Deterring a ‘Dirty Bomb’

Why we need nuclear forensics.

It's great that the 47 nations at last week's nuclear--security summit pledged to safeguard their nuclear materials so terrorists can't, say, nuke Times Square or set off a dirty bomb in the Tokyo subway. After all, as President Obama said, "we know that organizations like Al Qaeda are...trying to secure a nuclear weapon," and there is "a lot of loose nuclear material around the world"; nations must start "locking [it] down."

But even people who lock their doors can be robbed. To keep loose nukes out of the hands of terrorists, the world needs the successor to the deterrent power of mutual assured destruction (MAD) during the Cold War, when a nuclear attack would have come by missile with a de facto return address. Today, although a group like Al Qaeda might gloat about a successful attack, to rat out its source would be self-defeating, since it would assure retaliation against that source and eliminate it as a supplier.

The needed successor to MAD is, therefore, nuclear forensics and attribution: the science of inferring the source of nuclear materials from their chemical and isotopic properties. "If we could pinpoint the origin of nuclear material used in a terrorist attack, it would deter countries from allowing poor security at their nuclear facilities," says Benn Tannenbaum of the American Association for the Advancement of Science (AAAS). More bluntly, if hostile regimes perceive that the U.S. has an effective nuclear attribution capability, they might be deterred from helping terrorists obtain nuclear materials. (At least 40 countries have enough highly enriched uranium to build a crude atomic weapon, and 13 have enough reactor-grade plutonium.)

The U.S. has not stood out in advancing the cause of nuclear attribution even at home. A 2008 analysis by physicists and nuclear chemists for the American Physical Society and AAAS concluded the U.S. has too few experts (about 50) in nuclear attribution, and many are close to retirement. Training programs for the minimum of 35 new Ph.D.s needed over the next decade are "inadequate and underfunded," it warned; equipment to analyze debris after a detonation falls short of "the most modern and effective standards that prevail" in such countries as Japan and France.

Yet a 2008 bill was stripped of $4 million authorized to train experts. Even the Nuclear Forensics and Attribution Act that Obama signed in February "doesn't have any money" to remedy the shortfall of equipment and expertise, says Tannenbaum.

That is breathtakingly stupid, for nuclear forensics has made real advances lately. Nuclear materials have atomic and chemical properties that may survive detonation and "serve as unequivocal markers of specific sources, production processes, or transit routes," Michael Kristo of Lawrence Livermore National Lab told a U.S.-Russia workshop on nuclear security. Uranium can have a unique signature depending on where it was processed; the kind of reactor used to make weapons-grade plutonium leaves a telltale ratio of isotopes. The fraction of particles of different sizes in uranium-oxide powder can indicate what uranium conversion process was used, and therefore where it was carried out, while the dimensions of nuclear fuel pellets are often unique to a manufacturer. The ratio of oxygen-18 to oxygen-16 varies by region, and so can be used to pinpoint where nuclear fuel pellets were produced. Pollen and spores can indicate the route a nuke traveled (though only if it's intercepted; if it goes off, pollen and spores are incinerated).

Even so, it may be impossible with today's techniques to unambiguously trace nuclear materials to a unique source. That, Kristo warned, underlines the need to discover "new signatures...of the material that reveal the creator."

An even bigger need is a full library of nuclear samples. Only with a database of uranium or plutonium isotope ratios, trace impurities such as erbium or iron residues, and the molecular form of uranium or plutonium from reactors and fuel-making facilities worldwide could "day after" debris be matched uniquely to a source. The U.S., U.K., and France agree on the need for a nuclear library; China has been having "quiet conversations" about cooperating, says a physicist who did not wish to be identified discussing sensitive talks; and the nuclear summit emphasized the need for national libraries. But that is a far cry from a global library, which Russia and others in Europe are balking at (much of the information would come from commercial sources, which regard this as a trade secret). Failing to move ahead on a global library would be a serious missed opportunity; not even DNA fingerprinting can catch a murderer if you don't have samples to match it to.