A NEW RADIOACTIVE SERIES

As is known, there are three radioactive families in nature (the uranium–radium, thorium, and actinium families), for which the mass numbers are expressed, respectively, by the formulas \(4n+2\), \(4n\), \(4n+3\). A fourth family, with mass numbers expressed by the formula \(4n+1\), has been obtained artificially (the neptunium family). From time to time, individual artificially produced isotopes have been added to these four families. Studier and Hyde *) succeeded in obtaining and identifying an entire chain of five isotopes, beginning with the isotope of protactinium \(Pa^{230}\), and constituting a side branch of the uranium–radium family. Figure 1 shows the position of the elements of this “protactinium” series in the uranium–radium family.

When thorium targets (disks 8 cm in diameter and 5 mm thick) were bombarded in the 60-inch cyclotron of the University of California with deuterons of energy 19 MeV (100 microampere-hours) or with \(\alpha\)-particles of energy 38 MeV (39 microampere-hours), among other reactions there also occurred the reactions \((d, 4n)\) and \((\alpha, p5n)\), leading to the formation of protactinium.

However, bombardment also leads to a whole series of other spallation reactions. In addition, induced fission of \(Th^{232}\) also takes place. All this complicates the determination of the decay scheme and the identification of the elements obtained. To determine the isotopes of the protactinium series, the upper layer about 15 microns thick was removed from the part of the thorium target that had been bombarded, dissolved, and the protactinium and uranium fractions were separated chemically from one another and from the fission products. Further investigations were carried out on these two fractions.

1. Uranium.
   Investigation of the uranium fraction showed its decay with a period of about three weeks. Such a period can be assigned only to isotopes of uranium

) M. H. Studier and E. K. Hyde, Phys. Rev. 74*, 591 (1948).

with atomic number 231 and below. Since the uranium fraction with such properties is obtained, in particular, by bombardment of thorium with deuterons, mass number 231 should be excluded.

An accurate determination of the half-life of uranium was carried out using uranium extracted from the protactinium fraction. The presence of Pa \(^{230}\) and Pa \(^{231}\) did not affect the measurements, since the first of these listed isotopes gives the long-lived U \(^{230}\), while the second is itself long-lived. In order to exclude the influence of the isotope Pa \(^{232}\) (half-life 1.33 days; as a result of decay U \(^{232}\) is formed with a half-life of 70 years), the protactinium fraction was kept for 16 days, as a result of which the amount of Pa \(^{232}\) decreased by 4000 times. After this separation, the uranium formed was placed in a counter and its radioactivity was studied. From the curve of \(\alpha\)-decay a half-life was found equal to 0.8 days.

Fig. 1.

2. Thorium.
A study of the change in the activity of uranium with time shortly after its separation showed that over the course of four hours it practically reaches a maximum, increasing fivefold in doing so. This means that uranium has four rapidly decaying daughter isotopes emitting \(\alpha\)-particles. From the curve of increase of the activity the half-life for the first product of the decay of uranium, i.e. for Th \(^{226}\), was determined to be 30 minutes. Chemical separation of thorium confirmed the presence of a period of 31 minutes in the separated thorium fraction.

A more accurate determination of the period was made from the change in activity of recoil nuclei remaining in the counter after removal of the uranium sample. The period proved to be 30.9 minutes. It should be noted that in this case an additional activity with a period of \(\sim 0.5\) minute was also observed. As will be seen below, it should be ascribed to Ra \(^{222}\).

3. Radium.
In chemically separated radium, \(\alpha\)-radioactivity was not detected. It follows from this that the half-life of Ra \(^{222}\) is not more than one minute. The activity of recoil nuclei with a period of \(\sim 0.5\) minute noted above should be ascribed precisely to Ra \(^{222}\), since the subsequent decay products Rn \(^{218}\) and RaC′ have considerably shorter periods.

Separating Th \(^{226}\), introducing it into the counter and then removing it, it was possible to find, from the change in the activity of the recoil nuclei, the half-life of Ra \(^{222}\). It proved to be 38.0 seconds.

