Berichte der Deutschen Chemischen Gesellschaft
Vol. 46, p. 422-439 (1913).
The following selection is from pages 423-430 and 432
The starting point for these considerations is the connection that I have established between the type of a radioactive transformation and the electrochemical character of the radio element being considered. It should be emphasized here that it is possible to divide all radioactive transformations into two classes: a-ray transformations in which a helium atom with a double positive charge is expelled; the atomic weight of the resulting element is thus smaller by the atomic weight of helium (3.99 or about 4.0) than that of its direct mother substance; and b-ray transformations, in which only an electron is emitted: thus, by such a transformation the atomic weight will not be altered; there is merely a rearrangement of the constituent components of the atoms.1
This relationship now says that by an a-ray transformation the resulting product, electrochemically more positive, is lighter than its mother substance, while by a b-ray transformation exactly the opposite occurs, that is, the alteration product is electrochemically more negative than its mother substance. It can be shown that this relationship is valid for all transformations in which it can be tested, without exception. Since in the periodic system in a horizontal row, the electronegative character of the elements increases from left to right, we can express the opinion that in an a-ray transformation there results an element which belongs to a lower group of the same horizontal row of the periodic table, while in a b-ray transformation an element of a higher group (vertical row) results. This principle is also valid for all those cases in which the chemical nature of the elements considered is already known from a direct investigation (this is much more rarely possible than electrochemical characterization). The question still remains to be answered, by how many groups to the left (in a-ray transformation), or to the right (in b-ray transformation) this transition occurs. F. Soddy has already indicated as to this that in a-ray transformations in well-studied cases a transfer to the next second group is seen; thus, for example, from the fourth to the second, the sixth to the fourth, etc. I have accepted this Soddy rule as generally valid and will give a plausible meaning for it later. For the b-ray transformation I can show on the basis of several cases that we can assume a jump of only one group, and this principle is also accepted as generally valid. Since for several radioactive elements, the chemical character has already been known with assurance, with the aid of these two rules we can give the alteration of the groups for a- and b-ray transformations for all known radioelements so as to tell to which group of the periodic table they belong. The results obtained in this way are given in the following tables. The first table [Table I] contains the three known radioactive series (uranium-radium, thorium, and actinium series) in which the genetic relation of the individual products appears. The letters a and b indicate the type of transformation, the times under the symbols of the elements show their half-lives, while the upper numbers give the group in the periodic system to which the elements belong; the numbers in parentheses were derived in the above way. The table also shows the course of some transformations so arranged that here they have the positions most likely on the basis of the new rule.2
Table I
It should now be remarked that, beginning with ionium, radiothorium, and radioactinium, the transformations in the three series take place in a completely analogous manner, and that the corresponding members of the three series from the radioactive viewpoint also agree completely in chemical and radioactive respects. The groups to which the short-lived products of the radium series belong will be found by following the genetic series to the left, starting from radium D. Thus, it is known that RaD belongs in the fourth group. The results obtained in this way should be applied to the analogous products of the other series, and the result for ThB agrees completely with experience.
The two arrows from RaC1 and ThC1 express the fact, first observed by the author, that these products undergo two different transformations in which one part of the atoms is disintegrated in one way, the other in another. In this case we speak of a branching of the series. Such a branching will play a role in what follows for the interpretation of the periodic system.
Attention should be directed to one point: the fact that the three radioactive series are so extensively analogous shows clearly that the sequence in which the transformations of the elements in the groups (vertical rows) of the periodic system occurs in the cases already known to us is the same. We can already suspect from this fact that the periodic character of the transformations forms the basis of the periodic law. Thus, if we consider the uranium-radium-lead series, the periodic character of the transformations becomes clear: they pass through the groups
The series 6 4 5 6 4 is thus repeated three times.
In [Table II] all the radioactive elements are arranged according to decreasing atomic weight in the groups to which they belong. For calculation of the atomic weights we use as the basis the atomic weight of uranium (UrI) = 238.5 and that of thorium 232.4; the others are calculated on the assumption that in an a-ray transformation the atomic weight is decreased by 4 and that in a b-ray transformation no change in atomic weight occurs. The atomic weight of actinium and its transformation products is still unknown. The values given in the tables have only a hypothetical character. In this work they will be derived from the atomic weight of uranium on the basis of the assumption made probable by the new rule of a sort of relation of the uranium-radium series with the actinium series. It is also still uncertain whether actinium belongs in the second or third group of the periodic system.
