Estimation methods — group contribution

Every estimated value on Rareytec Thermo comes from a published group-contribution method. This page says in a paragraph each what they do, and gives the primary reference for all of them. All values are Rareytec Thermo's own estimations from the molecular structure — no third-party databank is embedded.

The idea

A molecule is broken into structural groups — –CH₃, –OH, an aromatic carbon, and so on — and each group contributes a fitted amount to the property. The sum then goes through an equation that also accounts for molecular size. The Rarey-Nannoolal methods add two refinements that plain additivity lacks: second-order corrections for structural features a group list cannot see (ring size, ortho/meta/para substitution, steric crowding), and group interactions — an –OH next to another –OH does not behave like two isolated ones. Fragmentation follows a fixed priority order, so a given structure always yields the same groups.

The properties

• Normal boiling point T_b — the foundation. It is estimated first and then used as the reference temperature by nearly every other method here, so its error propagates into them. If you have an experimental boiling point, estimates built on it are markedly better.
• Critical properties T_c, P_c, V_c — the same group scheme with its own parameters; T_c and P_c are expressed relative to T_b.
• Vapour pressure — a group-contributed slope on a reduced-temperature form anchored at the boiling point, plus a correction for the characteristic bowing of mono-alcohol curves.
• Saturated liquid viscosity — two group sums: a slope dB_v, and a reference temperature T_v at which the liquid reaches 1.3 mPa·s. T_v is to viscosity what the boiling point is to vapour pressure.
• Thermal conductivity (Rarey-Govender) — linear in reduced temperature, with the value at T_b and the slope each from their own group sums.
• Surface tension (Rarey-Olivier) — a Guggenheim-type form with the exponent 11/9, again referenced to T_b.
• Activity coefficient at infinite dilution in water γ^∞ (Rarey-Moller) — a constant plus group contributions at 25 °C, with ortho/meta/para corrections for NH₂, OH, COOH and NO₂ pairs on an aromatic ring and an interaction term between OH, C=O and COOH groups that is scaled by molecular size. It was developed for complex multifunctional solutes — pharmaceuticals, pesticides and the like — and reproduces ln γ^∞ of its training set to 7.3 % (9.1 % on an independent test set). The first members of homologous series (methanol, ethylene glycol, acetone) are outside its focus. The temperature dependence depends on how the solute is solvated: there is one correlation for small polar solutes and one for non-polar and complex solutes, recommended between 280 and 350 K. The class is chosen by the user, not derived from the structure.

Thermal conductivity, surface tension and γ^∞ in water use a structural group set derived from Moller's vapour-pressure work rather than the boiling-point set — which is why the group table under a result can differ between properties for the same molecule. Each method is applied only within its measured range of validity; outside it a value is flagged or withheld.

Primary references

If you use estimated results in published work, please cite the method papers below and state that the values are estimates.

Journal articles

1. Y. Nannoolal, J. Rarey, D. Ramjugernath, W. Cordes, Estimation of pure component properties. Part 1: Estimation of the normal boiling point of non-electrolyte organic compounds via group contributions and group interactions, Fluid Phase Equilibria 226 (2004) 45–63. doi:10.1016/j.fluid.2004.09.001
2. Y. Nannoolal, J. Rarey, D. Ramjugernath, Estimation of pure component properties. Part 2: Estimation of critical property data by group contribution, Fluid Phase Equilibria 252 (2007) 1–27. doi:10.1016/j.fluid.2006.11.014
3. Y. Nannoolal, J. Rarey, D. Ramjugernath, Estimation of pure component properties. Part 3: Estimation of the vapor pressure of non-electrolyte organic compounds via group contributions and group interactions, Fluid Phase Equilibria 269 (2008) 117–133. doi:10.1016/j.fluid.2008.04.020
4. Y. Nannoolal, J. Rarey, D. Ramjugernath, Estimation of pure component properties. Part 4: Estimation of the saturated liquid viscosity of non-electrolyte organic compounds via group contributions and group interactions, Fluid Phase Equilibria 281 (2009) 97–119. doi:10.1016/j.fluid.2009.02.016
5. O. Govender, J. Rarey, D. Ramjugernath, Estimation of Pure Component Properties, Part 5: Estimation of the Thermal Conductivity of Nonelectrolyte Organic Liquids via Group Contributions, Journal of Chemical & Engineering Data 65 (2020) 1300–1312. doi:10.1021/acs.jced.9b00741

Theses

The theses carry the full derivations, the complete parameter tables and the group definitions in far more detail than the papers. These are open access in the University of KwaZulu-Natal repository.

1. Y. Nannoolal, Development of a group contribution method for the prediction of normal boiling points of non-electrolyte organic compounds, M.Sc.Eng. thesis, University of Natal, Durban, 2004. hdl.handle.net/10413/4337
2. Y. Nannoolal, Development and critical evaluation of group contribution methods for the estimation of critical properties, liquid vapour pressure and liquid viscosity of organic compounds, Ph.D. thesis, University of KwaZulu-Natal, Durban, 2006. hdl.handle.net/10413/3766
3. B. Moller, Development of an improved group contribution method for the prediction of vapour pressures of organic compounds, M.Sc.Eng. thesis, University of KwaZulu-Natal, Durban, 2007. hdl.handle.net/10413/1459 — the structural group set that the thermal conductivity and surface tension methods are built on.
4. B. Moller, Activity of complex multifunctional organic compounds in common solvents, Ph.D. thesis, University of KwaZulu-Natal, Durban, 2009. hdl.handle.net/10413/749 — the activity coefficient at infinite dilution in water (Chapter 7).

The thermal conductivity work also exists as an M.Sc. thesis by O. Govender (University of KwaZulu-Natal), but it is not findable in the public repository, so the journal article above is the citable reference.

Surface tension — a note on provenance

The surface tension method was developed by E. Olivier at the University of KwaZulu-Natal but never published, so there is no paper or thesis to cite; it is credited here as “based on E. Olivier (2011)”. Its parameter set contains no groups for alcohols or acids. Those were regressed by Rareytec Thermo on the openly licensed NIST/TRC ThermoML archive, and any result that uses one says so on the result card. They are our values, not the author's, and carry a larger uncertainty than the rest of the method.

← Help · Data Sources & Licensing →