Talk:Nonmetal

From Wikipedia, the free encyclopedia

More information Article milestones, Date ...
Former featured article candidateNonmetal is a former featured article candidate. Please view the links under Article milestones below to see why the nomination was archived. For older candidates, please check the archive.
Good articleNonmetal has been listed as one of the Natural sciences good articles under the good article criteria. If you can improve it further, please do so. If it no longer meets these criteria, you can reassess it.
Article milestones
DateProcessResult
August 21, 2013Good article nomineeListed
July 26, 2021Featured article candidateNot promoted
October 5, 2021Peer reviewReviewed
October 18, 2021Featured article candidateNot promoted
January 18, 2022Featured article candidateNot promoted
February 5, 2022Featured article candidateNot promoted
February 10, 2022Good article reassessmentKept
August 2, 2022Peer reviewReviewed
September 26, 2022Featured article candidateNot promoted
June 24, 2023Featured article candidateNot promoted
October 22, 2023Featured article candidateNot promoted
May 9, 2024Featured article candidateNot promoted
June 4, 2024Featured article candidateNot promoted
Current status: Former featured article candidate, current good article
Close

Dubious tag: As and Sb

The article currently says:

One widely recognized physical property is the temperature coefficient of resistivity—that is, the way an element’s conductivity changes with temperature. In metals, conductivity typically decreases with increasing temperature, whereas in nonmetals it increases. However, there are notable exceptions. For instance, plutonium, although a metal, exhibits increased conductivity when heated from −175 °C to +125 °C. Conversely, carbon (as its graphite allotrope), often described as nonmetallic, behaves as a semimetal and shows decreased conductivity with temperature. [Atkins et al. 2006, pp. 320–21] Arsenic and antimony, sometimes classified as nonmetals, behave in the same way. [Zhigal'skii & Jones 2003, p. 66][dubious – discuss]

The reason given for the tag is: "No such statements appear in the more recent 9th and other editions, so unverifiable."

AFAIK, since As and Sb are semimetals in the physics-based sense and do behave in the same manner as graphite, and since there is no 9th edition of either Atkins et al. or Zhigal'skii & Jones, I have removed the dubious tag. Sandbh (talk) 05:11, 5 April 2025 (UTC)

Do the sources present plutonium, arsenic and antimony as exceptions to the rule? In statements such as In metals, conductivity typically decreases with increasing temperature we usually consider temperatures close to absolute zero. The variation from −175 °C to +125 °C thus seems irrelevant with respect to the classification. Jähmefyysikko (talk) 18:21, 4 May 2025 (UTC)
I guess this has been removed already from the article so nevermind. Jähmefyysikko (talk) 18:25, 4 May 2025 (UTC)

Concerns about Ashcroft & Mermin's [A&M] Solid state physics (1976)

The title and publication details of this book are:

Solid state physics, Saunders College Publ., Fort Worth, Philadelphia

It is cited four times in the Structure, quantum mechanics, and band structure section of our Nonmetal article.

My observations follow.

1. A&M write at p. 2:

"Even to understand nonmetals one must also understand metals..."

A fair point, which presumably ought to be reflected in our article.

2a. A&M write at p. 60:

"Why Are Some Elements Nonmetals? A more acute inadequacy of our rule of thumb for determining the number of conduction electrons is posed by the existence of insulators. Why, for example, is boron an insulator while its vertical neighbor in the periodic table, aluminum, an excellent metal?"

Rather than as an insulator, boron is well known as a semiconductor. See: Berger LI 1997, Semiconductor Materials, CRC Press, Boca Raton, pp. 37–43.

Steudel (2020, p. 224), who is also cited in the same section of our article, similarly writes:

"Boron is a semiconductor in all of its modifications with bandgaps for α- and β-rhombohedral boron ranging from 1.6 to 2.0 eV."

2b. On a related note, our article goes on to say:

"Band structure definitions of metals and nonmetals are widely used in current research into materials, and apply both to single elements such as insulating boron[215]..."

In this case, citation [215], notes the semiconducting properties of boron about ten times, and not at all as an insulator.

