Artificial Radioactivity Produced by Neutron Bombardment*

E. Fermi, E. Amaldi, O. d’Agostino, F. Rasetti, and E. Segrè

Rome

1. Introduction

In the present article we intend to give a complete account of all the experiments carried out in the physics laboratory of the University of Rome on new radioactive elements obtained by means of bombardment with neutrons. We have already had occasion to present the preliminary results in various short communications[^1]–[^8].

Curie and Joliot[^9] first discovered that atoms produced as a result of artificial disintegration need not necessarily correspond to stable isotopes, but may begin to decay with the emission of light particles, possessing at the same time a fairly considerable mean lifetime. Curie and Joliot used polonium for bombardment with α-particles. It was found in this case that the emitted light particles are chiefly positrons. Similar results were obtained also by other experimenters on various elements, the bombardment being carried out both with α-particles and with protons and deuterons to which a considerable acceleration had been artificially imparted[^10]–[^14].

The use of charged particles for bombardment compels one to restrict the excitation of radioactivity only to light elements, since by this method it is possible to activate only about ten elements whose atomic number does not exceed 15.

It therefore seemed to us very expedient to try to produce the same effect by means of bombardment with neutrons, which are able to approach closely the nuclei of even the heaviest atoms. It is true that the neutron sources at our disposal are considerably less intense than sources of α-particles, protons, or deuterons; however, there was sufficient reason to suppose that this factor would be compensated by the greater eff—

* Proc. Roy. Soc. A. 146, 483, 1934, translated by E. L. Feinberg.

effectiveness of decomposition by means of neutrons. Indeed, our experiments have shown that, of the sixty elements investigated, more than forty can be activated in this way.

2. Description of the experiment

In our experiments the source of neutrons was a sealed glass tube about 6 mm in diameter and 15 mm long, containing about 800 millicuries (mC) of beryllium powder and radon. According to the usually accepted figures for the neutron yield from beryllium, the number of neutrons emitted by this source should be of the order of 1000 per second per 1 mC. The energy of most of these neutrons varies within very wide limits—from zero to 7–8 million V. There is, however, also present a certain very small percentage of particles whose energy almost twice exceeds this upper limit.

The neutron radiation is accompanied by very intense γ-radiation, which, however, causes no difficulty, since, as it has been possible to show, radon without beryllium does not excite radioactivity at all. The emission of neutrons from beryllium is accompanied by the emission of γ-radiation harder than the γ-radiation of any of the decay products of radon (from 5 to 6 million V in the amount of one quantum per neutron). It is, however, very improbable that the observed effect should be in any way connected with this γ-radiation, if γ-radiation of much greater intensity and only slightly lower energy produces no effect.

The electrons emitted by the activated substance were registered with a Geiger—Müller counter, about 5 cm long and 1.4 cm in diameter. The walls of the counter were made of thin aluminum foil from 0.1 to 0.2 mm thick. The applied voltage was varied within the limits from 1000 to 1500 V. The amplified pulses were counted by an automatic device operating with the aid of a thyratron.

The substance under investigation was usually prepared in the form of a cylinder, with which, in order to reduce to a minimum the losses of intensity caused by geometrical relations, the counter could be surrounded. During irradiation the substance was placed as close as possible to the source. Those substances which, after irradiation, had to be subjected to chemical investigation were often taken in the form of a concentrated solution enclosed in a test tube.

The decay curves for the induced radioactivity in many elements revealed a simple exponential dependence. Sometimes, however, they could be resolved into two or more exponential curves. Therefore, in order to attain different degrees of activation in the different components, it was convenient to irradiate various substances for intervals of time of different length. The presence of several values of the mean lifetime in some cases is explained, of course, by the presence of several-

…of radioactive isotopes. In those cases where only one isotope is present, it may be ascribed to the existence of alternative decay processes, and sometimes (uranium) to an entire chain of such processes. The intensity of activation for different elements varies within wide limits. Sometimes the effect could be measured only with difficulty, because the number of pulses produced by the irradiated substance was of the same order as the number of spontaneous discharges in the counter. In other cases, however, the excitation was so strong that, when the object was placed too close, the number of pulses reached several thousand per minute, and consequently they could not be counted because of the insufficient resolving power of the counter.

We did not undertake exact measurements of the activation intensity of the various elements, because this would have required us to carry out experiments with a precisely defined arrangement of all the apparatus, and also to know both the efficiency of our counter with respect to counting electrons and the absorption in the substance under investigation and in the aluminum foil. However, we made a rough estimate of all these factors, as a result of which we were able to give a numerical expression for the activation intensity \((i)\). By this expression is meant the number of decays occurring in 1 sec. in 1 g of the given substance, placed at a distance of 1 cm from a neutron source containing 1 mC of radon (in radioactive equilibrium with its decay products) and beryllium powder.

