Can phosphorus prevent rusting?

Is phosphorus the alumina of the ancient world?


Keith S. Taber


An ancient iron column: Did "a very thin layer of phosphorus formed, between the rust and the fresh metal and basically stop… it from rusting any more"

What do you need to build a skyscraper?

I was listening to a podcast from the Royal Institution (where Humphrey Davy and Michael Faraday were based). I must confess I had downloaded the 'Recipe for a Skyscraper' episode some time ago but it had been passed over for other titles.


Royal Institution podcast: Recipe for a Skyscraper – with Roma Agrawal


My mistake. In this talk "structural engineer Roma Agrawal delves into the history of the materials that enable immense construction and the developments that have made our structures what they are today. All while noting the accomplishments of key visionary engineers of the past". This proved to be an engaging and fascinating talk.

A 'mega badass engineer'

On her website, Roma Agrawal , "a structural engineer, author and broadcaster, with a physics degree" describes herself as a "mega badass engineer". She is not above being a little mischievous.

The crumbly ages

For example, she has her own take on what historians used to call the 'dark ages', 1

"So, oddly enough, once the Roman empire fell, the use of concrete basically ended for nearly a thousand years, so that we call it the dark ages, or the crumbly ages as I like to call it, because they went back to using slightly older [construction materials], you know, mud and brick and things like that."

Roma Agrawal talking at the Royal Institution

But while the Romans may have championed the use of concrete, the Indians were outperforming them in the production of high quality iron: "The Romans actually used to import Indian steel at the time and they never knew how to make it because that secret was closely guarded…"

Iron is too reactive to be found 'native' but has to be produced by roasting its ores (that contain compounds of iron) with materials that will reduce the iron compounds to iron, and produce, as a by-product, slag – a complex mixtures of substances. The iron produced will contain some slag mixed into the metal unless this is carefully removed. 2

The Delhi column

As an example of the Indian expertise, Roma Agrawal referred to an old iron column near Delhi which "had not rusted" despite having been erected 1500 years ago.3 The column had originally been a stand for a statue of Garuda, the divine winged creature/demigod who acted as the vehicle for Vishnu. Garuda seems to have flown, but the iron column remains.


The (not quite 4) 'rustless wonder' (Srinivasan & Ranganathan, 2013): the Qtub Iron Pillar

(Photograph taken by Mark A. Wilson, available at https://en.wikipedia.org/wiki/Iron_pillar_of_Delhi#/media/File:QtubIronPillar.JPG)


Lord Vishnu on his mount Garuda (wood carving). It is thought the iron pillar near Delhi once supported a statue of Garuda.

(Image by waradet from Pixabay)


Iron is the main constituent of alloys known as steels, and by mixing other elements (principally, but not only, carbon) with iron it is possible to create steels with various properties, including corrosion resistance. 2 But iron itself readily rusts. The rust formed when iron corrodes is permeable and crumbly, exposing the unreacted metal beneath, which in turn forms rust that again fails to protect the iron beneath it. So, over time, a piece of iron can simply 'rust away' as the reacted material will simply fall off, or be eroded by weather.

Yet this iron column, erected around the time of the final collapse of the Roman Empire, seems to have survived throughout 'the crumbly ages' and through to the present day. Although, it is not that it never started rusting 4, but rather,

"it did initially rust, but then because of the climate in Delhi, the phosphorus, a very thin layer of phosphorus formed, between the rust and the fresh metal and basically stopped it from rusting any more…"

Roma Agrawal talking at the Royal Institution

Corrosion (as with tarnishing) is a generic term. Corrosion leads to structural damage to metal objects (whereas tarnishing is a surface effect).

Rusting is specific to iron as it refers to the material produced when iron corrodes – i.e., rust.


Unreactive phosphorus?: An alternative conception

Roma Agrawal's claim seems incredible to a chemist or science teacher because phsophorus is a very reactive element, and a very reactive element does not seem a good choice of material to protect iron from reacting! Even if the phosphorus did not itself react with the iron and so corrode it, it would soon react with air. In the laboratory, some forms of phosphorus can burst into flames spontaneously, suggesting it is very unlikely to remain intact very long exposed to the elements in India. Certainly not many centuries.