4. Radon.
The upper limit for the half-life of radon was established as follows. The gas coming from a boiling solution of U \(^{230}\) was rapidly fed directly into an \(\alpha\)-counter. Nevertheless, no activity was detected.

was found. This indicates that the half-life of Rn\(^{218}\) is no more than one second.

For its precise determination, a simple electronic circuit was constructed, serving to measure the time interval between two successive emissions of \(\alpha\)-particles (Fig. 2). The mean value of the interval between the decay of Ra\(^{222}\) and Rn\(^{218}\) gives the mean lifetime of Rn\(^{218}\), whence, by multiplication by 0.693, the half-life is found. The pulse from the \(\alpha\)-counter locks the tube \(T\) (see Fig. 2), the capacitor \(C\) is charged, and the spot on the oscilloscope is deflected in the direction up to the point at which the next pulse from the \(\alpha\)-counter unlocks the tube, after which the oscilloscope spot returns back. The time of deflection of the oscilloscope spot is the time between two \(\alpha\)-decays. Although the number of random coincidences can be

Fig. 2.

reduced by carrying out measurements at low intensities and using counters with a low background, it is nevertheless impossible to eliminate them completely. Therefore the influence of random coincidences was specially calculated and was excluded from the results obtained. To measure the half-life period of Rn\(^{218}\), a sample of U\(^{230}\) was placed in the counter. Since time intervals down to 0.2 sec were measured, the half-life of RaC′ could not be registered, as being too small, and the period of Ra\(^{222}\) as being too large. From the results of measurements by two methods, the half-life of Rn\(^{218}\) was found to be equal to 0.019 sec.

1. Protactinium.

The investigation of Pa\(^{230}\) is made difficult by the presence of the isotope Pa\(^{233}\), for which the period is 27.4 days, which is close to the period of Pa\(^{230}\). Therefore indirect methods of determining the half-life were used. A rough estimate (\(\sim 2\) weeks) was obtained from the change in the activity of uranium fractions separated from Pa at equal intervals of time. For a more accurate determination, a sample of Pa\(^{230}\) was deposited on a platinum disk and the course of the \(\alpha\)-activity of the sample was investigated. Knowing the half-life for U\(^{230}\) found from experiment, it is possible to calculate from the graph of the change in \(\alpha\)-activity the half-life of Pa\(^{230}\). It proved to be equal to 17 days.

1. Energy of \(\alpha\)-particles.

The determination of the energies of the emitted \(\alpha\)-particles was carried out with the aid of a 48-channel differential pulse analyzer, each channel of which registers pulses from \(\alpha\)-particles whose energy lies within certain narrow limits. The dependence of the number of pulses on the number

channel is shown in Fig. 3 (the energy increases with increasing channel number). The figure clearly shows five groups of alpha particles, corresponding to energies of 5.85 MeV, 6.30 MeV, 6.51 MeV, 7.12 MeV, and 7.7 MeV. The last group of alpha particles is identified with alpha particles from RaC'. The detection of alpha particles from RaC', which is the final isotope of the protactinium branch of the uranium-radium family, confirms the correctness of the identification of the preceding isotopes. The first group (5.85 MeV) corresponds to alpha particles from \(U^{230}\). This is shown by two methods: 1) analysis of the spectrum of alpha particles from a \(U^{230}\) sample in which daughter isotopes had not yet had time to form shows

Fig. 3.

the presence of a large peak for the energy 5.85 MeV; 2) analysis of the recoil-nucleus activity from \(U^{230}\) shows the presence of alpha particles of all the energies listed above, except 5.85 MeV.

Identification of the other three energy values is hindered by the short half-lives of the isotopes. If it is assumed that the alpha-particle energy increases as the half-life decreases, then \(Th^{226}\) should be assigned an energy of 6.30 MeV, \(Ra^{222}\) an energy of 6.51 MeV, and \(Rn^{218}\) an energy of 7.12 MeV.

E. Burshtein