Irradiation was always continued until saturation was reached. The efficiency of our counter (allowing for absorption in the aluminum foil and for geometrical factors) was approximately \(1/20\), as determined by measuring the number of pulses produced by known amounts of potassium and uranium.

Knowing this number expressing the intensity, it is easy to obtain the effective cross section for the exciting impact of a neutron, if the number of neutrons emitted in 1 sec. by a 1 mC source is known. Taking this number to be 1000, we immediately find that the effective cross section is

\[ \sigma = 2 \cdot 10^{-26} \cdot i \cdot A, \]

where \(A\) is the atomic weight of the element.

In order to be able to consider the nuclear reactions as a result of which the given active element appears, it is first necessary to determine what this element is. One may assume that the atomic number of the active element is close to the atomic number \(Z\) of the bombarded one. Since the amount of the active substance obtained is exceedingly small (in the most favorable cases about \(10^9\) atoms), there is no hope of isolating it by ordinary methods. Therefore one must proceed as follows: dissolving the irradiated substance, we add to it a small amount

of some inactive substance which, by assumption, could be an isotope of the radioactive decay product. After this, the irradiated substance is chemically separated from the added one, and each of them is tested separately for radioactivity. In most cases it turns out that the activity passes to one particular element. In this case one may consider that the active decay product is an isotope of this element.

We also carried out a preliminary investigation of the penetrating power of the $\beta$ rays of the new radioactive elements. In our experiments a counter of the standard type was used, and the object under investigation was again given the form of a cylinder, whose internal diameter somewhat exceeded the diameter of the counter, so that cylindrical aluminum screens of various thicknesses could be inserted between them. Since the arrangement of the instruments and of the objects in absorption measurements of this kind differs somewhat from the usual one, and, moreover, since we measured not the total ionization but the number of pulses, we had to check our method by measuring the absorption coefficient for already studied radioactive substances. As was to be expected, we found a discrepancy of the order of 20%. All the data have been corrected by this amount.

In some cases it was not possible to achieve complete absorption, even when a layer of lead 2 mm thick was inserted. This fact was regarded as evidence for the existence of $\gamma$ radiation.

It was very important to determine whether the emitted light particles are positive or negative electrons. Since the radiation is extremely weak, the convenient method for this purpose proved to be the method of Thibaud’s inhomogeneous magnetic field$^{15}$. However, even in this case we had to confine ourselves to investigating only those elements that can be strongly activated (Al, Si, P, S, Cr, As, Br, Rh, Ag, J, Ir, U). In all cases we found only negative electrons. This, however, does not exclude the possibility of the presence of a small quantity (up to 15%) of positive electrons.

In the case of a very small number of strongly activatable elements, the emitted electrons were successfully photographed in a Wilson chamber.

3. Results of the Experiments

Below are given the results of the investigation of various elements, arranged according to atomic number. The outcome of all the investigations is brought together in the table placed at the end of the article.

1. Hydrogen. Shows no effect even after 14-hour irradiation of (water) with a 670 mC source.

3. Lithium. Lithium hydroxide hydrate remains inactive after 14-hour irradiation with a 750 mC source.

4. Beryllium. Shows an extremely weak activity, which may successfully be ascribed to contamination of the preparation.

1. Boron. The same as beryllium.

2. Carbon. Paraffin irradiated for 15 hours with a source of 220 mC shows no activity.

3. Nitrogen. Guanidine carbonate (containing about 35% N), after 14-hour irradiation with a 500 mC source, shows no activity.

4. Oxygen. Shows no activity; see hydrogen.

5. Fluorine. By irradiating calcium fluoride, it is possible to obtain a strongly active product \((i = 0.7)\). Since calcium has been shown not to possess activity, this effect must be attributed entirely to fluorine. Its activity decreases rapidly, showing a half-life of 9 sec. In this case no chemical separation can be carried out. However, since it is known that under neutron bombardment fluorine disintegrates, emitting an \(\alpha\)-particle, it may be concluded that the active nucleus is, in all probability, \(N^{16}\). This unstable isotope, emitting one electron, passes into stable \(O^{16}\). The greater stability of this latter nucleus agrees well with the observed very high energy of the \(\beta\)-rays. Their intensity decreases to half the initial value after passing through \(0.24\ \mathrm{g}/\mathrm{cm}^2\) Al (all subsequent figures, like this one, refer to aluminum).

6. Sodium. This element was irradiated in the form of the carbonate. Sodium shows a fairly large activity. The half-life is approximately 40 sec.

7. Magnesium. Magnesium can be activated rather strongly. The decay curve shows the existence of two periods, equal to approximately 40 sec. and 15 hours. For the long period the half-attenuation layer is \(0.06\ \mathrm{g}/\mathrm{cm}^2\).