Sacrificial elements

Now, sometimes a more valuable metal is protected by connecting it physically to a more reactive but less valuable metal which preferentially corrodes. As the metals are in electrical contact, the one that loses electrons and releases cations more readily reacts first. The metal allowed to corrode is called a 'sacrificial' metal. For example, bars of sacrificial metal may be dangled from piers or oil rigs to protect the structural metal. The sacrificial metal will slowly 'dissolve' away into the sea 5 – but not that slowly that it would not need replacing for over a millennium. In any case, phosphorus is a non-metal, where the sacrificial element of the pair needs to be the more electropositive. So, there is no helpful explanation there.

Alumina – when tarnishing prevents corrosion

Aluminium is a more reactive metal than iron, yet does not readily undergo substantive corrosion. This is because the surface of an aluminium object readily reacts with oxygen from the air to form a layer of aluminium oxide (alumina). This then protects the aluminium because the alumina formed is a fairly inert substance (unlike the highly reactive phosphorus), and it forms an impermeable layer (preventing oxygen from the air reaching the metal beneath).

Any layer that were to form on iron protect it from rusting also needs to be impermeable and relatively inert. Unlike reactive phosphorus.

Phosphorus would not protect iron

Phosphorus is a fire hazard that burns to produce toxic fumes. In the laboratory, the direct reaction of iron and phosphorus usually requires heating to initiate reaction. Without active heating, the rate of reaction would be too low for a useful laboratory process. However, a very low rate of reaction would not prevent reaction over the centuries since the iron column was erected.

Even if phosphorus was able to form a layer that coated over the iron, using it as a means to prevent corrosion would be like fireproofing a wooden building by coating it with petroleum jelly (e.g., Vaseline). [A correspondent to the British Dental Journal (Brewer, 2017) warned of "the death of a bedbound patient who smoked following application of E45 cream…a paraffin-based product, the residue of which can act as an accelerant when ignited". Smoking kills. And even more rapidly if you smother yourself in flammable oil products prior to lighting up.]

So, it seems we have a mystery.

Or, Roma Agrawal simply got it wrong.

Or, perhaps, more likely, when Roma Agrawal refers to a 'layer of phosphorus' she is using the term loosely, and is actually referring to something else. That is, the protective layer may contain one or more phosphorus compounds, but not phosphorus – just as a layer of the unreactive aluminium compound alumina stops corrosion, although aluminium itself is reactive. Is this distinction just being pedantic? Not to a science educator.

An elementary misconception

The claim that a layer of phosphorus could protect iron from corrosion is therefore not credible to the scientifically literate, but might seem perfectly reasonable to a person with limited science background. One of the great challenges of learning chemistry is making sense of the set of ideas that:

  • the compound of an element is a completely different substance to the element itself
  • the properties of compounds are often quite different (sometimes contrastingly so) to those of the elements the compound was formed from
  • although the compound does not behave like the elements, and does not 'contain' the elements in any straightforward way, there is a sense in which something of the elements persists in (and so the element may be recovered from) the compound.

So, sodium is a reactive metal that burns in air, and chlorine is a green, toxic, choking gas; and both should be avoided unless taking very careful precautions; yet they react, very energetically, to give the relatively unreactive compound sodium chloride – which people readily use in cooking, and to season their food, and to dissolve in water to gargle with, or to soak tired feet. Chlorine would destroy the lining of your throat. Yet sodium chloride solution (despite its chlorine 'constituent') will help ease a sore throat! Still, the sodium chloride has the potential to be 'separated' into the elements with their dangerous properties intact.

Although the distinction between elements and compounds is a lot easier to understand once students learn about molecules and atoms (at least, if avoiding the alternative conception that compounds comprise of molecules and elements comprise of atoms!) this topic is fraught with complications and hang-overs from historical ideas about atoms (Taber, 2003).

If not a layer of phosphorus?

