Sir

We welcome the views expressed in the Commentary by M. May, J. Davis and R. Jeanloz on source attribution of nuclear materials (“Preparing for the worst”, Nature 443, 907–908; 2006), in particular the suggestion to establish an international databank of such material to support the interpretation of nuclear forensics analyses.

We would like to comment on a few points, in particular those where reference was made to the Institute for Transuranium Elements (ITU) in Karlsruhe, part of the European Commission's Joint Research Centre.

The ITU has been involved since the early 1990s in the analysis of seized nuclear materials from cases of illicit trafficking. Its database, mentioned in the Commentary, contains materials from several western European and Russian fuel manufacturers, and was established in collaboration with the Bochvar Institute in Moscow. For reasons of commercial sensitivity, its content is largely confidential. Nevertheless, we assess information requests on a case-by-case basis.

Most nuclear or radioactive material has been seized in quantities too small for the production of a nuclear device. However, it is conceivable that small quantities could be used by terrorists for a radiological dispersal device, or 'dirty bomb', which would also require source attribution.

As your Commentary mentions, the Nuclear Smuggling International Technical Working Group discusses scientific and technical aspects of nuclear forensics analyses. Its 2006 meeting discussed various options for information sharing, including the creation of a 'super' database containing information about other organizations' databases. This decentralized approach may be easier to implement than a single centralized database. Non-nuclear states also need to be involved in this effort.

We agree that incidents involving nuclear material call for a comprehensive response, which relies on an organizational and technical infrastructure. This includes 'crime scene' management to preserve forensic evidence — nuclear as well as traditional — and maintain the chain of custody.

Credible nuclear forensics is based on representative samples and on high-quality analyses carried out in specialized multidisciplinary laboratories. Some parameters (such as age, intended use and production mode) are self-explanatory. But others (such as impurities and pellet dimensions) can best be understood by comparison to references, which calls for information on a wide variety of materials or the ability to pose queries in databases. Source attribution is essential in the fight against illicit trafficking and in the prevention of malicious acts.