The active element having a half-life of 15 hours was successfully isolated chemically. For this purpose the irradiated magnesium was dissolved and sodium salt was added to the solution. Then the magnesium was precipitated in the form of phosphate and was found to show no radioactivity, while the sodium remaining in solution proved to be active. This establishes that the active atoms are not isotopes of magnesium; and since neon must in general be excluded from consideration, it remains to suppose that they are isotopes of sodium formed in accordance with the reaction

\[ \mathrm{Mg}_{12}^{24} + n_{0}^{1} = \mathrm{Na}_{11}^{24} + H_{1}^{2}. \]

1. Aluminum. As a result of neutron bombardment this element becomes very active. The decay curve indicates the existence of two periods, equal to approximately 12 min. \((i = 0.8)\) and 15 hours \((i = 0.5)\). The half-attenuation layer is respectively \(0.07\) and \(0.06\ \mathrm{g}/\mathrm{cm}^2\).

The active element with the long period can be chemically separated in the following way: dissolve the irradiated aluminum and add to the solution a small amount of sodium and magnesium. Then

we precipitate aluminum and magnesium in the form of hydroxide and phosphate. The precipitate proves inactive. After this, the solution containing sodium is evaporated, and we find considerable radioactivity with a half-life of 15 hours.

The active substance probably is the same element that we obtain also in the case of magnesium. This is indicated to us by the identity of the periods and of the half-attenuation layers. In the present case the nuclear reaction has the form:

\[ \mathrm{Al}_{13}^{27}+\mathrm{n}_{0}^{1}=\mathrm{Na}_{11}^{24}+\mathrm{He}_{2}^{4}. \]

After this the active isotope \(\mathrm{Na}^{24}\), emitting one electron, passes into the stable isotope \(\mathrm{Mg}^{24}\).

The active decay product with a 12-minute period was not isolated; however, we are inclined to regard it as \(\mathrm{Mg}^{27}\), since the other two possibilities, \(\mathrm{Al}^{28}\) and \(\mathrm{Al}^{26}\), are in all probability excluded: the first because \(\mathrm{Al}^{28}\), as we shall see later, is a radioactive isotope with a period of 3 min., and the latter because \(\mathrm{Al}^{26}\) should, probably, decay with emission of a positron.

1. Silicon. Silicon is also very active (\(i=0.7\)); its period is a quantity apparently less than 3 min. The half-attenuation layer is equal to \(0.16\ \mathrm{g/cm^2}\).

Chemical separation of the active element was carried out by heating the irradiated silicic acid together with hydrofluoric and sulfuric acids, after aluminum and magnesium had been added to it. In this process the silicon volatilized in the form of a fluorine compound, while the aluminum precipitated from the residue proved to be the carrier of the activity. Hence we conclude that the active substance is an isotope of aluminum and that the nuclear reaction has the form:

\[ \mathrm{Si}_{14}^{28}+\mathrm{n}_{0}^{1}=\mathrm{Al}_{13}^{28}+\mathrm{H}_{1}^{1}. \]

This also agrees with the hypothesis of Curie, Joliot, and Preiswerk\({}^{16}\), which asserts the identity of this active isotope with that which is formed upon collision of an \(\alpha\)-particle with \(\mathrm{Mg}^{25}\) and has in fact the very same period.

1. Phosphorus. This element exhibits very strong activity (\(i=0.6\)), decreasing with a period of about 3 hours, and also activity of 10 times smaller intensity, whose period is equal to 3 min., first observed by Curie, Joliot, and Preiswerk. The half-attenuation layer of \(\beta\)-rays for the element with the 3-hour period is equal to \(0.09\ \mathrm{g/cm^2}\).

The active element with the 3-hour period was successfully separated chemically. For this purpose phosphorus was irradiated in the form of a concentrated solution of phosphoric acid. The solution was then diluted with water, sulfuric acid and a small quantity of sodium silicate were added. All this was dried until the separa-

of insoluble silica and then dissolved in water and filtered. The activity was thereby transferred to the silica.

Consequently, the nuclear reaction must have the form:

\[ \mathrm{P}_{15}^{31}+\mathrm{n}_{0}^{1}=\mathrm{Si}_{14}^{31}+\mathrm{H}_{1}^{1}. \]

The element with a 3-minute period was not chemically isolated. The coincidence of the periods suggests that this may be the same \(\mathrm{Al}^{28}\), which is also obtained from silicon.

1. Sulfur. Sulfur exhibits a rather strong activity with a half-life of approximately 13 days (the figure was measured rather inaccurately). The half-attenuation layer is \(0.10\ \mathrm{g/cm^2}\).