The chemist or science teacher hearing about a protective 'layer of phosphorus' preventing rusting will immediately thinks this is not viable…but a compound of phosphorus might well have the necessary properties. Indeed, generally, the more reactive the elements, the more stable the compounds they form when reacting.

It seems that the layer that formed on the iron column contains the phosphorus compound iron hydrogen phosphate hydrate (FePO4·H3PO4·4H2O),

"Several theories have been postulated regarding corrosion resistance of the Delhi iron pillar. Some of those refer to the inherent nature of the construction material, such as the selection of pure iron, presence of slag particles and slag coatings, surface finishing using mechanical operation, phosphate film formation, or the Delhi's climate…

Earlier studies have delineated the formation of crystalline iron hydrogen phosphate hydrate (FePO4·H3PO4·4H2O), 𝛼-, 𝛾-, 𝛿-FeOOH and magnetite in the case of Delhi iron pillar"

Dwivedi, Mata, Salvemini, Rowles, Becker & Lepková, 2021

The claim re-interpreted

Yet this critical, and somewhat counter-intuitive, distinction between elements qua elements and elements as in some sense 'components' (or 'ingredients') of compounds needs to be acquired. Novices have to learn this. A common alternative conception is to assume that the properties of elements are carried over into their compounds.

So, if students hear that

  • phosphorus is essential in our diet, and that
  • phosphorus is important for healthy bones and teeth,

they can draw the obvious and reasonable conclusion – that phosphorus must be a pretty innocuous substance as it is part of our bodies and we eat it quite safely in our food. Actually, we need compounds of phosphorus in our food to allow our metabolisms to build and repair tissues that contain phosphorus compounds – and anyone misguided enough to try to eat any actual (elemental) phosphorus risks a nasty burn.

In conclusion, as a science graduate, Roma Agrawal presumably appreciates the key distinction between (i) elements as substances and (ii) elements as chemically combined components of other substances, and, as a structural engineer knowledgeable about different material properties, is using 'layer of phosphorus' as a shorthand for a layer of material that includes one or more phosphorus compounds.

That is fine as long as those hearing her talk appreciate that. Another scientist would likely automatically hear 'phosphorus layer' as meaning 'phosphorus compound containing layer'. A science teacher, however, might suspect that the reference to how "a very thin layer of phosphorus formed, between the rust and the fresh metal and basically stopped it from rusting" is likely to be misunderstood, and indeed to mislead, some listening to the podcast.

Minding your Ps…

One of the sources referred to reported how:

"P is found present in slag whereas the presence of P in iron was not detected within the limit of the analytical techniques used in this study. On the basis of this result, we speculate application of lime and other basic compounds during the iron making process which would have led to the transfer P to slag."

Dwivedi, Mata, Salvemini, Rowles, Becker & Lepková, 2021

P is the symbol for phosphorus, the element. However, someone with a sufficient scientific background appreciates from the context that references to

  • P found in slag
  • P in iron
  • transfer [of] P to slag

cannot refer to P as phosphorus the element, but rather some compound or compounds of phosphorus. As a reactive element, phosphorus is not found native and so would not be present (as an element) in the raw materials and, in any case, could certainly not survive (as an element) the high temperature conditions of the processes of iron smelting. Therefore the relevant 'context' for reinterpreting 'P' as not standing for the element itself would be any set of circumstances other than the special conditions where phosphorus can be safely stored without risk of reaction.

This is the prerequisite background knowledge that prevents an audience member misinterpreting what must be meant by a "thin layer of phosphorus [sic]" protecting an exposed iron column – as it cannot possibly refer to a thin layer of [actual, elemental] phosphorus.


