The radiations produced by uranium and thorium were quite feeble, and difficult to work with. This situation was corrected by Mme. Curie. Investigating the radioactivity of uranium minerals, she found some samples of ore of low uranium content that were nevertheless intensely radioactive - more so than if they had been pure uranium.
She reached the conclusion that the ore must contain some element other than uranium that was radioactive. Since she knew all the components of the ore that were present in significant amounts, and since all were known to be non-radioactive, the unknown element could be present only in very small quantities and must be extremely radioactive indeed.
During 1898, she and her husband slaved away over quantities of ore, trying to concentrate the radioactivity and isolate the new element. In July, one new element was located and named polonium, after Mme. Curie's native Poland, and, in December, a second new element, radium.
Radium was extremely radioactive, giving off radiations in 300,000 times the quantity that the same weight of uranium did. Furthermore, it was very rare. Out of tons of ore, the Curies managed to obtain only about 1/300th of an ounce of radium.
Other strongly radioactive elements were discovered in tiny traces. In 1899, the French chemist Andre Louis Debierne (1874-1949) discovered actinium. In 1900, the German physicist Friedrich Ernst Dorn (1848-1916) found a radioactive gas which eventually received the name radon. It was one of the inert gases, fitting below xenon in the periodic table. Finally, in 1917, the German chemist Otto Hahn (1879-1968) and Lise Meitner (1878-1968) discovered protactinium.
Experimenters could use these rare but extremely radioactive elements in "particle guns". Lead absorbs the radiation. If a bit of material containing one of these elements is placed in a lead-lined box with a hole in it, almost all the particles that come flying off are absorbed by the lead. Some, however, make their way through the hole to compose a tin stream of very many, very energetic particles which can be directed at some target.
It was Rutherford who used such "particle guns" most effectively. Beginning in 1906, he bombarded thin sheets of metal (such as gold) with speeding alpha particles. Most of the alpha particles passed clear through, unaffected and undiverted, recording themselves on the photographic plate behind. There were, however, some particles that were scattered - even through large angles.
Since the gold foil that served as target was two thousand atoms thick, and since most alpha particles passed through untouched, it would seem that the atoms were mostly empty space. Since some alpha particles were deflected sharply, it meant that somewhere in the atom must be a massive, positively charged region, capable of turning back the positively charged alpha particle.
Rutherford therefore evolved the theory of the nuclear atom. The atom, he decided, contains a very tiny nucleus at its center, which is positively charged and which contains all the the protons (and, it was later discovered, the neutrons, too) of the atom. The atomic nucleus has to be very tiny in order to account for the very small fraction of the alpha particles that were deflected, but it must also contain virtually all the mass of the atom.
In the outer regions of the atom are the negatively charged electrons, which are too light to interpose an important barrier to the passage of the alpha particles. Although the protons and alpha particles are as massive as atoms, they are actually bare atomic nuclei. They take up so little room in comparison with the atom that they, too, despite their large mass, may be considered subatomic particles.
Rutherford's nuclear atom lent a new subtlety to the question of the indivisibility of the atom. The central nucleus, which was the heart of the atom, was surrounded and protected by a cloud of electrons. It remained untouched and intact through all chemical changes. It was this seeming permanence of the nucleus that led all experimental evidence prior to the 1890's to appear to suggest the notion of an indivisible atom.
However, the atom did undergo one type of change in ordinary chemical reactions. Much of the electron cloud remained intact, but not all. Some electrons could be removed from the "surface" of the atom, or added to that surface. In this way, the problem of ions, which had puzzled three generations of chemists, was finally solved.
If the nuclear atom is accepted, the next question is: How does the nuclear atom of one element differ from that of another?
Since Dalton's time, different atoms had been known to differ in mass, but how is this difference reflected in the subatomic particles making up the nuclear atom?
The beginnings of an answer came through a study of x-rays. The German physicist Max Theodor Felix von Laue (1879-1960) began, in 1909, to bombard crystals with x-rays. These classic experiments established two vital facts: Crystals consist of atoms arranged in a geometrical structure of regular layers, and these layers scatter x-rays in a set pattern. From the manner in which the x-rays are scattered (or diffracted), the size of the tiny waves (wavelength) making up the x-rays can be determined.