The chemical separation of the active decay product was carried out as follows: the irradiated sulfuric acid was diluted with water and an exceedingly small amount of sodium phosphate was added to it; then the phosphorus was precipitated in the form of phosphomolybdic acid by adding ammonium molybdate. The activity was found in the precipitate.

We therefore suppose that the nuclear reaction is as follows:

\[ \mathrm{S}_{16}^{32}+\mathrm{n}_{0}^{1}=\mathrm{P}_{15}^{32}+\mathrm{H}_{1}^{1}. \]

This active \(\mathrm{P}^{32}\), emitting an electron, passes into the stable isotope \(\mathrm{S}^{32}\).

1. Chlorine. The half-life and the penetrating power of the \(\beta\)-rays for this element are expressed by approximately the same figures as for sulfur. The intensity \(i=0.1\). The active substance was separated in exactly the same way as in the case of sulfur. The irradiated ammonium chloride was dissolved in dilute nitric acid, phosphorus was added, which was then separated by the same method as in the preceding case. This element, as before, was the carrier of the activity.

The nuclear reaction giving the very same active phosphorus as in the case of sulfur has the following form:

\[ \mathrm{Cl}_{17}^{35}+\mathrm{n}_{0}^{1}=\mathrm{P}_{15}^{32}+\mathrm{He}_{2}^{4}. \]

1. Calcium. It was not possible to detect any activity.

2. Titanium. Only a very weak effect could be observed, the period being found to be several minutes. It is not excluded that it too was caused by foreign impurities.

3. Vanadium. This element exhibits medium activity. The half-life is approximately 4 min. and, within the error of observation, coincides with the period of the decomposition products of chromium and manganese, which are both isotopes of vanadium. This suggests the assumption that the active element is none other than \(\mathrm{V}^{52}\). The half-attenuation layer is the same as for chromium, \(0.16\ \mathrm{g/cm^3}\).

4. Chromium. Metallic chromium, as a result of neutron bombardment, acquires a rather strong activity. The period of ...

half-life, as in the preceding case, is equal to 4 min. The layer of half-attenuation of the β-rays is \(0.16\ \mathrm{g/cm^2}\).

To determine the chemical nature of the active element we proceeded as follows: to the irradiated ammonium dichromate a little sodium metavanadate was added, and then the vanadium was precipitated by adding ammonium chloride. The activity was found in the precipitate. Consequently, it belongs to a substance which is not an isotope of chromium. To check whether this substance is an isotope of titanium, to the irradiated chromium compound, in addition to the vanadium salt, a titanium salt was added, and then the titanium was precipitated by hydrolysis. The precipitate showed no activity. Consequently, we must regard the active substance, in all probability, as the same one as in the preceding case, the isotope \(V^{52}\), which appears as a result of the nuclear reaction:

\[ \mathrm{Cr}_{24}^{52} + \mathrm{n}_{0}^{1} = \mathrm{V}_{23}^{52} + \mathrm{H}_{1}^{1}. \]

1. Manganese. We subjected manganese dioxide to irradiation, as a result of which we were able to detect a rather considerable activity, decreasing in accordance with the presence of two half-life periods: about 4 min and about 2.5 hours. The layer of half-attenuation for the electrons of the element with the period of two and a half hours is equal to \(0.16\ \mathrm{g/cm^2}\).

It was not possible to separate the active element with the long period from manganese, although for this purpose chromium and vanadium were added, which were subsequently precipitated in the form of lead chromate and lead metavanadate. Consequently, we are dealing, in all probability, with an isotope of manganese, which is also obtained upon irradiation of iron and cobalt, as is indicated by the agreement of the half-life periods.

Wishing to determine the active element with the 4-minute period, we subjected manganese nitrate to irradiation; a vanadium compound was then added to it, and the vanadium was precipitated in the form of lead metavanadate. The activity passed into the precipitate. A similar experiment was carried out with the addition of chromium. In this case as well the precipitate proved active, although to a definitely lesser degree. Apparently, the principal activity belongs to the isotope \(V^{52}\), the nuclear reaction being of the form:

\[ \mathrm{Mn}_{25}^{55} + \mathrm{n}_{0}^{1} = \mathrm{V}_{23}^{52} + \mathrm{He}_{2}^{4}. \]

One and the same active vanadium can thus be obtained from chromium, vanadium, and manganese (Fig. 1).

1. Iron. This element shows a rather considerable activity (\(i = 0.05\)), the half-life period being 2.5 hours[^17]. The layer of half-attenuation is equal to \(0.16\ \mathrm{g/cm^2}\). The active decay product can be separated in the following way: the irradiated iron is dissolved in nitric acid, a small amount of a soluble manganese salt is added to the solution, and then the manganese is precipitated in the form of \(\mathrm{MnO_2}\) by adding chlorate-

of sodium hydroxide. Activity is detected in the manganese precipitate. The active element is probably formed according to the reaction

\[ \mathrm{Fe}_{26}^{56}+\mathrm{n}_{0}^{1}=\mathrm{Mn}_{25}^{56}+\mathrm{H}_{1}^{1}, \]

and this same element is also obtained from manganese.

