Aston 1922
ISOTOPES[1]
F. W. ASTON, M.A., D.Sc, A.I.C., F.R.S.
Fellow of Trinity College, Cambridge
LONDON
EDWARD. ARNOLD & CO.
1922
[All rights reserved]
Printed in Great Britain
PREFACE
I have undertaken the preparation of this book on isotopes in response to many requests made to me by teachers of physics and chemistry and others working in these subjects that I should publish the results obtained by means of the mass spectrograph in a form more convenient to the public than that in which they first appeared. This is one of the reasons why the space allotted to the inactive isotopes may appear, in the light of the general title of the book, somewhat disproportion- ately large. Another is that the subject of radioactive isotopes really requires a book to itself, and I am in the hope that the inadequacy of my account may stimulate the production of such a volume by hands more competent than mine to deal with this very special and remarkable field of modern science. The logical order of exposition of a scientific subject is to start with the simple and from that build up the more complex. Unfortunately the sequence of events in experimental research is the exact opposite of this so that a compromise must be effected, unless one is content to sacrifice historical treatment altogether. The latter seems very undesirable in a new subject. I have endeavoured in Chapters I, II and IV, and elsewhere when possible, to adhere strictly to the historical order of events even at the cost of some reiteration.
I wish to take this opportunity of expressing my indebted- ness to Mr. C. G. Darwin for his timely criticism and unfailing assistance throughout the work, and also to Mr. R. H. Fowler for help with the proofs. My thanks are also due to Professor Soddy for his diagram of the radioactive isotopes, to Mr. A. J. Dempster for kindly sending me the illustrations of his work, to the proprietors of the Philosophical Magazine and to the Council of the Chemical Society for permission to use the plates and figures of my original papers, and to Messrs. Macmillan & Co., for the diagram of the radioactive trans- formations.
- F. W. Aston
- Cambridge,
- January, 1922.
CONTENTS
CHAPTER I - INTRODUCTION
CHAPTER II - THE RADIOACTIVE ISOTOPES
CHAPTER III - POSITIVE RAYS
CHAPTER IV - NEON
CHAPTER V - THE MASS-SPECTROGRAPH
CHAPTER VI - ANALYSIS OF THE ELEMENTS
CHAPTER VII - ANALYSIS OF THE ELEMENTS (Continued)
CHAPTER VIII - THE ELECTRICAL THEORY OF MATTER
84. The Whole Number rule
By far the most important result of the measurements detailed in the foregoing chapters is that, with the exception of hydrogen, the weights of the atoms of all the elements measured, and therefore almost certainly of all elements, are whole numbers to the accuracy of experi- ment, in most cases about one part in a thousand. Of course, the error expressed in fractions of a unit increases with the weight measured, but with the lighter elements the divergence from the whole number rule is extremely small.
This enables the most sweeping simplifications to be made in our ideas of mass, and removes the only serious objection to a=
unitary theory of matter.
85. The Unitary Theory of the constitution of matter
From the very earhest times it has been a favourite hj'pothesis that all matter is really composed of one primordial substance, Air, Fire, Earth and Water have all been suggested in the past. The first definite theory of the constitution of the atoms of the elements out of atoms of a primordial element (Protyle, Urstoff, etc.) was made by Prout in 1815. ProuVs Hypothesis was that the atoms of the elements were different aggregations of atoms of hydrogen . On this view it is obvious that the atomic weights should all be expressed by whole numbers when the atomic weight of hydrogen itself is taken as unity. Owing to the roughness of the methods available and the considerable inaccuracies of the atomic weight determinations made at that time there was fittle to disprove the hypothesis, and its marked=
simplicity gained it many adherents. But as time went on chemical methods grew more precise and it became more and more impossible to reconcile experimental results with integral
90
THE ELECTRICAL THEORY OF MATTER 91
combining weights until the evidence against it was strong enough to cause J. S. Stas (1860-1865) to state: " I have arrived at the absolute conviction, the complete certainty, so far as it is possible for a human being to attain to certainty =
in
such matters, that the law of Prout is nothing but an illusion, =
a
mere speculation definitely contradicted by experience."
Nevertheless, though abandoned temporarily by the chemist as impracticable, the idea of primordial atoms appealed strongly to the mind of the philosopher and the physicist. Herbert Spencer, in his hypothesis of the constitution of matter, says :=
"All material substances are divisible into so-called elementary substances composed of molecular particles of the same nature as themselves ; but these molecular particles are complicated structures consisting of congregations of truly elementary atoms, identical in nature and differing only in position, arrangement, motion, etc, and the molecules or chemical atoms are produced from the true or physical atoms by processes of evolution under conditions which chemistry has not been able to reproduce."
The discovery of the electron, the proof that it was the same whatever the atom from which it was detached and, most important of all, the demonstration by Sir J. J. Thomson and others that electricity could simulate the known properties of matter, gave us the key to the riddle of what these primordial atoms really are. The only serious obstacle, the fractional atomic weights, has now been removed so that there is nothing to prevent us accepting the simple and fundamental conclusion : The atoms of the elements are aggregations of atoms of positive and negative electricity.
86. The Atom of Negative Electricity, or Electron
The fundamental unit of negative electricity makes its appear- ance in physical phenomena in many guises, such as the cathode ray of electrical discharge, the beta ray of radioactive change, the thermion of the wireless valve. A very complete account of it has recently been published by Millikan^ so that it is not proposed to describe its history and properties at any length
^ The Electron, by R. A. Millikan, University of Chicago Press, = 1917.
92 ISOTOPES
here. It will be sufficient to note a few of its more important constants.
Its charge e is given by Millikan as 4-774 X 10 " lo E.S.U. The most reliable measurements of e/m for the electron, at low velocity, give the value 5-30 x 10" E.S.U. Hence its mass is almost exactly 9*00 X 10~ ^s grg,^ i845 times less than the mass of the hydrogen atom, or 0-00054 on the ordinary scale of atomic=
weights (Oxygen =3D16).
87. The atom of Positive Electricity, or Proton
Our physical knowledge of this body is not nearly so complete as that of its counterpart the electron. It is very significant that in no analysis of positive rays so far performed have we been able to discover a particle of mass less than that of the hydrogen atom. This direct result, supported as it is by many less direct lines of evidence, leads logically to the conclusion that the hydrogen positive ray, i.e. the positively charged part remain- ing when an electron is detached from a neutral hydrogen atom, is the atom of positive electricity itself. The name "proton" was suggested for it by Sir Ernest Rutherford at the Cardiff meeting of the British Association in 1920. The charge on a proton is, of course, equal and of opposite sign to that on the electron. Its mass in the free state has been measured directly and is practically identical with that of the neutral atom of hydrogen 1-66 x 10" ^^ grs., or 1-007 on the oxygen scale.
88. The Nucleus Atom
Certain experimental results, notably the scattering of alpha rays, led Sir Ernest Rutherford in 1911* to formulate an atom model which has resulted in the most remarkable advances in both physics and chemistry, and is now almost universally accepted as correct in fundamental principle. This is that an atom of matter consists of a central massive nucleus carrying a positive charge which is surrounded, at distances relatively great compared with its diameter, by "planetary" electrons. The central nucleus contains aU the positive electricity in the atom, and therefore practically all its mass. The weight of the atom and its radioactive properties are associated with the nucleus; its chemical properties and
^ From Greek Trpwrov first the primary substance.
=C2=BB P. 67. ' V. p. 105.
- Rutherford, Phil. Mag. 41, 669, 1911.
THE ELECTRICAL THEORY OF MATTER 93
spectrum, on the other hand, are properties of its planetary electrons. It is clear that in a neutral atom the positive charg= e on the nucleus must be equal to the sum of the negative charges=
on the planetary electrons.
89. Moseley's Atomic Numbers
The scattering experiments mentioned above indicated that the net positive charge on the nucleus (expressed in terms of the natural unit e) was roughly equal to half the atomic weight. Now if we arrange the elements in order of atomic weight, starting with hydrogen, each element will have a position the number of which will be about half its atomic weight. It was suggested by Van den Broek that this atomic number might be equal to the charge on the nucleus. Two years after the formulation of the nucleus atom theory Moseley undertook an investigation of the changes which took place in the wave-length of the X-rays given off when various elements were used in turn as anticathodes. The result of this piece of work, now classical,^ was the estabhshment of the most important generaUsation in the history of chemistry since Mendeleef 's Periodic Law. Discussing a quantity Q related to wave-length Moseley writes:
"It is at once evident that Q increases by a constant amount as we pass from one element to the next, using the chemical order of the elements in the periodic system. Except in the case of Nickel and Cobalt ,2 this is also the order of the atomic weights. While, however, Q increases uniformly the atomic weights vary in an apparently arbitrary manner, so that an exception in their order does not come as a surprise. We have here a proof that there is in the atom a fundamental quantity, which increases by regular steps as we pass from one element to the next. This quantity can only be the charge on the central positive nucleus, of the existence of which we already have definite proof. Rutherford has shown, from the magnitude of the scattering of a particles by matter, that the nucleus carries a positive charge approximately equal to that of A/2 electrons when A is the atomic weight. Barkla, from the scattering of X-rays by matter, has shown that the number of electrons in an=
1 Moseley, Phil. Mag., 26, 1031, 1913.
=C2=BBCf. Barkla, Phil. Mag., 14, 408, 1907.
94 ISOTOPES
atom is roughly A/2, which for an electrically neutral atom is the same thing. Now the atomic weights increase on the average by about 2 units at a time, and strongly suggest the view that N increases from atom to atom always by a single electronic unit. We are therefore led by experiment to the view that N is the same as the number of the place occupied by=
the element in the periodic system. This atomic number is then for H 1, for He 2, for Li 3 for Ca 20 for Zn 30, etc. This theory was originated by Broek ^ and since used by Bohr. We can confidently predict that in the few cases in which the order of the atomic weights A clashes with the chemical order of the periodic system the chemical properties are governed by N, while A itself is probably some compHcated function of N."
Subsequent work has supported in an unquestionable manner the ideas so expressed by Moseley. That the number of the element in the order of the periodic table is actually the same as the positive charge on the nuclei of its atoms, expressed of course in terms of the natural unit of electric charge e, has been proved by direct experiment for some of the heavier elements. The recent work of Chadwick^ leaves little room for doubt on that point. At the other end of the scale all the known pro- perties of hydrogen point to the conclusion that its atomic number is 1 ; its exceptional atomic weight, as will be seen later, is itself strong corroborative evidence of this.
A complete table of the elements with their Atomic Numbers, Atomic Weights and isotopes (where these are known) is given on page 142.
From Moseley's law of atomic numbers the explanation of the empirical rule of radioactive transformation given on page 11 follows at once. An alpha particle carries two positive charges, a beta particle one negative one. If therefore the atom of a radioactive substance emits one alpha particle from its nucleus it naturally descends two units in atomic number, that is moves back two places in the periodic table. If on the other hand it emits one beta particle it clearly moves forward one place, for by the operation the nucleus has acquired one additional charge.
1 Van den Brock, Phys. Zeit. 14, 33, 1913. 'Chadwick, Phil. Mag.,l40, 734, 1920.
THE ELECTRICAL THEORY OF MATTER 95
90. The Bohr Atom
In this atom model the electrons outside the nucleus are supposed to be in a state of continual revolution about it, like planets round the sun. This rotation is considered to take place in orbits defined in a very special manner by means of a " quantum relation." This postulates that when, and only when, an electron changes its orbit, radiation is given out and the energy acquired by the change of orbit is entirely given off as radiation of frequency v where the change of energy equals hv where h is Planck's quantum or element of action (6-55 x 10" ^^ C.G.S.).^ This theory lends itself to exact mathematical analysis but unfortunately it can only be worked out adequately for the two simplest cases, the neutral hydrogen atom and the singly charged heUum atom. Here, however, its success is most remarkable; for not only is itpossible to calculate by its means the wave length of the chief series lines of the hydrogen spectrum, to an accuracy almost unprecedented in physics, but, by applying the relativity correction for change of mass with velocity to the rotating electron, the fine structure of the fines and the effects of electric and magnetic fields have been predicted with the most astonishing exactness by Sommerf eld, Epstein and others. Bohr has recently expressed the hope of extending his theory to heavier atoms by means of a new device which he terms the principle of "correspondence."^
91. The Lewis Langmuir Atom
This form of atom model was primarily designed to afford some theoretical basis for the numerous general quahtative properties of elements and their compounds. In it the electrons outside the nucleus are supposed to be at rest at, or vibrating about, definite points. The first two electrons wiU form a pair, the next eight will tend to set in positions corresponding to the eight corners of a cube, or some other soUd figure, and so on. In this way we shall get a series of shells or sheaths one outside the other. Langmuir has recently * reduced his postulates to the following three:
(1) The electrons in atoms tend to surround the nucleus in
iBohr, Phil Mag. 36, 1, 476, 857, 1913.
2 Sommerfeld, Atombau and SpektralUnien, Brunschweig, 1921.
3 Bohr, Nature, 107, 104, 1921.
- Langmuir, Brit. Assoc. Edinburgh meeting, 1921.
96 ISOTOPES
successive layers containing 2, 8, 8, 18, 18, 32 electrons re- spectively.
(2) The atoms may be coupled together by one or more " duplets " held in common by the complete sheaths of the atoms.
(3) The residual charge on the atom and on each group of atoms tends to a minimum.
This atom model is not amenable to mathematical treatment, but it has been exceedingly successful in accounting for the general chemical qualitative properties of many of the elements and in predicting those of their compounds.
92. Diagrammatical representation of atoms of Isotopes and Isobares
The accompanying diagrams (Fig. 16) are intended to indicate the sort of arrangements which may take place in atoms. The smaU dark circle is the nucleus, the number of protons and electrons comprising it being indi- cated by the numerals. The electrons outside the nucleus are indicated by small Light circles.
