Reliably predicting material properties with the help of Computational Chemistry is of critical importance in the Research & Development of many industries including the energy & utilities, consumer products & retail, medical, and pharmaceutical industries. Likewise, quickly, and accurately obtaining fluid properties with basic chemical knowledge becomes increasingly important for other simulation and engineering disciplines, like system simulation or Computational Fluid Dynamics.


Simcenter Culgi 2211 is a dedicated calculation and simulation tools to perform such material property calculations quickly and reliably.

Be productive immediately with dedicated simulation tools tailored to industry specific applications or the prediction of specific material properties.

With tailored frontends directly embedded into the Culgi user interface and a back-end incorporating simulation best practices, like e.g. solver settings, etc. you can now leverage Computational Chemistry simulations with ease.

Designed as a transparent architecture built directly on Simcenter Culgi technology, expert users can easily adjust templated simulations.

List of apps available in Simcenter Culgi 2211:

  • Boiling point and vapor pressure
  • Diffusion coefficient
  • Heat of evaporation
  • Interfacial tension
  • Liquid-liquid phase diagram
  • Mixing free energy
  • Mobility and conductivity
  • Partition coefficient
  • Solubility parameters
  • Solvation free energy
  • Vapor-liquid phase diagram
  • Zero shear viscosity

Automatically run simulations at the required level of detail with fully integrated automated parametrization and fragmentation (AFP)

Many industrial applications happen at the microns length scale and milliseconds timescale, and yet they heavily depend on the chemical and molecular processes at smaller sales. Therefore, it is of critical importance to bridge insights and predictions from computational chemistry to the continuum scale.

Simcenter Culgi has been built around the idea of seamlessly switching between the right simulation method to capture the chemistry and molecular dynamics from quantum mechanics up to the mesoscale and ultimately bridge the gap to continuum descriptions.

With the introduction of the fully integrated automated parametrization and fragmentation (AFP) method in Simcenter Culgi , we take multiscale modeling to the next level. Thanks to the automatic switching of methods you can now run your computational chemistry simulations at the required level of detail without worrying about the appropriate level of length and time scales. Generally applicable for any given application, Simcenter Culgi 2211 will automatically go to a larger length and timescale where the relevant things happen. This will save you significant simulation time while ensuring accurate results without the risk of missing out on critical details. 

  • Gain detailed insights into reactivity with integrated orbital visualization for electronic structure characterization

For many industrial applications and in the quest for novel materials and products chemists and material scientists need to understand and leverage the reactivity of molecules. Whether it is the mitigation of off-flavors resulting from undesired oxidation reactions of a product ingredient in the food and beverage industry or the optimization of detergent ingredients for better performance and sustainability, an in-depth understanding of the reactivity of molecules is critical.

Simcenter Culgi 2211 enables you to understand the reactivity stemming from Quantum Mechanics simulations of a given molecule by visualizing the electronic orbital structure in 3-dimensional space.

Fully embedded into the Simcenter Culgi interface this spacial displaying functionality enables you to identify the highest occupied and lowest unoccupied orbitals with ease, and hence understand quickly what part of the molecule will react. The functionality complements existing Infrared calculation visualization to gain detailed insights and further explore the possibilities of novel materials.  

  • Automatically generated SMILES bead names after fragmentation of atomistically detailed molecules

The mapping of atomic fragments to coarse-grained beads now also makes it possible to describe the beads as SMILES strings. The SMILES notation is a powerful way to capture the chemistry of molecules in a string. With this option, the user can directly understand how the molecule is fragmented and at the same time keep track of the underlying chemistry and topology when running coarse-grained simulations.

Simcenter CULGI is a multi-layered software that can be used from non-expert to the expert by using

  • Predefined packages: for example designing new molecules (lab technicians)
  • Developing or using our tailored-made scripts (from the lab technician to the well-versed engineer)
  • Take advantages of other python packages by exporting your scripts, using CULGIPy and any other packages to develop your solution (Experts)

Simcenter CULGI: Key technology to achieve desired results

  • Graphical Programming Environment (GPE): An intelligent scripting editor that encompasses all relevant molecular modeling algorithms
  • Automated Bond Correction (ABC): algorithms to calculate COSMO charge envelop of a molecule by bond-corrected semi-empirical quantum calculation.
  • Automated Fragmentation and Parameterization (AFP): algorithms to coarse-grain molecules automatically, fast and accurately
  • Stokesian Particle Dynamics (SPD): algorithms to simulate creeping-flow particle dynamics on a microscopic scale, including reactions (Kinetic Monte Carlo)

Quantum (NWChem), ABC

Molecular Modeling (MD, etc.)