Next, the English physicist Charles Glover Barkla (1877-1944) found, in 1911, that when x-rays are scattered by particular elements, they produce beams of x-rays that penetrate matter by characteristic amounts. Each element gives rise to a particular set of characteristic x-rays. Another English physicist, Henry Gwyn-Jeffreys Moseley (1887-1915), used Laue's method to determine the wavelengths of these characteristic x-rays. He found, in 1913, that the wavelength of these x-rays decreased smoothly with the increasing atomic weight of the elements emitting them. This inverse relationship, Moseley argued, depended on the size of the positive charge on the nucleus of the atom. The larger the charge, the shorter the wavelength of the characteristic x-rays.
From the wavelength, in fact, it was possible to calculate what the charge must be for the atoms of any particular element. Thus, as was eventually shown, hydrogen had a nuclear charge of +1, helium of +2, lithium of +3, and so on all the way up to +92 for uranium. (These numbers are based on a standard according to which the charge on a proton is arbitrarily set equal to +1, and that on an electron to -1.)
The size of the nuclear charge is called the atomic number. For the first time it was understood that when Mendeleev had arranged his elements in order of what was taken to be atomic weight, he really was arranging them in order of atomic number. In the couple of cases in which he had placed the more massive atoms ahead of less massive ones, the less massive one nevertheless had the larger atomic number for reasons which will shortly be discussed.
Now it was finally possible to replace Boyle's operational definition of an element, as a substance that could not be broken down into simpler substances, with a structural definition. The twentieth-century definition of an element would be: An element is a substance consisting of atoms that all possess an identical and characteristic atomic number.
For the first time it became possible to predict exactly how many elements remained to be discovered. All the atomic numbers from 1 to 92 were already occupied by known elements in 1913, except for seven - atomic numbers 43, 61, 72, 75, 85, 87, and 91. In 1917, protactinium (atomic number 91) was discovered. In 1923, hafnium (atomic number 72) was discovered, and in 1925, rhenium (atomic number 75). Exactly four gaps were left then in the periodic table: 43, 61, 85, and 87. Only four elements, it would seem, remained to be discovered; but those gaps remained well into the 1930's.
Since the proton is the only positively charged particle to be found in the nucleus, the atomic number is equal to the number of protons in the nucleus. Aluminum, with an atomic number of 13, has to contain 13 protons in its nucleus. But since its atomic weight is 27, it must also contain (as was later discovered) 14 neutrons in its nucleus. The neutrons contribute to the mass but not to the charge. In the same way, a sodium atom with an atomic number of 11 and an atomic weight of 23 must have a nucleus with 11 protons and 12 neutrons. (Because protons and neutrons are both found in the nucleus, they are lumped together as nucleons.)
The atom in its normal state is electrically neutral. This means that for every proton in the nucleus there must be an electron in the outskirts. Therefore, the number of electrons in the neutral atom is equal to the atomic number. A hydrogen atom contains 1 electron, a sodium atom 11 electrons, a uranium atom 92 electrons, and so on. (Of course, positive ions have lost electrons and negative ions have gained them. A sodium ion, therefore, has fewer electrons than its atomic number, while a chloride ion has more electrons than its atomic number.)
When two atoms collide and react, they either cling together, sharing a number of electrons, or separate again after having transferred one or more electrons from one atom to the other. It is this sharing of transferring of electrons that results in the changes of property noted in the substances undergoing chemical reactions.
A certain amount of order with respect to the manner in which such electrons changes occur began to appear from the careful work that was done with the characteristic x-rays. Out of this work arose the concept that the electrons in an atom existed in groups that might be pictured as electron shells. The shells can be visualized as enclosing the nucleus like the rings in an onion, each successive shell capable of holding more electrons than the ones within. The shells were lettered K, L, M, N, O, P, and Q.