(1) is an atom of atomic weight 6. Its nucleus contains 6 protons and 3 electrons, hence its atomic number is 3. It is in=
fact the atom of the lighter isotope of lithium of atomic weight=
6, Li^. To be electrically neutral it must have 3 electrons out-=
side the nucleus. Now the principles underlying Langmuir's first postulate are derived from the Periodic Table and are certainly correct. Langmuir explains this by saying that the first two of=
these electrons will form an innermost ring or shell of two. This sheU being now complete, any more electrons will go out- side and start the next shell of eight, so we indicate this by=
putting the third electron in a circle of greater diameter.
Now suppose we add one electron and one proton to this atom. If both enter the nucleus we shaU get the configuration repre- sented by (2) The nuclear charge is unaltered, so that the arrangement of the exterior electrons will be precisely the same-=
It follows that all properties depending on these electrons such=
as atomic volume, spectrum, chemical properties, etc., will be quite unaltered. But the weight of this atom is now 7, so it is an isotope of lithium ; ' it is actually the atom of the he= avier constituent Li^ (1) and (2) are Isotopes.
THE ELECTRICAL THEORY OF MATTER 97
But now suppose we add a proton and an electron to (1) so that the proton only enters the nucleus and the electron remains=
outside as shown at (3) We shall now have a charge 4 on the nucleus and two electrons in the outer ring. The chemical
0) Atom of Li^
CZ) Atom of Li^
(3) Atom of hypothetical isotope of Beryllium
(4--) Atom of F
(5) Atom of Ne^=C2=B0
re) Atom of Ne^^
(7) Atom of Na O =3D Electron
Fig. 16. Diagrammatic Representation of Nucleus Atoms. The =
planetary
electrons are shown as lying on plane circles, the first containi= ng 2, the second 8 and so on. The dark circle is the nucleus and the + = and charges within it are indicated by figures. (1) and (2) are Isot= opes. (2) and (3) are Isobares and (5) and (6) are Isotopes.
properties of such an atom, if it could exist, would be completely
different from those of Uthium, but would be identical with those
of beryUium, of which it would be an isotope. But its mass is
clearly identically the same as that of (2) so that (2) and (3)
ARE Isobares.
H
98 ISOTOPES
In the same way (4) will be recognised as the atom of fluorine=
(5) and (6) as the atoms of the two isotopes of neon and (7) = as the atom of sodium.
93. The relation between Isotopes and Elements in the same Group
As far as can be seen the chemical properties by which the elements are divided into groups depend practically entirely on the outermost shell of electrons, which are therefore called valency electrons. Now consider all that part called by Langmuir the kernel of the atom lying within the shell of these valency electrons. The movements or configuration of the outermost electrons will depend in the first degree on the charge on the kernel, which may be looked upon as a virtual nucleus. The kernels of atoms (1) and (7) both have the same net charge 1, so that the elements they represent should have many chemi- cal similarities. These they certainly have as both are alkaU metals. In general the atoms of elements belonging to the same group chemically have the same number of electrons in the outer shell and the same net charge on the kernel. On this view it will be seen that the similarity of isotopes may be regarded as the extreme limiting case of the similarities long observed between elements of the same chemical group.
94. Abnormal compounds formed by charged atoms
The tendency of elements to form compounds with each other, that is the property with which the idea of valency is associated, is ascribed to the tendency of the atom to complete its outer shell. This it can do either by parting with the electrons in this sheU and so promoting the next inner completed shell to outer position, or by sharing the electrons in the atom of another element so that they fill the gaps in its own outer shell. We have already alluded to the success which has attended this idea in explaining valency and the properties of chemical compounds. For the present argument it will be enough if it is understood that lithium and sodium (1) (2) (7) will very readily part with their solitary valency electron and become positively charged, i.e., will be strongly electropositive elements with valency + 1 ; whereas a fluorine (4) wiU have an equally powerful tendency to take up an electron and become nega- tively charged and so will be a strongly electronegative
THE ELECTRICAL THEORY OF MATTER 99
element with valency 1, Both of these tendencies will b= e satisfied if (4) and (7) combine forming the compound molecule NaF, for the outer electron of (7) will enter the outer shell o= f (4) thus forming two complete shells of eight (Langmuir's octets). We trace the tendency of the atom of fluorine, or any=
other halogen, to form compounds with the atom of an electro- positive element, or with the atom of hydrogen, to the fact that it has one too few electrons in its outer shell.
Now the only way we can give a positive charge to an atom of neon (5) or (6) is by knocking one or more electrons out of=
its outer shell. Suppose we remove one from (5) as indicated by the dotted line. (5) now will have a similar outer shell to=
(4) and a valency 1, so we may expect that atoms of = the inert gases carrying a single positive charge will behave chemically in a similar manner to neutral halogen atoms and will therefore be capable so long as they are charged of forming=
hydrides. This very important idea was first suggested by Sir J. J. Thomson in connection with the charged atoms of chlorine ^ and certainly supplies a very satisfactory explana- tion of the very abnormal hydrides of inert gases and com- pounds such as OH3 discovered in positive rays. The line at 41 (Spectrum VI, Plate III) is probably to be put down to a charged hydride of argon of this type. Exceedingly faint lines at 5 in the case of hehum, and 21 in the case of=
neon, are probably to be ascribed to similar abnormal compound (HeH) and (NeH) respectively. In the case of atoms carrying more than one charge it can be generally stated that each positive charge given to an atom wiU increase its negative valency by one.
95. The failure of the additive law in respect to mass
We have seen that, for velocities smaU compared with that of light, the masses of the proton and the electron may be regarded as universal constants. If the additive law were strictly true as regards the summation of their masses it is clear that any mass whatever, whether it were that of an atom or a molecule, a planet or a star, or even the universe itself=
could be expressed in the form NM where N is a pure integer
1 J. J. Thomson, Proc. Roy. Soc. 99A, 90, 1921,
100 ISOTOPES
and M the mass of the neutral system 1 proton + 1 electron ( =3D the atom of hydrogen). The simplicity of this idea, which=
is Prout's theory in the language of modern physics, is extremely attractive ; but we know it to be false, for although the discovery of isotopes has removed the difficulty of the grosser fractions associated with such elements as neon and chlorine, we are still left with the more minute but none the less real one associated with hydrogen itself. To explain this the additive law must be qualified by some such reasoning as is contained in the following paragraph.
96. The explanation of the fractional mass of the hydrogen atom by the hypothesis of packing"
According to generally accepted views the proton and the electron possess mass, or what on the relativity theory is regarded as the same thing, weight, by virtue of the energy in the electromagnetical field which surrounds them. It can easily be shown on classical lines that if we give it a spherical form a charge e spread uniformly over the surface of the sphere will have a mass m when its radius a is such th= at
2 e^ m =3D ; hence to give the electron its proper=
mass its charge
must be compressed to a sphere of diameter about 3-8 x lO" ^^= cm. By the same argument the proton will be nearly two thousand times smaller and have a diameter 2-06 x 10" ^^ cm. The extreme range of the diameter of atoms themselves is 1 5 X 10~^ cm., so that it will at once be realised = that the structure of an atom is an exceedingly open one, even more so than that of our solar system.
Now it can be shown that if we bring two charges of opposite sign sufficiently close together their fields will aflfect each other in such a way that the mass of the system will be reduced. This effect is quite inappreciable for distances comparable with the diameter of an atom, but begins to make itself felt when the distance apart is of the order of the size of the ele= ctron itself as given above. The nucleus of the atom of an ordinary element (not hydrogen) contains both protons and electrons and is very small compared with the atom itself. Its dimen- BJons can be roughly determined by actual experiment in the
THE ELECTRICAL THEORY OF MATTER 101
case of the heavy elements and are found to be so small that even to get in the electrons alone these would have to be packed very closely together. Such a nucleus wiU contain more protons than electrons, roughly twice as many, so that it may be regarded as practically certain that : In the = nuclei of normal atoms the packing of the electrons and protons is so=
close that the additive law of mass will not hold and the mass =
of
the nucleus will be less than the sum of the masses of its con= - stituent charges.
The nucleus of a hydrogen atom consists of one single free proton, its planetary electron is too far away to cause any effect so that it is clear that we shaU find the mass associate= d with the atom of hydrogen greater than one -fourth the mass of a hehum atom or one-sixteenth the mass of an oxygen atom. The mass lost when four free protons and two free electrons are packed close together to form a helium nucleus (No. 9, p. 106) is roughly 0-7 per cent, of the whole and it = can be calculated that, if we take the value of the diameter of the electron given above, the protons must approach nearer than half of this to give so large a reduction. This means that=
the charges must be so closely packed that the electrons are actually deformed.
The whole number rule may now be simply translated into a statement that the mean packing effect in all atoms is approximately constant, and the unit of mass 1 when 0=3D16 wiU be (mass of a packed proton) -f | (mass of free electron)=
+ I (mass of packed electron). The whole number rule is not, and never was supposed to be, mathematically exact, for this would imply an identical packing effect in the case of all atoms, an exceedingly improbable supposition. It is almost certain that atoms of some elements, such as nitrogen, weigh slightly more than a whole number (looser packing) while those of others such as caesium or iodine may weigh slightly less (closer packing). The Umit of accuracy so far attained in mass-spectrum measurement is not sufficient to detect a change of the order expected, except in the case of hydrogen, where the variation in mass is exceptionally high.
97. The structure of the nucleus
The manner in which
102 ISOTOPES
the units of electricity are arranged in the nucleus of an atom has received a good deal of attention from theorists but ideas on this subject are almost entirely of a conjectural character. Thus Harkins ^ has proposed a constitutional formula for the nuclei of all the elements. In this, besides electrons and protons, he uses as building units a particles (4 protons + 2 electrons) of mass 4, and hypothetical units of mass 3 with a single positive charge (3 protons + 2 electrons). The matter has been more recently discussed and nucleus models sug- gested by Rutherford, 2 E. Gehrcke ^ and others.
The fact that the helium nucleus is almost exactly an integer on the oxygen scale that is to say helium has a= pproxi- mately normal packing gives a distinct balance of probabili= ty that helium nuclei actually exist as such in the nuclei of norma= l elements. In support of this idea it has been stated that the presence of heUum nuclei inside the nuclei of radioactive atoms is definitely proved by the ejection of a particles by the latter. In the writer's opinion this is much the same as saying that a pistol contains smoke, for it is quite possible that the a particle, Hke the smoke of the pistol, is only forme= d at the moment of its ejection. Brosslera * defends this view and points out that if the alteration from looser to closer packing of the charges forming the particle is at all large energy will be liberated amply sufficient for the purpose of detaching it and giving it the energy of an a ray. The reason to expect that this energy will be set free will be described in the next section. Brosslera's suggestion that in the nuclei of radioactive atoms there are loosely bound protons and electrons and that these, given something of the nature of a certain exact and instantaneous correlation, might combine to form an a particle is in good accordance with the most reasonable theory of radioactive disintegration, which was first put forward by Lindemann.^
There are therefore two different ideas which we may regard
1 Harkins, Phys. Rev., 15, 73, 1920.
2 Rutherford, Proc. Roy. Soc, 97A, 374, 1920.
3 Gehrcke, Phys. ZeiL, 22, 151, 1921.
- Brosslera, Rev. Chwi., 1, 42, 74, 1921.
=C2=BB Lindemann, Phil. Mag., 30, 560, 1915.
THE ELECTRICAL THEORY OF MATTER 103
as working hypotheses. According to the first the nuclei of atoms consist of helium nuclei, or a particles, held together in some way so that their packing effect wpon each other is small ; and, in the case of atoms not having a mass of the type 4w, additional protons and electrons. According to the second we only have to suppose that the mean packing of all the charges in the nucleus is such as will account for = the whole number rule with sufficient exactness, but that the actual arrangement of the protons and electrons need not necessarily be at aU similar to that in a helium nucleus.
The experimental evidence is, so far, definitely in favour of the first of these views. In their remarkable work on the disintegration of light atoms by the collision of swift a rays Rutherford and Chadwick i show that as the result of such collisions swift hydrogen rays, i.e. free protons, are liberated from the atoms of boron, nitrogen, fluorine, sodium, aluminium and phosphorus. They point out that the masses of the atoms of aU these elements are of the types 4w + 2 and 4w -|- 3. The effect is not obtained from atoms of the type 471 so that this result suggests that in these the protons are=
already all bound together to form helium nuclei.
98. Cosmical effects due to change of mass
It has long been known that the chemical atomic weight of hydrogen was greater than one quarter of that of helium, but so long as fractional weights were general there was no particular need to explain this fact, nor could any definite conclusions be drawn from it. The results obtained by means of the mass- spectrograph 2 remove all doubt on this point, and no matter whether the explanation is to be ascribed to pacldng or not, we may consider it absolutely certain that if hydrogen is transformed into helium a certain quantity of mass must be annihilated in the process. The cosmical importance of this conclusion is profound and the possibilities it opens for the future very remarkable, greater in fact than any suggested before by science in the whole history of the human race. We know from Einstein's Theory of Relativity that mass
1 Rutherford and Chadwick, Phil. Mag., 42, 809, 1921. ^ V. p. 70.
104 ISOTOPES
and energy are interchangeable i and that in C.G.S. units a mass m at rest may be expressed as a quantity of energy mc"^, where c is the velocity of light. Even in the case of th= e smallest mass this energy is enormous. The loss of mass when a single helium nucleus is formed from free protons and electrons amounts in energy to that acquired by a charge e falling through a potential of nearly thirty million volts. A swift a ray has an energy of three to four million volts so th= at the change of packing suggested by Brosslera need not be nearly so great to provide the energy needed. If instead of considering single atoms we deal with quantities of matter in ordinary experience the figures for the energy become prodigious.