3. In our article, A&M chapters 8 and 19 are cited in the following following passage:

"From this the concept of nonmetals as "not-a-metal" originates. The original approach to describe metals and nonmetals was a band-structure with delocalized electrons (i.e. spread out in space). A nonmetal has a gap in the energy levels of the electrons at the Fermi level.[206]: Chpt 8 & 19  In contrast, a metal would have at least one partially occupied band at the Fermi level;[206] in a semiconductor or insulator there are no delocalized states at the Fermi level, see for instance Ashcroft and Mermin.[206] (A semimetal is similar to a metal, with a slightly more complex band structure.) These definitions are equivalent to stating that metals conduct electricity at absolute zero, as suggested by Nevill Francis Mott,[207]: 257 and the equivalent definition at other temperatures is also commonly used as in textbooks such as Chemistry of the Non-Metals by Ralf Steudel[208] and work on metal–insulator transitions.[209][210]"

Here, the citation of Chpts 8 & 19 appears to be misplaced—neither chapter discusses nonmetals specifically. Chapter 8 discusses semiconductors and insulators. Chapter 19 (p. 377) says, "Ionic crystals, such as sodium chloride are compounds composed of a metallic and nonmetallic element."

4. A&M at p. 726 present a periodic table in which the "nonmetallic elements" are shaded as:

H; B; C; N, P, As; O, S, Po; F to At; He to Rn.

From their periodic table, the status of Si, Ge, Se, and Te is unclear. In real life these elements in their most stable forms in ambient conditions are well known as semiconductors, but so too are B, P, and I—elements A&M classify as nonmetallic.

Note that C is shaded by them as a nonmetal even though, in its most stable form of graphite, it is a semimetal in a physics-based sense and hence a metal in the physics-based sense.

Note further that while As and Sb are both semimetals in a physics-based sense and hence metals in the physics-based sense, As is classed by A&M as a nonmetal whereas on their periodic table the status of Sb is unclear. Elsewhere in their book they show Sb as a metal (pp. 5, 21, 38, 48, 364). The logic here is unclear.

Concluding observations
Given the concerns raised above, we may wish to consider the following improvements to our article:

  • Clarify that A&M’s discussion of boron as an insulator (p. 60) is not consistent with its classification as a semiconductor in more recent and widely accepted sources, including Berger (1997) and Steudel (2020). Citation [215], currently supporting boron's status as an insulator, also needs revision.
  • Re-express or qualify the reference to A&M Chapters 8 and 19 in citation [206], as these chapters do not directly address the classification of nonmetals and do not appear to support the statements made in the paragraph in question.
  • Acknowledge the confusion in A&M’s classification of elements such as silicon, arsenic, and antimony, with the latter two variably treated in physics- and chemistry-based schemes. Their depiction of carbon, arsenic, and boron also merits comment, given the physical forms in which these elements commonly occur and their electronic properties in those forms.

Overall, while A&M may have been regarded as a landmark text in solid-state physics, some of its examples and classifications may not reflect contemporary chemical understanding or may require contextualization when used in a general encyclopedia article.

I welcome further thoughts on how best to proceed. Sandbh (talk) 08:21, 4 May 2025 (UTC)