The innermost shell, the K-shell, can hold only two electrons, the L-shell can hold eight, the M-shell as many as eighteen, and so on. This concept served to explain the periodic table.
The three electrons of the lithium atom are arranged 2,1 among the electron shells; the eleven electrons of the sodium atom are arranged 2,8,1; the nineteen electrons of the potassium atoms are arranged 2,8,8,1; and so on. Each of the alkali metals has the electrons of its atoms so arranged that the outermost occupied electron shell contains just one electron.
Since it is the outermost electron shell that makes contact in collisions between atoms, it is the number of electrons in that outermost shell that would be expected to determine the chemical activity of an element. Different elements with the outermost electron shells similar would have related properties. It is for this reason that the various alkali metals are so similar in their properties.
In the same way, the alkaline earth elements (magnesium, calcium, strontium, and barium) are all similar, for each possesses two electrons in the outermost shell. The halogens (fluorine, chlorine, bromine, and iodine) all possess seven electrons in the outermost shell; while the inert gases (neon, argon, krypton, and xenon) all possess eight.
Indeed, Mendeleev, in arranging his periodic table, had - without knowing it - placed the elements into rows and columns in accordance with the arrangement of their atoms among the electron shells.
As more and more electrons are to be found in the heavier atoms, the electron shells begin to overlap. Atoms of successive atomic numbers have added electrons to an inner shell, but the number of electrons in the outermost shell has remained constant. This configuration happens, in particular, with the rare earth elements, the atomic numbers of which range from 57 to 71 inclusive. While we find an increase in inner shell electrons as we go up the periodic table, all the rare earths retain three electrons in their outermost shell. That similarity of outermost shells explained, at last, why the elements of this group were so unexpectedly similar in their properties.
Mendeleev had arranged his periodic table by considering the valence of the different elements, rather than their electronic arrangements, which were unknown to him. It seemed reasonable to suppose that the valence of an element was governed by its electronic arrangement.
The German chemist Richard Abegg (1869-1910) had pointed out, in 1904, that the inert gases must have a particularly stable electronic configuration. The inert gas atoms had no tendency to add to or subtract from this number, and that was why they did not participate in chemical reactions. It followed that other atoms might give up or accept electrons in order to achieve the inert gas configuration.
Thus, sodium's eleven electrons are 2,8,1 while chlorine's seventeen electrons are 2,8,7. If sodium gives up an electron and chlorine accepts one, the former achieves the 2,8 configuration of neon and the latter the 2,8,8 configuration of argon.
The sodium atom, in giving up a negatively charged electron, is left with a positive charge and becomes the sodium ion. The chlorine atom in gaining an electron gains a negative charge and becomes the chloride ion. The two tend to cling together by virtue of electric attraction between positive and negative, as Berzelius had suspected a century earlier.
It is clear, from this consideration, why sodium should have a valence of 1. It cannot give up more than one electron without breaking up the stable 2,8 arrangement. Nor can the chlorine atom accept more than one electron. On the other hand, calcium, with a 2,8,8,2 arrangement, and thus a valence of 2, tends to give up two electrons, and oxygen, with a 2,6 arrangement, tends to accept two electrons.
It is these electron shifts that make it possible to set up concentrations of charge in one place or another, so that chemical reactions can serve as a source for electric current, as Volta had discovered over a century earlier.
From the electronic view, equivalent weight turned out to represent the relative weights of elements involved in a single electron shift of this sort. The equivalent weight is, after all, the atomic weight divided by the valence or, in other words, the atomic weight divided by the number of electrons transferred.
Abegg's suggestion only considered complete transfers of electrons from one atom to another, producing electrically charged ions which then held together by electrostatic attraction. This is electrovalence. Two American chemists, Gilbert Newton Lewis (1875-1946) and Irving Langmuir (1881-1957), independently extended this notion in the years following 1916. They suggested an explanation, for instance, for the structure of the chlorine molecule, in which two chlorine atoms are tightly bound together. Surely, there is not reason for one chlorine atom to transfer an electron to another chlorine atom, and surely they could not hold together by ordinary electrostatic attraction. Both Berzelius's and Abegg's theories of interatomic attraction fall short in this situation.