Take the case of one gramme atom of hydrogen, that is to say the quantity of hydrogen in 9 c.c. of water. If this is entirely transformed into helium the energy liberated will be
0077 X 9 X 1020 =3D 6-93 X lO^^ ergs.
Expressed in terms of heat this is 1-66 x 10^^ calories or in=
terms of work 200,000 kilowatt hours. We have here at last a source of energy sufficient to account for the heat of the Sun.2 In this connection Eddington remarks that if only 10 per cent, of the total hydrogen on the Sun were trans- formed into helium enough energy would be liberated to main- tain its present radiation for a thousand million years.
Should the research worker of the future discover some means of releasing this energy in a form which could be employed, the human race will have at its command powers beyond the dreams of scientific fiction ; but the remote possibility must always be considered that the energy once liberated will be completely uncontrollable and by its intense violence detonate all neighbouring substances. In this event the whole of the hydrogen on the earth might be transformed at once and the success of the experiment published at large to the universe as a new star.
^Eddington, Time, Space and Gravitation, p. 146, Cambridge, 1920.=
Eddington, Brit. Assoc, address, 1920 ; Perrin, Scientia,=
Nov., 1921.
THE ELECTRICAL THEORY OF MATTER 105
===99. The stable systems of protons and electrons known to occur===. Starting with our standard bricks, the protons and electrons, we may make, theoretically at least, an infinity of systems by the combination of any number of each. It is interesting to consider the systems actually occur- ring in practice, that is to say those which are sufficiently sta= ble to give definite evidence of their existence. The follow- ing table gives, in order of mass, the first twenty-four known. Where the circles representing the charges touch each other, to form nuclei, the packing is extremely close, where they do not touch they are to be taken as distant thousands of times further from each other. The masses of the first twelve are deduced as follows : The most accurate value for the che= mical atomic weight of hydrogen is 1-0077 (0 =3D 16), and as it is v= ery improbable that it consists of isotopes we take this as the mass of a neutral hydrogen atom. The mass of the electron is 0*00054 and as the packing effect is nil we arrive at the figure 1-0072 for the mass of the proton, and this agrees within=
the experimental error with that directly determined by the mass-spectrograph. The most probable value of the mass of a neutral helium atom is 4-00(0) we wiU assume the last figure=
for the sake of simplicity. The masses of (13) to (24) are less=
accurately known.
The stability, where known, is expressed in volts and represents the potential through which a charge e must faU in order to acquire sufficient energy to disrupt the particular configuration concerned. This is the ionisation potential in the case of atoms.
106
ISOTOPES
^:s
i-e =C2=A3
Nuclear
institution.
6
r
a
i^
O
0
-1
0-00054
1 +
+1
1-0072
1 +
0
1-0077
1 +
-1
1-0082
1 +
+1
2-0149
1 +
0
2-0154
1 +
+1
3-0226
1 +
0
3-0231
4+2-
+2
3-999
4+2-
+ 1
3-999
4+2-
0
4-000
Description.
o
o
o o
o
o o
o o
o o o
10
11
o
O CO o
14
4-3
small
small
>3x 106
55
25
Electron
Proton or
positively
charged H
atom
Neutral H atom
Negatively- charged H atom
Positively charged H molecule
Neutral Hg molecule
Positively charged H3
Neutral H,
Doubly
charged
heUum
atom or
alpha ray
Singly charged helium atom
Neutral helium atom.
THE ELECTRICAL THEORY OF MATTER 107
^i
a 3
sl
>>
o
=C2=A71 ^1
II
S
Description.
12
m
lo.
2
4+2-
+ 1
5-007
Positively charged HeH
13
=C2=B0l
%
3
6 + 3-
+ 1
6-0
Positively charged Li8 atom
14
J
io o
3
6 + 3-
0
6-0
4-9*
Neutral Li atom
15
6\
i
^o
3
7+4-
+ 1
7-0
Positively charged Li' atom
16
M
^ 0 o
3
7+4-
0
7-0
4-9*
Neutral Li' atom
17
O O J
3
6+3-
0
6-0(07)
Neutral
Li^H
{
l^k 0
molecule
18
O O j
^^
3
7+4-
0
7-0(07)
Neutral
{
Li'H
- 04
IC#
molecule
19
=C2=B08
<^o o
4
9 + 5-
+1
9-0
Positively charged Be atom
20
4
9+5-
0
9-0
3-3*
Neutral Be atom
21
5
10 + 5-
+2
10-00
Doubly charged B^o atom
22
oo|
^ o )#0=C2=A9 o
5
10 + 5-
0
10-00
Positively charged B^ atom
23
o ot
1
5
10 + 5- 0
10-00
Neutral B^"
-
atom
24
=C2=B0^
3=C2=B0 =C2=B0
5
11+6-
+2
11-00
Doubly charged B" atom
- Calculated from frequency of radiation.
CHAPTER IX - ISOTOPES AND ATOMIC NUMBERS
100. The relation between chemical atomic weight and atomic number. Inasmuch as it is now recognised to be in general merely a statistical mean value the importance=
of the chemical atomic weight has been greatly reduced by the discovery of isotopes. Its position as the natural numerical=
constant associated with an element has been taken by the atomic number, though from the point of view of chemical analysis the chemical atomic weight is just as important as it ever was.
The possibility of anomalies in the order of the elements in the periodic table when their chemical atomic weights are considered, is now obvious enough. The true weights of the atoms as directly determined, are so intermingled in the order of the natural numbers and the proportions present in complex elements so varied that such anomalies are bound to occur, indeed it is rather surprising there are not more.
The following table (Fig, 17) shows the masses of the isotopes of three groups of elements now completely investigated. The approximate proportions present are indicated by the heights of the columns ; plain for the alliali metals, black for the in= ert gases, and hatched for the halogens. The anomalous order of argon and potassium is at once seen to be due to the fact=
that whereas the heavier constituent of argon is present in much the greater proportion, in potassium the reverse is the case. Had the proportions of heavier and Ughter isotopes been similar in each case the atomic weight of potassium would have been greater instead of less than that of argon.
108
ISOTOPES AND ATOMIC NUMBERS
109
- s
19
20
21
22 23
Fluorine (9) Neon (10)
Sodium (11)
35 36 37 38 39 40 41
(Chlorine 17) Argon (18) Potassium (19)
=C2=A5
1
I-D
78
I
127
79 80 81 82 83 84 85 86 =
87
Bromine (35) Krypton (36) Rubidium (37).
I n
W
136
128 129 130 131 132 133 134 135
Iodine (53) Xenon (54) Caesium (55)
Fig. 17. Isotopes of the Halogens, the inert gases and t= he alkali metals.
101. Statistical relations exhibited by elements and their isotopes. -Although our knowledge of true atomic weights is far from complete, for out of eighty-seven existing elements only twenty-seven have been analysed, of which thirteen are simple, interesting relations have already become clear which are stated in the form of rules as follows : =E2=80= =94
In the nucleus of an atom there is never less than one electron=
to
every two protons. There is no known exception to this law. It is the expression of the fact that if an element has an ato= mic number N the atomic weight of its lightest isotope cannot be less than 2N. Worded as above, the exception in the case of hydrogen is avoided. True atomic weights corresponding exactly to 2N are known in the majority of the Ughter elements=
]jp to A^^ Among the heavier elements the difference between
110 ISOTOPES
the weight of the lightest isotope and the value 2N tends to increase with the atomic weight ; in the cases of mercury it amounts to 37 units. The corresponding divergence of the mean atomic weights from the value 2N has of course been noticed from the beginning of the idea of atomic number.
The number of isotopes of an element and their range of atomic weight appear to have definite limits. Since the atomic number only depends on the net positive charge in the nucleus there is no arithmetical reason why an element should not have any number of possible isotopes. An examination of the tables of results given on p. 89 and at the end of the book show that so far the largest number determined with certainty is 6 in the case of krypton. It is possible that xenon has even more, but the majority of complex elements have only two each. The maximum differ- ence between the lightest and heaviest isotope of the same element so far determined is 8 units in the cases of krypton and xenon. The greatest proportional difference, calculated on the lighter weight, is recorded in the case of lithium, where=
it amounts to one-sixth. It is about one-tenth in the case of boron, neon, argon and krypton.
The number of electrons in the nucleus tends to be even. This rule expresses the fact that in the majority of cases even atomic number is associated with even atomic weight and odd with odd. If we consider the three groups of elements, the halogens, the inert gases and the alkah metals, this tendency is very strongly marked. Of the halogens od= d atomic numbers all 6 ( + 1 ?) atomic weights are odd. = Of the inert gases even atomic numbers 13 (+ 2 ?) = are even and 3 odd. Of the alkali metals odd atomic numbers 7=
are
odd and 1 even. In the few known cases of elements of the other groups the preponderance, though not so large, is still very marked and nitrogen is the only element yet discovered to consist entirely of atoms whose nuclei contain an odd number=
of electrons.
A further interesting result is the absence of isobares. So far none have been definitely identified, but it is quite obvious=
that in the cases of elements such as calcium and selenium they=
must exist, for the supply of integers in the region of their
ISOTOPES AND ATOMIC NUMBERS HI
atomic weights have been exhausted by the needs of other elements.
A table of the first 40 natural numbers and the true atomic weights corresponding to them is given in Fig. 18. The gaps
H
He
Li
Li
Be
B
B
c
N
O
F
Ne
1
2 3 4^ 5 6 7 8 9 10 11 =
12
13 14
15 16 17
18
19 20
Ne
Na
Mg
m
Mg
Al?
Si
Si Si?
P
s
CI
A
CI
K
A
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Fig. 18. The first 40 natural numbers, showing those occ= upied by atomic weights of known elements.
are particularly interesting and seem to show no semblance of regularity. It is very clear that many more experimental results will have to be obtained before any satisfactory theory for the occurrence of these, or of the other laws, is to be fo= rmu- lated.
102. The preponderance of elements of even atomic number. In discussing the nuclear structure of elements the question of their relative abundance in nature is one of great interest. This may be estimated by direct chemical analysis of the Earth's crust, and such extra-terrestrial sources=
as are available in the form of meteorites. The spectroscope will teU us what elements are present in the stars, but unfortu-=
nately it does not give much direct information as to their relative quantities.
On this question we can classify to use biological terms=
either by individuals or by species. We may examine the=
percentage composition, which wiU give a measure of the total number of individual atoms of each element present, or we may inquire into the number of different nuclear species which occur and classify them without respect to their individual abundance.
A very valuable discussion from the first point of view has been published by Harkins,^ who considers the percentage composition of meteorites and of parts of the Earth's crust. He demonstrates in a most convincing manner that there are
^Harkins, Jour. Amer. Chem. Soc, 39, 856, 1917.
112 ISOTOPES
immensely more atoms of elements of even atomic number. This interesting preponderance can, with a reasonable amount of probability, now be extended to even atomic weight, by the statistics given in the preceding paragraphs, but it will not be certain until the constitution of certain abundant elements such as iron has been actually determined.
The second point of view can be examined by means of the atomic weights of the radioactive isotopes and also by the true=
atomic weights given by the mass-spectra. In both cases nuclear systems of even atomic number are found to predomi- nate. The mass-spectra of 13 elements of even, and 14 ele- ments of odd atomic number indicate 32 isotopes of even atomic number and 20 of odd. The average element of even atomic number has therefore 2-5 isotopes to 1-4 for each element=
of odd atomic number.
The table on p. 15 shows that among the radioactive isotopes the preponderance is greater 32 as against 10 but=
it is
possible that the former figure may include some atomic systems absolutely identical though of different origin.
103. The constancy of chemical atomic weights.
One of the first difficulties in the way of accepting the idea = of the complex constitution of an element such as chlorine was the constancy of its atomic weight. This had been determined by many different observers using different methods and the results were always the same within a very small experimental error. This difficulty may be met, in the first place, by noting=
that the vast majority, if not all, of the really accurate value= s were obtained from chlorine which must have been originally derived from the sea. The sea has been mixed for so long that it would be absurd to expect to find chlorines of different=
chemical atomic weights in it. Had ordinary galena been the only source of lead used in the atomic weight determina- tions given on page 16 no difference would have been found. It was only by examining the lead from extraordinary radio- active sources that the results were obtained which gave such definite and valuable support to the theory of isotopes.
The atomic weight of chlorine from sources other than the sea is now receiving the attention of chemists, though it is
ISOTOPES AND ATOMIC NUMBERS 113
naturally very difficult to be at all sure that any known source=
of chlorine is not of marine origin. Mile. Irene Curie ^ has examined the atomic weight of chlorine from three minerals whose marine origin seems unlikely. The values obtained from a sample of sodalite (sodium aluminium chlorosilicate) from Canada, and from a sample of calcium chlorophosphate from Norway agree with the value for chlorine from sea-water. The value 35-60, for chlorine from a sample of sodium chloride from a desert region in Central Africa was slightly high.
The comparison of the atomic weights of terrestrial and meteoric nickel made by Baxter and Parsons ^ is interesting in this connection. As a mean of nine determinations with the terrestrial material the figure 58-70 was found, whilst three experiments with meteoric nickel gave 58-68. The standard value found by Richards and Cushman was 58-68 (Ag =3D 107-88). The difference found between terrestrial and meteoric nickel is considered to be within the limits of experi-=
mental error, but further comparisons are to be made.
The writer regards these negative results as having a cause probably much more fundamental than the mere mechanical mixing of the different constituent isotopes during the history of the body containing them, namely a constancy of proportion during the evolution of the elements themselves. This will be considered later. The case of the radioactive leads is entirely exceptional. These substances have been produced continuously during the history of the earth's crust and are being so produced to-day. Although ordinary lead may con- sist of isotopes which is practically certain and = these isotopes may be identical in every respect with those produced in the last stage of radioactive disintegration, yet there is no reason=
whatever to assume that ordinary lead is itself the accumulated result of these processes. It takes its place among the other ordinary elements and would doubtless have done so had thorium and uranium never existed.
104. The agreement between the chemical atomic weight and the mean atomic weight deduced from the mass spectrum. The mean atomic weight of the isotopes
II. Curie, Compt. Retid. 172, 1025, 1921.
- Baxter and Parsons, Jour. Amer. Chem. Soc, 43, 607, 1921.