Dear Sandbh,
It is really important to understand the science, not just pick a few words and phrases. There is a reason that we do not typically teach solid state physics to year 1 or 2 undergrads in chemistry, engineering, materials science and physics -- it is not simple. It is absolutely critical to understand the details, not just the words. With much (perhaps most) of solid state physics it is the math that matters, the words are just attempts to express concepts.
One of the trickiest is the difference between an insulator and a semiconductor, which is useage, not a fundamental property. If you check the latest version you will see that there is mention of diamond as a semiconductor, despite the fact that pure diamond is the most insulating of all the pure elemental solids! All insulators are intrinsic semiconductors, those used for applications are extrinsic semiconductors. Everything is in the doping. This is why, for instance, most industrial semiconductor chips are built using a substrate of pure silicon, with various doping to create devices ( the article Semiconductor device fabrication talks about some of this but you have to dig deeper). Pure silicon is used as an electrically insulating base.
The current text is appropriate, and I have tried to clearly indicate the issue of semiconductor versus insulator. Ashcroft and Mermin is correct, and has been used for almost fifty years to teach senior undergraduate and graduate students in chemistry, engineering, materials science and physics. You may want to check Kittel or any of the texts at Condensed matter physics#Further reading, paying attention to the equations etc not just the words.
N.B., please note that I deliberately included chemistry students in those who take a class in solid state physics. At Northwestern it is 1/3-1/2 the graduate chemistry class, and probably the same around the world.
N.N.B., I hit reply too quickly, so had to continued this response. Ldm1954 (talk) 12:32, 4 May 2025 (UTC)
In my opinion your post wonderfully illustrates that "nonmetal" is small, minor concept primarily notable via "not a metal". Nonmetal has no unique qualifying characteristic. We can't even find sources to agree that which elements are nonmetals and yet the vast majority of readers know quite well that "nonmetal" is a valuable way to characterize materials. Using this word for a whole long article about elements is inappropriate.
So yes, put a bunch of "other sources say" in the article. That will clarify the character of the topic. Johnjbarton (talk) 17:26, 4 May 2025 (UTC)
@Johnjbarton, I really don't think adding more "other sources say" will do anything positive. There is already enough of this and more would just create more confusion for novices. We would both rewrite this article differently, but the concensus appears to be to keep some of the pre-QM empirical categorization, plus of course the QM and modern band-structure interpretation. I don't think you would support removing Ashcroft and Mermin and modern "metal versus not-a-metal", or am I wrong? Ldm1954 (talk) 19:35, 4 May 2025 (UTC)
I do not support removing Ashcroft and Mermin unless someone has a reliable source that explicitly says their content is incorrect. For example, Berger LI 1997, Semiconductor Materials, CRC Press, says Boron possesses semiconductor properties, it does not claim that Boron is not an insulator. The quote from A&M on page 60 contrasts B with Al in the context of the inadequacy of a rule of thumb. A&M is not wrong, they are simply making a different point. Johnjbarton (talk) 19:43, 4 May 2025 (UTC)

Some further concerns about A&M (1976)

I preface these concerns by respectfully presuming it is important to both understand the maths and to get the words and phrasing right. The examples that follow relate primarily to inconsistencies and overgeneralisations in the textual presentation found in Ashcroft & Mermin, which affect how their work is used as a source in the Nonmetal article. My concern is not with the mathematical integrity of their work, but with the implications of their wording and classification choices when interpreted by a general readership.

5. On p. 2 of their book, Ashcroft and Mermin flatly write:

Metals occupy a rather special position in the study of solids, sharing a variety of striking properties that other solids (such as quartz, sulfur, or common salt) lack. They are excellent conductors of heat and electricity, are ductile and malleable, and display a striking luster on freshly exposed surfaces.

(The relevance of this to the Nonmetal article is that, as A&M write later on the same page, “even to understand nonmetals one must also understand metals.”)

I referred to the passage as being stated flatly, since it is neither qualified as a generalisation nor presented as a being typical summation.

In this light, it is in fact demonstrably false: several metals — including beryllium, chromium, tungsten, manganese, osmium, bismuth, and plutonium — are brittle.

6. In the same passage quoted above, A&M also assert that metals display a striking lustre on freshly exposed surfaces. However, arsenic — a semimetal in the physics-based sense, and one which they classify (p. 726) as a nonmetal — displays precisely this kind of metallic lustre. This undermines the implied uniqueness of metallic lustre as a property exclusive to metals and highlights further ambiguity in A&M’s treatment.

7. On p. 563, A&M write:

The conductivity of a metal...declines with increasing temperature... The most striking feature of semiconductors is that, unlike metals, their electrical resistance declines with rising temperature.

This statement is false. Plutonium, widely recognised as a metal, exhibits increasing conductivity with temperature in the range −175 °C to +125 °C. See: Russell AM & Lee KL 2005, Structure–Property Relations in Nonferrous Metals, Wiley-Interscience, New Jersey, p. 466. This is not a trivial anomaly but a well-documented counterexample to their binary characterisation of metals versus semiconductors.