Sources cited
  • Anantharaman, T. R. (1997). The iron pillar at Delhi. In S. Ranganathan (Ed.), Iron and Steel Heritage of India (pp. 1-28). Indian Institute of Metals and Tata Steel.
  • Brewer, E. Patient safety: Paraffin-based products. British Dental Journal 223, 620 (2017). https://doi.org/10.1038/sj.bdj.2017.936
  • Dwivedi, D., Mata, J. P., Salvemini, F., Rowles, M. R., Becker, T., & Lepková, K. (2021). Uncovering the superior corrosion resistance of iron made via ancient Indian iron-making practice. Scientific Reports, 11(1), 4221. doi:10.1038/s41598-021-81918-w
  • Falk, S. (2020). The Light Ages. A Medieval journey of discovery. Allen Lane.
  • Srinivasan, S., & Ranganathan, S. (2013). Minerals and Metals Heritage of India. Bangalore: National Institute of Advanced Studies.
  • Taber, K. S. (2003). The atom in the chemistry curriculum: fundamental concept, teaching model or epistemological obstacle? Foundations of Chemistry, 5(1), 43-84. (The author's manuscript versions is available here.)

Notes:

1 A simplistic view was that advancing civilisation underwent something of a relapse during the middle ages, until the gains of the classical age (the Greeks, the Romans) were rediscovered in the Enlightenment. Thus, the term 'dark ages' applied to the 'middle ages'.

There were no dark ages:
as a matter of fact, they are all dark

with apologies to Pink Floyd

That is clearly a great simplification, and ignores many medieval achievements, as well as being a rather Eurocentric view. Some historians have been seeking to redress this impression: for example, Seb Falk (2020) has renamed this period 'the light ages'.


2 To suggest that steel deliberately contains impurities added to iron could give the impression that iron artefacts are made of purer materials than steel ones. This is misleading. Basic iron smelting produces iron that is impure (sometimes known as 'pig iron') and which can contain quite high levels of impurities. Pig iron typically has a high level of carbon – more than is usually used in steels.

Wrought iron is produced by physical working of pig iron which expels much of the slag content, giving purer iron. Wrought iron has long been widely used in structures, but still does not have a high level of purity.

Alloys are mixtures of different metals, or of metallic elements with other elements. 'Metal' here is ambiguous as it can refer to

  • an electropositive element (the usual meaning in chemistry) or
  • a material with certain properties (the usual meaning in engineering) – i.e., malleable, ductile, high electrical and thermal conductivities, lustre, sonorous.

Steels are metals in the 'materials' sense, but 'chemically' are mixtures of the metallic element iron with other elements.

As the properties of steels are sensitive to the levels of other elements, making steel requires using high quality iron that has been treated to remove most of the impurities. This is similar to doping a semiconductor such as silicon to produce electronic components. Very pure silicon is needed as a starting point, so that just the right amount of a specific dopant can be added.

The Indian iron manufacture of Roman times tended to produce iron with a significant phosphorus content.


3 The column was made of wrought iron,

"The forging of wrought iron seems to have reached its zenith in India in the first millennium AD. The earliest large forging is the famous iron pillar with a height of over 7 m and weight of about 6 tons at New Delhi ascribed to Chandragupta Vikramaditya 400- 450 CE… the absence of corrosion is linked to the composition, the high purity of the wrought iron and the phosphorus content and the distribution of slag."

Srinivasan & Ranganathan, 2013

4 The lack of rusting may have been exaggerated,

"The first impression in 1961 was that the portion of the Pillar below the earth was "superficially rusted". However, on detailed examination, the buried portion of the Pillar was found covered with thick crusts of rust and, in fact, copious rust scales could be collected, ranging in thickness from a few millimeters (mm) to no less than 15 mm in some portions. Further, the bulbous base of the Pillar was found riddled with numerous cavities and hollows caused by deep corrosion and mineralization of the iron.

Anantharaman, 1997

Even so, the survival of an iron column exposed to weathering for this length of time is still worthy of note.


5 I thought I should put 'dissolve' into 'scare quotes' here. Corrosion is a chemical change, whereas dissolving refers to what is generally considered a physical change. As the sacrificial metal reacts, it releases cations into solution in the sea, in much the same was as, say, dissolving salt releases sodium ions when common salt is added to water. The metal reacts and enters solution – dissolves, if you are comfortable with that word in this context.