I
114
ISOTOPES
of a complex element can be calculated if the relative intensitie=
s
of their lines in the mass-spectrum is known. This has been
directly measured by Dempster. ^ The charged particles of
isotopes of the same element are practically certain to afEect
the photographic plate to the same extent as each other,
hence we can obtain a rough estimate of their relative pro-
portion by comparing the intensities of the lines. If this is
done it is found that the great majority of the elements so
far tested give mean results in good agreement with the
accepted chemical values. The following table gives the data
concerning four in which the difference is noteworthy : =
Element.
Atomic Weight.
Mean from Mass-spectrum.
Difference.
Per cent. Difference.
Boron .... Krypton . Xenon Caesium
10-90 82-92 130-2 132-81
10-75^0-07 83-5 =C2=B10-3 131-3 =C2=B10-3 133 =C2=B10-3
0-15 0-6 1-1 0-2
1-37
0-72 0-85 0-05
The case of boron is the most difficult to account for. The masses of its isotopes 10 and 11 certainly do not differ from integers by more than one or two parts in a thousand. The ratio of the intensities of their second order lines 5 and 5-5=
(and there were no other substances present which could possibly give such lines) is equally certainly not as high as 9:1. It was for this reason that a third isotope 12 was suspected, but as no evidence of this has been found it seems most probable that the chemical atomic weight is still slightly too high.
The atomic weights of krypton and xenon are not of course chemical in the ordinary sense, as they are deduced direct from density determinations. Any trace of the impurity most likely to be present, argon in the first case, krypton in the second, would tend to make the densities too low, and this appears the most hkely explanation.
In the case of caesium the chemical result may be correct, for the probable error in the determination of mass is at least=
as large as the discrepancy. On the other hand caesium
1 V. p. 81.
ISOTOPES AND ATOMIC NUMBERS 115
appears to be a simple element, in which case its chemical atomic weight must represent the true weight of its atoms. Any error in this figure would probably be of the sign suggested= , for it is the heaviest member of its chemical group. If, how- ever, as is possible, the true mass of its atom differs from an=
integer by as much as 0-2 it is a fact of the greatest interes= t.
105. The meaning of the word " element." The
exact idea conveyed by the word " element " in chemistry and physics has given rise to endless difficulties in the past.=
In this connection Crookes in 1886 sums up the matter as follows : " Of the attempts hitherto made to define or = explain an element, none satisfy the demands of the human intellect. The textbooks tell us that an element is ' a body which has not been decomposed ' ; that it is ' a something to which we=
can add, but from which we can take away nothing,' or ' a body which increases in weight with every chemical change.' Such definitions are doubly unsatisfactory : they are provi- sional, and may cease to-morrow to be applicable to any given case. They take their stand, not on any attribute of things to be defined, but on the limitations of human power ; they are confessions of intellectual impotence."
There was good reason for this dissatisfaction. The dis- covery ten years later of the electron, and the subsequent electrical theory of matter robbed the word of any pretence to its original meaning ; for although Ramsay attempted to introduce into chemistry electricity itself as an element, it soon became obvious that this extension was unsuitable. The discovery of isotopes brings us face to face with two possible alternatives. The first is to call each isotope, as it is dis-=
covered, a new element. The second is to fix the word pre- cisely, now and for the future, as meaning a substance with definite chemical and spectroscopic properties which may or may not be a mixture of isotopes in other words to asso= ciate it exclusively with the conception of atomic number. On this view there would be, corresponding to Moseley's numbers, 92 possible elements, of which 87 are known.
If we adopt the first of these alternatives a new word will be necessary to express such substances as chlorine or mag-
116 ISOTOPES
nesium, hitherto called elements, and also the word element would mean something entirely different from what it has meant in all the chemical and physical Uterature of the past century. It would moreover be still subject to alterations in the future.
In the opinion of the writer the second alternative the=
association of element with atomic number is much the more preferable. The difficulties arising from it are practi- cally confined to the radioactive substances which can differ from one another even when their atomic numbers and atomic weights are identical. This is not very serious, for the radio-=
active elements are in a class by themselves and the special nomenclature already applied to them could be retained or re- vised as convenient without affecting that of general chemistry.
106. Disintegration theory of the evolution of the elements. A theory has been put forward by some writers=
that all the elements occurring in nature are the result of radioactive disintegrations of the ordinary type, but continued far beyond the ordinary limit observed at present. For instance, if we continue the a ray changes of the thorium series=
far enough we shall ultimately reach helium. The emission of an a particle is the only change known to occur which alters=
the atomic weight and it always does so by 4 units at a time.=
Hence from thorium we shall get a series of elements or iso- topes of atomic weights from 232 to 4 of the general type 4w. Uranium in the same way will yield a similar series of the type=
4:71 + 2. In order to obtain isotopes of odd atomic weight it is necessary to postulate parent elements of the type 4n + 1=
and 4:% -f 3.
Using hypotheses based on this general idea Van den Broek, ^ Harkins,^ Kohlweiler,^ Kirchoff ^ and others have built up the most elaborate systems of isotopes,
1 Van den Broek, Phys. Zeit., 17, 260, 579, 1 916 ; 23, = 164, 1921.
"Harkins and Wilson, Jour. Am. Chem. Soc, 37, 1367, 1915 ; Har-=
kins and Hall, ibid., 38, 169, 1916 ; Harkins, Phys. Rev., 15, = 73, 1920; Nature, 105, 230, 1920; Jour. Amer. Chem. Soc, 42, 1956, 1920 ; PhU. Mag., 42, 305, 1921.
- Kohlweiler, Zeit. fur physikal. Chem. , 94, 51 3, 1 920 ; Phys.=
Zeit. , 21,
311, 543; 22, 243, 1921. * Kirchoff, idid., 21, 711, 1920.=
ISOTOPES AND ATOMIC NUMBERS 117
The writer regards this view as unhkely and misleading. In the first place it does not appear to succeed in its objects= . As an explanation of how the elements may have been evolved it starts with at least fom: elements as complicated as any known to exist, which does not advance the inquiry very much. On the other hand it may be used to predict the atomic weights of the isotopes composing known elements, and a great many predictions of this kind have been made. Here, though the measure of its success has varied to some extent with the particular modification of the theory employed, it has never been worthy of serious consideration. In cases where two or three isotopes of a given element were pre- dicted they proved as often wrong as right, and when the number of isotopes of integral atomic weights was so large that some agreements were inevitable the argument obviously loses all its force.
Another objection is that radioactive transformations do not continue, as far as we can see, beyond the stage (lead) indicated in the diagrams on p. 15. The lighter elements are definitely not radioactive. The radioactivity of potassium and rubidium is exceedingly small and its nature doubtful ; in any case it is best ascribed to minute vestiges of radioactiv= e isotopes, not to feeble radioactivity of the main constituents. It seems therefore more reasonable, for the present, to regard the property of radioactivity as absent entirely from the inactive elements than to suppose it present but too weak to be detected. It must not be gathered from these remarks that it is considered impossible to imagine physical conditions violent enough to disrupt the nuclei of light atoms, but rather=
that the mechanism causing such disruption need not be similar in any way to that causing normal radioactivity.
107. Grookes' theory of the evolution of the elements.
A more attractive theory than the one given abo=
ve is that the complex atoms of matter have been evolved by the aggre- gation of simpler atoms. This idea has received a good deal of=
attention in the past. Crookes^ remarks on it as follows : =E2= =80=94 " Let us picture the very beginnings of time, before geological=
^ Crookes, Brit. Assoc, address, 1886.
118 ISOTOPES
ases, before the earth was thrown off from the central nucleus of molten fluid, before even the sun himself had consolidated from the original pivtyle. Let us still imagine that at this primal stage aU was in an ultra-gaseous state, at a temperature=
inconceivably hotter than anything now existing in the visible universe ; so high indeed that the chemical atoms could not yet have been formed, being still far above their dissociation point. In so far as protyle is capable of radiating or reflectmg=
light, this vast sea of incandescent mist, to an astronomer in a=
distant star, might have appeared as a nebula, showing in the spectroscope a few isolated hues, forecasts of hydrogen, carbon and nitrogen spectra.
" But in due course of time some process akin to cooling, probably internal, reduces the temperature of the cosmic protyle to a pomt at which the first step in granulation takes=
place ; matter as we know it comes into existence, and atoms are formed."
This vivid picture may be brought up to date by the sub- stitution of free protons and electrons for the hypothetical protyle. We can imagme regions containing matter where the temperature is so high that not only is the dissociation of=
atoms from atoms and nuclei from planetary electrons com- plete but also protons and electrons are in a state of agitation=
so violent that even the most stable nuclei cannot be formed. We should have here matter of the simplest form we can imagine, or rather of no form at all, simply a more or less neutral electric gas. Such a condition is by no means impos- sible in our miiverse and may actually occur during one of those=
excessively violent catastrophes occurring in far distant space and observed by us as new stars.
By some such cooling process as that suggested by Crookes we easily imagine the free charges combining to form the nuclei of elements. Whether those of heavier elements are formed direct by the charges getting into particular geometrical relations with each other, or whether hehum nuclei are formed first and then subsequently coalesce depends on which theory of nuclear structure is adopted. In any case vast quantities of=
energy will have to be radiated off and this radiation may be of such extremely high frequency that it is capable of dis-
ISOTOPES AND ATOMIC NUMBERS 119
rupting nuclei themselves, so that there might be at this stage=
rapid and continuous transformations from heavier to lighter nuclei and vice versa.
For the present we are interested in the number of each type of atom which survives. It is obvious that if the con- ditions of cooling are practically identical throughout the whole mass there is no reason why the composition of the matter produced should vary. If 3 atoms of CP^ are formed to every 1 of CP^ at any one point the same ratio must hold at every point so that a complex element of constant atomic weight will be formed. But it is much more likely that different parts=
of this primordial mass will undergo their transformations under different rates of cooling, etc., so it is worth while inquiring if variation in the mean atomic weight of a complex element is to be expected.
The quantity of one particular atomic nucleus formed will probably depend (a) on the probability of a certain configura- tion of charges happening as a chance event ; (b) the stability o= f the particular nucleus formed as the result of that event. Again to take the case of chlorine each isotope may be regarded=
as completely stable and the relative quantities formed will simply depend on condition (a). Now it is not unreasonable to suppose that this is not seriously affected by different rates=
of cooUng, and in this case the isotopes will be evolved in constant proportion. As we know of no natural process by which the proportion of isotopes can be altered appreciably the complex elements will have to-day the same chemical atomic weight as when they were first formed.
The above argument is of course purely a speculative one, and the conclusion drawn from it would fall to the ground at once if noteworthy differences of atomic weight in a single complex element were found supposing that element was not=
the product of a radioactive change at different points o= n the earth's surface. It may be worth noting that condition (a) suggests that, in general, the lighter atoms will outnumber the heavier ones. In aU matter available in nature this preponderance is actually enormous.
If the matter forming the earth ever went through a prim- ordial stage such as that suggested above it certainly did so
120 ISOTOPES
more than 10^ years ago. It follows that of the radioactive elements then formed only two, thorium and uranium, wiU now be found on the earth, for the other radioactive elements existing to-day are of such short period that they must have been formed since. Hence we may divide the original elements very simply and definitely into two groups : (1) All the inactive elements, whose nuclei are sufficiently simple to be stable ; (2) Thorium and Uranium, whose nuclei are so complex that they are only partially stable.
Other less stable elements vfiay have been formed then but there can be no proof of this for they would, in any case, hav= e disappeared long ago, and it is clear that the other radioactive=
elements now found can all be regarded as formed from the two parent elements in comparatively recent times.
CHAPTER X - THE SPECTRA OF ISOTOPES
108. The Spectra of isotopes. As has already been stated^ the first experimental work on the spectra of isotopes was that of Russell and Rossi in 1912 who failed to distinguish=
any difference between the spectrum of thorium and that of a mixture of thorium and ionium containing a considerable percentage of the latter. The same negative result was obtained by Exner and Haschek.^ During the fractional diffusion of neon^ no spectroscopic difference was detected between the heaviest and the lightest fraction, though as the separation was small this negative evidence was not very strong. In 1914 Soddy and Hyman showed that the spectrum of lead derived from thorium was identical with that of ordinary=
lead.* Furthermore in the same year the experiments of Richards and Lembert,^ Honigschmidt and HoroAvitz,*^ and Merton proved the same result. Merton concluded from his 1914 experiments that the difference in wave-length for the A 4058 line must be less than 0-003 A. Before going on to consider the more recent results it will be as well to discuss = the magnitude of the difference to be expected from theory.
109. The magnitude of the Gravitational effect. In
the Bohr theory of spectra the planetary electrons of the atom rotate round the central positively charged nucleus in various
1 F. p. 9.
2 Exiier and Haschek, Sitz. Akad. Wiss. Wien, iia, 121, 175, = 1912.
3 V. p. 39.
- Soddy and Hyman, Jour. Chem. Soc, 105, 1402, 1914.
^ Richards and Lembert, Jour. Amer. Chem. Soc, 36, 1329, 1914.=
^ Honigschmidt and Horowitz, Sitz. Akad. Wiss. Wien, iia, = 123, 1914.
' Merton, Proc. Roy. Soc, 91A, 198, 1914.
121
122 ISOTOPES
stable orbits. The frequencies of the spectral lines emitted by the element are associated in an absolutely definite manner with the rotational frequencies of these orbits which are calculated by what is known as a " quantum " relation. Without going further into the theory it will be seen at once that if we alter the force acting between the central nucleus and its planetary electrons these orbits will change and with them the frequency of the light emitted. It is therefore of interest to examine the magnitude of the change, to be expected=
from this theory, when we alter the mass of the nucleus without=
changing its charge, and so pass from one isotope to another.