8a. I earlier noted (see item 2a above) that A&M referred to boron as an insulator (p. 60). On p. 564, they go on to write:

Semiconducting crystals come primarily from the covalent class of insulators. The simple semiconducting elements are from column IV of the periodic table, silicon and germanium being the two most important elemental semiconductors. Carbon, in the form of diamond, is more properly classified as an insulator, since its energy gap is of order 5.5 eV. Tin, in the allotropic form of grey tin, is semiconducting, with a very small energy gap... The other semiconducting elements, red phosphorus, boron, selenium, and tellurium, tend to have highly complex crystal structures, characterized, however, by covalent bonding.

Even though A&M clarify the semiconducting status of boron, it is referred to in our Nonmetal article as an insulator. Clearly this is only part of the story. Likewise, other sources consistently classify boron as a semiconductor.

8b. While the crystal structures of boron and red phosphorus are indeed highly complex, this is not the case for selenium and tellurium, which consist of relatively simple parallel covalent spiral chains. These chains are weakly bonded to one another via metallic interactions. There is thus no general tendency for semiconducting nonmetals to possess highly complex crystal structures, as A&M imply.

9. After asserting that diamond is “more properly classified as an insulator,” A&M then include it (at p. 566) in a list of "selected semiconductors". This again illustrates internal inconsistency in their treatment of nonmetallic semiconductors.

Conclusion

These contradictions render A&M’s descriptions not only inaccurate, but at times careless or overly idealised — especially surprising in a text that elsewhere emphasises physical rigour.

If A&M is to be cited in the Nonmetal article, we should do so with caution and avoid reproducing such contradictions without clarification or qualification.

Ideally, while age alone does not disqualify a source, the several shortcomings of A&M (1976) suggest the need for a more up-to-date and element-informed source — particularly for an encyclopaedic treatment intended for a general audience. --- Sandbh (talk) 11:43, 7 May 2025 (UTC)

@Sandbh, you appear to be unwilling to accept solid state chemistry/physics. Yes, it is both, a large amount of band structure was and continues to be developed by chemists. Your comments above are a combination of WP:SYNTH, WP:OR and extreme WP:Cherry picking. You are pulling words and phrases out of context, and using them in ways that have nothing to do with the article, the book or where it is used as a source. If you read carefully what is already in the existing page much of what you claim as anomalies is already asked and answered, for instance the band structure many body problem, relativistic quantum mechanic terms for the actinides and why it is completely legitimate to refer to diamond as both an insulator and a semiconductor.
Please stop. Ldm1954 (talk) 12:11, 7 May 2025 (UTC)
At least in condensed matter physics contexts, the binary division into metals and insulators is very common. Semiconductors are considered a subclass of insulators. For example, a quote from Steven H. Simon (2013) Oxford Solid State Basics p.174: If the band gap is below about 4 eV, then these type of insulators are called semiconductors, since at room temperature a few electrons can be thermally excited into the conduction band, and these electrons then can move around freely, carrying some amount of current. Thus, there is no inconsistency here, except that textbook definitions do not always align with those used on Wikipedia. The above definition by Simon is compatible with what A&M write on page 562. Regarding 7., it is implicitly clear in the text that A&M refer to asymptotic behavior at low temperatures: they refer to power laws and exponential dependencies. Jähmefyysikko (talk) 15:55, 7 May 2025 (UTC)
@Jähmefyysikko Thank you. I agree that in condensed matter physics contexts, a binary division into metals and insulators is common.
The confusion arises when comparing the situation at absolute zero versus room temperature. At absolute zero, there are only metals and semimetals (representing the metallic side) and insulators (representing the nonmetallic side). At room temperature, the metallic side remains unchanged, but on the nonmetallic side, semiconductors emerge as a distinct class. However, there is no universally agreed cut-off between semiconductors and insulators at room temperature (see also below re A&M, p. 562). That said, it is fair to regard semiconductors as a subclass of insulators, just as semimetals are often treated as a subclass of metals — while noting that there are no semiconductors at absolute zero, only insulators.
This gives rise to complications. For example, at absolute zero, carbon as graphite is a semimetal, while diamond and C60 are insulators. At room temperature, graphite remains a semimetal; C60 becomes a semiconductor; and diamond remains an insulator (re diamond, Kittel 2005, in Intoduction to Solid State Physics, 8th ed., p. 167, writes, "Diamond itself is more an insulator rather than a semiconductor.") Even A&M appear inconsistent: they refer to graphite as a semimetal (p. 303), yet label carbon as a nonmetal on their periodic table (p. 726). The same inconsistency arises with arsenic, which they describe as a semimetal (p. 304) yet classify as a nonmetal on their periodic table.
On a comparability between Simon and A&M as to a c. 4 eV cut-off between semiconductors and insulators, A&M nuance the situation on p. 562:

