Documentation / ChemSep data

ChemSep data

Cheméo serves values from the ChemSep v8.3 pure component data, by Harry Kooijman and Ross Taylor (2021, 2022). The data are distributed under the Artistic License 2.0. Every value taken from them is cited as "ChemSep v8.3" on the compound pages, and the original data are at chemsep.org.

The values on Cheméo are Cheméo's conversion of the ChemSep files, a modified version: they are converted to Cheméo's units and correlation forms, some ranges are narrowed, and part of the files is left out. This page documents every difference, as the licence asks.

The files

DWSIM's copies of ChemSep's chemsep1.xml ("ChemSep v8.3 pure component data", 2021) and chemsep2.xml ("ChemSep v8.31 pure component data 2", 2022), from the DWSIM repository at commit 1abf72d:

file	sha256
`chemsep1.xml`	`31e72c40b2b3b084161314c3262e227fcdd773c0bf2070bacbae9c5da203533f`
`chemsep2.xml`	`3dfc57f4d5f0c9c0fb4dfb6caffe860e0dffe44618eb091a91f9ca5dd932cd2a`

They hold 558 compound records, 556 distinct CAS numbers. Benzoic acid and cumene hydroperoxide appear in both files with different values: Cheméo uses the chemsep1.xml record and leaves the other out.

A record is matched to a Cheméo compound by its structure (its SMILES). A record that cannot be matched is not shown. When two records (two CAS numbers) have the same structure, only one of them is shown.

Constants

Each constant is shown as a point from ChemSep, converted to Cheméo's unit:

ChemSep	unit in the file	Cheméo property	conversion
Critical temperature	K	Tc	none
Critical pressure	Pa	Pc (kPa)	÷ 1,000
Critical volume	m³/kmol	Vc	none
Critical compressibility		Zc	none
Acentric factor		ω	none
Normal boiling point	K	Tboil	none
Melting point	K	Tfus	none
Triple point temperature	K	Ttriple	none
Triple point pressure	Pa	Ptriple (kPa)	÷ 1,000
Heat of formation (ideal gas, 298.15 K)	J/kmol	ΔfH°gas (kJ/mol)	÷ 10⁶
Gibbs energy of formation	J/kmol	ΔfG° (kJ/mol)	÷ 10⁶
Rackett parameter		Zra	none
Dipole moment	C·m	debye	÷ 3.33564 × 10⁻³⁰
UNIQUAC r and q		UNIQUAC Ri, Qi	none

Left out:

zeros used as placeholders: a value of exactly 0 for the critical temperature, pressure or volume, or for a boiling, melting or triple point;
a heat or Gibbs energy of formation of 0, except for an element. In the files: two heats and one Gibbs energy.

Temperature-dependent properties

Each is shown as a correlation, with ChemSep's validity range:

ChemSep	unit in the file	Cheméo property	unit factor f
Vapour pressure	Pa	Pvap (kPa)	10⁻³
Ideal gas heat capacity	J/kmol/K	Cp,gas (J/mol×K)	10⁻³
Liquid heat capacity	J/kmol/K	Cp,liquid (J/mol×K)	10⁻³
Heat of vaporization	J/kmol	ΔvapH (kJ/mol)	10⁻⁶
Liquid density	kmol/m³	ρl (kg/m³)	the record's molecular weight
Liquid viscosity	Pa·s	η	1
Liquid thermal conductivity	W/m/K	λ, liquid	1
Vapour thermal conductivity	W/m/K	λ, gas	1
Surface tension	N/m	γ	1

Equation forms

ChemSep numbers its equations as DIPPR does. The unit factor f is folded into the coefficients, which is exact: no curve is refitted.

ChemSep equation	Cheméo form	coefficients
1 to 5, polynomials	`dippr-100`	equation n keeps its first n coefficients, each × f; the others are 0
100	`dippr-100`	A to E × f
101	`dippr-101`	A + ln f; B to E unchanged
10, ln Y = A − B/(T + C)	`antoine`, ln Y = A + B/(T + C)	A + ln f, −B, C
102	`dippr-102`	A × f
105	`dippr-105`	A × f
106	`dippr-106`	A × f; Tc is the record's own critical temperature
107	`dippr-107`	A, B and D × f
16, Y = A + exp(B/T + C + D·T + E·T²)	`chemsep-16`	A × f, C + ln f

Equation 10 is read with the natural logarithm: it gives 1 atm at the normal boiling point. Four liquid heat capacities in equation 4 carry a fifth coefficient E. It is not part of the cubic equation and is dropped.

Left out:

equations with no Cheméo form: 116 (41 surface tensions), 200 (5 vapour pressures), 118 (1 liquid viscosity);
every other element of the files:
- properties: absolute entropy, heat of combustion, heat of fusion, vapour viscosity, the second liquid viscosity (RPS), the second (Antoine) vapour pressure, the RPP ideal gas heat capacity, second virial coefficient, solid heat capacity and density, relative permittivity, liquid volume at the normal boiling point, specific gravity, parachor, solubility parameter, radius of gyration, Lennard-Jones diameter and energy, van der Waals area and volume, Fuller and Wilson volumes, charge;
- model parameters: Mathias-Copeman, COSTALD, Chao-Seader, API-SRK, UNIQUAC q', and the group assignments (UNIFAC VLE and LLE, modified UNIFAC, ASOG, UMR, GC method);
- the molecular weight, which only converts the liquid density (Cheméo computes its own from the structure).

Ranges, narrowed and never extended

Capped at the critical temperature. A dippr-106 curve (or a dippr-105 curve, at its C) whose range runs past it ends there: 52 heats of vaporization, 13 surface tensions, 1 liquid density. Water's heat of vaporization, given to 647.28 K, ends at its critical temperature, 647.14 K. For a few records the critical temperature of the file lies 9 to 24 K below the range of its own curves (octanal to tridecanal, ethylacetylene, acetyl chloride, biphenyl), so their curves end that much earlier.
Cut before a negative tail. 190 surface tensions turn slightly negative (up to about −0.7 mN/m) as they approach the critical temperature. The range ends at the last temperature before the first negative value, on a grid of 1,001 temperatures.
The original upper limit is kept in the comment. A narrowed curve says "ChemSep Tmax … K" there.
Refused curves, not shown:
- an empty range, or one starting at 0 K: 2 liquid heat capacities, 1 surface tension;
- a curve that cannot be evaluated over its range: 1 heat of vaporization (dimethylacetylene, undefined at the file's critical temperature);
- a curve that is negative at the start of its range: 1 vapour thermal conductivity.

What ChemSep does not say

ChemSep does not flag which of its values are measured and which are estimated. Cheméo shows them as ChemSep's values, cited to ChemSep, beside the other sources of each compound.