The difference in the system which will first occur to one is that although the electrical force remains the same the gravi- tational force must be altered. The order of magnitude of the change expected in the total force will clearly be given by=
considering the ratio between the electrical and gravitational forces acting, to take the simplest case, between the protou and the electron in a neutral hydrogen atom.
Assuming the law of force to be the same in both cases, this ratio is simply e^/GMm ; where e is the electronic charge 4-77 X 10~i", G the universal gravitational constant 6-6 x 10"^,=
M the mass of the proton 1-66 x lO"^*^ and m the mass of the=
electron 9-0 x 10~ 2^. Putting in these numerical values we obtain the prodigious ratio 2-3 x 10 ^9. In other words the effect of doubling the mass of the nucleus without altering its=
charge would give the same percentage increase in the total pull on the planetary electron, as would be produced in the pull between the earth and the moon by a quantity of meteoric dust weighing less than one million millionth of a gramme falling upon the surface of the former body. The gravitational effect may therefore be dismissed as entirely negligible.
110. Deviation of the Bohr orbits due to change in the position of the centre of gravity of the rotating system. Although we may neglect the gravitational effect there is another, of quite a different order, which arises in th= e following manner. The mass of the electron compared with that of the nucleus is small but not absolutely negligible, hence=
it will not rotate about the nucleus as though that were a
THE SPECTRA OF ISOTOPES 123
fixed point, but both will rotate about their common centre of gravity. The position of this centre of gravity will be shifted by any alteration in the mass of the nucleus. If E, M=
and e, m are the respective charge and mass of the nucleus and=
the rotating electron, the equation of motion is
rM , Ee
M + m r^
where r is the distance between the two charges and w the angular velocity. Bohr ^ introduced this effect of the mass of the nucleus in order to account for the results obtained by Fowler. 2 The Bohr expression for the frequency then becomes
where e, E and m, M are the charges and masses of the electron=
and nucleus respectively. If we suppose that the atomic weight of lead from radium to be one unit less than that of ordinary lead, this theory predicts a difference in wave-length, for the principle line, of 000005 A between the two, a quantity=
beyond the reach of the most delicate methods of spectrum analysis used up to the present.
111. Later experiments of Aronberg and Merton.
In 1917 Aronberg,^ applying the extremely high dispersion derived from the spectrum of the sixth order of a Michelson 10-inch grating to the line A 4058 emitted from a specimen of radio-lead of atomic weight 206-318, observed a difiference of 0-0044 A between this and ordinary lead, of atomic weight 207-20. This remarkable result has been since confirmed by Merton of Oxford* who gives the difference of wave-length between radio-lead from pitchblende and ordinary lead as 0-0050^2 0-0007, Merton made use of a totally different optical system, namely a Fabry and Perot etalon, so that the agreement is very striking.
It is to be noticed that the effect observed was not a mere
1 Bohr, Nature, 92, 231, 1913.
2 Fowler, Nature, 92, 95, 1913.
3 Aronberg, Proc. Nat. Acad. Sci., Z, 710, 1917, and Ast= rophys, Jour., 47, 96, 1918.
4 Merton, Proc. Boy. Soc, 96A, 388, 920.
124
ISOTOPES
broadening of the line but a definite shift, and that, though
of the same sign, it is about one hundred times greater than
that predicted by the Bohr theory, Merton also found a shift
of 0-0022 =C2=B10-0008 A between the wave-length of thorite-lead
and ordinary lead, differing in atomic weight by about 0-6.
The heavier atom shows the higher frequency in all cases.
This remarkable discrepancy between the shift predicted by
theory and that actually observed has been discussed by
Harkins and Aronberg.^
At a recent discussion on isotopes at the Royal Society ^ Merton commented upon the line 6708 A emitted by the element lithium, which consists of two components 0-151 A apart. If lithium is accepted as a mixture of isotopes 6 and 7,= ^ he calculated that each of these components should be accom- panied by a satellite, some sixteen times as faint, displaced by=
0-087 A. So far he had not been able to observe such satellites= . Previous experiments of Merton and Lindemann* on the expected doubling in the case of neon had given no conclusive results on account of the physical width of the lines. It was hoped that this difficulty could be overcome by the use of liquid hydrogen temperatures.
StiU more recently Merton^ has repeated his experiments on lead, using a very pure sample of uranium lead from Australian Carnotite. His final results are indicated in the following table :
A
(Carnotite lead)"! . ^(ordinary lead) J
r Wave niimber (ordinary lead) ' . Wave-number (Carnotite lead).
4058 3740 3684 3640 3573
0-011 =C2=B10-0008 0-0074=C2=B10-0011 0-0048=C2=B10-0007 0-0070=C2=B10-0003 0-0048=C2=B10-0005
0-065=C2=B10-005 0-053=C2=B10-008 0-035=C2=B10-005 0-C52=C2=B10-002 0-037=C2=B10-004
1 Harkiiis and Aronberg, Jour. Am. Chem. Soc, 42, 1328,
Merton, Proc. Roy. Soc.=C2=BB 99A, 87, 1921.
=C2=BB V. p. 86.
- Lindemann, ibid.
Merton, Roy. Soc. Proc, lOOA, 84, 1921.
1920.
THE SPECTRA OF ISOTOPES 125
It will be noticed that the shift for the line A 4058 is rathe= r more than twice that obtained before. Merton suggests that the most probable explanation of this difference is evidently that the Carnotite lead used is a purer sample of uranium lead=
than that obtained from the pitchblende residues. It is also apparent that the differences are not the same for different lines, an interesting and somewhat surprising result.
112. "Isotope" effect on the Infra-red spectrum of molecules. The extreme smaUness of the isotope " shift "=
described above in the case of line spectra emitted by atoms is=
due to the fact that one of the particles concerned in the vibration is the electron itself, whose mass is minute compared with that of the nucleus. Very much larger effects should be expected for any vibration in which two atoms or nuclei are concerned, instead of one atom and an electron. Such a vibration would be in the infra-red region of the spectrum.
This effect was first observed by Imes^ when mapping the fine structure of the infra-red absorption bands of the halogen acids. In the case of the HCl " Harmonic " band at 1-76^, mapped with a 20,000 line grating, the maxima were noticed to be attended by satellites. Imes remarks : " The apparent tendency of some of the maxima to resolve into doublets in the=
case of the HCl harmonic may be due to errors of observation, but it seems significant that the small secondary maxima are all on the long-wave side of the principal maxima they accom- pany. It is, of course, possible that still higher dispersion applied to the problem may show even the present curves to be composite."
Loomis^ pointed out that these satellites could be attributed to the recently discovered isotopes of chlorine. In a later paper ^ he has shown that, if mi is the mass of the hydrogen nucleus, and ma the mass of the charged halogen atom, the
difference should be expressed by the quanity ^ = ~ the
square root of which occurs in the denominator of the expression=
^ Imes, Astrophysical Journal, 50, 251, 1919.
2 Loomis, Nature, Oct. 7, 179, 1920.
^ Loomis, Astrophysical Journal, 52, 248, 1920.
126 ISOTOPES
for frequency. " Consequently the net difference between the spectra of isotopes will be that the wave-lengths of lines in the spectrum of the heavier isotope will be longer than the=
corresponding lines for the lighter isotope in the ratio 1 + 1/1330 : 1 for chlorine and 1 -f 1/6478 : 1 for bromine.=
Since the average atomic weight of chlorine is 35-46 the amounts=
of CP^ and CP' present in ordinary chlorine must be as 1-54 : 0-46 or as 3-35 : 1 and, if the lines were absolutely = sharp and perfectly resolved, the absorption spectrum of ordinary HCl should consist of pairs of lines separated by 1/1330 of their frequency and the one of shorter wave-length should have about 3-35 the intensity of the other. The average atomic weight of bromine is 79-92, hence the two isotopes are present in nearly equal proportions and the absorption spectrum of HBr should consist of lines of nearly equal intensity separated by 1/6478 of their frequency."
The latter will be too close to be observed with the dispersion=
employed. In the case of the HCl band at IIQ ju the difference=
of wave number on this view should be 4-3. The mean differ- ence of wave number given by Loomis' measurements of 13 lines on Imes' original curves for this band is 4-5 ^ 0-4 corre= - sponding to 14 A in wave-length.
The spectroscopic confirmation of the isotopes of chlorine has also been discussed by Kratzer,! who considers that the oscillation-rotation bands of hydrogen chloride due to Imes^ are in complete accordance with the theory.
1 H. Ivratzer, Zeit. Physik., 3, 60, 1920.
- Loc. cit.
CHAPTER XI - THE SEPARATION OF ISOTOPES
113. The Separation of Isotopes
The importance, from purely practical and technical points of view, of the theory of isotopes would have been insignificant had its application been confined to the radioactive elements and their products, which are only present in infinitesimal quantities on the Earth. But now that the isotopic nature of many elements in everyday use has been demonstrated, the possi- bility of their separation, to any reasonable extent, raises questions of the most profound importance to applied science. In physics all constants involving, e.g., the density of mercury=
or the atomic weight of silver may have to be redefined, while=
in chemistry the most wholesale reconstruction may be necessary for that part of the science the numerical founda- tions of which have hitherto rested securely upon the constancy of atomic weights.
It is therefore of great interest to consider in turn the various methods of separation proposed and examine how far they have been successful in practice.
114. Separation by Diffusion
he subject of the separation of a mixture of two gases by the method of Atmolysis or has been thoroughly investigated by the late Lord Rayleigh. The diffusion is supposed to take place through porous material. The conditions under which maximum separation is to be obtained are that " mixing " is perfect, so that there can be no accumulation of the less diffusible gas at the surface of the porous material, and that the apertures in the material through which the gases must
iRayleigh, Phil. Mag., 42, 493, 1896. 127
128 ISOTOPES
pass are very small compared with the mean free path of the molecules. If these conditions are satisfied he obtains as an expression for the effect of a single operation :
X + y _ ^ . _^ Y
r '^
X + Y X + Y "-'^ X + Y "-'-
where (X Y) {x, y) are the initial and final volumes of the gases, /I, V, the velocities of diffusion, and r the enrichment=
of the residue as regards the second constituent.
The velocity of diffusion of a gas is proportional to the square root of the mass of its molecules, so that if a mixture=
of two isotopes is allowed to diffuse a change in composition must be brought about. Now no known isotopes differ from each other much in mass, so the difference between their rates of diffusion will also be small, hence the above equation=
may be written in the approximate form
^- =3D rTc where h =3D ^ a small quantity and,
and, finally, the enrichment by diffusion of the residue as regards the heavier constituent may be expressed with sufficient accm'acy by the expression
mi-m /Initial volume
Final volume
where Wi, mg are the molecular masses of the lighter and
heavier isotope respectively. In the most favourable case
known at present, that of the isotopes of neon, the number
over the root is 21 so that the change in composition obtain-
able in a single operation will in practice be very small.
If we take the density of the original mixture as unity, the increase in density of the residual gas to be expected from the=
operation of diffusion will be approximately
(r 1) X ^ X 2 ^
X Wg + Wi
Now neon consists of monatomic molecules differing between each other in mass by 10 per cent, and the heavier is present=
to the extent of 10 per cent. In the diffusion experiments described on p. 39 the effective ratio of the initial volume to=
THE SEPARATION OF ISOTOPES 129
the final volume was estimated as certainly greater than 500 and probably less than 10,000, so that r lies between 1-3 and 1-5. Hence the increase of density of the heavier residue should have been between -003 and -005. It was actually 004.
115. The separation of the isotopes of chlorine by the diffusion of HCl
In the case of other isotopic gaseous mixtures the numerical obstacles in the way of practical separation wiU be correspondingly greater. Thus in the case of HCl the 36th root is involved, and in that of HBr the 80th root. The only way by which measurable increase in density may be hoped for wiU clearly be by increasing the effective ratio of the initial to final volumes to an heroic degree. This can be done by experiments on a huge scale or by a vast number of mechanical repetitions.
Harkins started to attack the HCl problem in 1916 using the first of these two alternatives. In 1920 he mentions a quantity of 19,000 litres of HCl as having been dealt with in these experiments. 2 In the following year^ he published numerical results indicating that a change in atomic weight of 0-055 of a unit had been achieved.
At the recent discussion on isotopes * Sir J. J. Thomson pointed out that a change in the molecular weight of HCl should be caused by allowing a stream of the gas to flow over=
the surface of a material which absorbed it. The higher diffusion coefficient of the lighter isotope would result in it being absorbed more rapidly than the heavier one, so that the residue of unabsorbed gas should give a higher molecular weight. This " free diffusion " without the interposition of porous material has been recently tried in the Cavendish Laboratory by E. B. Ludlam, but no measurable difference has so far been detected.
116. Separation by Thermal Diffusion
It has been
^ Harkins, Jour. Amer. Cheni. Soc, Feb., 1916.
2 Harkins, Science, Mar. 19, 1920 ; Nature, Apl. 22, 1920 ; see=
also Phys. Rev., 15, 74, 1920 ; Science, 51, 289, 1920 ; Jour. = Amer, Chem. Soc, 42, 1328, 1920.
3 Harkins, Science, Oct. 14, 1921 ; Nature, Oct. 3, 1921.=
- J. J. Thomson, Proc. Roy. Soc, 99A, 98, 1921.
K
shown on theoretical grounds independently by Enskog ^ and Chapman ^ that if a mixture of two gases of different molecular weights is allowed to diffuse freely, in a vessel of which the ends are maintained at two different temperatures T,T', until equilibrium conditions are reached, there will be a slight excess of the heavier gas at the cold end, and of the=
lighter gas at the hot end. The separation attained depends on the law of force between the molecules and is a maximum if they behave as elastic spheres. The effect was experi- mentally verified for a mixture of CO2 and Ha by Chapman and Dootson,^ and recently Ibbs * has demonstrated that the separation can be carried out continuously and that the time for equilibrium to be established is quite short.
