The majority of substances to be registered under REACH will require a full set of physico-chemical data depicting physico-chemical properties. Physico-chemical data are used to assess the physical hazards (e.g. flammability) and help predict possible toxicological or environmental hazards, fate and behaviour. They are used mainly for the purposes of safe handling but also, in the determination of the risk posed to humans and the environment from all stages of a substance lifecycle (see table Impacts and uses of physico-chemical data under REACH).

Physico-chemical data requirements for substances ≥ 1 tonne per year (per registrant)

• Melting/freezing point (°C or K): temperature at which the phase transition from solid to liquid state at normal atmospheric pressure takes place. Information on the melting point will impact the choice of method for flash point, flammability, autoflammability, oxidising properties and explosive properties.
  
• Boiling point (°C or K): temperature at which the pressure of the saturated vapour of a liquid equals the standard atmospheric pressure. This data is one of the criteria used in assigning a substance to an appropriate flammability category.
  
• Relative density (dimensionless): ratio between the mass of a volume of the substance, determined at 20°C, and the mass of the same volume of water, determined at 4° C. Relative density is not used for Classification and Labelling (C&L) but information on it is used in the determination of viscosity (as required for the classification criteria for aspiration hazard).
  
• Vapour pressure (Pa or N/m2): saturation pressure above a solid or liquid substance. This data is not used as a C&L criterion or to define persistent, bioaccumulative and toxic (PBT) properties, but is a key parameter in determining the environmental fate and behaviour, for environmental and human health risk assessments.
  
• Surface tension (N/m): surface free energy per unit of surface area. It corresponds to the minimum work required to expand the surface by one unit area. Surface tension is not used as a C&L criterion, to define PBT properties, or as a specific property in the chemical risk assessment. It can be used to provide guidance as to whether a chemical would be considered a surfactant under EU Regulation 648/2004 (last modified by Regulation 907/2006 - Detergents).
  
• Water solubility (kg/m3 or g/l): specified by the saturation mass concentration of the substance in water at a given temperature. This property is not a C&L criterion as such. It applies to substances unless there is existing additional scientific evidence concerning degradation and/or toxicity, sufficient to provide an adequate assurance that neither the substance nor its degradation products will constitute a potential long-term and/or delayed danger to the aquatic community.
  
• Partition coefficient n-octanol/water (Kow, dimensionless): defined as the ratio of the equilibrium concentrations of a dissolved substance in a 2-phase system consisting of n-octanol and water. It is a critical parameter for the chemical risk assessment, C&L, and PBT assessment.
  
• Flash point (°C or K): lowest temperature, corrected to a standard pressure of 101.3 kPa, at which a liquid evolves vapours, under the conditions defined in the test method, in such an amount that a flammable vapour/air mixture is produced. This data is used to allocate a substance into the appropriate flammability class.
  
• Flammable properties: they include pyrophoricity, flammability and flammability on contact with water.
• Pyrophoricity: a substance is pyrophoric if it ignites spontaneously within five minutes of being exposed to air under the conditions of a standardised test;
• flammability:
  • a flammable gas is a gas having a flammable range with air at 20° C and 101.3 kPa,
  • a flammable liquid is one with a flash point below the upper limit set in the C&L criteria,
  • a flammable solid is a readily combustible solid (powdered, granular or pasty substance). It can be easily ignited by brief contact with an ignition source (such as a burning match) and the flame spreads rapidly;
• flammability on contact with water: substances which, in contact with water, are liable to become spontaneously flammable or emit flammable gases in dangerous quantities.
The flammable properties tests are designed to allocate a substance into the appropriate hazard class.

• Explosive properties: tendency of a substance to undergo violent and rapid decomposition, under appropriate conditions, to produce heat and/or gas. The explosive properties tests are designed to allocate an explosive substance into the appropriate hazard class. There are some substances, which although they do not fall into one of the explosive hazard classes, are on the borderline of being explosive. Consideration should be given to applying a suitable hazard statement to these substances.
  
• Self-ignition temperature (°C or K): auto-ignition temperature for gases & liquids and relative self-ignition temperature for solids.
• Gases & liquids: lowest temperature at which a substance will ignite when mixed with air under the conditions defined in the test method;
• solids: minimum temperature at which a certain volume of a substance will ignite under defined conditions.
This data is not used directly for C&L, but they can be used for safety handling and risk assessment.

• Oxidising properties: while in themselves not necessarily combustible, substances with oxidising properties may cause or contribute to the combustion of other material. These properties tests are designed to allocate an oxidising substance into the appropriate hazard class by comparison to one or more reference substance(s).
  
