Putting atoms in the balance, one by one.

When a proton captures a neutron to create a nucleus of deuterium, the interaction releases energy in the form of gamma rays. The mass of the resulting nucleus ends up slightly less than the sum of the masses of the individual particles involved. This mass loss is presumed equivalent to the energy released.

Researchers have now measured the atomic masses of various isotopes with sufficiently high accuracy to permit direct, high-precision comparisons of mass differences with gamma-ray energies in nuclear processes. These improved mass measurements may also contribute to an ongoing effort to produce an atomic standard for mass, replacing the platinum-iridium cylinder currently representing a kilogram (SN: 4/24/93, p.264).

To make the measurements, David E. Pritchard and Frank DiFilippo of the Massachusetts Institute of Technology and their coworkers trapped different ions, one at a time, in a uniform magnetic field. The captive ion would circulate within the field at a characteristic frequency dependent on the ion's mass and electrical charge.

By comparing the frequencies of two different ions, the researchers could work out a mass ratio for the pair. From measurements involving 20 different pairs of alternately trapped ions, they extracted atomic masses for nine isotopes and the neutron (see table).

Atom                           Mass
Hydrogen-1                1.007 825 031 6
Neutron                   1.008 664 923 5
Hydrogen-2 (deuterium)    2.014 101 777 9
Carbon-13                13.003 354 838 1
Nitrogen-14              14.003 074 004 0
Nitrogen-15              15.000 108 897 7
Oxygen-16                15.994 914 619 5
Neon-20                  19.992 440 175 4
Silicon-28               27.976 926 532 4
Argon-40                 39.962 383 122 0

As reported in the Sept. 12 PHYSICAL REVIEW LETTERS, these masses are at least 20 times more precise than earlier values using other methods. The MIT results also agree closely with those achieved recently by Robert S. Van Dyck Jr. and his colleagues at the University of Washington in Seattle, who used a similar magnetic-trap technique. However, the uncertainties in the MIT measurements are smaller.

"They have put out some absolutely superb results," comments Ernest G. Kessler Jr. of the National Institute of Standards and Technology (NIST) in Gaithersburg, Md.

As one consequence, the atomic mass of silicon-28 is now known accurately enough to encourage further work on determining the density and atomic spacing of an ultrapure silicon crystal. Precisely characterized, such a crystal could serve as an alternative to the present kilogram standard.

The improved atomic masses have also added impetus to an effort by NIST researchers to increase the precision of measurements of gamma-ray wavelengths in nuclear processes. Kessler and his colleagues are preparing such an experiment at a nuclear research facility in Grenoble, France.

Meanwhile, Pritchard envisions the possibility of improving atomic mass determinations by another factor of 20. "If we can do that, we could 'weigh' chemical bonds and determine chemical binding energies," he says. "What we have to do is figure out how to make measurements on two different ions in the same trap."