Evidently the distinction between a semiconductor and an insulator is not a sharp one, but roughly speaking the energy gap in most important semiconductors is less than 2 eV and frequently as low as a few tenths of an electron volt. Typical room temperature resistivities of semiconductors are between 10–3 and 109 Ω·cm (in contrast to metals, where ρ ≈ 10–6 Ω·cm, and good insulators, where ρ can be as large as 1022 Ω·cm).

On point 7., I agree that in many contexts, the temperature dependence of conductivity is analysed in the low-temperature limit, often with reference to power-law or exponential behaviour. However, in this case I feel that’s not the intent of Ashcroft & Mermin on pp. 562–563.
Their wording reads as a general claim:

The most striking feature of semiconductors is that, unlike metals, their electrical resistance declines with rising temperature.

This follows:

The electrical conductivity of a semiconductor should be a very rapidly increasing function of temperature. This is in striking contrast to the case of metals.

Nowhere in this passage do they indicate that the contrast refers to low-temperature asymptotics or limiting behaviour as T → 0. The tone and phrasing suggest a broad, generalised comparison that readers would reasonably interpret as applying over a typical temperature range.
That’s why I pointed to plutonium as a counterexample. It is widely regarded as a metal, yet its conductivity increases with temperature between –175 °C and +125 °C — directly contradicting A&M’s generalisation. If the statement had been explicitly restricted to low-temperature behaviour, I would agree the contrast holds. But as written, I think it is too strong and lacks the necessary qualification.
For this reason, I suggest that any use of this passage in the article should be carefully framed or clarified.
I further note that Simon does not use the term "nonmetal" in his book. This seems to reflect solid-state physics' predominant focus on metals and their electronic properties; in contrast, he uses "metal" 70 times and "metals" 80 times.
Simon also writes (p. 19):

The defining characteristic of a metal is that it conducts electricity.