(reactive) Mesoscale Modeling (DPD, SPD)

Scale Mappers (AFP)

Engineering Thermodynamics (COSMO-SAC/RS, UNIFAC)

Chemical Informatics and AI

  • Molecular Density Functionals approaches (RISM 1D, Mesodyn 3D)

Markush structures

Open data structure

Apps: Nanorheometer, IR Spectra, Solubility,…

Library of examples and demo-workflows

Library of COMSO-ed molecules

All atom model

The idea of model building in molecular modeling is to explicitly represent each atom in a system in the model. This is known as an all-atom model. This is the most common approach to molecular modeling.

However to speed up molecular simulations, a united-atom approach is sometimes used, in which several atoms are grouped together to create a pseudo-atom (or united atom). With the proper parameterization of force fields, properties predicted from a united-atom model can be as accurate as those from an all-atom model.

In spite of its ability to accurately reproduce physical, chemical, and thermodynamic properties of materials, all-atom (AA) molecular dynamics (MD) is constrained by its characteristic length and time scales, and by the computational resources at one’s disposal.

Coarse graining

  • Coarse-grained (CG) modelling is a process of representing atomistic systems with fewer degrees of freedom. Mechanics, classical thermodynamics, and kinetic theories can be considered as CG models with a large number of degrees of freedom eliminated, to provide a considerable degree of mathematical simplicity
  • From the MD point-of-view, CG models consist of interacting mass points (CG beads) that each correspond to a group of atoms in an AA simulation of the same system. They can thus be viewed as relatively low resolution models, fairly consistent with their more detailed AA counterparts. Because these models have fewer degrees of freedom, there are fewer particles to compute and fewer neighbours to consider per particle. CG models with a reduction factor in the number of particles, N, exhibit a computational speedup of the order of N2

The CG models provide computational efficiency while successfully reproducing several properties of AA models in MD simulations. Coarse-graining also enables a simplification of interaction potentials and provides a smoothed potential energy surface for molecules to move on, enabling an acceleration in MD calculations, thus permitting the use of larger integration time steps (5–20 fs)


The Monte Carlo method was first introduced by Metropolis et al. [43] to study a system of molecules in the liquid state. The Metropolis Monte Carlo method aims at generating a trajectory in phase space that samples a desired statistical mechanical ensemble.

In Culgi, the Monte Carlo method has been implemented for simulation of atomistically detailed (molecular) models, as well as for soft-core particle models. In the current version, Culgi supports MC simulations in the following ensembles (termed as MC-mode in the software):


semiempirical quantum chemistry methods

In the 1970s and 1980s, accurate ab initio quantum chemistry calculations for molecules containing more than a few non-hydrogen atoms were generally not feasible on the computer systems of that time. As a way around that, semiempirical quantum chemistry methods were invented. These methods use a number of severe approximations that are then partially compensated for by introducing parameters that are fitted to experimental data. This gives a class of methods that are much faster than ab initio and DFT methods, and in many cases still give relevant results.

Culgi provides the AM1 method of Dewar, Zoebisch, Healy and Stewart, which falls within the MNDO family of methods pioneered by Dewar and Thiel.

Culgi also contains RM1, a 2006 reparameterization of AM1 by Rocha, Freire, Simas and Stewart.


COSMO (COnductor-like Screening MOdel) is a dielectric continuum solvation model that is effective in including solvent effects in a quantum chemistry calculation. The advantage of COSMO and other continuum solvaton models is that they provide a good description of properties of molecules in a solvent, with computational costs comparable to gas phase calculations.

However, one of the major limitations of such theories is that they assume a linear behavior in solvent polarizability with respect to the strength of the solute-induced electric fields.

To overcome this deficiency, a COSMO-based thermodynamics method (COSMO-RS, COnductor-like Screening MOdel for Real Solvents) was introduced by Klamt [80], where the ideally screened molecule from COSMO is taken as the primary description of a molecule and the deviation from ideality is considered by including pairwise misfit interactions of the ideal screened charges between molecules in the fluid.

Culgi focuses on COSMO-RS-like thermodynamics methods based on quantum COSMO calculations from external quantum chemistry packages.