The Lewis-Langmuir suggestion was, instead, that each atom could contribute an electron to a shared pool. The two electrons in the shared pool remained in the outermost electron shell of both atoms. The electron arrangement in the chlorine molecule might therefore by pictured as 2,8,6,(1:1),6,8,2 with both shared electrons counting as part of the electron complement in each atom. Each atom would then have the 2,8,8 configuration in place of the much less stable 2,8,7 arrangement of the individual chlorine atoms. It is for that reason that the chlorine molecule is much more stable than are the free atoms.
In order to keep the electron pool in the outermost electron shell of both atoms, the two atoms had to remain in contact, and it takes considerable energy to tear them apart. Each electron contributed to such a pool represents a valence of 1 for the atom doing the contributing. Such valence, requiring the action of two atoms in cooperation, is covalence.
The Lewis-Langmuir theory was especially convenient for organic compounds, since the bonds between one carbon atom and another or between one carbon and a hydrogen atom were easily explained in this fashion. Most organic molecules could therefore easily be represented by electronic formulas where the old dash of the Kekule formula was replaced by a shared electron pair.
In fact, the English chemist Nevil Vincent Sidgwick (1873-1952) was able, in the 1920's, to extend the notion of electron-pair covalence to inorganic compounds. In particular, he applied them to Werner's coordination compounds where the ordinary Kekule representations were difficult to apply.
In all these chemical changes only electrons are being shifted. The protons (in all but one case) are safely protected in the central nucleus. The exceptional case is that of hydrogen, which has a nucleus made up of a single proton. If the hydrogen atom is ionized through removal of its single electron, the proton is left bare. (Such a bare proton is very active and does not remain bare for long. In water solution, it immediately attaches itself to a water molecule, adding a positively charged hydrogen atom to that molecule. Thus is formed the hydronium ion, H3O+.)
In 1923, the Danish chemist Johannes Nicolaus Bronsted (1879-1947) introduced a new view of acids and bases. An acid was defined as a compound tending to give up a proton (or hydrogen ion), while a base was one tending to combine with a proton. This new view accounted for all the facts already satisfactorily accounted for by the old view of Svante Arrhenius. In addition it represented a greater flexibility that made it possible to extend acid-base notions into areas in which the old view was inadequate.
The relatively small molecules and rapid, ionic reactions in inorganic chemistry had proven comparatively easy to study. Chemists, from Lavoisier's time onward, could predict the course of such reactions and the manner of modifying them to suit needs. The complicated molecules and slow reactions in organic chemistry were much harder to analyze. Often there were several ways in which two substances could react; guiding the reaction into some desired path was a matter of art and intuition rather than of secure knowledge.
The electronic atom offered organic chemists a new look at their field. In the late 1920's, men such as the English chemist Christopher Ingold (1893- ) began to try to interpret organic reactions in terms of electron shifts from point to point within a molecule. The methods of physical chemistry began to be applied intensively in an attempt to interpret the directions and tendencies of such shifts. Physical organic chemistry became an important discipline.
In proved insufficient to attempt to interpret organic reactions in terms of hard little electrons moving here and there, and it did not long remain necessary to do so.
For the first quarter-century after the discovery of the electron, it was taken for granted that the particle was a tiny, hard sphere. In 1923 Louis Victor de Broglie, Prince de Broglie, a French physicist (1892-1987), had presented theoretical reasons for considering electrons (and all other particles as well) to possess wave properties. Before the end of the 1920's this view had been confirmed by experiment.
Pauling (the first to suggest the helical shape of proteins and nucleic acids), developed methods, in the early 1930's, for taking into account the wave nature of electrons in considering organic reactions. He showed that the Lewis-Langmuir electron pool could be interpreted as wave-interactions. Electron waves paired off in reinforcement, resonating with each other to form a stabler situation in combination than in separation.
