History of the periodic table

None of the proposals was accepted immediately, and many contemporary chemists found it too abstract to have any meaningful value. Of those chemists that proposed their categorizations, Mendeleev stood out as he strove to back his work and promote his vision of periodicity. In contrast, Meyer did not promote his work very actively, and Newlands did not make a single attempt to gain recognition abroad.

In addition to the predictions of scandium, gallium, and germanium that were quickly realized, Mendeleev's 1871 table left many more spaces for undiscovered elements, though he did not provide detailed predictions of their properties. In total, he predicted eighteen elements, though only half corresponded to elements that were later discovered.[30]

Even as Mendeleev corrected positions of some elements, he thought that some relationships that he could find in his grand scheme of periodicity could not be found because some elements were still undiscovered, and thus he believed these elements that were still undiscovered would have properties that could be deduced from the expected relationships with other elements. In 1870, he first tried to characterize the yet undiscovered elements, and he gave detailed predictions for three elements, which he termed eka-boron, eka-aluminium, and eka-silicium,[31] as well as more briefly noted a few other expectations.[32] In 1871, he expanded his predictions further.

Compared to the rest of the work, Mendeleev's 1869 list misplaces seven then-known elements: indium, thorium, and the five rare earth metals—yttrium, cerium, lanthanum, erbium, and didymium (the latter two were later found to be mixtures of different elements); ignoring those would allow him to restore the logic of increasing atomic weight. These elements (all thought to be divalent at the time) puzzled Mendeleev in that they did not show gradual increase in valency despite their seemingly consequential atomic weights.[33] Mendeleev grouped them together, thinking of them as of a particular kind of series.[lower-alpha 2] In early 1870, he decided the weights for these elements must be wrong and that the rare earth metals should be trivalent (which accordingly increases their weights by half). He measured heat capacity of indium, uranium, and cerium to demonstrate their increase of accounted valency (which was soon confirmed by Prussian chemist Robert Bunsen).[34] Mendeleev considered the change by assessing each element an individual place in his system of the elements.

Mendeleev noticed that there was a significant difference in atomic mass between cerium and tantalum with no element between them; his consideration was that between them, there was a row of yet undiscovered elements, which would display similar properties to those elements which were to be found above and below them: for instance, an eka-molybdenum would behave as a heavier homolog of molybdenum and a lighter homolog of wolfram (the name under which Mendeleev knew tungsten).[35] This row would begin with a trivalent lanthanum, a teravalent cerium, and a pentavalent didymium. However, the higher valency for didymium had not been established, and Mendeleev tried to do that himself.[36] Having had no success in that, he abandoned his attempts to incorporate the rare earth metals in late 1871 and embarked on his grand idea of luminiferous ether. His idea was carried on by Austrian-Hungarian chemist Bohuslav Brauner, who sought to find a place in the periodic table for the rare earth metals;[37] Mendeleev later referred to him as to "one of the true consolidators of the periodic law".[lower-alpha 3]

Eventually, the periodic table was appreciated for its descriptive power and for finally systematizing the relationship between the elements,[39] although such appreciation was not universal.[40] In 1881, Mendeleev and Meyer had an argument via an exchange of articles in British journal Chemical News over priority of the periodic table, which included an article from Mendeleev, one from Meyer, one of critique of the notion of periodicity, and many more.[41] In 1882, the Royal Society in London awarded the Davy Medal to both Mendeleev and Meyer for their work to classify the elements; although two of Mendeleev's predicted elements had been discovered by then, Mendeleev's predictions were not at all mentioned in the prize rationale.

