Amine CO2 Solubility — Kent-Eisenberg Model

NeqSim provides a fast, screening-level model for the solubility of carbon dioxide in aqueous alkanolamine solvents. It implements the classic Kent-Eisenberg approach: acid-gas speciation is governed by a small set of apparent (lumped) equilibrium constants, and the physically dissolved free CO2 is related to its partial pressure through Henry’s law.

The model is intended for process screening — solvent selection, loading and circulation-rate estimates, and qualitative comparison of primary, secondary and tertiary amines. It is not a custody-grade VLE package.

Supported amines

Amine Type Class enum Molar mass (g/mol) Typical loading window (mol CO2 / mol amine)
MEA Primary MEA 61.08 0.2 – 0.5
DEA Secondary DEA 105.14 0.2 – 0.5
MDEA Tertiary MDEA 119.16 0.1 – 1.0
aMDEA Activated MDEA (MDEA + piperazine) AMDEA 119.16 0.1 – 1.0

Primary and secondary amines (MEA, DEA) form a stable carbamate, bind CO2 strongly at low loading, and saturate near a stoichiometric limit of about 0.5 mol/mol. Tertiary MDEA cannot form a carbamate; it promotes bicarbonate formation, binds CO2 more weakly, and can be loaded toward 1.0 mol/mol.

AmineKentEisenberg (the core static model) supports MEA, DEA and MDEA. AmineSystem (the convenience wrapper) additionally accepts AMDEA, which is mapped to the MDEA correlation for the screening partial-pressure path.

Theory

CO2 absorbed into an aqueous amine is partitioned between a small amount of physically dissolved (free) molecular CO2 and a much larger pool of chemically bound species (bicarbonate, carbonate, carbamate). Only the free CO2 sets the vapor-phase partial pressure through Henry’s law:

\[p_{CO_2} = H_{CO_2}(T)\,[CO_2]_{\text{free}}\]

The free-CO2 concentration is found from the loading (total absorbed CO2 per mole of amine) by solving the apparent-equilibrium speciation. For a tertiary amine the dominant reaction is bicarbonate formation:

\[CO_2 + R_3N + H_2O \;\rightleftharpoons\; R_3NH^+ + HCO_3^-\]

For primary and secondary amines, carbamate formation dominates at low loading:

\[CO_2 + 2\,R_2NH \;\rightleftharpoons\; R_2NCOO^- + R_2NH_2^+\]

The temperature dependence of the Henry coefficient drives the absorber / stripper duality: cold solvent holds CO2 at low partial pressure (absorption), while hot solvent rejects it at high partial pressure (regeneration).

Validated screening API

The design-default, validated entry point is the partial-pressure calculation. There are two ways to call it.

1. Static model (AmineKentEisenberg)

import neqsim.thermo.util.amines.AmineKentEisenberg;
import neqsim.thermo.util.amines.AmineKentEisenberg.AmineType;

// Solvent molarity from mass fraction and amine molar mass
double molarity = AmineKentEisenberg.amineMolarity(0.50, 119.16); // 50 wt% MDEA -> ~4.36 mol/L

// CO2 partial pressure [bara] at 40 C and loading 0.4 mol CO2 / mol MDEA
double pCO2 = AmineKentEisenberg.partialPressureCO2Bara(
    AmineType.MDEA, 313.15, molarity, 0.40); // ~0.229 bara

partialPressureCO2Bara(type, temperatureK, amineMolarity, loading) returns the equilibrium CO2 partial pressure in bara. It returns 0.0 at zero loading and throws IllegalArgumentException for non-physical inputs (negative temperature, molarity, or loading).

2. Convenience wrapper (AmineSystem)

import neqsim.thermo.util.amines.AmineSystem;
import neqsim.thermo.util.amines.AmineSystem.AmineType;

AmineSystem solvent = new AmineSystem(AmineType.MDEA, 313.15, 1.0);
solvent.setAmineConcentration(0.50); // 50 wt% MDEA
solvent.setCO2Loading(0.40);          // mol CO2 / mol amine

double pCO2 = solvent.getCO2PartialPressure(); // validated screening path, ~0.229 bara

AmineSystem.getCO2PartialPressure() is the validated screening path used in design workflows. For AMDEA it uses the MDEA correlation.

Heat of absorption (optional)

import neqsim.thermo.util.amines.AmineHeatOfAbsorption;
import neqsim.thermo.util.amines.AmineHeatOfAbsorption.AmineType;

AmineHeatOfAbsorption hoa = new AmineHeatOfAbsorption(AmineType.MEA, 0.30, 0.30, 313.15);
double dH = hoa.calcHeatOfAbsorptionCO2(); // kJ/mol CO2 (negative = exothermic)

Validation status

The screening model is calibrated and regression-tested against the engineering loading windows of each amine class. In the validated windows it reproduces literature isotherms to within roughly a factor of two on CO2 partial pressure — adequate for solvent screening and loading/circulation estimates, but not for final tower rating.

Selected regression anchors (verified in AmineCO2SolubilityTest):

Case Condition Verified result
50 wt% MDEA molarity mass fraction 0.50, MW 119.16 ≈ 4.36 mol/L
30 wt% MEA molarity mass fraction 0.30, MW 61.08 ≈ 5.03 mol/L
MDEA partial pressure 40 °C, loading 0.10 ≈ 0.0093 bara
MDEA partial pressure 40 °C, loading 0.40 ≈ 0.229 bara
Carbamate vs bicarbonate below half loading MEA binds CO2 tighter than MDEA
Stripping ratio 100 °C vs 40 °C, same loading hot / cold pCO2 > 5×

Experimental rigorous path. AmineSystem.getCO2PartialPressureRigorous() runs a full electrolyte-CPA equilibrium (SystemElectrolyteCPAstatoil, mixing rule 10, amine physical-property model, chemical reactions enabled). It is experimental and not yet calibrated — it may return NaN when the equilibrium does not converge. Use getCO2PartialPressure() for design work.

References