27. Cobalt. Cobalt can also be activated, and the decay curve gives the same period as for iron. The active element can be chemically separated together with manganese by the same method as described for iron. This suggests that it may again be the isotope \(\mathrm{Mn}^{56}\). In that case the nuclear reaction should have the form:

\[ \mathrm{Co}_{27}^{59}+\mathrm{n}_{0}^{1}=\mathrm{Mn}_{25}^{56}+\mathrm{He}_{2}^{4}. \]

Thus the active isotope \(\mathrm{Mn}^{56}\) can apparently be obtained from iron, cobalt, and manganese (Fig. 1).

28. Nickel. No activity is detected even after 13 hours of irradiation with a 250 mC source.

29. Copper. It exhibits a moderate activity, corresponding to a period of approximately 6 min. We consider it quite possible that the active nucleus appearing in this case is identical with the nucleus encountered in the case of zinc, i.e., that it is an isotope of copper.

Fig. 1. The figure gives a possible scheme of transformations for elements whose atomic numbers lie in the interval from 23 to 27. Along the abscissa is plotted the number of protons in the nucleus; along the ordinate, the number of neutrons. Stable isotopes are denoted by circles: black circles (\(\bullet\)) if they correspond to more than 20% of the atoms of the elements, and white circles (\(\circ\)) in the remaining cases. Active isotopes are denoted by squares. The arrows indicate the direction of the transformations.

30. Zinc. It exhibits weak activity, and the decay curve indicates the existence of two periods, one of which is approximately 6 min, while the other, considerably longer, has not yet been measured.

The active element with the 6-minute period was chemically separated,

for which purpose the irradiated metal was dissolved, copper and nickel were added to the solution, and then the copper was precipitated in the form of the sulfide or collected electrolytically on a zinc plate. In both cases the copper carried the activity with it. The nickel precipitated with dimethylglyoxime proved inactive.

Thus the active element is probably an isotope of copper.

1. Gallium. This element shows activity of moderate strength, decreasing in accordance with a half-life of about 30 min.

2. Arsenic. Gives a strong effect. The activity decreases with a period of about 1 day. The half-attenuation layer is equal to 0.16 kg/cm². We tried to separate the active substance by adding gallium and germanium and precipitating the first in the form of the ferrocyanide, and the second in the form of the sulfide, having first removed the arsenic. In this, the precipitation of gallium was apparently complete, whereas we are not sure that the separation of germanium was perfect. In both cases the separated substance proved inactive, which leads to the supposition that the active element is identical with an isotope of arsenic. However, since the reaction for germanium gave somewhat uncertain results, we cannot regard this supposition as strictly proven.

3. Selenium. This element was only weakly activated. Its period is about 35 min.

4. Bromine. Ammonium bromide, as a result of neutron bombardment, proved strongly activated. The decay curve gave two periods, equal to about 30 min and 6 hours. To determine the active element we added arsenic and selenium and then precipitated them: the first in the form of the sulfide, the second—electrolytically. Both proved inactive. Thus the active element is apparently an isotope of bromine. To check this supposition arsenic and selenium were added and then bromine was precipitated in the form of silver bromide. The precipitate proved active. We believe that the two periods may correspond to two active isotopes, Br⁸⁰ and Br⁸², formed from the two ordinary isotopes Br⁷⁹ and Br⁸¹.

5. Rubidium. Rubidium nitrate showed very weak activity with a half-life of about 20 min.

6. Strontium. This element, irradiated for 13 hours with a 400 mC source, showed no activity.

7. Yttrium. This element showed no activity, despite irradiation lasting about 20 min with a 750 mC source.

8. Zirconium. Zirconium oxide showed extremely weak activity with a period of several minutes. It is possible that this too should be attributed to extraneous impurities.

9. Molybdenum. This element proved very slightly active. The decay curve indicates the presence of at least two periods, one of which is equal to about 15 min, and the other to more than one day.

1. Ruthenium. No activity was detected after 40 minutes of irradiation with a 750 mC source.

2. Rhodium. Metallic rhodium can be strongly activated; moreover, the decay curve is composed of two curves with periods equal to 50 sec and 5 min. The half-absorption layer for the electrons of the element with a period of 5 min is equal to 0.10 g/cm².

3. Palladium. Under the influence of neutron bombardment it acquires moderate activity. The decay curve, determined, to be sure, not very accurately, indicates that the period is of the order of 6 hours. The half-absorption layer is equal to 0.03 g/cm².

