The origin and usage of the term metalloid is convoluted. Its origin lies in attempts, dating from antiquity, to describe metals and to distinguish between typical and less typical forms. It was first applied to metals that floated on water (lithium, sodium and potassium), and then more popularly to nonmetals. Only recently, since the mid-20th century, has it been widely used to refer to elements with intermediate or borderline properties between metals and nonmetals.
Pre-1800
Ancient conceptions of metals as solid, fusible and malleable substances can be found in Plato's Timaeus (c. 360 BCE) and Aristotle's Meteorology.Cornford 1937, pp. 249–50Obrist 1990, pp. 163–64
More sophisticated classification arrangements were proposed by Pseudo-Geber (in the Geber corpus, c. 1310), Paracelsus (De Natura Rerum libri nonem, 1525–6; and later works), Basil Valentine (Conclusiones, 1624), and Boerhaave (Elementa Chemiæ, 1733). They attempted to separate the more characteristic metals from substances having those characteristics to a lesser degree. Such substances included zinc, antimony, bismuth, stibnite, pyrite and galena. These were all then called semimetals or bastard metals.Paul 1865, p. 933Roscoe & Schorlemmer 1894, pp. 3–4Partington 1961, pp. 148, 192–193
In 1735, Brandt proposed to make the presence or absence of malleability the principle of this classification. On that basis he separated mercury from the metals. The same view was adopted by Vogel (1755, Institutiones Chemiæ) and Buffon (1785, Histoire Naturelle des Minéraux). In the interim, Braun had observed the solidification of mercury by cold in 1759–60. This was confirmed by Hutchins and Cavendish in 1783.Jungnickel & McCormmach 1996, p. 279–281 The malleability of mercury then became known, and it was included amongst the metals.
In 1789, FourcroyFourcroy, p. 380 highlighted the weakness of this distinction between metals and semimetals. He said it was evident from the fact that
between the extreme malleability of gold and the singular fragility of arsenic, other metals presented only imperceptible gradations of this character, and because there was probably no greater difference between the malleability of gold and that of lead, which was considered to be a metal, than there was between lead and zinc, which was classed among semi-metals, while in the substances intermediate between zinc and arsenic the differences were slight.
This idea of a semimetal, as a brittle (and thereby imperfect)Craig 1849Roscoe & Schorlemmer 1894, pp. 1–2 metal, was gradually discarded after 1789 with the publication of Lavoisier's 'revolutionary' Strathern 2000, p. 239 Elementary Treatise on Chemistry.Roscoe & Schormlemmer 1894, p. 4
1800–1959
In 1808, Erman and Simon suggested using the term metalloid to refer to the newly discovered elements sodium and potassium.Erman and Simon (1808) "3. Dritter Bericht des Hrn. Prof. Erman und des Geh. Oberbauraths Simon über ihre gemeinschaftlichen Versuche" (Third report of Prof. Erman and State Architect Simon on their joint experiments), Annalen der Physik, 28 (3) : 347-367. After studying the new metallic elements that Humphry Davy had isolated, physicist Paul Erman and architect Paul Ludwig Simon concluded that potassium and sodium didn't conform to the traditional criteria of metals and therefore should be classified as "metalloids". From p. 347: "Hierin liegt aber eine sehr grosse Schwierigkeit, weil veilleicht keine der bekannten Substanzen einen so hohen Grad von Oxydabilität besitzt, als diese beiden Metalloide, wie man sie wohl an füglichsten vor der Hand nennen sollte." (However, here lies a very great difficulty because perhaps none of the known substances possess so high a degree of oxidizability as these two metalloids, as one should perhaps call those [that are] most conveniently available.) These elements were lighter than water and many chemists did not regard them as proper metals. Erman and Simon's proposal may have been made '[in] an attempt to revive this old distinction between metals and substances resembling metals'.Tweney & Shirshov 1935 Their suggestion was ignored by the chemical community.Goldsmith 1982, p. 526
In 1811, Berzelius referred to nonmetallic elements as metalloids,Berzelius 1811, p. 258 in reference to their ability to form oxyanions.Partington 1964, p. 168Bache 1832, p. 250 A common oxyanion of sulfur, for example, is the sulfate ion SO. Many metals can do the same. Chromium, for instance, can form the chromate ion CrO. Berzelius' terminology was widely adopted although it was subsequently regarded by some commentators as counterintuitive, misapplied, incorrectGlinka 1959, p. 76 or invalid.Hérold 2006, pp. 149–150 In 1825, in a revised German edition of his Textbook of Chemistry,Partington 1964, pp. 145, 168Jorpes 1970, p. 95 Berzelius subdivided the metalloids into three classes. These were: constantly gaseous 'gazolyta' (hydrogen, nitrogen, oxygen); real metalloids (sulfur, phosphorus, carbon, boron, silicon); and salt-forming 'halogenia' (fluorine, chlorine, bromine, iodine).Berzelius 1825, p. 168
In 1844, Jackson gave the meaning of 'metalloid' as 'like metals, but wanting some of their properties.'Jackson 1844, p. 368 In 1845, in A dictionary of science, literature and art, Berzelius' classification of the elementary bodies was represented as: I. gazolytes; II. halogens; III. metalloids ('resemble the metals in certain aspects, but are in others widely different'); and IV. metals.Brande & Cauvin 1845, p. 223
In 1864, calling nonmetals 'metalloids' was still sanctioned 'by the best authorities'The Chemical News and Journal of Physical Science 1864 even though this did not always seem appropriate. The greater propriety of applying the word metalloid to other elements, such as arsenic, had been considered.