This theory of resonance was particularly useful in establishing the structure of benzene, which had been puzzling in Kekule's day and which had retained questionable points ever since. As usually drawn, the structure of benzene is that of a hexagon with alternating single and double bonds. By the Lewis-Langmuir system, two-electron pools and four-electron pools alternated. Benzene lacked almost completely the characteristic properties of other compounds which contained double bonds, or four-electron pools.
Pauling showed that if electrons were regarded as wave-forms, the individual electrons need not be considered as occupying a single point, but could "smear out" over a considerable area. The "electron waves" could spread out to take up far larger areas than a tiny "billiard ball" electron could be expected to take up. The tendency to "smear" in this fashion was accentuated if a molecule was quite flat and symmetrical.
The benzene molecule is flat and symmetrical, and Pauling showed that the electrons "smeared out" in such a fashion that all six carbon atoms of the benzene ring were found in equal fashion. The bonds connecting them could not be represented as either single bonds or double bonds, but as a kind of particularly stable average, or resonance hybrid, between the two extremes.
Other points besides the structure of benzene were clarified by the theory of resonance. For instance, the four electrons in the outermost shell of the carbon atom are not all equivalent from the standpoint of energy characteristics. It might have been assumed, then, that bonds of slightly different type would be formed between a carbon atom and its neighbor, depending on which of carbon's electrons was involved.
It could be shown, though, that the four electrons, as wave-forms, interacted and formed four "average" bonds that were precisely equivalent, and directed toward the apices of a tetrahedron. Thus, the Van't Hoff-Le Bel tetrahedral atom was explained in electronic terms.
Resonance also helped to explain a group of strange compounds that had first impinged on the chemical consciousness at the opening of the twentieth century. In 1900, the Russian-American chemist Moses Gomberg (1866-1947) was trying to prepare hexaphenylethane, a compound with a molecule consisting of two carbon atoms to which six benzene rings were attached (three per carbon atom).
He obtained, instead, a colored solution of a very reactive compound. For various reasons, he was forced to conclude that he had obtained triphenylmethyl, a "half-molecule" consisting of a carbon atom with three benzene rings attached. The fourth valence bond of the carbon atom remained unused. Such a compound resembled one of the old radicals torn loose from a molecule. It was therefore termed a free radical.
Once the electronic atom was introduced, a free radical such as triphenylmethyl was understood to contain an unpaired electron in the place where the old Kekule view would have put an unused bond. Ordinarily, such an unpaired electron is highly unstable. However, if a molecule is flat and highly symmetrical, as triphenylmethyl is, the unused electron can be "smeared out" over the entire molecule. The free radical is then stabilized.
When organic reactions came to be studied in electronic terms, it became clear that there were usually stages where a free radical had to be formed. Such free radicals, generally not stabilized by resonance, could exist only momentarily and could be formed only with difficulty. It was the difficulty of forming free radical intermediates that made the most organic reactions so slow.
In the second quarter of the twentieth century, organic chemists were beginning to get considerable insight into the detailed steps that made up organic reactions - the reaction mechanism, in other words. It was this insight, more than anything else, which has guided contemporary organic chemists in their synthetic work and has led to the syntheses of molecules whose complexities had defeated earlier generations.
Nor were resonance considerations confined to organic chemistry alone. The boron hydrides possessed molecules that could not be neatly represented by older views. The boron atom possessed too few valence bonds (or electrons) for the purpose. Yet if the electrons were properly "smeared" as wave forms, a reasonable molecular structure could be proposed.
Again, in 1932, Pauling reasoned that the inert gas atoms could not be as resistant to forming bonds as had been assumed for the third of a century that had elapsed since their discovery. Under sufficient pressure by an extremely reactive atom such as that of fluorine, compounds might be formed.
This suggestion of Pauling's went unheeded at first, but in 1962, xenon fluoride was formed by reacting the inert gas xenon with fluorine. In short order a number of xenon compounds with fluorine and with oxygen were formed, as well as one or two of radon and of krypton.
If the studies of the internal atomic structure had led to new insights and understandings, they also posed a normal share of new problems.