4. Silver. An element possessing high activity. The decay curve gives periods of approximately 20 sec and 2 min (the intensity of the latter is \(i = 0.5\), the half-absorption layer is 0.08 g/cm²).

5. Cadmium. Metallic cadmium could be activated only to a very small degree. The decay curve indicates that the period somewhat exceeds 1 hour.

6. Tin. This metal exhibits extremely weak activity, which, perhaps, should be ascribed to extraneous impurities.

7. Antimony. The same as tin.

8. Tellurium. It exhibits weak activity, decreasing in accordance with a period equal to approximately 1 hour.

9. Iodine. Both iodine and ammonium iodide were subjected to irradiation. In both cases strong activity with a period of 30 min was detected. The intensity is \(i = 0.4\). The half-absorption layer is 0.09 g/cm². After the addition to a solution of irradiated ammonium iodide of tellurium and antimony, the active element passes into the precipitate together with iodine separated by the addition of nitric acid. It follows iodine also in the case when iodine is precipitated in the form of silver iodide in the presence of tellurium and antimony. It is very probable that here we are dealing with the formation of an active isotope of iodine.

10. Cesium. Cesium (tartrate) exhibits very weak activity. The corresponding period has so far been measured very inaccurately.

11. Barium. As a result of irradiation of barium hydroxide hydrate, weak activity was detected. The decay curve indicates that the period is close to 3 min.

12. Lanthanum. It continues to remain inactive after 40 minutes of irradiation with a 400 mC source.

13. Cerium. This element, after 30 minutes of irradiation with a 400 mC source, showed no activity.

14. Praseodymium. Preparations both of this element and of other rare earths—Nd, Sm, and Pr—were kindly provided to us by Prof. Rolla of the University of Florence. These preparations are very pure. They were investigated in the form of oxides.

Praseodymium exhibits rather weak activity. The decay curve—

decay indicates the existence of a decay period of 5 min., and also, perhaps, the existence of another, longer one.

1. Neodymium. Exhibits rather weak activity. Period about 1 hour.

2. Samarium. Also exhibits weak activity. Period about 40 min.

3. Tantalum. This element, after 2.5 hours of irradiation with a 750 mC source, showed no activity.

4. Tungsten. Exhibits extremely weak activity, apparently with a long period.

5. Rhenium. This element, despite irradiation for 10 min. with a 750 mC source, showed no activity.

6. Osmium. Despite 15 hours of irradiation with a 450 mC source, it acquired no activity.

7. Iridium. This element acquires strong activity, decreasing in accordance with a period equal to approximately 20 hours. The half-attenuation layer is equal to 0.13 g/cm².

To establish the chemical nature of the active element we irradiated iridium tetrachloride, then added osmium and rhenium to it and separated the first of them by distillation in the form of the tetroxide, and the second in the form of the sulfide. Both proved inactive. The iridium, however, continued to retain the activity, which, therefore, should be ascribed to its isotope.

1. Platinum. Pure platinum (obtained from Heraeus) showed extremely weak activity, corresponding to a short period.

2. Gold. This metal exhibits rather considerable activity. The half-life period is about two days. The half-attenuation layer is equal to 0.33 g/cm².

We tried to separate the active substance by dissolving the irradiated gold in aqua regia, adding platinum and iridium to the solution and precipitating them in the form of chloroplatinate and chloroiridate. Both precipitates showed no activity, while the gold retained it throughout. This shows that the active substance is its isotope. Gold continues to remain active also in the case where one attempts to add mercury and then evaporate it.

1. Mercury. After very strong irradiation it exhibits only extremely weak activity, which may be explained by the presence of foreign impurities.

2. Thallium. The same as mercury.

3. Lead. This element did not acquire activity even after 10 hours of irradiation with a 500 mC source.

4. Bismuth. The same as lead.

5. Thorium. Up to the present time it has been studied insufficiently. Thorium nitrate, purified of its usual β-active decay products, became very active after irradiation. The decay curve indicates the presence of at least two periods, one of which does not exceed 1 min., and the other is equal to 15 min.

92. Uranium. We shall give here only the main results of investigations of this element, since its behavior has recently already been discussed elsewhere[^18].

In addition to periods of 10 sec., 40 sec., and 13 min., we subsequently succeeded in determining the presence of one more, approximately equal to 1.5 hr. The intensity for each of them is of the order of 0.5. We have already once had occasion to consider experiments which apparently prove that the active decay product with a period of 13 min. is not an isotope of any of the elements whose atomic numbers lie in the interval from 86 to 92 (emanation, eka-caesium, radium, actinium, thorium, protactinium, uranium). These experiments were subsequently repeated under various conditions, the chief aim being to obtain a negative answer in identifying the element corresponding to the 13-minute period with protactinium. This answer was extremely difficult to obtain because of the small amount of the protactinium isotope available to us—UX$_2$. The reaction on manganese already described by us gives a yield of the 13-minute decay product of approximately 15%. Its yield for UX$_2$ depends very strongly on the conditions of the experiment and may be varied from 2 to 10%, allowance also being made for the natural decay of this substance.

