『Abstract
　Soil organic matter consists of a highly complex and diversified blend of organic molecules, ranging from low molecular weight organic acids (LMWOAs), sugars, amines, alcohols, etc., to high apparent molecular weight fulvic and humic acids. The presence of a wide range of functional groups on these molecules makes them very reactive and influential in soil chemistry, in regards to acid-base chemistry, metal complexation, precipitation and dissolution of minerals and microbial reactions. Out of these functional groups, the carboxylic and phenolic ones are the most abundant and most influential in regards to metal complexation. Therefore, chemical equilibrium models have progressively dealt with organic matter in their calculations.
　This paper presents a review of six chemical equilibrium models, namely NICA-Donnan, EQ3/6, GEOCHEM, MINTEQA2, PHREEQC and WHAM, in light of the account they make of natural organic matter (NOM) with the objective of helping potential users in choosing a modelling approach. The account has taken various faces, mainly by adding specific molecules within the existing model databases (EQ3/6, GEOCHEM, and PHREEQC) or by using either a discrete (WHAM) or a continuous (NICA-Donnan and MINTEQA2) distribution of the deprotonated carboxylic and phenolic groups.
　The different ways in which soil organic matter has been integrated into these models are discussed in regards to the model-experiment comparisons that were found in the literature, concerning applications to either laboratory or natural systems. Much of the attention has been focused on the two most advanced models, WHAM and NICA-Donnan, which are able to reasonably describe most of the experimental results. Nevertheless, a better knowledge of the humic substances metal-binding properties is needed to better constrain model inputs with site-specific parameter values. This represents the main axis of research that needs to be carried out to improve the models. In addition to humic substances, more non-humic compounds should also be introduced in model databases, notably the ones that readily interact with the soil microorganisms. Thermodynamic data are generally available for most of these compounds, such as low molecular-weight organic acids. However, the more complex non-humic substances, exhibiting a ratio of hydrophobic versus hydrophilic bonds lower than humic substances, need to be further characterised for a comprehensive implementation in chemical equilibrium models.

Keywords: Modelling; Metal complexing; Organic acids; Humic acids; Soils; Aqueous solutions

1. Introduction
2. Main chemical equilibrium processes
　2.1. Basic speciation and phase equilibrium
　2.2. Surface interaction equilibrium
3. Why should NOM be accounted for in chemical equilibrium speciation models?
　3.1. The nature of NOM
　3.2. The properties of NOM
　3.3. The functions of NOM
4. How do the existing models cope with NOM?
　4.1. Non-humic substances
　4.2. Humic substances
　　4.2.1. Continuous site distribution−the NICA-Donnan model
　　4.2.2. Discrete site distribution−the WHAM model
5. Has organic matter been well accounted for? Examples of model applications
　5.1. Laboratory applications
　5.2. Model applications to natural systems
　　5.2.1. Analogue models
　　5.2.2. Humic ion-binding models
　　　5.2.2.1. On the need and difficulties to constrain field parameters
　　　5.2.2.2. Lack of knowledge of NOM speciation and binding properties
　　　5.2.2.3. Applications supported by a better knowledge of HS-binding properties
　　　5.2.2.4. Applications with some consideration on non-humic substances
6. How could organic matter be better accounted for in chemical equilibrium models? Some perspectives
Acknowledgements
References