Using water to feed the fire

How NOT to heat up your blast furnace


Keith S. Taber


"From one of the known ingredients of steam being a highly inflammable body, and the other that essential part of the air which supports combustion, it was imagined that [steam] would have the effect of increasing the fire …"


Producing iron requires high temperatures: adding H2O does not help
(Image by zephylwer0 from Pixabay)

The challenge of chemical combination

School science teachers are likely aware of how chemistry poses some significant leaning challenges for learners. One of these is the nature of chemical compounds. That is, compounds of chemical elements.

It may seem obvious to learners that when we 'mix' two components with different properties we should get a mixture with a combination of the component properties. So far, so good. But of course, in chemical reactions we do not just mix different substances, but rather they chemically react. So, sodium will react with chlorine, which can be understood in terms of processes occurring at the nanoscopic scale where molecules of a gas interact with the metallic lattice of sodium cations and delocalised electrons.

Sodium and chlorine behaving badly

Although we can model this process, we cannot observe it directly, or even the starting structures at that scale. Understandably, students often struggle to relate the macroscopic and molecular:

As Sodium is a reactive meterial [sic] and chlorine is a acid. When Sodium is placed in Chlorine, Sodium react badly making a flame and maybe a noise. I think why this reaction happen is because as Sodium reactive metal meaning that it atomic configuration is unstable make the metal danger And as Chlorine is a dangerous acid. When sodium is placed in Chlorine, the sodium start dissolving in the acid due to all the particle rushing around quickly pushing together with Chlorine atom. Producing Sodium chloride.

Student setting out on Advanced level chemistry, quoted in Taber, 1996

So, for example, if we do burn sodium in chlorine we end up with sodium chloride which is a new substance that has its own properties – properties which are not simply some mixture of, or intermediate between, the properties of the substances we start with (the reactants).

Indeed, sodium is a dangerous material to handle: it will react vigorously with water (in a person's sweat for example!) and burns violently in air. Chlorine is so nasty that it has been used as a weapon of war (and since banned as an 'unacceptable' weapon, even in war). In the 'great' war ('great' only because of its scale) the way men died in agony from breathing chlorine was much reported, as well as the effects on those who survived the gas – being blinded for example.

"In all my dreams before my helpless sight,

He plunges at me, guttering, choking, drowning."

Wilfred Owen, Dulce et Decorum Est 1

Sweet and honourable? 1 (Image by Bruce Mewett from Pixabay)

Sodium chloride certainly has its associated hazards – if eaten in excess it is a risk factor for high blood pressure for example – but is certainly not dangerous in anything like the same sense. Many people put sodium chloride on their chips (often along with ethanoic acid solution). No one would want sodium on their food, or to eat in a canteen with a chlorine atmosphere!

When is something both present and not present?

Why this is especially challenging is that the chemistry teacher tells the students that although, at one level, the new substance does not contain its precursors – there is no sodium (substance) or chlorine (substance) in the substance sodium chloride – yet it is a compound of these elements and in some some sense the elements remain 'in' the compound.


Learning chemistry requires understanding how disciplinary concepts explained in terms of submicroscopic level models (After Figure 5, Taber, 2013)

This links to that key theoretical framework in chemistry where we can explain macroscopic (bench scale) phenomena in terms of models of matter at the submicroscopic (indeed nanoscopic or even subnanoscopic) scale. The sense in which sodium chloride 'contains' sodium and chlorine is that it is comprised of a lattice of sodium ions and chloride ions – species which include the specific types of nuclei (those of charge +11 and +17 respectively) that define those elements.

So, when we ask whether the elements are in some sense 'in' the compound we have to think in terms of these abstract models at a tiny scale – there is no sodium substance or chlorine substance present, but there is something that is inherently identified with these two elements. In a sense, but a very abstract sense, the elements are still present. Or, perhaps, better, something intrinsic to those elements is still present.

"We are working here with two complementary meanings for the idea of element, one at the (macroscopic) level of phenomena we can demonstrate to students (substances, and their reactions); the other deriving from a theoretical model in terms of conjectured submicroscopic entities ('quanticles'…).

