Substances of unknown or variable composition, complex reaction products, and biological materials (UVCBs) account for an estimated 20–25% of chemicals registered globally and pose unique challenges for environmental risk assessment. A tripartite group within the Health and Environmental Sciences Institute (HESI) is developing an exposure-based, tiered framework to support the prioritization and evaluation of UVCBs.
The first tier (Tier 0) emphasizes basic characterization to help identify potential concerns related to hazard, persistence, or bioaccumulation. However, this step depends heavily on the availability and accuracy of foundational data, including exposure indicators, compositional information, and ecotoxicity summaries.
While UVCBs represent a significant proportion of commerce, publicly available data for these substances remains sparse. Many existing databases contain incomplete, outdated, or incorrectly assigned substance characterizations, complicating efforts to link information to the correct UVCB and limiting transparency for registrants, researchers, and regulators. These limitations highlight the need for a purpose-built, publicly accessible UVCB database.
To address this need, a subgroup of the HESI UVCB Committee has developed this pilot database compiling Tier 0 information. The effort aggregates identifiers, structural descriptors, registration and use information, physicochemical properties, process details, and ecotoxicity data.
Data for both whole substances and representative constituents are being collected from online databases, registration dossiers, and the scientific literature. Challenges encountered—such as inconsistent identifiers, data gaps, and harmonization requirements—have been documented to support reproducibility and guide improvements, including opportunities to automate aspects of data mining.
This pilot demonstrates the feasibility of establishing a curated UVCB database and underscores the complexities involved. Continued collaboration across industry, government, and academia will be essential to support data collection and to realize the full potential of this effort.
If you have any questions about the database or would be interested in contributing, please reach out to
jkrzykwa@hesiglobal.org.
| Designation in the Database | Example Data Sources | Detailed Description |
|---|---|---|
| Regulatory Assessment / Regulatory Dossier |
chem.echa.europa.eu services.industrialchemicals.gov.au canada.ca echa.europa.eu epa.gov ec.gc.ca atsdr.cdc.gov cdxapps.epa.gov regulations.gov |
Regulatory reviews and registration dossiers are sources produced for regulatory decision-making. They typically draw on a mix of proprietary and non-proprietary studies and supporting information, and they often cover a broad range of relevant chemistries that may be present in a UVCB. Manufacturer and consortium registration dossiers typically include substantial technical details that regulators can audit. As a result, these sources are generally medium to high quality, though quality can vary by dossier. Because companies frequently participate in consortia, the underlying data is often shared across multiple registrants, reducing the likelihood of capturing manufacturer-specific composition details. Some data, especially composition data, may not be publicly available due to confidentiality reasons. |
| Online Database |
pubchem.ncbi.nlm.nih.gov canadachemicals.oecd.org hpvchemicals.oecd.org chem-info.nite.go.jp comptox.epa.gov chemview.epa.gov guidechem.com ncbi.nlm.nih.gov chemicalbook.com |
Online database sources include publicly accessible databases and regulatory repositories. These sources are assumed to have undergone some level of initial screening and substance mapping. Still, uncertainties remain for UVCBs due to manufacturing variability, limited access to proprietary data, and potential limitations in substance mapping. Reliability varies with the amount of supporting documentation: sources with detailed documentation are more reliable, whereas repositories that provide limited metadata are less reliable because it is difficult to evaluate data quality from public information alone. |
| Expert Judgement |
assumed from name read-across from similar substance weight of evidence |
Expert judgement entries are estimates made by the Tier 0 DB team rather than values directly extracted from an external source. These carry added uncertainty because they depend on the basis used to generate the estimate:
Assumed from name: Reliability is uncertain because the value is inferred from the substance name rather than confirmed against a referenced primary source. Read-across from a similar substance: Reliability is uncertain because the value is inferred from a substance judged to be qualitatively similar; applicability depends on how similarity was determined. Weight of evidence: Reliability is uncertain because the value reflects a synthesis across multiple pieces of information rather than a single directly cited record. |
| Scientific Literature |
doi.org pure.tue.nl sciencedirect.com |
Peer-reviewed scientific literature sources report results from specific studies conducted under a defined testing scenario. For UVCBs, this typically provides greater confidence than generic data compilations because the data reflect observations made for a UVCB in a particular experiment. However, uncertainties remain because UVCB composition can vary substantially across manufacturers and batches. Overall, there is a reasonable likelihood the findings apply to the target UVCB, but applicability should be treated as conditional on how well the tested material aligns with the intended substance definition. |
| Commercial / Supplier / Manufacturer |
cleaninginstitute.org assets.publishing.service.gov.uk augustus-oils.ltd.uk bocsci.com h4rconsortium.com ventos.com |
Commercial, supplier, or manufacturer sources are information shared by companies about their own products or materials. For UVCBs, these sources may reflect a single producer’s formulation or production scenario and may not represent the full range of variability across manufacturers. Detail can be useful, especially for identifiers, product descriptions, or composition claims, but confidence is limited by the risk of incomplete scope and by the possibility that supporting data may be selective or not fully auditable. Treat as manufacturer-specific unless explicitly shown otherwise. |
| Other |
cpsc.gov hdl.handle.net iris.who.int img.perfumerflavorist.com |
“Other” is a catch-all for sources that do not fit into the other categories. For UVCBs, confidence and applicability vary and should be evaluated case by case based on the source’s actual content and whether it clearly pertains to the UVCB of interest under a defined scenario. |
hesi@hesiglobal.org
Phone: +1-202-659-8404
Fax: +1-202-659-8403
740 15th Street NW, Suite 600
Washington, DC 20005
Sign up for our monthly e-newsletter.