Mendeleev's eka-aluminium was discovered in 1875 and became known as gallium; eka-boron and eka-silicium were discovered in 1879 and 1886, respectively, and were named scandium and germanium.[13] Mendeleev was even able to correct some initial measurements with his predictions, including the first prediction of gallium, which matched eka-aluminium fairly closely but had a different density. Mendeleev advised the discoverer, French chemist Paul Emile Lecoq de Boisbaudran, to measure the density again; de Boisbaudran was initially skeptical (not least because he thought Mendeleev was trying to take credit from him) but eventually admitted correctness of the prediction. Mendeleev contacted all three discoverers; all three noted the close similarity of their discovered elements with Mendeleev's predictions, with the last of them, German chemist Clemens Winkler, doing so prior to correspondence with Mendeleev. Some contemporary chemists were not convinced by these discoveries, noting the dissimilarities between the new elements and the predictions or claiming those similarities that did exist were coincidental,[40] but later chemists used the successes of these Mendeleev's predictions to justify his table.[10]

After the three discoveries closely matched Mendeleev's three detailed predictions, his table was widely recognized. In 1889, Mendeleev noted at the Faraday Lecture to the Royal Institution in London that he had not expected to live long enough "to mention their discovery to the Chemical Society of Great Britain as a confirmation of the exactitude and generality of the periodic law".[42]

British chemist Henry Cavendish, the discoverer of hydrogen in 1766, discovered that air is composed of more gases than nitrogen and oxygen.[44] He recorded these findings in 1784 and 1785; among them, he found a then-unidentified gas less reactive than nitrogen. Helium was first reported in 1868; the report was based on the new technique of spectroscopy and some spectral lines emitted by the Sun did not match those of any of the known elements. Mendeleev was not convinced by this finding since variance of temperate led to change of intensity of spectral lines and their location on the spectrum;[45] this opinion was held by some other scientists of the day. Others believed the spectral lines could belong to an element that occurred on the Sun but not Earth; some believed it was yet to be found on Earth.

In 1894, British chemist William Ramsay and British physicist Lord Rayleigh isolated argon from air and determined that it was a new element. Argon, however, did not engage in any chemical reactions and was—highly unusually for a gas—monatomic;[lower-alpha 4] it did not fit into the periodic law and thus challenged the very notion of it. Not all scientists immediately accepted this report; Mendeleev's original response to that was that argon was a triatomic form of nitrogen rather than an element of its own.[47] The next year, Ramsay tested a report of American chemist William Francis Hillebrand, who found a steam of an unreactive gas from a sample of uraninite. Wishing to prove it was nitrogen, Ramsay analyzed a different uranium mineral, cleveite, and found a new element, which he named krypton. This finding was corrected by British chemist William Crookes, who matched its spectrum to that of the Sun's helium.[48] Following this discovery, Ramsay, using fractional distillation to separate air, discovered several more such gases in 1898: metargon, krypton, neon, and xenon; detailed spectroscopic analysis of the first of these demonstrated it was argon contaminated by a carbon-based impurity. Ramsay's remaining five unreactive substances were called the inert gases (now noble gases). Although Mendeleev's table predicted several undiscovered elements, it did not predict the existence of such inert gases, and Mendeleev originally rejected those findings as well.[49]

In 1898, when only helium, argon, and krypton were definitively known, Crookes suggested these elements be placed between the hydrogen group and the fluorine group.[50] In 1900, at the Prussian Academy of Sciences, Ramsay and Mendeleev discussed the new inert gases and their location in the periodic table; Ramsay proposed that these elements be put in a brand new group 0, to which Mendeleev agreed.[51] Two weeks before that discussion, Belgian botanist Léo Errera proposed the same idea, that of a group 0, to the Royal Academy of Science, Letters and Fine Arts of Belgium; after Mendeleev learned about this proposal, he referred to Errera as to the first person to suggest the idea.[51] Mendeleev himself added these elements to the table as group 0 in 1902, without disturbing the basic concept of the periodic table.[51][52]

In 1905, Swiss chemist Alfred Werner resolved the dead zone of Mendeleev's table. He determined that the rare earth elements (lanthanides), 13 of which were known, lay within that gap. Although Mendeleev knew of lanthanum, cerium, and erbium, they were previously unaccounted for in the table because their total number and exact order were not known; Mendeleev still could not fit them in his table by 1901.[49] This was in part a consequence of their similar chemistry and imprecise determination of their atomic masses. Combined with the lack of a known group of similar elements, this rendered the placement of the lanthanides in the periodic table difficult.[53] This discovery led to a restructuring of the table and the first appearance of the 32-column form.[54]