  • In the COSMO method (originally published by by Andreas Klamt and Gerrit Schüürmann) the molecule of interest is not considered in vacuum, but in a conductive shell. This shell represents a polarizable solvent with a particular dielectric constant. It is approximated by a large number of point charges such that as little of the electron cloud as possible extends outside the shell (that is, the electrostatic potential at the surface is zero) thus approximating an ideal conductor (infinite dielectric constant).
  • After charges have been determined in this way, and the electron distribution is known, one can calculate the electrostatic energy due to the conductive shell. For ‘real’ solvents, the dielectric constant is of course not infinite, and a scaling is applied to get point charges corresponding to the dielectric permittivity of the solvent of interest. The total energy can then be calculated both with and without conductive shell, and the difference can be taken as a solvation energy (after taking into account an additional cavitation/dispersion energy).
  • The COSMO-based thermodynamics methods (COSMO-RS and later reimplementations such as COSMOSAC) extend the COSMO concepts to describe solvation and mixing in realistic solvents.

Here, the relevant charges are also calculated for a perfectly conductive shell. Then, after averaging these charges over a small area, a histogram is made of the surface charge densities according to the total surface area with a particular surface charge 𝜎. These so-called sigma profiles give important and detailed information on how a molecule would interact electrostatically with other molecules. On the basis of sigma profiles for different molecules, it is in many cases possible to obtain a very decent estimate of their mixing thermodynamics.


  • To bridge the gap between atomistic simulations and macroscopic network simulations, and to overcome the inherent difficulties faced by conventional methods when applied to complex fluid systems, we need an intermediary technique focused at a length scale larger than the atomistic scale, but smaller than the macroscopic connection scale
  •  Mesoscopic simulations aim at identifying characteristic physical lengths and times in the system in order to use them for simplification of complex models. In particular, for soft matter and polymeric systems, the hydrodynamic behaviors are captured easily using continuum methods while it is expensive to handle them at atomistic levels.

Dissipative particle dynamics was first introduced by Hoogerbrugge and Koelman [104]. The coarse graining of the structure and the soft interactions used in DPD simulations allow larger systems to be modelled over significantly longer times than is possible with Molecular Dynamics.

The advantage of DPD over Brownian Dynamics is the inclusion of hydrodynamics. Brownian Dynamics may provide a faster method for studying dilute systems, since no explicit solvent is needed. When hydrodynamics does not play a role in the physics, BD and DPD should give similar results.

DPD should certainly be used instead of BD for any study involving shear or flow.

  • The dissipative particle dynamics (DPD) is a potentially very powerful and simple mesoscopic approach, which facilitates the simulation of the statics and dynamics of complex fluids and soft matter systems at physically interesting length and time scales. Since 1990, when the method was first developed in Europe, DPD has been applied in the study of the dynamical properties of a wide variety of systems and applications,
  • DPD, as an off-lattice technique, does not suffer from some of the restrictions imposed by the lattice as in the LGA or LB methods. We believe that presently, the DPD is arguably one of the best mesoscale simulation techniques, and in the near future, it has the potential to emerge as an even more widely used modeling and simulation technique for many complex fluid systems.

This method can be perceived as the clustering a number of molecules into single particle where the number of molecules per DPD particle is known as the coarse-graining parameter and is usually denoted by Nm. This coarse graining parameter plays a vital role and has significant impact on the speed of simulation


  • Multiscale modeling techniques promise access to longer time and length scales in simulation and analysis by leaving out certain non-essential details of a system and looking at the system on a coarser scale. An important tool for these techniques is a mapper that allows us to go from a specific molecular model to a coarse grained model (or the other way around) in a convenient way, preferably automatically. The idea is that a mapper takes a system on one level of detail and generates a system on another level of detail, while preserving as much of the essential structure as possible.

One use of such a structure-preserving conversion is obtaining reasonable simulation parameters for coarsegrained systems. An example: suppose we represent a polymer by a chain of ‘blobs’ that each represent one or more monomers, and we have a way of mapping the coordinates of an atomistic chain onto the coarsegrained chain (schematically indicated in figure 25.1). We can then take a set of realistic conformations of atomistically detailed chains, map them onto coarse-grained chains, and obtain typical distributions of bond lengths and angles for the ‘blobs’. These distributions can then be used to obtain parameters for use in e.g. DPD simulations.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Ut elit tellus, luctus nec ullamcorper mattis, pulvinar dapibus leo.