In 1900, Crookes had discovered that freshly prepared pure uranium compounds were only very slightly radioactive, but that their radioactivity strengthened on standing. By 1902 Rutherford and a co-worker, the English chemist Frederick Soddy (1877-1956), proposed that as a uranium atom gave off an alpha particle, its nature changed. It became a new type of atom, with different radioactive characteristics, producing stronger radiations than uranium itself (thus accounting for Crooke's observations).
This second atom in turn broke down, forming still another type of atom. Indeed, the uranium atom was the parent of a whole series of radioactive elements, a radioactive series, that included radium and polonium and ended finally with lead, which was not radioactive. It was for this reason that radium, polonium, and other rare radioactive elements could be found in uranium minerals. A second radioactive series also began with uranium, while a third series began with thorium.
(This breakdown of uranium into lead would, by Boyle's definition of elements, have made it necessary to view uranium as not being an element. By the new atomic number definition it was still an element. It is just that since atoms are not really indivisible particles after all, elements are not necessarily entirely unchangeable. This represents a return - on a much higher level of sophistication - to the old alchemical concept.
It is reasonable to ask why, though, if radioactive elements are constantly breaking down, any remained in existence at all? It was Rutherford who, in 1904, solved the matter. In studying the rate of radioactive breakdown, he was able to show that after a certain period, which was different for each element, half of any given quantity of a certain radioactive element would have broken down. This period, which is characteristic for each particular type of radioactive substance, Rutherford called the half-life.
The half-life of radium is just under 1600 years. Over the geological eras any radium in the earth's crust would certainly have long since vanished, were it not that new supplies are constantly being formed through the breakdown of uranium. The same is true for other breakdown products of uranium, some of which have half-lives of only fractions of a second.
The half-life of uranium is 4.5 X 109 years. This is a tremendous period of time, and in all the history of the earth, only a fraction of the original supply of uranium has had a chance to break down. Thorium breaks down even more slowly, its half-life being 1.4 X 1010 years.
Such huge stretches of time can be determined by counting the number of alpha particles produced by a given mass of uranium (or thorium). The alpha particles were counted by Rutherford, by noting the small flashes they made when they struck a screen of zinc sulfide. (This was a scintillation counter.)
Each alpha particle given off meant a uranium atom breaking down so that Rutherford could determine how many atoms were breaking down per second. From the mass of the uranium he was dealing with, he knew the total number of uranium atoms present. With this information, he could easily calculate how long it would take for half the uranium atoms present to break down, and it turned out to be a matter of billions of years.
So constant and characteristic is the majestically slow decay of uranium that it can be used to measure the age of the earth. In 1907, the American chemist Bertram Borden Boltwood (1870-1927) suggested that the lead content of uranium minerals would serve as guide in this respect. If it is assumed that all the lead in the mineral originated from uranium decay, it would be easy to calculate how long a time must have elapsed to bring that amount of lead into existence. It was eventually calculated in this way that the solid crust of the earth must have been in existence for at least four billion years.
Meanwhile, Soddy had gone on to describe the exact manner in which an atom changed as it gave off subatomic particles. If an atom lost an alpha particle, with a charge of +2, the total charge on its nucleus was decreased by two. The atom moved two places to the left in the periodic table.
If an atom lost a beta particle (an electron with a charge of -1), the nucleus gained an additional positive charge (In Soddy's time, it was felt that there were electrons in the nucleus and that the loss of a beta particle from the nucleus left an additional proton unbalanced, hence raised the positive charge. Nowadays, it is felt that the nucleus contains only protons and neutrons, but that an electron is formed and expelled when a neutron is converted into a proton, for the gain of a positive charge is equivalent to the loss, by expulsion, of a negative charge.) and the element moved one place to the right in the periodic table. If an atom lost a gamma ray (uncharged), its energy content was altered but there was no change in its particle makeup, so that it remained the same element.
Using these rules as a guide, chemists could work out the details of the various radioactive series.
But all this raised a serious problem. What was one to do with the various breakdown products of uranium and thorium? Dozens of these were discovered, but there were at most only nine places in the periodic table (from polonium at atomic number 84 to uranium at atomic number 92) in which to place them.