By 1904, Mendeleev's table rearranged several elements, and included the noble gases along with most other newly discovered elements. It still had the dead zone, and a row zero was added above hydrogen and helium to include coronium and the ether, which were widely believed to be elements at the time.[54] Although the Michelson-Morley Experiment in 1887 cast doubt on the possibility of a luminiferous ether as a space-filling medium, physicists set constraints for its properties.[55] Mendeleev believed it to be a very light gas, with an atomic weight several orders of magnitude smaller than that of hydrogen. He also postulated that it would rarely interact with other elements, similar to the noble gases of his group zero, and instead permeate substances at a velocity of 2,250 kilometres (1,400 mi) per second.

Mendeleev was not satisfied with the lack of understand of the nature of this periodicity; this would only be possible with the understanding of composition of atom. However, Mendeleev firmly believed that future would only develop the notion rather than challenge it and reaffirmed his belief in writing in 1902.[56]

• Early developments of Mendeleev's table
• 
  Mendeleev's 1904 table. It includes the noble gases in group 0, and scandium, gallium, germanium, and radium are added. It has gaps in row 0 (hypothesized elements lighter than hydrogen) and row 9 (lanthanides).
• 
  Werner's 32-column 1905 table. This table left spaces for many then-unknown elements, and several elements had their positions revised following advances in atomic theory.

Four radioactive elements were known in 1900: radium, actinium, thorium, and uranium. These radioactive elements (termed "radioelements") were accordingly placed at the bottom of the periodic table, as they were known to have greater atomic weights than stable elements, although their exact order was not known. Researchers believed there were still more radioactive elements yet to be discovered, and during the next decade, the decay chains of thorium and uranium were extensively studied. Many new radioactive substances were found, including the noble gas radon, and their chemical properties were investigated.[13] By 1912, almost 50 different radioactive substances had been found in the decay chains of thorium and uranium. American chemist Bertram Boltwood proposed several decay chains linking these radioelements between uranium and lead. These were thought at the time to be new chemical elements, substantially increasing the number of known "elements" and leading to speculations that their discoveries would undermine the concept of the periodic table.[30] For example, there was not enough room between lead and uranium to accommodate these discoveries, even assuming that some discoveries were duplicates or incorrect identifications. It was also believed that radioactive decay violated one of the central principles of the periodic table, namely that chemical elements could not undergo transmutations and always had unique identities.[13]

Frederick Soddy and Kazimierz Fajans found in 1913 that although these substances emitted different radiation,[57] many of these substances were identical in their chemical characteristics, so shared the same place on the periodic table.[58][59] They became known as isotopes, from the Greek isos topos ("same place").[13][60] Austrian chemist Friedrich Paneth cited a difference between "real elements" (elements) and "simple substances" (isotopes), also determining that the existence of different isotopes was mostly irrelevant in determining chemical properties.[30]

Following British physicist Charles Glover Barkla's discovery of characteristic X-rays emitted from metals in 1906, British physicist Henry Moseley considered a possible correlation between x-ray emissions and physical properties of elements. Moseley, along with Charles Galton Darwin, Niels Bohr, and George de Hevesy, proposed that the nuclear charge (Z) or atomic mass may be mathematically related to physical properties.[61] The significance of these atomic properties was determined in the Geiger-Marsden experiment, in which the atomic nucleus and its charge were discovered.[62]

In 1913, amateur Dutch physicist Antonius van den Broek was the first to propose that the atomic number (nuclear charge) determined the placement of elements in the periodic table. He correctly determined the atomic number of all elements up to atomic number 50 (tin), though he made several errors with heavier elements. However, Van den Broek did not have any method to experimentally verify the atomic numbers of elements; thus, they were still believed to be a consequence of atomic weight, which remained in use in ordering elements.[61]