However, there is also a sense in which an element is considered to be present, in a virtual or potential sense, within its compounds. This use is more common among French-speaking chemists, and in the English-speaking world we normally consider it quite inappropriate to suggest that sodium is somehow present in sodium chloride, or hydrogen in water. Yet, of course, chemical formulae (NaCl, H2O, etc) tell us that the compounds somehow 'contain' the elements."

Taber, 2012, p.19

Figure 1.9 from Taber, 2012

A source of alternative conceptions

This is easy to understand for someone very familiar with molecular level models – but is understandably difficult for novice learners. Thus we can reasonably understand why there are common alternative conceptions along the lines of students thinking that, for example, a compound of a dangerous element (say chlorine) must also be dangerous. Yet we 'mix' and react a soft, reactive, metal and a choking green gas – and get hard white crystals that safely dissolve in water to give a solution we can use in cooking, or to soak our feet, or to gargle with.

An historical precedent

Because science teachers and chemists are so used to thinking in models at the molecular level, we can forget just how unfamiliar this perspective is to the novice, and so the challenge of acquiring the scientific ways of thinking that have become 'second nature' through extensive application.

I was therefore fascinated to see an example of this same alternative conception, assuming a compound will show the properties of its constituent elements, reported by the scientist Sir John Herschel (astronomer, chemist, mathematician, philosopher…), not in a school science context, but rather an industrial context.

"The smelting of iron requires the application of the most violent heat that can be raised, and is commonly performed in tall furnaces, urged by great iron bellows driven by steam-engines. Instead of employing this power to force air into the furnace through the intervention of bellows, it was, on one occasion, attempted to employ the steam itself in, apparently, a much less circuitous manner; viz. by directing the current of steam in a violent blast, from the boiler at once into the fire. From one of the known ingredients of steam being a highly inflammable body, and the other that essential part of the air which supports combustion, it was imagined that this would have the effect of increasing the fire to tenfold fury, whereas it simply blew it out; a result which a slight consideration of the laws of chemical combination, and the state in which the ingredient elements exist in steam, would have enabled any one to predict without a trial."

Herschel, J. F. W. (1830/1851/2017), §37 2

So, here, instead of dropping marks on a test, this misunderstanding of the chemistry leads to a well-intentioned industrialist trying to generate heat in a blast furnace by adding water to the fire. But this does remind us just how counter-intuitive some of the things taught in science are. It might also be a useful anecdote to share with students to help them appreciate that that their errors are by no means unusual, or necessarily a reflection on their ability.

Perhaps this might even be a useful teaching example that could be built up into a historical anecdote which students might readily recall and that will help them remember that compounds have new properties that may be quite different from their constituent elements. So, while a mixture of the flammable gas hydrogen and oxygen can be explosive, a combination (that is, a chemical combination – a compound), of hydrogen and oxygen will not 'feed' a fire but dampen it down. Just as well, really, as otherwise emergency fire and rescue services would need to find an alternative to the widely available, inexpensive, recyclable, non-toxic, agent they widely use in fighting fires.


Compounds and mixtures are not interchangeable (Image by David Mark from Pixabay)

Work cited:

Notes:

1 Wilfred Owen was famous for his war poetry written about the horrors of the trench fighting in the 'first world war'. Owen was killed a week before the war ended. 'Dulce Et Decorum Est' referred to a Latin phrase or motto (dulce et decorum est pro patria mori) that Owen labelled as 'the old lie', that it was sweet and honourable to die in the service of one's country.


2 For some reason, "…it was imagined that this would have the effect of increasing the fire to tenfold fury, whereas it simply blew it out…" puts me in mind of

"the mighty ships tore across the empty wastes of space and finally dived screaming on to…Earth – where due to a terrible miscalculation of scale the entire battle fleet was accidentally swallowed by a small dog."

Douglas Adams, The Hitchhiker's Guide to the Galaxy

She'd never thought about whether ionic bonding is the same thing as chemical bonding

Keith S. Taber

Amy was a participant in the Understanding Science Project. When I talked to her near the start of her GCSE 'triple science' course in Y10 she told me that ionic bonding was "atoms which have either lost or gained electrons so they are either positively or negatively charged" and that chemical bonding was "like in a compound, where two or more elements are joined together", but she seemed unsure how the two concepts were related.