Chapman has suggested ^ that thermal diffusion might be used to separate isotopes. He shows that the separating power depends on a constant ^x. And when the difference between the molecular masses mi, ma is smaU the value of this is approximately given by
, _ 17 ma mi AiAj
^^ ~~ 3 ma + mi 9-15 8-25 AiAa where ^1,^2 denote the proportions by volume of each gas in the mixture ; thus Ai -f Aa =3D=3D1. The actual separation =
is
given by
Ai A'l =3D (Ai A'a) =3DA;t log T'/T.=
He gives the following numerical example : " Suppose that it is=
desired to separate a mixture of equal parts of Ne^" and Ne^^,=
then, writing mi =3D 20, ma =3D 22, Ai =3D A3 =3D ^, we find =
that
Ic,^ =3D 0-0095. Suppose that the mixture is placed in a vessel=
consisting of two bulbs joined by a tube, and one bulb is maintained at 80=C2=B0 absolute by liquid air, while the other is=
heated to 800=C2=B0 absolute (or 527=C2=B0 C). When the steady st= ate has been attained the difference of relative concentration between the two bulbs is given by the equation
1 Enskog, Phys. Zeit., 12, 538, 1911 ; Ann. d. Phys., 38, 75= 0, 1912.
2 Chapman, Phil. Trans., 217A, 115, 1916; Phil. Mag., 34, 146, 1917.
3 Chapman and Dootson, Phil. Mag., 34, 248, 1917.
- Ibbs, Proc. Boy. Soc, 99A, 385, 1921.
^Chapman, Phil Mag., 38, 182, 1919.
THE SEPARATION OF ISOTOPES 131
Ai A'l =3D (A 2 A' 2) =3D 0-0095 lo= g, 800/80
=3D 0-022
or 2-2 per cent. Thus the cold bulb would contain 48-9 per cent. Ne^" to 51-1 per cent. Ne^^, and vice versa in the hot bulb. By drawing o=C2=A3f the contents of each bulb separately, and by repeating the process with each portion of the gas, the=
difference of relative concentrations can be much increased. But as the proportions of the two gases become more unequal, the separation effected at each operation slowly decreases. For instance, when the proportions are as 3 : 1, the variation=
at each operation falls to 1-8 per cent. ; while if they are a= s 10 : 1 the value is 1-2 per cent. This assumes that the mole-=
cules behave like elastic spheres : if they behave like point centres of force varying as the inverse nth. power of the distan= ce, the separation is rather less; e.g., ii n=3D9, it is just over=
half the above quantities."
Chapman points out that for equal values of log p/p and log T/T pressure diffusion (centrifuging) is about three times as powerful as thermal diffusion but suggests that it may be more convenient to maintain large differences of temperature than of pressure.
117. Separation by Gravitation or "Pressure Diffusion"
When a heterogeneous fluid is subjected to a gravitational field its heavier particles tend to concentrate in the direction of the field, and if there is no mixing to co= unter- act this a certain amount of separation must take place. If therefore we have a mixture of isotopes in a gaseous or liquid=
state partial separation should be possible by gravity or centrifuging.
The simplest case to consider is that of the isotopes of neon in the atmosphere and, before the matter had been settled by the mass-spectrograph, analysis of the neon in the air at very great heights was suggested as a possible means of proving its isotopic constitution. 1 The reasoning is as follows: =E2=80= =94
If M be the atomic weight, g the gravitational constant, p the pressure, and p the density, then if no mixing takes place dp =3D gpdh, h being the height. In the isother= mal
1 Lindemann and Aston, Phil. Mag., 37, 530, 1919.
132 ISOTOPES
layer convection is small. If it is small compared with diffusion the gases will separate to a certain extent. Since T is constant
RTp , dp Mp ,,
whence p =3D pffi Rt ,
Po being the density at the height Jiq at 'which mixing by convection ceases, about 10 kilometres, and A^ the height above this level. If two isotopes are present in the ratio 1 to Ko, so that the density of one is po and of the other Kopo=
at height Jiq, then their relative density at height h^ + /SJi = is given by
Putting T =3D 220 as is approximately true in England,
XT
A^ being measured in kilometres. If Mi Ma =3D 2, th= erefore
It might be possible to design a balloon which would rise to 100,000 feet and there fill itself with air. In this case the relative quantity of the heavier constituent would be reduced from 10 per cent, to about 8-15, so that the atomic weight of=
neon from this height should be 20-163 instead of 20-2. If one could get air from 200,000 feet, e.g. by means of a long-=
range gun firing vertically upwards, the atomic weight of the neon should be 20-12.
A more practicable method is to make use of the enormous gravitational fields produced by a high speed centrifuge.
In this case the same equation holds as above except that g varies from the centre to the edge. In a gas therefore <ip__Mv2 dr _ _Mo)'^ ~^ ~ Rf "y ~ RT '
whence p =3D poe 2rt,
Vq being the peripheral velocity. Here again, if Kq is the
THE SEPARATION OF ISOTOPES 133
ratio of the quantities present at the centre, the ratio at the=
edge will be
A peripheral velocity of 10^ cm,/s. or perhaps even 1-3 x 10^ cm./s. might probably be attained in a specially designed
rr
centrifuge, so that:^^ might be made as great as e"=C2=B0'2^^'^'~^=
- ^ or
even e ~0'^'^^'^>~^2),
If Ml M2 is taken as 2 a single operation would there= fore give fractions with a change of K of 0-65. In the case of neon=
the apparent atomic weight of gas from the edge would be about 0-65 per cent, greater than that of gas from the centre,=
i.e. a separation as great as the best yet achieved in practice=
by any method could be achieved in one operation. By centrifuging several times or by operating at a lower tempera- ture the enrichment might be increased exponentially.
Centrifuging a liquid, e.g. liquid lead, would not appear so favourable, though it is difficult to form an accurate idea of the quantities without a knowledge of the equation of state. If compression is neglected and the one lead treated as a solution in the other, a similar formula to that given above holds. On assumptions similar to these Poole ^ has calculated that a centrifuge working with a peripheral velocity of about 10^ cm. /sec should separate the isotopes of mercury to an extent corresponding to a change of density of 0-000015.
The only experiments on the separation of isotopes by the use of a centrifuge, so far described, are those of Joly and Poole 2 who attempted to separate the hypothetical isotopic constituents of ordinary lead by this means. No positive results were obtained and the check experiments made with definite alloys of lighter metals with lead were by no means encouraging.
118. Separation by Chemical Action or Ordinary Fractional Distillation. The possibility of separating iso-=
topes by means of the difference between their chemical affinities or vapour pressures has been investigated very fully
1 Poole, Phil. Mag., 41, 818, 1921.
2 Joly and Poole, Phil. Mag., 39, 372, 1920.
134 ISOTOPES
from the theoretical standpoint by Lindemann. The thermo- dynamical considerations involved are the same in both cases. The reader is referred to the original papers ^ for the details=
of the reasoning by which the following conclusion is reached : =
" Isotopes must in principle be separable both by fractiona- tion and by chemical means. The amount of separation to be expected depends upon the way the chemical constant is calculated and upon whether ' NuUpunktsenergie ' is assumed. At temperatures large compared with ^v,^ which are the only practicable temperatures as far as lead is concerned, the difference of the vapour pressure and the constant of the
Bv law of mass action may be expanded in powers of ^. The
Bv most important term of the type log "^ is cancelled by the
chemical constant if this is calculated by what seems the only
Bv reasonable way. The next term in is cancelled by the=
' NuUpunktsenergie ' if this exists. All that remains ar=
e
Bv terms containing the higher powers of ^. In practice there- fore fractionation does not appear to hold out prospects of success unless one of the above assumptions is wrong. If the first is wrong a difference of as much as 3 per cent, should occur at 1200 and a difference of electromotive force of one miUivolt might be expected. Negative results would seem to indicate that both assumptions are right."
As regards experimental evidence it has already been pointed out that the most careful chemical analysis, assisted by radio- active methods of extraordinary delicacy, was unable to achieve the shghtest separation of the radioactive isotopes. The laborious efforts to separate the isotopes of neon by a differ-=
ence of vapour pressure over charcoal cooled in hquid air also gave a completely negative result.
119. Separation by evaporation at very low pressure
If a liquid consisting of isotopes of different mass is allowed
1 Lindemann, Phil. Mag., 37, 523, 1919 ; 38, 173, 1919.
- (iv is the " characteristic " and T the " Absolute " tempera=
ture.
THE SEPARATION OF ISOTOPES 135
to evaporate it can be shown that the number of Hght atoms escaping from the sm'face in a given time will be greater than=
the number of heavier atoms in inverse proportion to the square roots of their weights. If the pressure above the surface is kept so low that none of these atoms return the concentration of the heavier atoms in the residue will steadily increase. This method has been used for the separation of isotopes by Bronsted and Hevesy, who appUed it first to the element mercury.
The mercury was allowed to evaporate at temperatures from 40=C2=B0 to 60=C2=B0 C. in the highest vacuum attainable. The eva= porat- ing and condensing surfaces were only 1 to 2 cms. apart, the latter was cooled in liquid air so that all atoms escaping reached it without coUision and there condensed in the sohd form.
It will be seen that the Uquid surface acts exactly Uke the porous diaphragm in the diffusion of gases. ^ The diffusion rate of mercury can be obtained approximately from the diffusion rate of lead in mercury ^ and is such that the mean=
displacement of the mercury molecule in Uquid mercury is about 5 X 10"^ cm. sec."^. It follows that if not more than 5 X 10"^ c.cm. per cm.^ surface evaporate during one second no disturbing accumulation of the heavier isotope in the surface layer takes place.
The separation was measured by density determination. Mercury is particularly well suited for this and a notable feature of this work was the amazing deUcacy with which it could be performed. With a 5 c.cm. pyknometer an accuracy of one part in two millions is claimed. The first figures pubhshed ^ were :
Condensed mercury. . . . 0-999981
Residual mercury .... 1-000031
The densities being referred to ordinary mercury as unity.
The later work was on a larger scale.* 2700 c.cm. of mercm-y were employed and fractionated systematically to about
1 V. p. 127.
- Groh and Hevesy, Ann. der Phys., 63, 92, 1920.
^ Bronsted and Hevesy, Nature, Sept. 30, 1920.
- Bronsted and Hevesy, Phil. Mag., 43, 31, 1922.
136 ISOTOPES
1/100,000 of its original volume in each direction. The final=
figures were :
Lightest fraction vol. 0-2 c.c. . . 0-99974
Heaviest fraction vol. 0-3 c.c. . . 1-00023
Mercury behaves as though it was a mixture of equal parts of two isotopes with atomic weights 202-0, 199-2 in equal parts or of isotopes 201-3, 199-8 when the former is four times=
as strong as the latter, and so on.
120. Separation of the isotopes of chlorine by free evaporation
The same two investigators were able to announce the first separation of the isotopes of chlorine ^ by applying the above method to a solution of HCl in water. This was allowed to evaporate at a temperature of 50=C2= =B0 C. and condense on a surface cooled in hquid air. Starting with 1 litre 8-6 mol. solution of HCl 100 c.c. each of the lightest=
and heaviest fraction were obtained.
The degree of separation achieved was tested by two difiEerent methods. In the first the density of a saturated solution of NaCl made from the distillate and the residue respectively was determined with the following results :
Density (salt from distillate) =3D 1-20222 Density (salt from residue) =3D 1-20235
These figures correspond to a change in atomic weight of 0-024 of a unit.
In the second method exactly equal weights of the isotopic NaCls were taken and each precipitated with accurately the same volume of AgNOg solution, in shght excess. After pre- cipitation and dilution to 2,000 c.c. the approximate concen- tration of the filtrate was determined by titration, also the ratio of Ag concentration of the two solutions was measured in a concentration cell. Calculation showed that the difference in atomic weight of the two samples was 0-021 in good agree- ment with the density result.
121. Separation by Positive Rays
The only method which seems to offer any hope of separating isotopes completely, and so obtaining pure specimens of the constituents of a com-
1 Bronsted and Hevesy, Nature, July 14, 1921.
THE SEPARATION OF ISOTOPES 137
plex element, is by analysing a beam of positive rays and trapping the particles so sorted out in different vessels. It is=
therefore worth while inquiring into the quantities obtainable by this means.
Taking the case of neon and using the parabola method of analysis with long parabolic slits as collecting vessels we find=
that the maximum separation of the parabolas corresponding to masses 20 and 22 (obtained when electric deflexion d is haK the magnetic) is approximately
^ 1 M,-M, _ d_ V2 Ml 28"
Taking a reasonable value of 0 as -3 the maximum angular width of the beam for complete separation =3D 0-01. If the canal-ray tube is made in the form of a slit at 45=C2=B0 to ax= es, i.e. parallel to the curves, the maximum angular length of the beam might be say 5 times as great, which would collect the positive rays contained in a solid angle of -0005 sq. radian= .
The concentration of the discharge at the axis of the positive ray bulb is considerable, and may be roughly estimated to correspond to a uniform distribution of the entire current over a |- sq. radian. One may probably assume that half the current is carried by the positive rays, and that at least half=
the positive rays consist of the gases desired. If neon is analysed by this method therefore the total current carried by the positive rays of mass 20 is
0005 x4:Xixlxi=3D -0005 i.
If i is as large as 5 miUiamperes this =3D 1-5 x 10* E.S.U.=
1-5 X 10*
or
2-7 X 1019 X 4-77 X 10-1"
=3D 1-2 X 10"^ c.c./sec.
i.e. one might obtain about one-tenth of a cubic millimetre of
Ne2o and 1/100 cubic miUimetre of Ne^^ per 100 seconds run.
It is obvious that even if the difficulties of trapping the rays=
were overcome, the quantities produced, under the most favourable estimates, are hopelessly small.