Moseley was determined to test Van den Broek's hypothesis.[61] After a year of investigation of the Fraunhofer lines of various elements, he found a relationship between the X-ray wavelength of an element and its atomic number.[63] With this, Moseley obtained the first accurate measurements of atomic numbers and determined an absolute sequence to the elements, allowing him to restructure the periodic table. Moseley's research immediately resolved discrepancies between atomic weight and chemical properties, where sequencing strictly by atomic weight would result in groups with inconsistent chemical properties. For example, his measurements of X-ray wavelengths enabled him to correctly place argon (Z = 18) before potassium (Z = 19), cobalt (Z = 27) before nickel (Z = 28), as well as tellurium (Z = 52) before iodine (Z = 53), in line with periodic trends. The determination of atomic numbers also clarified the order of chemically similar rare earth elements; it was also used to confirm that Georges Urbain's claimed discovery of a new rare earth element (celtium) was invalid, earning Moseley acclamation for this technique.[61]

Swedish physicist Karl Siegbahn continued Moseley's work for elements heavier than gold (Z = 79), and found that the heaviest known element at the time, uranium, had atomic number 92. In determining the largest identified atomic number, gaps in the atomic number sequence were conclusively determined where an atomic number had no known corresponding element; the gaps occurred at atomic numbers 43, 61, 72, 75, 85, and 87.[61]

In 1914, Swedish physicist Johannes Rydberg noticed that the atomic numbers of the noble gases was equal to doubled sums of squares of simple numbers: 2 = 2·1², 10 = 2(1² + 2²), 18 = 2(1² + 2² + 2²), 36 = 2(1² + 2² + 2² + 3²), 54 = 2(1² + 2² + 2² + 3² + 3²), 86 = 2(1² + 2² + 2² + 3² + 3² + 4²). This finding was accepted as an explanation of the fixed lengths of periods and led to repositioning of the noble gases from the left edge of the table to the right.[51] Unwillingness of the noble gases to engage in chemical reaction was explained in the alluded stability of closed noble gas electron configurations; from this notion emerged the octet rule.[51] Among the notable works that established the importance of the periodicity of eight were the valence bond theory, published in 1916 by American chemist Gilbert N. Lewis[64] and the octet theory of chemical bonding, published in 1919 by American chemist Irving Langmuir.[65][66]

In the 1910s and 1920s, pioneering research into quantum mechanics led to new developments in atomic theory and small changes to the periodic table. The Bohr model was developed during this time, and championed the idea of electron configurations that determine chemical properties. Bohr proposed that elements in the same group behaved similarly because they have similar electron configurations, and that noble gases had filled valence shells;[67] this forms the basis of the modern octet rule. This research then led Austrian physicist Wolfgang Pauli to investigate the length of periods in the periodic table in 1924. Mendeleev asserted that there was a fixed periodicity of eight, and expected a mathematical correlation between atomic number and chemical properties;[68] Pauli demonstrated that this was not the case. Instead, the Pauli exclusion principle was developed. This states that no electrons can coexist in the same quantum state, and showed, in conjunction with empirical observations, the existence of four quantum numbers and the consequence on the order of shell filling.[67] This determines the order in which electron shells are filled and explains the periodicity of the periodic table.

British chemist Charles Bury is credited with the first use of the term transition metal in 1921 to refer to elements between the main-group elements of groups II and III. He explained the chemical properties of transition elements as a consequence of the filling of an inner subshell rather than the valence shell. This proposition, based upon the work of American chemist Gilbert N. Lewis, suggested the appearance of the d subshell in period 4 and the f subshell in period 6, lengthening the periods from 8 to 18 and then 18 to 32 elements.[69]

As early as 1913, Bohr's research on electronic structure led physicists such as Rydberg to extrapolate the properties of undiscovered elements heavier than uranium. Many agreed that the next noble gas after radon would most likely have the atomic number 118, from which it followed that the transition series in the seventh period should resemble those in the sixth. Although it was thought that these transition series would include a series analogous to the rare earth elements, characterized by filling of the 5f shell, it was unknown where this series began. Predictions ranged from atomic number 90 (thorium) to 99, many of which proposed a beginning beyond the known elements (at or beyond atomic number 93). The elements from actinium to uranium were instead believed to form part of a fourth series of transition metals because of their high oxidation states; accordingly, they were placed in groups 3 through 6.[71]