Table II
0 | I | II | III | IV | V | VI |
Au 197.2 | Hg 200.6 | Tl 204.4 | ||||
ActD 206.5 | Pb 206.5 | |||||
ThD 208.4 | ThD2 208.4 | Bi 208.4 | ||||
RaC2 210.5 | RaD 210.5 | RaE 210.5 | ||||
ActB 210.5 | ActC 210.5 | RaF 210.5 | ||||
ThB 212.4 | ThC1 212.4 | ThC2 212.4 | ||||
RaB 214.5 | RaC 214.5 | RaC 214.5 | ||||
ActA 214.5 | ||||||
ThA 216.4 | ||||||
RaA 218.5 | ||||||
ActEm 218.5 | (ActX2) 218.5 | |||||
ThEm 220.4 | (ThX2) 220.4 | |||||
RaEm 222.5 | (RaX) 222.5 | |||||
ActX 222.5 | RadAct 226.5 | |||||
ThX 224.4 | Act 226.5 | RadTh 228.4 | ||||
Ra 226.5 | ||||||
MesThI 228.4 | MesThII 228.4 | Io 230.5 | ||||
Th 232.4 | ||||||
UrX 234.5 | (UrX2) 234.5 | UrII 234.5 | ||||
UrI 238.5 |
It happens that in this table places that are already occupied in the periodic system are here occupied by several elements. If we compare the chemical behavior of the elements which occupy these same positions it appears that this is much more similar than for that of any other elements. Such elements cannot be separated from each other either by chemical methods or by crystallization. The similarity here is thus much greater than among the rare earths. There is much more trouble in separating ionium from thorium, or radium D from lead, or mesothorium I from radium, and actually it would be meaningless to obtain a separation. If we consider all these elements as separate individuals it is only because of their radioactive properties such as the different life periods, the different rays, and the genetic relations. They are not separable by the usual chemical methods. This fact is of fundamental significance for the arrangement of the radioelements in the general periodic system. In order to arrange the last two horizontal rows properly by analogy with the others we must naturally proceed in the same way as for the other elements. Thus, only chemical methods can be used. However, this would give us only one element from a mixture of such elements in a group of the same horizontal row, and so it is that only one place in the Mendeleev system can, in fact, be assigned to this complex element. We must therefore answer only the question of which atomic weight should be ascribed to this complex element. Here also we must use experimental methods that we use for the ordinary elements, that is, the result of direct atomic weight determinations of this complex element isolated from a mineral chosen from the suitable system. The resulting value will depend on the mass ratio in which the individual components of this complex are combined. If the radioactive substances are in the stationary state, the individual products will be present in so much the greater amount, the longer-lived they are. If, however, one of the elements of such a group is much longer lived than another, we can simply take its atomic weight as the one to be used, or else choose a corresponding intermediate value. The first procedure is exact enough for all the radioactive elements. If we proceed in this way, we get the arrangement of Table III for the two last rows of the periodic system.
Table III
The previously empty spaces in the 0, I, III, and V groups of the last row and in the sixth group of the next row are occupied by short lived (still partly hypothetical) elements, which explains completely why they have not been discovered by ordinary chemical methods. The places in groups II, IV, and VI of the last row belong to known elements.
II. The End Products of the Transformation Series
The consequences that result from the places occupied by bismuth, lead, and thallium are of especially great importance. It is practically certain that lead is the end product of the radium series and indeed the direct transformation product of RaF (polonium). Besides lead, for which the theoretical atomic weight from uranium is calculated at 206.5, and which with other very short-lived elements is in the fourth group of the next row down according to the new rule, we must also assume the existence of ThD2, whose atomic weight should be 208.4. This results from the fact that ThC2, which according to the new rule belongs in the sixth group, undergoes an a-ray transformation and thus must yield an element of the fourth group. Since the existence of such an element is not recognizable by radioactive means, we must conclude that it is very long-lived. Then its chemical properties will be those of the lead obtained from uranium minerals since it belongs to the same group of the same horizontal row. It will thus appear to us as lead. However, there will be an apparent difference between the two leads: the atomic weight of one differs from that of the other by two full units. We must therefore obtain a different value for the atomic weight of lead from thorium-free uranium minerals than for lead from uranium-free thorium minerals. If ordinary lead is a mixture of these two types of lead, this would explain why the experimentally determined atomic weight of lead, 207.1, is greater than that calculated from the atomic weight of uranium on the basis of the genetic relationship, 206.5.
Similar considerations are applied to bismuth and thallium.
The determination of the atomic weights of lead, bismuth, and thallium from different sources would be of the very greatest significance for our understanding of the elements.