For separating the 13-minute decay product, the following reaction is considerably more effective. The irradiated uranyl nitrate is dissolved in dilute hydrochloric acid. Then rhenium nitrate is added, after which the rhenium is precipitated as sulfide by the addition of hyposulfite. This precipitate carries about 50% of the initial activity, and sometimes even more. The percentage of UX$_1$ and UX$_2$ found in the rhenium precipitate also varies depending on the conditions under which the reaction proceeds (in particular, on the acidity), but it can be made very low, perhaps even less than 1%. In this way we succeeded in separating the 13-minute decay product and measuring its period, using a uranium preparation which had previously undergone no purification whatever from UX impurities. The 90-minute product apparently possesses chemical properties analogous to those of the 13-minute product; this we may conclude from the fact that in all reactions both substances are always obtained in one and the same percentage ratio. Thus, in both cases the activity must be ascribed to elements whose atomic number is greater than 92, and which may perhaps be isotopes of one and the same element.

4. Theoretical Consideration of the Results Obtained

Let us now discuss, from the theoretical point of view, those processes which may occur under the influence of neutron bombardment. It must be noted that, in the present state of nuclear theory, all statements of this kind can have only an entirely preliminary character.

All the results of the experiments set forth in the preceding section can be summarized in the following few propositions.

1. In most cases elements can be activated independently of their atomic weight. From this point of view, in general no distinction can be made between light and heavy elements.

2. The cross section for collision with a neutron for the most intensely activated elements is of the order of the geometrical cross section of the nucleus. This means that collision of a neutron with a nucleus in most cases causes the appearance of an active atom.

3. The active substance is sometimes an isotope of the original atom (of atomic number \(Z\)), while sometimes its atomic number is lower by one or two units. In this respect there apparently is a difference between light and heavy elements. Thus, in the light elements the atomic number of the active substance is usually lower than \(Z\), whereas in the five heavy elements investigated by us that do not possess spontaneous radioactivity, the active product of the reaction is always an isotope of the bombarded element.

4. The emitted electrons always have a negative charge; at least we have never succeeded in detecting the presence of positrons.

The explanation of the general mechanism of excitation of the light elements apparently presents no specific difficulties. Evidently, this process usually consists in the capture by the nucleus of a neutron striking it, immediately followed by the ejection of an \(\alpha\)-particle or a proton. In the case where the energy of the ejected \(\alpha\)-particle or proton is of the order of several million volts, the time required for emission of the particle, according to Gamow’s theory, is very small. Consequently, there is a fairly considerable probability that this process will occur before the neutron leaves the nucleus. As a result of this process, which may last for a time interval of the order of \(10^{-20}\) sec, the original nucleus is transformed into another one having a greater atomic weight than follows from the value of its nuclear charge, since the process of neutron absorption and emission of an \(\alpha\)-particle or proton increases the ratio of the number of neutrons to the number of protons in the nucleus. This probably is the reason why we always observe emission of negative electrons. The process of electron emission restores the correct value of the ratio of the number of neutrons to the number of protons and corresponds to the formation of a stable isotope.

Since the atomic weight of the bombarded element increases, the potential barrier around the nucleus becomes an even more serious obstacle to the emission of heavy positively charged particles. From this point of view it becomes clear why the emission of protons and \(\alpha\)-particles becomes very improbable.

However, the theoretical interpretation encounters difficulties

in considering a reaction of another type, usually occurring in the case of heavy elements, namely when the active atom is an isotope of the initial one. It would be simplest of all to suppose that we are dealing with the capture of the colliding neutron, as a result of which an unstable isotope of the bombarded element appears, with an atomic weight exceeding the atomic weight of this initial element by one unit. This hypothesis, which agrees well with the observed fact of the emission of negative electrons, leads, however, to serious difficulties when we have to proceed to an explanation of the process of the capture of a neutron by a nucleus that is in a stable or quasi-stable state. It is usually considered that a neutron is attracted by a nucleus only in the case when its distance from the nucleus is of the order of \(10^{-12}\) cm. Consequently, a neutron with an energy of several million volts can be in the nucleus (i.e., experience a strong interaction with the constituent parts of the nucleus) only for a time interval of the order of \(10^{-21}\) sec, which is the classically calculated time for the passage of the neutron through the nucleus. A neutron can be captured by the nucleus only if during this interval of time it succeeds in losing all its excess energy (for example, by emitting a \(\gamma\)-quantum). If we calculate by ordinary methods the probability of this emission process, we obtain a value too small for it to explain the observed value of the cross section. Wishing to preserve the capture hypothesis, one may assume either that the probability of emission of a \(\gamma\)-quantum (or of the occurrence of the equivalent process of formation of an electron–positron pair) is considerably greater than is usually accepted, or else that, owing to reasons still unclear to us in the present state of the theory, the nucleus can, at least for \(10^{-16}\) sec, remain in an energetic state sufficiently high to permit the emission of a neutron.