As a specific example, the uranium atom (atomic number 92) emitted an alpha particle and the atomic number of what was left of the atom therefore became 90, by Soddy's rule. This meant that a thorium atom had been formed. However, whereas ordinary thorium had a half-life of 14 billion years, the thorium produced from uranium had a half-life of 24 days.
Differences existed even in the case of non-radioactive elements and in properties not involving radioactivity. For instance, Richards (the expert on atomic weights) was able to show, in 1913, that the lead produced by the decay of uranium did not have quite the same atomic weight as ordinary lead.
Soddy advanced the bold suggestion that more than one kind of atom could fit into the same place in the periodic table. Place number 90 might hold different varieties of thorium, place number 82 different varieties of lead, and so on. He called these atom-varieties occupying the same place isotopes, from the Greek word meaning "same place".
The different isotopes in a given place in the table would have the same atomic number, therefore the same number of protons in the nucleus and the same number of electrons in the outskirts. The isotopes of an element would have the same chemical properties, since these properties depend on the number and arrangement of the electrons in the atoms.
But in that case, how does one explain differences in radioactive properties and in atomic weight?
Atomic weight might represent the key to the difference. A hundred years earlier, Prout had advanced his famous hypothesis that all atoms are composed of hydrogen so that all elements should have integral atomic weight. The fact that most atomic weights are not integers seemed to have destroyed his hypothesis.
But now the atom, in its new nuclear guise, had to be made up of protons (and neutrons). Protons and neutrons are about equally massive, and therefore, all atoms had to have weights that were integral multiples of the weight of hydrogen (made up of a single proton). Prout's hypothesis was reinstated, and a look of new suspicion was directed at the atomic weights instead.
In 1912, J. J. Thomson (the discoverer of the electron) had subjected beams of positively charged neon ions to the action of a magnetic field. The field deflected the neon ions and caused them to fall on a photographic plate. If all the ions had been identical in mass they would all have been deflected by the same amount, and a single discolored spot on the photographic film would have appeared. However, two spots were located, one some ten times as dark as the other. A co-worker, Francis William Aston (1877-1945), later improved the device and confirmed the results. Similar results were uncovered for other elements. Since this device separated chemically similar ions into a kind of spectrum of dark spots, it was called the mass spectrograph.
The extent of deflection of ions of identical charge by a magnetic field depends upon the mass of the ion; the more massive the ion, the less it is deflected. From the results obtained by Thomson and Aston it would seem that there were two kinds of neon atoms, one more massive than the other. One type had a mass number of 20 and the other, a mass of 22. Since the neon-20 was ten times as common as neon-22, judging from the relative darkness of the spots (in later years tiny quantities of neon-21 were also isolated), it was reasonable that the atomic weight of neon was about 20.2.
In other words, individual atoms had masses that were an integral multiple of that of the hydrogen atom, (Not quite a multiple, in actual fact. The small deviations in mass are of no importance in chemistry but are a reflection of the huge energies involved in nuclear forces - energies that have been made manifest in nuclear bombs.) but a particular element, being made up of atoms of different mass, would have an atomic weight that was a weighted average of these integers and would therefore not necessarily be an integer itself.
The weighted average of the isotopes of a particular atom may be greater, in some cases, than the weighted average for an atom of higher atomic number.
For instance, tellurium, with an atomic number of 52, consists of seven isotopes. Of these, the two most massive isotopes, tellurium-126 and tellurium-128, are the most common. The atomic weight of tellurium therefore comes to 127.6. Iodine has the next higher atomic number, 53, but it is made up of iodine-127 only and therefore has the atomic weight of 127. When Mendeleev placed iodine after tellurium in his periodic table, reversing the order dictated by atomic weight, he was, without knowing it, following atomic number instead; and this was the correct thing to do.
Here's another example. Potassium (atomic number 19) is made up of three isotopes, potassium-39, potassium-40, and potassium-41, but the lightest isotope, potassium-39, is by far the most common. Hence, the atomic weight of potassium is 39.1. Argon has a lower atomic number (18) and is made up of three isotopes also, argon-36, argon-38, and argon-40. Here, however, it is the most massive isotope, argon-40, which is most common. Therefore the atomic weight of argon is about 40. When Ramsay placed argon before potassium instead of after in defiance of atomic weights, he, too, without knowing it, was following atomic number and was doing the correct thing.