In 1940, neptunium and plutonium were the first transuranic elements to be discovered; they were placed in sequence beneath rhenium and osmium, respectively. However, preliminary investigations of their chemistry suggested a greater similarity to uranium than to lighter transition metals, challenging their placement in the periodic table.[72] During his Manhattan Project research in 1943, American chemist Glenn T. Seaborg experienced unexpected difficulties in isolating the elements americium and curium, as they were believed to be part of a fourth series of transition metals. Seaborg wondered if these elements belonged to a different series, which would explain why their chemical properties, in particular the instability of higher oxidation states, were different from predictions.[72] In 1945, against the advice of colleagues, he proposed a significant change to Mendeleev's table: the actinide series.[71][73]

Seaborg's actinide concept of heavy element electronic structure proposed that the actinides form an inner transition series analogous to the rare earth series of lanthanide elements—they would comprise the second row of the f-block (the 5f series), in which the lanthanides formed the 4f series. This facilitated chemical identification of americium and curium,[73] and further experiments corroborated Seaborg's hypothesis; a spectroscopic study at the Los Alamos National Laboratory by a group led by American physicist Edwin McMillan indicated that 5f orbitals, rather than 6d orbitals, were indeed being filled. However, these studies could not unambiguously determine the first element with 5f electrons and therefore the first element in the actinide series;[72] it was thus also referred to as the "thoride" or "uranide" series until it was later found that the series began with actinium.[71][74]

In light of these observations and an apparent explanation for the chemistry of transuranic elements, and despite fear among his colleagues that it was a radical idea that would ruin his reputation, Seaborg nevertheless submitted it to Chemical and Engineering News and it gained widespread acceptance; new periodic tables thus placed the actinides below the lanthanides.[73] Following its acceptance, the actinide concept proved pivotal in the groundwork for discoveries of heavier elements, such as berkelium in 1949.[75] It also supported experimental results for a trend towards +3 oxidation states in the elements beyond americium—a trend observed in the analogous 4f series.[71]

Seaborg's subsequent elaborations of the actinide concept theorized a series of superheavy elements in a transactinide series comprising elements from 104 to 121 and a superactinide series of elements from 122 to 153.[72] He proposed an extended periodic table with an additional period of 50 elements (thus reaching element 168); this eighth period was derived from an extrapolation of the Aufbau principle and placed elements 121 to 138 in a g-block, in which a new g subshell would be filled.[76] Seaborg's model, however, did not take into account relativistic effects resulting from high atomic number and electron orbital speed. Burkhard Fricke in 1971[77] and Pekka Pyykkö in 2010[78] used computer modeling to calculate the positions of elements up to Z = 172, and found that the positions of several elements were different from those predicted by Seaborg. Although models from Pyykkö and Fricke generally place element 172 as the next noble gas, there is no clear consensus on the electron configurations of elements beyond 120 and thus their placement in an extended periodic table. It is now thought that because of relativistic effects, such an extension will feature elements that break the periodicity in known elements, thus posing another hurdle to future periodic table constructs.[78]

The discovery of tennessine in 2010 filled the last remaining gap in the seventh period. Any newly discovered elements will thus be placed in an eighth period.

Despite the completion of the seventh period, experimental chemistry of some transactinides has been shown to be inconsistent with the periodic law. In the 1990s, Ken Czerwinski at University of California, Berkeley observed similarities between rutherfordium and plutonium and dubnium and protactinium, rather than a clear continuation of periodicity in groups 4 and 5. More recent experiments on copernicium and flerovium have yielded inconsistent results, some of which suggest that these elements behave more like the noble gas radon rather than mercury and lead, their respective congeners. As such, the chemistry of many superheavy elements has yet to be well-characterized, and it remains unclear whether the periodic law can still be used to extrapolate the properties of undiscovered elements.[2][79]