio investigations in the framework of density functional theory (DFT) have been carried out. This MOF exists in two crystalline structures (DUT-8(Ni)_open /DUT-8(Ni)_closed), where both of them exhibit low-spin states, thus an antiparallel alignment of the spins at the magnetic centers.

(a) Unit cells of DUT-8(Ni)closed	(b) Supercell for DUT-8(Ni)open	(c) Model system
Structure of unit cells, a supercell and a model system for DUT-8(Ni).

Further investigations included several model systems. Those models consist of the two magnetic centers and their chemical environment. Different alterations of this chemical environment were introduced to see the corresponding change in the magnetic behaviour. This allows an insight into what could happen to the crystalline systems if such changes were implied into the periodic structures. Some alterations to the model systems, especially the ones including a removal of the NH₃ groups, show that the ground changes from low-spin to high-spin. This provides a starting point for further investigations which might lead to a stable high-spin MOF.
Another interesting research field is the structural determination of MOFs using NMR. A favourable canditate for NMR is ¹²⁹Xe as it is mononuclear, chemically inert and has a spin of I = 1/2. Experiments are carried out for the MOFs UiO-66 and UiO-67, which are Zr based structures. The corresponding NMR parameters for the chemical shift of Xe δ_Xe inside such structures can be determined by means of DFT. This can help to understand the experimental spectrum and to predict trends of δ_Xe for different conditions.

References

MoS₂ Nanotubes and Layers

References

Properties of boron-rich systems

The electron-deficient element boron represents a challenge to both experiment and theory. The list of unknowns and questions is long: How can we understand the chemical bonding? Why are boron-rich crystal structures so complex? How many elemental phases exist? What is the ground state structure? How does the phase diagram look like? Are there non-icosahedral nanostructures? etc.

Left: a boron nanotube and its current-voltage relation (calculated). Right: the crystal structure of α-tetragonal boron, consisting of four icosahedral B12 units (grey) and partially occupied interstitial sites (colored).

Our main objective is to tackle these questions by theoretically studying structural, chemical and physical properties of boron bulk and nanostructures. Recent results focused on boron sheets, boron nanotubes and α-tetragonal boron. We are closely collaborating with experimentally working partners and also organize an annual, international boron symposium within the MS&T conference.

References

Optical and electronic properties of molybdenum disulfide monolayers

Molybdenum disulfide (MoS₂) is a compound among the rich family of layered transition-metal dichalcogenides that has been studied most intensively in the past 40 years. Whereas bulk MoS₂ is a semiconductor with an indirect band gap of 1.3 eV, monolayer MoS₂ is a direct-gap semiconductor with a band gap of 1.8 eV. Monolayer MoS₂ has been used to build transistors with very promising electron mobility. Furthermore, MoS₂ has been recently exploited for biosensing and energy-storage applications, and strong absorption and photoluminescence bands were observed in the visible range at 1.9 eV (~650 nm).
Recently, monolayer MoS₂ has been in the focus of research because of its exceptional electronic and optical properties. Due to this reason we studied as an example Optical and electronic properties of triangular MoS₂ nanoflakes with edge lengths ranging from 1.6 to 10.4 nm. The optical spectra were calculated using the time-dependent extension of the density-functional tight-binding method (TD-DFTB). The size effect in the optical absorption spectra is clearly visible. With decreasing length of the nanoflakes edges, the long-wavelength absorption in the range of visible light is shifted toward short-wavelength absorption, confirming a quantum-confinement-like behavior of these flakes. In contrast, the edges of the nanoflakes exhibit a distinct metallic-like behavior.

Fig. 1: Calculated absorption spectra of a continuous set of MoS2 nanoflakes with different edge lengths. The spectra are shifted in the y direction for a better visibility.		Fig 2: (a) Projected Mo-4d density of states of a single Mo153 S342 nanoflake resulting from groups of Mo atoms triangularly. (b) Color code of the groups of Mo atoms (labeled pn, n = 1-6) as used for the PDOS representation in panel a.

References