The use of the mass spectrograph made it possible to determine atomic weight by actually measuring the mass of the individual isotopes and the quantity of each present - and then taking the average. This method surpassed chemical methods for measuring atomic weight in accuracy.
Different isotopes of a given element have the same atomic number but different mass numbers. The different isotopes would have the same number of protons in their nucleus but different numbers of neutrons. Thus, neon-20, neon-21, and neon-22 all have 10 protons in the nucleus, so that all have an atomic number 10, and all have an electron arrangement of 2,8. However, neon-20 has a nucleus containing 10 protons plus 10 neutrons; neon-21, one containing 10 protons plus 11 neutrons; and neon-22, one containing 10 protons plus 12 neutrons.
Most elements (but not all) could be divided into isotopes in this manner. In 1935, the Canadian-American physicist Arthur Jeffrey Dempster (1886-1950) found that uranium, as it occurred in nature, was a mixture of two isotopes even though its atomic weight (238.07) was close to a whole number. It was just that one isotope existed in overwhelming proportion. Fully 99.3 per cent of the uranium atoms had nuclei made up of 92 protons and 146 neutrons or a total mass number of 238. These were uranium-238 atoms. The remaining 0.7 per cent had three fewer neutrons and were uranium-235 atoms.
Since radioactive properties depend upon the constitution of the atomic nucleus, and not upon electron arrangement, the isotopes of an element might be similar chemically, but quite different from the standpoint of radioactivity. Thus, whereas uranium-238 has a half-life of 4.5 X 1010 years, that of uranium-235 was only 7.0 X 108 years. (This accounts also for the difference, mentioned earlier in the half-lives of natural thorium (thorium-232) and the thorium formed from the breakdown of uranium (thorium-234) which contains two additional neutrons in each nucleus.) Both are parents of separate radioactive series as well.
There were theoretical reasons for suspecting that hydrogen itself, the simplest element, might be made up of a pair of isotopes. Ordinary hydrogen atoms, with nuclei composed of a single proton, make up hydrogen-1. In 1931 the American chemist Harold Clayton Urey (1893-1981) slowly evaporated four liters of liquid hydrogen on the presumption that if any heavier isotope of hydrogen existed, it would have a higher boiling point and would evaporate more slowly. This meant it would remain behind and accumulate in the residue.
Sure enough, in the final cubic centimeter of hydrogen Urey was able to detect unmistakable signs of the existence of hydrogen-2, the nucleus of which consisted of one proton plus one neutron. Hydrogen-2 received the special name of deuterium.
Nor was oxygen immune. In 1929, the American chemist William Francis Giauque (1895-1982) succeeded in showing that oxygen was made up of three isotopes. The most common variety, comprising nearly 99.8 per cent of all the atoms, was oxygen-16. It nucleus contained 8 protons plus 8 neutrons. The rest were almost all oxygen-18 (8 protons plus 10 neutrons) with a trace of oxygen-17 (8 protons plus 9 neutrons).
This created a problem. Ever since the days of Berzelius, the atomic weights had been based on the arbitrary assignment of a weight of 16.0000 to the oxygen atom. But the atomic weight of oxygen could be only the weighted average of the three isotopes, and the proportion of the isotopes in oxygen might vary slightly from sample to sample.
The physicists took to determining atomic weights on the basis of oxygen-16 set equal to 16.0000, and this gave them a series of values (the physical atomic weight) that were uniformly greater, by a very small amount, than the values that had been used and gradually improved throughout the nineteenth century (the chemical atomic weights).
In 1961, however, international organizations of both chemists and of physicists agreed to adopt an atomic weight standard based on carbon-12 set equal to exactly 12.0000. This new standard was almost exactly that of the old chemical atomic weights and yet it was tied to a single isotope and not to the average of a group of them.