122. Separation by photochemical methods
A remarkably beautiful method of separating the isotopes of=
138 ISOTOPES
chlorine has been suggested by Merton and Hartley which depends upon the following photochemical considerations. Light falling on a mixture of chlorine and hydrogen causes these gases to combine to form hydrochloric acid. This must be due to the activation of the atoms of hydrogen or those of=
chlorine. Supposing it to be the latter it is conceivable that the radiation frequency necessary to activate the atoms of Cl^^ will not be quite the same as that necessary to activate those of CP'^. CaUing these frequencies 5^35 and V37 respectively=
it would seem possible, by excluding one of these frequencies entirely from the activating beam, to cause only one type of chlorine to combine and so to produce pure HCI^^ or HCI^'. Now ordinary chlorine contains about three times as much CP^ as CP^ and these isotopes must absorb their own activat- ing radiation selectively. In this gas therefore light of frequency V35 will be absorbed much more rapidly than that of frequency V37, so that if we aUow the activating beam to pass through the right amount of chlorine gas V35 might be completely absorbed but sufficient V37 radiation transmitted to cause reaction. On certain theories of photo-chemistry light containing ^37 but no V35 would cause only atoms of CP^ to combine so that a pure preparation of HCP^ would result. Pure CP'^ made from this product could now be used as a filter for the preparation of pure HCP^, and this in its turn would yield pure CP^ which could then be used as a more efficient filter for the formation of more HCP^
Had this very elegant scheme been possible in practice it would have resulted in a separation of a very different order to those previously described and the preparation of un- limited quantities of pure isotopes of at least one complex element. There is however little hope of this, for so far the results of experiments on this method have been entirely negative.
123. Other methods of separation and general conclusions
The following methods have also been suggested. By the electron impact in a discharge tube, in the case of the inert gases, the Ughter atoms being more strongly urged towards
THE SEPARATION OF ISOTOPES 139
the anode ;^ by the migration velocity of ions in gelatine ; ^=
by the action of light on metallic chlorides,^
A survey of the separations actually achieved so far shows that from the practical point of view they are very small. In cases where the method can deal with fair quantities of the substance the order of separation is small, while in the case of complete separation (positive rays) the quantities produced are quite insignificant. We can form some idea by considering the quantity
Q =3D (difference in atomic weight achieved) X (average quantity of two fractions produced in grammes). As regards the first of these factors the highest figure so far was 0-13 obtained by the writer in the original diffusion experiments on neon, but as the quantities produced were only a few milli- grams Q is negligibly small. The highest values of Q have been obtained by Bronsted and Hevesy by their evaporation method.* It is 0-5 in the case of Hydrochloric Acid, 0-34 in that of Mercury.
When we consider. the enormous labour and difficulty of obtaining this result it appears that unless new methods are discovered the constants of chemical combination are not likely to be seriously upset for some considerable time to come.=
1 Skaupy, Zeitsch. Phys., 3, 289, 460, 1920.
2 Lindemann, Proc. Roy. Soc, 99A, 104, 1921.
3 Renz, Zeit. Anorg. Chem., 116, 62, 1921.
- V. p. 134.
APPENDIX I
Table of atomic weights and isotopes of the elements.
The elements are given in order of their atomic= numbers. The
different periods are indicated by gaps after the inert gases. A curious relation, pointed out by Rydberg, is that the atomic numbers of all the inert gases are given by taking the series 2 (P + 2^ + 22 + 3^ + 3^ + 4^ + = ) and stoppmg the summation at any term. This gives the numbers used by Langmuir (p. 95).
The atomic weights given are the International ones except in the cases marked with an asterisk, where the figures are taken f= rom some of the recent determinations given below.
The isotopes where known are given in order of their atomic masses. The proportion of an isotope in a complex element is indicated by the index letters a, 6, c ... in descending order.=
In the case of isotopes of the radioactive elements 81-92 the ro= man numeral gives the number of them believed to exist. The nomen- clature of some of the rare earths 69-72 is not yet standardised.=
The names here are those used by Moseley. Some of these elements= , though detected by their X-ray spectra, have never been isolated.=
The elements corresponding to atomic numbers 43, 61, 75, 85, 87=
(all odd) have not yet been discovered.
Recent atomic weight determinations. The following is a list of some of the elements whose atomic weights have been re-=
determined quite recently, together with references to the papers in which they were published. Where more than one value is given different methods were used :
Fluorine 19-001. Moles and Batuecas, Jour. Chim. Phys., 18, 35= 3,
1920. Aluminium 26*963. Richards and Krepelka, Journ. Am. Chem. Soc,=
42, 2221, 1920.
Silicon 28-111. Baxter, Weatherelland Holmes, ibid., 42, 1194, =
1920.
Scandium 45-10. Honigschmid, Zeit. Electrochem., 25, 93, 1919.=
Tin 118-703. Baxter and Starkweather, Journ. Am. Chem. Soc, 42,=
905, 1920.
118-699. Brauner and Krepelka, ibid., 42, 917, 1920.
141
142
APPENDIX I
Tellurium 127-73, 127-79. Bruylants and Michielsen, Bull=
. Acad.
Bdg., 119, 1919. Samarium 150 "43. Owens, Balke and Kremers, Journ. Am. Chem= .
Soc, 42, 515, 1920. Thtdium 169-44, 169-66. James and Stewart, ibid., 42, 2022, = 1920. Bismuth 209-02. Honigschmid, Zeit. Electrochem., 26, 403, 1920= .
208-9967. Classen and Wey, Ber., 53, 2267, 1920. Antimony 121-773. Willard and McAlpine, Jouryi. Am. Chem. Soc, = 43,
797, 1921. Lanthanum 138-912. Baxter, Tani and Chapin, Journ. Am. Chem.=
Soc, 43, 1085, 1921.
Germanium 72-418. Miller, Journ. Am. Chem. Soc, 43, 1085, 19=
21.
Zinc 65-38. Baxter and Hodges, i&id., 43, 1242, 1921.
Cadmium 112-411. Baxter and Wilson, ibid., 43, 1230, 1921.
-Q
" m
o^
Element.
2
a
if
Masses of isotopes.
=C2=A3 -2 *^ Hydrogen . .
H
1
1-008
1
1-008
f^^'o Helium . . .
He
2
4-00
1
4
&> 1"
00 Lithivim .
Li
3
6-94
2
-
" Beryllium
Be
4
91
1
9
r^ Boron
B
5
10-9
2
10=C2=BB 11"
3 Carbon .
C
6
12-00
1
12
S Nitrogen .
N
7
14-008
1
14
^ Oxygen . . .
0
8
16-00
1
16
0 Fluorine .
F
9
19-00
1
19
^ Neon ....
Ne
10
20-20
2
20" 22* 23
oQ Sodium .
Na
11
2300
1
^ Magnesium .
Mg
12
24-32*
3
24-=3D 25* 26^
Aluminium .
Al
13
26-96*
_o Silicon
Si
14
28-3
2
28" 29* (30)
3 Phosphorus .
P
15
31-04
1
31
^ Sulphur . . .
s
16
3206
1
32
'S Chlorine . . .
CI
17
35-46
2
35" 37* (39)
^ Argon . . .
A
18
39-9
2
36* 40" 39" 41*
Potassium
K
19
39-10
2
Calcium .
Ca
20
40-07
(2)
40 (44)
Scandium
Sc
21
45-1*
Titanium .
Ti
22
48-1
Vanadium
V
23
510
0
2 Chromium .
Cr
24
52-0
H Manganese .
Mn
25
54-93
' Iron ....
Fe
26
55-84
n
^ Cobalt . . .
Co
27
58-97
J Nickel
Ni
28
58-68
2
58" 60*
P
n Copper .
Cu
29
63-57
J
=3D Zinc ....
Zn
30
65-37
(4)
(64=C2=B0 66* 68 7O<0
- Galliimi . . .
Ga
31
70-10
Germanivmi .
Ge
32
72-5
Arsenic .
As
33
74-96
1
75
Seleniima .
Se
34
79-2
Bromine .
Br
35
79-92
2
79" 81*
Krypton .
Kr
36
82-92
6
78/ 80 82'^ 83-^ 84=C2=BB
86*
APPENDIX I
143
"S .
^
o *^
O^i
o ^^
Element
o
X!
E >,
00
Masses of Isotopes.
Rubidium
Rb
37
85-45
2
85" 87*
Strontium
Sr
38
87-63
Yttrium .
Y
39
89-33
Zirconium
Zr
40
90-6
Niobium .
Nb
41
93-1
00 Molybdenum
Mo
42
96-0
- H _ ~
43
'-' Ruthenium .
Ru
44
101-7
'o Rhodium.
Rh
45
102-9
=C2=A7 Palladium
Pd
46
106-7
An Silver ....
Ag
47
107-88
X Cadmium
Cd
48
112-40
"O Indiimi .
In
49
114-8
Tin ... .
Sn
50
118-7
Antimony
Sb
51
120-2
Tellurium
Te
52
127-5
Iodine
I
53
126-92
1
127
L Xenon
X
54
130-2
(7)5
(128) 129" (130) 13P 132=C2=BB 134 136"
Caesium .
Cs
55
132-81
1
133
Barium .
Ba
56
137-37
Lanthanum .
La
57
139-0
Cerium
Ce
58
140-25
Praseodymium .
Pr
59
140-6
Neodymiimi .
Nd
60
144-3
61
Samarium
Sm
62
150-4
Europium
Eu
63
152-0
Gadolinium .
Gd
64
157-3
Terbium .
Tb
65
159-2
Dysprosium .
Ds
66
162-5
c
5 Holmium
Ho
67
163-5
J, Erbium .
Er
68
167-7
=C2=B0 Thulium . . .
Tu
69
168-5
1 Ytterbiiun . .
Yb
70
173-5
'C Lutecuim
Lu
71
175
Pm (Keltium) . .
(Kt)
72
ji Tantalum
Ta
73
181-5
<=C2=BB Tungsten.
W
74
1840
75
Osmium .
Os
76
190-9
Iridium .
Ir
77
193-1
Platinimi .
Pt
78
195-2
1
Gold ....
Au
79
197-2
Mercury .
Hg
80
200-6
(6)
(197-200) 202 204
Thallium . . .
Tl
81
204-0
IV
Lead ....
Pb
82
207-2
XI
Bismuth .
Bi
83
209-0*
V
Poloniuna
Po
84 85
z
VII
L Emanation
Em
86
222-0
III
i
87
.2 Radium . =C2=AE Actinium.
Ra
88
226-0
IV
Ac
89
II
^ Thorium . . .
Th
90
23215
VI
^ Uranium X .
UX
91
II
t_ Uranium
Ur
92
238-2
II
APPENDIX II
The Periodic Table of the Elements. The atomic numbers ar= e given in bold type, the atomic weights in italics and the isotopes, where = known, in ordinary numerals. The roman ntmierals indicate the chemical groups and the most important associated valencies are given below them. Elem= ents are placed to the left or to the right of the columns according= to their chemical properties, those in the same vertical line as each other have s= trong chemical similarities. The Rare Earth group is surrounded by a thick line.= Elements 59-72 have no properties pronounced enough to give them definite = places in the table. The properties of the missing elements can be p= redicted with
PERIODIC TABLE OF
IH
1-008
Valency
0
I
+ 1
II
+ 2
III
+ 3
IV
+ 4
2 He
4-00 4
3 Li
6-94 6, 7
4 Be
9-1
9
5B 10-9 10, 11
60
12-00 12
10 Ne
20-2 20, 22
11 Na
23-00 23
12 Mg
24-32
24, 25, 26
13 AI
26-96
14 Si 28-3 28,29
18 A
39-9 36, 40
19 K
39-1 39, 41
29 Cu
63-57
20 Ca
40-07
30 Zn
65-37
21 Sc 45-1
31 G
70-1
22 Ti 48-1
32 Ge
72-5
36 Kr
82-92
78, 80, 82, 83, 84, 86
37 Rb
85-45
85, 87
47 Ag 107-88
38 Sr
87-83
48 Cd 112-40
39 Y
89-33
49 In
114-8
40 Zr
90-6
50 Sn
118-7
54 Xe
130-2
129, 131, 132, 134, 136
55 Cs
132-81
133
56 Ba
137-37
57 La 58 Ce 139-0 140-25
59 Pr eONd 61 62 Sm 63 Eu =
64 Gd 65 Tb
140-6 144-3 150-4 152-0 =
157-3 159-2
66 Ds 67 Ho 68 Ev 69 Tu 70 Yb 7= 1 Lu 72 (Kt) 162-5 163-5 1677 168-5 173-5 =
175
79 Au
197-2
80 Hg
200-6 197-204
81 Tl
204-0
82 Pb
207-2
86 Em
222-0
87-
88 Ra
226-0
89 Ac
90 Th
232-15
144
considerable certainty from the positions of their atomic numbers. From the point of view of the construction of the atom the inert gas= es should mark the end of the periods as they are shown to do ua the hst of = atomic weights in Appendix I, on the other hand it is more usual in chemistry = to start with valency 0. From principles of general convenience of arrangement t= he latter plan is adopted in this table, which is intended to give = the maximum amount of chemical information. Hydrogen, which belongs equally wel= l to group I or group VII, is best omitted from the. table altoget= her.
THE ELEMENTS
V
VI
VII
VIII
3
2
-
-1
7N
80
9F
14-01
16-00
1900
14
16
19
15 P
16 S
17 CI
31-04
32-06
35-46
31
32
35, 37
23 V
24 Cr
25 Mn
26 Fe
27 Co
28 Ni
Sl-O
33 As
74-96 75
52-0
34 Se
79-2
54-93
35 Br
79-92 79, 81
55-85
58-97
58-68 58.60
41 Nb
42 Mo
43
44 Ru
45 Rh
46 Pd
93-5
51 Sb 120-2
96-0
52 Te 127-5
531
126-92 127
101-7
102-9
106-7
73 Ta
74 W
7&-
76 0a
77 Ir
78 Pt
181-5
83 Bi
209-0
184-0
84 Po
85
190-9
1931
195-2
91 UX
ii
92 U
238-2
145
Recent results obtained by Dempster. Thanks to a private=
communication the writer is able to include some further results=
obtained by Dempster and a diagram of his apparatus for obtaining=
Fig. 19. Diagram of Anode in Dempster's latest apparatus.=
positive rays from metals. A full account is to appear in the
Physical Review. Fig. 19 shows the new arrangement of
vaporising furnace A and ionising filament C. The analysing
apparatus has already been described on p, 31 and the results wi=
th
.4F
5-9
f
'
1
k
Lithium.