If we adopt the criterion that "a metal is something that conducts electricity," then graphite certainly qualifies. But from a chemical perspective — which considers a range of physical and chemical properties — graphite lacks most metallic traits, and carbon remains classified as a nonmetal.
Returning to the use of the term "nonmetal" in solid-state physics, a similar pattern appears in A&M, who use "nonmetals" twice and "nonmetallic" three times, compared to 284 uses of "metal" and 478 of "metals".
Likewise, Kittel, in Introduction to Solid State Physics, 2005 (8th ed.) refers to "nonmetals" twice, compared to "metals", "metal" and "metallic" 172, 152, and 57 times respectively.
If the appearances of “nonmetal” and its derivatives, and “metal” and its derivatives are tallied across the three books, the result is 4:1293 or 1:323 or 0.3% to 99.69%. These ratios suggest a conceptual asymmetry in solid-state physics: the literature tends to emphasise metals, with “nonmetals” receiving significantly less systematic attention and definitional clarity.
I would welcome any further thoughts you may have in this matter.
— Sandbh (talk) [date] Sandbh (talk) 12:21, 15 May 2025 (UTC)
@Sandbh, please see prior response: #Asked and Answered. You are continuing to attack established science. Ldm1954 (talk) 12:32, 15 May 2025 (UTC)
@Sandbh You are just repeating a point of view that I don't agree with, the idea that the topic 'nonmetal' concerns every possible property of a poorly defined set of elements. That is a meaning you invented as far as I can tell. Applying your concept to A&M will of course result in many inconsistencies. That is exactly what we have been trying to explain to you over many months now. The problem is not with A&M or Kittel, but rather it is with the concept you have adopted without any source backing it up. Johnjbarton (talk) 16:39, 15 May 2025 (UTC)
Framing and article scope
@Ldm1954, @Johnjbarton — Thank you both.
There are a few earlier replies I’ve yet to respond to. I appreciate the points that were raised there and plan to return to them soon.
For now, I’d like to briefly respond.
I don’t believe I’ve attacked any science. I’m not questioning the validity of Ashcroft & Mermin or Kittel, nor am I advocating for removing solid-state physics from the article. My aim is to examine how those sources are being used in the context of the nonmetal article, which concerns a chemistry-based classification of elements.
It’s precisely because I respect these sources that I want to avoid misattributing meanings or generalisations they may not support — particularly where other reliable sources (e.g. Steudel, Berger) provide important nuance. For example, I cited Ashcroft & Mermin’s own wording on the fuzzy boundary between semiconductors and insulators (p. 562), and I pointed to plutonium as a valid counterexample to one of their unqualified generalisations.
As for the conceptual approach: the distinction I’ve drawn — between solid-state physics (which classifies forms of matter based on electronic properties) and chemistry (which classifies elements based on physical and chemical properties) — is well-established in the literature. This is not an invented perspective, but a disciplinary observation supported by sources like Steudel, Kean, and standard general chemistry textbooks.
I also don’t believe that the concept I’ve adopted is without source backing. The idea that nonmetals can be meaningfully discussed as a class of elements — based on their physical and chemical properties — is well represented in the literature. For example, the article List of nonmetal monographs lists 17 monographs that focus on the chemistry and properties of nonmetals. These works span more than a century and reflect serious and systematic attention to the topic. This clearly indicates that the nonmetal classification is not an arbitrary construct, but a sustained subject of study — particularly within chemistry.
If there are better ways to describe these disciplinary differences, I’m open to hearing them. But I don’t believe that examining the coherence, framing, and sourcing of article content constitutes original research or "attacks on science."
— Sandbh (talk) [date] Sandbh (talk) 06:58, 16 May 2025 (UTC)
The first sentence of Steudel Part II is
  • The strictest criterium to define nonmetals in contrast to metals is the electric conductivity. Typically, metals show a finite conductivity at ambient conditions, whereas the conductivity of nonmetals is close to zero. With this definition, 23 of the known chemical elements are nonmetals, and these are the subject of this textbook, namely hydrogen, boron, carbon, silicon, germanium, nitrogen, phosphorus, arsenic, the chalcogens, that is, oxygen through tellurium, as well as halogens and noble gases.
To me this is broadly similar to the solid-state physics definition.
I agree that there are inconsistencies between A&M and Kittel and other sources, just as there are inconsistencies among the 17 sources in List of nonmetal monographs. Yes, books like Steudel have chapter of elements, just like we have articles on elements, but Steudel does not associate topics like abundance and extraction with the "nonmetal" property.
You started this topic as "Concerns about Ashcroft and Mermin....". I think it is clear that there is no consensus to remove A&M as unreliable. Johnjbarton (talk) 15:38, 16 May 2025 (UTC)

Asked and answered

@Sandbh, I hope that your attacks on Ashcroft and Mermin are not a prelude to attacks on other sources and to your advocating to remove all the well sourced material on band structure definitions, and revert to empirical high-school textbook classifications such as this version from about a year ago. That was full of inaccurate sources and multiple problems. Multiple editors have improved upon that version, strongly disagreed with your views about what a metal is versus a not-a-metal and multiple other topics which can be found in the archive here and also at talk physics. This has all been hashed out, and it is time to move on. The current version is a viable compromise. Ldm1954 (talk) 13:22, 7 May 2025 (UTC)

Related Articles

Wikiwand AI