• Granulometry (effective hydrodynamic radius, m): the different particle sizes defined in EN 481 document1 are:
• inhalable fraction: mass fraction of particles that can be inhaled by nose and mouth;
• thoracic fraction: mass fraction of particles that passes the larynx;
• respirable fraction: mass fraction of particles that reaches the alveoli.
For particle size distribution, the parameter of interest is the effective hydrodynamic radius, or effective Stoke’s radius Rs. The particle size distribution is needed in order to decide which route of administration is most appropriate for animal toxicity studies (acute toxicity and repeated dose toxicity). The determination of the particle size fractions is used to assess the possible health effects resulting from inhalation of airborne particles in the workplace.

Physico-chemical data requirements for substances ≥ 100 tonnes per year (per registrant)

• Stability in organic solvent: percentage of the concentration of the test substance in the solvent extract at a particular time period compared with the initial starting concentration of the test substance at t = 0. Information on the stability of a substance in a solvent is desirable, particularly when samples are to be stored.
  
• Dissociation constant: ratio of concentrations of dissociated and undissociated forms of a substance in water at equilibrium. In the case of an ionisable organic substance, this data indicates which chemical species will be present at a particular pH (fate and toxicity of the ionised form of a substance may be markedly different from the corresponding neutral molecule).
  
• Viscosity: measure of the resistance of a fluid to being deformed by either shear stress or extensional stress (commonly perceived as resistance to pouring). This data is used in human health risk assessment (liquid substances and preparations may present an aspiration hazard in humans because of their low viscosity).

Fulfilling physico-chemical data requirements

Manufacturers and importers have to follow 4 steps to fulfil the information requirements for registration: 1) gather existing relevant information, 2) consider information needs, 3) identify information gaps and 4) if necessary generate new data or propose testing strategies (for more information see Scheme to fulfil the data requirements).
 Available physico-chemical data is experimental or non-experimental data. It is published in many sources (environmental handbooks, scientific journals, databases). These sources can be primary references (the best approach) or secondary. They can be historical data sources (appropriate if reliable and authoritative, should be used in a Weight of Evidence approach).The table Sources of physico-chemical data for REACH requirements contains a short list of useful sources of physico-chemical data. They are extracted from the Guidance on information requirements under REACH. Note that some properties are not recorded in standard textbooks or scientific publications: stability in organic solvent and degradation products, viscosity and granulometry.
Available data must be evaluated to determine if results are valid or not (sufficient quality, rigour and reproducibility). For instance, historical data must be checked due to the fact that sometimes the original test reports are not available or are incomplete.
Results are scientifically acceptable, if experimental data is generated from test(s) using an appropriate standardised method which is GLP2 (various OECD3 and EC4 test guidelines exist, see table Methods for determining the physico-chemical properties under REACH). If experimental data came from tests that have not been done under GLP, they will also be accepted, provided that they have been obtained using an appropriate test method and that there is sufficient documentation about quality procedures (i.e. compliance with ISO 170255). Sometimes expert judgement is necessary, where a non-standard test method is used, due to the wide range of modifications and variations that are possible.
In the case of non-experimental data, physico-chemical properties can be estimated using a computer program based on Quantitative Structure Property Relationship (QSPR) or read-across predictions. QSPR is a mathematical relationship between chemical structure and a specific physico-chemical property. Some reliable software is outlined in Guidance on information requirements under REACH. Some properties cannot be predicted by such models: flash point, flammability, explosivity, self-ignition temperature, oxidising properties, granulometry and stability in organic solvents. Predictions from transparent QSPR models can be accepted if they are supported by adequate and reliable documentation. If not, the use of QSPR estimation techniques requires some expert judgement. Provided that it is practicable, it is always better to obtain property predictions from at least three different methods. A read-across/analogue approach assesses a given property of one chemical structure and then makes some assessment (qualitative or quantitative) of this information for non-tested chemical(s). This approach needs an expert judgement evaluation. It must be pointed out that, in practice, read-across for physico-chemical properties is not generally recommended, since reliable data should normally be available or easily obtainable.
When testing is to be carried out there are advantages in considering the order in which testing is done. Ideally, when a full set of physico-chemical tests is to be performed they should be done according to the plan summarised in the diagram Tiered testing scheme on physico-chemical testing. Note that in certain cases, tests are not technically possible or are unnecessary. For more information, see table Adaptations of the standard physico-chemical testing requirements under REACH.
The experimental tests should be carried out according to recognised test methods and preferably under a Quality Assurance regime (possibly, under conditions of Good Laboratory Practice, although this is not a requirement for REACH). Methods and practices conform to GLP standards promote the transparency and credibility of the submitted data by ensuring their quality and integrity.