\
1
\
1
\
)
J
[
<=3D/
v..
^^
/
K
9
30
ZO
10
60
6-1
6-9
Atomic Weight.
7-0
7-1
Fig. 20. Curve for Lithium.
146
APPENDIX III
147
magnesium on p. 81. Fig. 20 shows one of the curves obtained
with lithium. It will be seen that the relative intensities of t=
he
isotopes is entirely different from that found by the writer (p. =
86)
and also disagrees very definitely with the chemical atomic weight= . Dempster describes these relative intensities as varying very considerably. This is a most remarkable phenomenon and further information upon it is very desirable. There seems just a possibi= lity that the 6 line is enhanced by doubly charged carbon but it is =
not
easy to see where such particles could be produced.
l/oltS 943 928 913-5 899-5 886 873 860 847-5=
J
\
Zinc.
1
t
\
1
\
1
\
f
\
r
\
\
1
1
\
\i
1
1
\
/
\
I
/
1
=C2=AE
l/
\
1
i^
\
^^
62 63 64 65 66 67 Atomic Weight.
Fig. 21. Curve for Zinc.
68 69
70
Fig. 21 gives a remarkable curve obtained from zinc. This
indicates three strong isotopes and a faint fourth. The absolute=
scale of atomic weight is not known with certainty, and the valu= es 63, 65, 67, 69 are given by Dempster as those in best agreement=
with the atomic weight 65-37. Considering that the error in th= e
148 APPENDIX III
mean atomic weight of lithium, when calculated on these lines, is about 5 per cent, it would appear possible that these might = be a unit too high or too low. The probability of this is strengthene= d very much by the rule given on p. 110 connecting even atomic number with even atomic weight.
Results with calcium show only one line. This makes it extremely=
probable that this is a simple element of atomic weight 40 and=
therefore an isobare of argon. ^
Note. In a still later communication Dempster states that =
he
has been successful in using an anode of calcium to which a sma= U quantity of zinc had been added. By this means he is able to compare the masses of the zinc isotopes with the strong calcium=
maximum, assumed as 40. This gives the atomic weights as 64, 66, 68 and 70. The intensities are quite different to those in = the curve given above for zinc. 64 is now the strongest, 66 and 68=
fainter, while 70 is very faint indeed. No explanation is yet advanced for these remarkable irregularities in relative intensity.=
He has also observed a small maximum at 44 invariably accom-
panying the strong calcium maximum 40. This he considers to be
probably due to an isotope of that element present in smaU quant=
ity
as suggested by the atomic weight 40 07.
The above values are included provisionally in the tables on pages 89 and 142.
" V. p. 88.
INDEX
Abnormal hydrides, 98
Abundance of the elements, 111
Accuracy of mass-spectrograph, 60
Actinivim chain, 14, 15
Additive law of mass, 99
Alkali metals, mass-spectra of, 83
Alpha ray changes, 13
Analysis of the elements, 63
Andrade and Rutherford, 11
Anode, composite, 80, 86
hot, 80, 83, 84
Anticathode, silica, 48
Antimony, 78
Argon, 66
Aronbeeg, 123
,, and Harkins, 124
Atmolysis, separation by, 127
Atomic number, 13, 93
theory, 2
,, volume of isotopes, 18
weights, tables of, 89, 141
weights of radio -elements, 13,
141
Atoms, structure of, 90
Balke, Owens and Kremers, 142 Barkla, 93
Batuecas and Moles, 141 Baxter and Hodges, 142 and Parsons, 113 and Starkweather, 141 and Wilson, 142 Tani and Chapin, 142 Weatherell and Holmes, 73, 142 Beryllium, 88 Beta ray change, 13 Bohr, 94, 95, 121, 122, 123
,, atom, 95 BOLTWOOD, 1, 7 Boron, 72
anomalous atomic weight of,
114
trifluoride, 73
Bracketing, method of, 59, 69 Brauner and Krepelka, 141 Broek, Van den, 93, 94, 116 Bromine, 76
Bronsted and Hevesy, 135, 136, 139
Brosslera, 102, 104
Bruylants and Michielson, 142
Caesium, 87
,, anomalous atomic weight of, 114 Calcium, 88, 148 Calibration curve, 55 Camera of mass-spectrograph, 51
positive ray, 26
Canalstrahlen, 22 Carbon, 63
Carnotite, lead from, 124 Cathode rays, 22, 24 Chadwick, 94
and Rutherford, 103
Chapin, Baxter and Tani, 142 Chapman, 130
and DooTSON, 130
Chemical action, separation by, 133
law of radioactive change,
11 Chlorine, 65, 113
separation of the isotopes
of, 136 Classen, 31
and Wey, 142 Claude, 35 Cleveite, lead from, 17 Coincidence, method of, 57 Composite anode, 80, 86 Constancy of chemical atomic weights,
22 Cosmical effect of change of mass, 103 Crookes, 3, 4, 24, 115, 117 ,, dark space, 24, 35
theory of the evolution of
elements, 117 Curie, Mlle. I., 113
M., 18
Dalton's hypothesis, 2 Darwin, 15
Davies and Horton, 68 Deflection of positive rays, 27 Dempster, 31, 80, 81, 86, 114, 146
149
150
INDEX
Dempster's method of analysis, 31,146
Density balance, 35
,, of isotopic leads, 17, 18 Diffusion of neon, 39
separation by, 127 velocity, determination of, 20 Disintegration theory of the evolu- tion of elements, 116 Distillation of neon, 37 Distribution of lines on mass-
spectrum, 64 DooTSON and Chapman, 130 Du Bois magnet, 61
Eddington, 104
Einstein's theory of relativity, 103 Electrical theory of matter, 90 Electric discharge in gases, 23
,, field of mass-spectrograph, 50 Electricity as an element, 115 Electrochemical properties of isotopes,
10 Electron, the, 91
Element, meaning of the word, 115 Enskog, 130 Epstein, 95 ExNER and Haschek, 121
Fa JANS, 11
First order lines, 61
Fleck, 12
Fluorine, 72, 97
Focussing positive rays, 44
FOWLEB, 123
and Aston, 45
Fractional distillation, separation by,
133 Fbanck and Knipping, 68
Gehrcke, 102
,, and Reichenheim, 80, 83, 88 Geigek and Nuttall, 10, 13 Goldstein, 22 Gravitation effect on spectra, 121
separation by, 131
Groh and Hevesy, 20, 135
Hahn, 8
and Meitner, 8
Halation effect, 60 Half-tone plates, 25 Hall and Harkins, 116 Harkins, 102, 111, 116, 129
and Aronberg, 124
and Hall, 116
,, and Wilson, 116 Haschek and Exner, 121 Helium, 67, 69, 106
Hevesy, 10, 12, 19
and Bronsted, 136, 136,
139
and Groh, 20, 135
and Paneth, 11
and Zechmeisteb, 20
Hodges and Baxter, 142 Holmes, Baxteb and Weathebell,
73, 141 Honigschmid, 17, 18, 141, 142
and Horovitz, 18,
121 Horovitz and Honigschmid, 18, 121 HoBTON and Davies, 68 Hot anode, 80, 83, 84 Hydrochloric acid, diffusion of, 129 Hydrogen, 67, 69, 106 Hyman and Soddy, 17, 121
Ibbs, 130
Imes, 125, 126
Indicators, radioactive, 19
Infra-red spectrum of isotopes, 125
Intensity of positive rays, 44
Iodine, 78
Ionic dissociation theory, proof of, 20
lonisation in discharge tube, 24
Ionium, 1, 7, 9, 18
,, atomic weight of, 18 Isobares, 12, 13, 97, 110 Isotopes, definition of, 12
diagrams of, 97
discovery of, 5
melting point of, 18
refractive index of, 18
separation of, 127
solubility of, 18
table of, 89, 141
James and Stewabt, 142 JoLY and Poole, 133
Keetman, 7
Kernel of atom, 98
Kibchoff, 116
Knipping and Franck, 68
kohlweiler, 116
Kratzer, 126
Kremers, Owens and Balke, 142
Krepelka and Bbaun, 141
,, and RiCHABDS, 141
Krypton, 70
,, anomalous atomic weight of, 114
Landaueb and Wendt, 70 Langmuib, 95, 96, 99 Lead, atomic weight of, 16
,, from carnotite, 124
,, from thorite, 17
isotopes of, 14, 15
INDEX
15)
Lembert and Richards, 17, 121
Lewis-Langmuir atom, 95
LmDEMANN, 102, 124, 134, 139
,, and Aston, 131
Lines of first and second order, 61, 76
of reference, 55, 64
Lithium, 86, 97, 146 LooMis, 125, 126
LUDLAM, 129
McAxpiNE and Willard, 142
Magnesimn, 80
Magnetic field of mass-spectrograph,
51 Marckwald, 7, 8 Mass, change of, 100
deduced from parabolas, 28
deduced from mass -spectrum,
55 Mass-spectrograph, 43 Mass-spectrum, 47, 54 Measurement of lines on mass-
spectrum, 59 Meitner, 21
,, and Hahn, 8 Melting point of isotopes, 18 Mercury, 72, 80
parabolas of, 30
separation of the isotopes
of, 134 Merton, 121, 123, 124, 125 Mesothorium, 8, 10 Meta-elements, 4
Metallic elements, mass-spectra of, 80 Meteoric nickel, 113 MiCHiELSON and Bruylants, 142 Microbalance for density, 35 MiLLIKAN, 22, 91
Molecular lines of second order, 75 Moles and Batuecas, 141 MOSELEY, 11, 93, 115 Mtjller, 142 Multiply charged rays, 30
Natural numbers and atomic weights,
111 Negatively charged rays, 29, 62 Negative mass-spectra, 62, 66 Neon, 1, 33, 64, 97 Neuberger, 21 Nickel, 79
meteoric, 113
Nitrogen, 67, 110 Nomenclature of isotopes, 61 Nucleus atom, 10, 92, 97, 125
structure of, 101
Ntjttall and Geiger, 10, 13
Order, lines of first and second, 61 Owens, Balke and Kremers, 142 Oxygen, 63
Packing effect, 100
Paneth and Hevesy, 11
Parabola method of analysis, 25
Parsons and Baxter, 113
Perforated electrodes, 22, 24
Periodic law, 11, 12, 34
table of the elements, 144,
145 Period of radio-elements, 13 Perrin, 104 Phosphonas, 77
Photochemical separation, 137 Photographic plates for positive rays,
25 Planck's quantum, 95 Planetary electrons, 92 Poole, 133
and JoLY, 133
Positive ray paraljolas, 28
rays, 22
separation by, 136
Potassium, 87 Pressure diffusion, 131 Proton, the, 92 Protyle, 90, 118 Prout's hypothesis, 2, 90, 100
Radioactive isotopes, 7, 14
classification of,
21
transformations, 13, 14,
15 Radium B and lead, 11
D and lead, 11
Ramsay, 115
and Collie, 39
and Travers, 33
Ratner, 24 Rayleigh, 127 Reference lines, 55, 64 Refractive index of isotopes, 18 Reichenheim and Gehrcke, 80, 83,
88 Renz, 139
Resolving power of mass-spectro- graph, 60 Richards 17
and Krepelka, 141
and Lembert, 17, 121
and Wads WORTH, 17
Richardson, 85 Rossi and Russell, 9, 120 Rubidium, 87 Russell, U
and Rossi, 9, 120
Rutherford, Sir E., 7, 9, 13, 92, 93, 102
and Chadwick, 103
and Andrade, 11
Rydberg, 141
162
INDEX
SCHUTZENBERGER, 3
Screens, willemite, 25
Secondary rays, 29
Second order, lines of the, 61
Selenium, 77
Separation of isotopes, 127
Silicon, 72
fluoride, 74
Skaupy, 139
Slit system of mass-spectrograph, 49 Smith and Van Haagen, 72 SoDDY, 6, 8, 10, 11, 12, 13, 14, 16, 17, 35
and Hyman, 17, 121
Sodium, 86 Solubility of isotopes, 18
SOMMERFEIiD, 95
Spectra of isotopes, 9, 121,
Spectrum lines, form of, 53
Spencer, 91
Starkweather and Baxter, 141
Stas, 91
Statistical relation of isotopes, 109
Stewart, 11, 12
and James, 142
Sulphur, 76
Tani, Baxter and Chapin, 142 Tellurium, 77 Thermal diffusion, 129 Third order line of argon, 67
lines of, 61
Thomson, G. P., 86, 88
Sir J. J., 1, 22, 29, 33, 62, 70, 72, 75, 84, 91, 129 Thorite, 17, 18 Thorium, 7, 9, 14, 15, 18, 120
Thorium, chain, 17, 18, 116
,, atomic weight of, 18
Tin, 78 Travers, 39
and Ramsay, 33
Triatomic hydrogen, 70
Unitary theory of matter, 90 Uranium, 10, 120 ,, chain, 15
Valency electrons, 98
Van Haagen and Smith, 72
Wadsworth and Richards, 17 Watson, 33
and Aston, 24, 35
Weatherell, Baxter and Holmes,
73, 141 Welsbach, 8
Wendt and Landaueb, 70 Wey and Classen, 142 Whole number rule, 90 WiEN, 22
WiLLARD and McAlpine, 142 Willemite screens, 25 Wilson and Baxter, 142
and Harkins, 116
Xenon, 70
- anomalous atomic weight of, 114
X-ray spectra of isotopes, 1 1
Zechmeister and Hevesy, 20 Zinc, 147
