my research

Time and length scales covered by different computational methods.
Showing the time and length scales covered by different computational methods. The intensity of the green colouring somehow reflects my expertise in the related field. 😉

I have conducted research in the field of computational (chemical) physics and biophysics. Here I utilized a broad range of computational methods, from ab-initio via particle based through to continuum methods. But my focus is on applying classical molecular dynamics (atomistic and coarse grained) simulations to study problems in chemical physics. Here I would like to give you a brief overview on the main topics I have work(ed) on.


actual research

Ionic Liquids and Ionanofluids as Heat-Transfer Fluids

describes in a short form the focus of my  research project, recently granted by the DFG.
More informations will be added here soon.


Homogeneous nucleation of metal vapour

“Nanoparticles are interesting in many applications because of their special properties which differ from those of the same substance in the bulk state. This is caused by the large surface-volume ratio for small clusters and particles as well as by the confinement of small systems. One way of producing nanoparticles is the inert gas aggregation process. This is the formation of particles in a supersaturated vapor. The first step in this particle formation process is homogeneous nucleation, an activated process in a metastable phase such as a supersaturated vapor. [1]”

zn_nucleationWe have studied the nucleation of zinc using molecular dynamics (MD) simulation. Zinc particles are of particular interest in a process for solar hydrogen formation [2, 3].
The interaction between the zinc atoms was modelled by an embedded atom method [4] (EAM) potential. The heat of condensation was removed by an inert carrier gas (argon) which was thermalized [5].

Here is a video showing a simulation (8 ns) of nucleation of zinc from the vapour phase [1]. The 512 red spheres represent zinc and the 1024 light blue ones argon atoms.  The temperature of the argon atoms is controlled and set to 300 K. The density of argon is given with 0.106 mol/L.

 


Structure of nano particles

zn_particleTo study the structure of the nano particles we gain from our nucleation simulation, we have extended  [6] the common neighbour analysis [7] (CNA) method to hcp-surfaces that significantly affect the properties and growth behaviour of zinc nano particles. The CNA is a geometric analysis of the next neighbours of each single atom. It allows a much better identification of the structural environment as the pair correlation function does.

 


 Calculation of spinodals by MD simulation

co2_film_260K

“The limit of metastability, the so-called spinodal, is calculated for pure carbon dioxide by molecular dynamics simulation. The determination of the spinodal is based on properties of the liquid vapor interface using a recently [8] developed method. This method relates the tangential pressure component through the vapor-liquid interface to the van der Waals loop in the two-phase region of the phase diagram. By application of the thermodynamic stability criteria, the location of the spinodal can be determined. [9]”

Here is a video we have “embedded” in our 1st animated poster.

 


Development of force fields

naproxen-charges

A proper suitable force field is crucial for success of a modelling.

In the context of the investigation of particle formation [1] we developed a potential model by means of the embedded atom method [4] (EAM) for the hexagonal close packed metal zinc [6].
We employed here a scheme I had developed in my Diploma thesis [10].

During my PhD in needed a force field model for naproxen, which is a pharmacologically active substance. We propose a united-atom potential model for naproxen suitable for molecular dynamics (MD) simulation [11]. The intramolecular interactions and point charges (CHELPG method [12]) were obtained from ab initio calculations using Gaussian [13].


MD simulation of particle formation
by rapid expansion of a supercritical solution (RESS)

RESS-MD

The particle formation processes based on supercritical fluids is of great interest, in particular for the formulation of pharmacological agents [14].

One of these processes is the rapid expansion of a supercritical solution [15]. We propose simulation technique to do MD studies on adiabatic expansions [16], which is a sub-process of RESS. The technique was verified on naphthalene and then applied to naproxen [17].

This was the main issue of my PhD thesis [18].

Here a video showing a simulation of the rapid expansion of a solution of (64) napthalene in (2424) supercritical carbon dioxide. The  starting conditions are: 330 K, 12 MPa and 0.6218 g/cm3 CO2 density. [16]

 


Fluids & solutions in a thermal gradient

icimagesThermal gradients are responsible for a whole range of transport phenomena, like

  • the motion of solute particles in suspensions (thermophoresis),
  • the separation of mixtures, and
  • thermal diffusion in aqueous solutions (thermodiffusion, Soret effect).
  • Also, the heat transport (thermal conductivity) its self is of interest.

We have investigated the microscopic mechanisms of heat transfer in water [19], as well as the thermal conductivity for different water models [20] and at extreme pressures [21] by means of non-equilibrium MD (NEMD).

Also, alkali halide aqueous solutions where studied with respect to Sorret effect and thermal conductivity. In cooperation with experimentalist from FZ Jülich (Zilin Wang and Simone Wiegand) we where able to compile a comparative paper [22] which gives new insights into heat transfer mechanisms in aqueous solutions.