By as early as 1866, some authors were instead using the term nonmetal, rather than metalloid, to refer to nonmetallic elements.Oxford English Dictionary 1989, 'nonmetal' In 1875, Kemshead observed that the elements had been subdivided into two classes—'non-metals or metalloids, and metals.' He added that '[t]he former term, although not so convenient, because a compound word, is more correct, and is now universally employed.'Kemshead 1875, p. 13
In 1876, Tilden protested against, 'the [still] too common though illogical practice of giving the name metalloid to such bodies as oxygen, chlorine or fluorine'. He instead divided the elements into ('basigenic') true metals, metalloids ('imperfect metals') and ('oxigenic') nonmetals.Tilden 1876, pp. 171–3; 198. He listed as metalloids: H; Ti(?), Zr; V, Nb(?), Ta(?); Mo, W, U; As, Sb, Bi; and Te. Non-metals included B, C, Si and Se. Graphitic C and Si were however noted as furnishing the only examples among the non-metals of electric conductivity.
As late as 1888, classifying the elements into metals, metalloids, and nonmetals, rather than metals and metalloids, was still regarded as peculiar and potentially confusing.The Chemical News and Journal of Physical Science 1888
Beach, writing in 1911, explained it this way:Beach 1911
Metalloid (Gr. "metal-like"), in chemistry, any nonmetallic element. There are 13, namely, sulfur, phosphorus, fluorin[e], chlorin[e], iodine, bromine, silicon, boron, carbon, nitrogen, hydrogen, oxygen, and selenium. The distinction between the metalloids and the metals is slight. The former, excepting selenium and phosphorus, do not have a "metallic" lustre; they are poorer conductors of heat and electricity, are generally not reflectors of light and not electropositive; that is, no metalloid fails of all these tests. The term seems to have been introduced into modern usage instead of nonmetals for the very reason that there is no hard and fast line between metals and nonmetals, so that "metal-like" or "resembling metals" is a better description of the class than the purely negative "nonmetals". Originally it was applied to the nonmetals which are solid at ordinary temperature.
In or around 1917, the Missouri Board of Pharmacy wroteMayo 1917, p. 55 that:
A metal may be said to differ from a metalloid [that is, a nonmetal] in being an excellent conductor of heat and electricity, in reflecting light more or less powerfully and in being electropositive. A metalloid may possess one or more of these characters, but not all of them ... Iodine is most commonly given as an example of a metalloid because of its metallic appearance.
During the 1920s, the two meanings of the word metalloid appeared to be undergoing a transition in popularity. Writing in A Dictionary of Chemical Terms, CouchCouch 1920, p. 128 defined 'metalloid' as an old, obsolescent term for 'nonmetal.' In contrast, Webster's New International Dictionary noted that use of the term metalloid to refer to nonmetals was the norm. Its application to elements resembling the typical metals in some way only, such as arsenic, antimony and tellurium, was recorded merely on a 'sometimes' basis.Webster's New International Dictionary 1926 p. 1359
Use of the term metalloid subsequently underwent a period of great flux up to 1940. Consensus as to its application to intermediate or borderline elements did not occur until the ensuing years, between 1940 and 1960.
In 1947, Pauling included a reference to metalloids in his classicLundgren & Bensaude-Vincent 2000, p. 409 and influentialGreenberg 2007, p. 562 textbook, General chemistry: An introduction to descriptive chemistry and modern chemical theory. He described them as 'elements with intermediate properties ... occupy[ing] a diagonal region [on the periodic table], which includes boron, silicon, germanium, arsenic, antimony, tellurium, and polonium.'Pauling 1947, p. 65
In 1959 the International Union of Pure and Applied Chemistry (IUPAC) recommended that '[t]he word metalloid should not be used to denote nonmetals'IUPAC 1959, p. 10 although it was still being used in this sense (around that time) by, for example, the French.Friend 1953, p. 68
1960–present In 1969, the classicAmerican Institute of Chemists 1969, p. 485 and authoritativeAmerican Chemical Society California section 1969, p. 55 Hackh's Chemical Dictionary included entries for both 'metalloid' and 'semimetal'. The latter term was described as obsolete.Grant 1969, pp. 422, 604: 'metalloid.—(1) having the physical properties of metals and the chemical properties of nonmetals, e.g., As. (2) a nonmetal (incorrect usage) ... semimetal.—an element midway in properties between metals and nonmetals, as arsenic (obsolete).'
In 1970, IUPAC recommended abandoning the term metalloid because of its continuing inconsistent use in different languages. They suggested using the terms metal, semimetal and nonmetal instead.IUPAC 1971, p. 11 Despite this recommendation, use of the term 'metalloid' increased dramatically. Google Ngram Viewer showed a fourfold increase in the use of the word 'metalloid' (as compared to 'semimetal') in the American English corpus from 1972 to 1983. There was a sixfold increase in the British English corpus from 1976 to 1983. As at 2011, the difference in usage across the English corpus was around 4:1 in favour of 'metalloid'.Google Ngram, viewed 11 February 2011
The most recent IUPAC publications on chemical nomenclature (the "Red Book", 2005) IUPAC 2005 and terminology (the "Gold Book", 2006–) IUPAC 2006– do not include any recommendations as to the usage or non-usage of the terms metalloid or semimetal.