In contrast to this hypothesis, it may be supposed that the colliding neutron is not captured by the nucleus, but merely causes the ejection of another neutron. This process may be described as follows: as a result of a kind of inelastic collision, the first neutron loses part of its energy, bringing the nucleus into an excited state. Theoretically it is easy to admit that precisely such a process occurs in a large number of cases of collision of a neutron with a nucleus. If the excitation energy is sufficiently high, the nucleus may, before giving it up by emission of a \(\gamma\)-quantum, emit a neutron. The atom formed as a result of such a process is an isotope of the initial one and has an atomic weight smaller by one unit. The objection that may be raised against this hypothesis is that, with the decrease (instead of increase) in the number of neutrons which then occurs, a priori the emission not of a negative electron, as actually takes place, but of a positron would be more probable. However, in several investigated cases of heavy elements that are activated under the influence of neutron bombardment and transform into their

TABLE 1

(continuation of Table 1)

isotopes, when the isotopes composing the neighboring elements are known, the always possible final product of the transformations is a stable isotope of the element \(Z - 1\), possessing an atomic weight one unit less than the initial one.

Moreover, it must be borne in mind that if, for an unstable nucleus, the emission of either an electron or a positron is energetically possible, then the theory of \(\beta\)-rays\({}^{19}\) indicates that, other conditions being equal, the emission of an electron is more probable.

In conclusion it should be noted that at present one cannot yet firmly adopt any one of the mutually exclusive hypotheses set forth above, and further experimental investigations are necessary in order to resolve the situation that has arisen.

5. Table of Results

In Table 1 (p. 948) the principal results of the investigations are summarized. In the first column are given the ordinal number and the designation of the elements investigated. In the second column are the isotopes of the given element, with those isotopes corresponding to more than 20% of the atoms of the element marked in boldface. The third column gives the half-life. A dash means that, as a result of the investigations, no radioactivity was detected. In the fourth column a rough estimate of the intensity is given: s. (strong), med. (medium), and w. (weak). In the fifth column the mean energy of the electrons is indicated, in millions of volts. These data were obtained as a result of a rather rough extrapolation based on the values of the absorption coefficient for ordinary \(\beta\)-active substances that do not give strong \(\gamma\)-radiation (\(\mathrm{RaE}\) and \(\mathrm{UX}_2\)). The sixth column indicates whether \(\gamma\)-rays are present. A dash indicates that, despite special attempts, it was not possible to detect the presence of \(\gamma\)-rays. In the seventh column the probable active decay product is indicated. For the sake of simplification we have everywhere assumed that neutron capture takes place. If, however, one assumes that a neutron was emitted, then the corresponding atomic weights should be lowered by 2 units. In the case of the presence of two or more periods, the data of the fourth, fifth, sixth, and seventh columns refer to the various periods in the order in which they are written.

References

1. Fermi, Ric. Scient. 1, 283, 330, 1934.
2. Amaldi, d’Agostino, Fermi, Rasetti, Segrè, Ric. Scient. 1, 452, 652, 21, 1934.
3. Fermi, Rasetti, d’Agostino, Ric. Scient. 1, 533, 1934.
4. Fermi, Nature 133, 757, 898, 1934.
5. Fermi, Nuovo Cim. 11, 1934.
6. Amaldi, Fermi, Rasetti, Segrè, Nuovo Cim. 11, 442, 1934.
7. Amaldi and Segrè, Nuovo Cim. 11, 452, 1934.
8. D’Agostino, Gz. Chim. Ital. (in press), 1934.
9. Curie et Joliot, C. r. Acad. Sci. Paris. 198, 254, 561, 1934.
10. Wertenstein, Nature 133, 564, 1934.
11. Frisch, Nature 133, 721, 1934.
12. Cockcroft, Gilbert and Walton, Nature 133, 328, 1934.
13. Neddermeyer and Anderson, Phys. Rev. 45, 498, 1934.
14. Lauritsen, Crane and Harper, Science 79, 234, 1934.
15. Thibaud, C. r. Acad. Sci. Paris. 197, 447, 1933.
16. Curie, Joliot et Preiswerk, C. r. Acad. Sci. Paris 198, 2089, 1934.
17. See also Fleischmann, Naturwiss. B. 22, 434, 1934.
18. Fermi, Nature 133, 898, 1934.
19. Fermi, Z. Physik. B. 88, 161, 1934.