By investigating symmetrical binary mixtures and corresponding diatomic fluids [23] we show a thermomolecular orientation (TMO) effect in nonpolar fluids [24].

Here is a video, showing a 6 mol/kg KCl solution in a thermal gradient. The middle of the oblong simulation box is cold and the right and left side is hot. [22]


Dense semi-flexible polymer brushes in shear flow

polybrushSemi-flexible polymers at high density stiff anchored onto a surface can be found in many biological systems. For example in our body

  • glycocalyx brush structure on the endothelial surface layer (ESL) [25] of our blood vessels, or
  • the periciliary layer (PCL) of the lung airway [26]

is exposed to shear flow.

We utilized Smoothed Dissipative Particle Dynamics (SDPD) [27] to study those brushes and developed a theoretical model to predict properties like effective brush hight with respect to elasticity, grafting density and shear flow. The results were recently published [28].

Here a video of DPD simulation of a polymer brush made of semi-flexible polymers grafted on a slit surface under good solvent conditions in flow. The solvent particles are invisible in order to show the polymers.

 


References

[1] [doi] F. Römer and T. Kraska, The Journal of Chemical Physics 127, 234509 (2007).
[2] [doi] A. Steinfeld, International Journal of Hydrogen Energy 27, 611-619 (2002).
[3] [doi] M. Karlsson, I. Alxneit, F. Rütten, D. Wuillemin, and H. R. Tschudi, Review of Scientific Instruments 78, 34102 (2007).
[4] [doi] M. S. Daw and M. I. Baskes, Phys. Rev. Lett. 50, 1285-1288 (1983).
[5] [doi] S. Braun, F. Römer, and T. Kraska, The Journal of Chemical Physics 131, 64308 (2009).
[6] [doi] F. Römer, S. Braun, and T. Kraska, Phys. Chem. Chem. Phys. 11, 4039-4050 (2009).
[7] [doi] D. Faken and H. Jónsson, Computational Materials Science 2, 279-286 (1994).
[8] [doi] A. R. Imre, G. Mayer, G. Házi, R. Rozas, and T. Kraska, The Journal of Chemical Physics 128, 114708 (2008).
[9] [doi] T. Kraska, F. Römer, and A. R. Imre, The Journal of Physical Chemistry B 113, 4688-4697 (2009).
[10] [doi] F. Römer, “Entwicklung eines EAM-Potentials für Ruthenium,” Master Thesis, 2006.
[11] [doi] F. Römer and T. Kraska, Molecular Simulation 38, 152-160 (2012).
[12] [doi] C. M. Breneman and K. B. Wiberg, Journal of Computational Chemistry 11, 361-373 (1990).
[13] M. J. Frisch et al., Gaussian 03, Revision C.02.
[14] J. G. Wagner, Biopharmaceutics and Relevant Pharmacokinetics, Drug Intelligence Pub., Hamilton, IL, 1971.
[15] J. B. Hannay and J. J. Hogarth, Proceeding of the Royal Society, London 30, 178-188 (1879).
[16] [doi] F. Römer and T. Kraska, The Journal of Physical Chemistry C 113, 19028-19038 (2009).
[17] [doi] F. Römer and T. Kraska, The Journal of Supercritical Fluids 55, 769-777 (2010).
[18] [doi] F. Römer, “MD Studien zur Partikelbildung von pharmazeutischen Wirkstoffen mit dem RESS-Verfahren,” PhD Thesis, 2010.
[19] [doi] J. Muscatello, F. Römer, J. Sala, and F. Bresme, Phys. Chem. Chem. Phys. 13, 19970-19978 (2011).
[20] [doi] F. Römer, A. Lervik, and F. Bresme, The Journal of Chemical Physics 137, 74503 (2012).
[21] [doi] F. Bresme and F. Römer, Journal of Molecular Liquids 185, 1-7 (2013).
[22] [doi] F. Römer, Z. Wang, S. Wiegand, and F. Bresme, The Journal of Physical Chemistry B 117, 8209-8222 (2013).
[23] [doi] F. Römer and F. Bresme, Molecular Simulation 38, 1198-1208 (2012).
[24] [doi] F. Römer, F. Bresme, J. Muscatello, D. Bedeaux, and M. J. Rub’i, Physical Review Letters 108, 105901 (2012).
[25] [doi] S. Weinbaum, J. M. Tarbell, and E. R. Damiano, Annu. Rev. Biomed. Eng. 9, 121-167 (2007).
[26] [doi] B. Button et al., Science 337, 937-941 (2012).
[27] [doi] P. Español and M. Revenga, Phys. Rev. E 67, 26705 (2003).
[28] [doi] F. Römer and D. A. Fedosov, EPL (Europhysics Letters) 109, 68001 (2015).