I followed up on Amy's use of the term 'compound' to explore how she understood the term:

How would you define a compound?

Erm Something which has erm two or more elements chemically bonded.

… So you give me an example of that, compound?

Erm, sodium oxide.

Sodium oxide, okay, so there are two or more elements chemically bonded in sodium oxide are there?

Uh hm

And what would those two or more elements be?

Sodium and oxygen.

Okay. Erm, so when we say sodium oxide is chemically bonded, what we are saying there is?

[pause, c 2s]

Erm – a sodium atom has been bonded with a oxygen atom to form erm a new substance.

So Amy's example of a compound was sodium oxide, which would normally be considered essentially an ionic compound, that is a compound with ionic bonding. So this gave me an opportunity to test out whether Amy saw the bonding in sodium chloride and sodium oxide as similar.


Okay, so that was chemical bonding,

Mm.

and that occurs with compounds?

Yeah.

And what did you say about ionic bonding?

Erm, it's the outer electrons they are transferred from one element to another.

Now what does that occur in? You gave me one example, didn't you?

Uh huh

Sodium chloride?

Yeah

Erm. Would sodium chloride be er an element?

[pause, c.2s]

Sodium chloride, no.

No?

It would be a compound.

You think that would be a compound?

Yeah.

And a compound is two or more elements joined together by chemical bonding?

Yeah.

So Amy had told me that sodium chloride, which had ionic bonding, was (like sodium oxide) a compound, and she had already told me that a compound comprised of "two or more elements chemically bonded", so it should be follow that sodium chloride (which had ionic bonding) had chemical bonding.

Do you think sodium chloride has chemical bonding?

Er – I think so

And it also has ionic bonding, or is that the same thing?

Erm,

[pause, c.2s]

I dunno, I've never thought about it that way, erm,

[pause c.3s]

I'm not sure, erm

[pause, c.2s]

I dunno, it might be.

Clearly, whatever Amy had been taught (and interviewing students reveals they often only recall partial and distorted versions of what was presented in class) she had learnt

  • (1) that ionic bonding was transfer of electrons (an alternative conception) as in the example of sodium transferring an electron to chlorine; and that
  • (2) a compounds was where two or more elements chemically bonded together, and an example was sodium oxide where the elements sodium and oxygen were chemical bonded.

Yet these two pieces of learning seemed to have been acquired as isolated ideas without any attempt to link them. Initially Amy seemed to feel ionic bonding and chemical bonding were quite separate concepts.

When taken through an argument that led to her telling me that sodium chloride, that she thought had ionic bonding, was a compound, which therefore had chemical bonding, there should have been a logical imperative to see that ionic bonding was chemical bonding (actually, a kind of chemical bonding – as the logic did not imply that chemical bonding was necessarily ionic bonding). Despite the implied syllogism:

  • sodium chloride has ionic bonding
  • sodium chloride is a compound
  • compounds have elements chemically bonded together
  • therefore ionic bonding …

Amy was unsure what to deduce, presumably because she had seen the two concepts of ionic bonding and chemical bonding as discrete notions and had had given no thought to a possible relationship between them. However explicit teaching had been on this point, it is very likely that the teacher had expected students to appreciate that ionic bonding was a type of chemical bonding – but Amy had not integrated these ideas into a connected conceptual structure (i.e., there was a learning bug that could be called a fragmentation learning impediment).

Ionic bonding – where the electron's transferred to complete the outer shell

Keith S. Taber

Amy was a participant in the Understanding Science Project. The first time I talked to Amy, near the start of her GCSE 'triple science' course in Y10 she told me that "in normal chemistry (i.e., the chemistry part of 'double science', as opposed to the optional additional chemistry lesson as part of 'triple science' that Amy also attended) we're doing about ionic bondingwhich was "atoms which have either lost or gained electrons so they are either positively or negatively charged" and

"how the outer electron's transferred…to complete the outer shell of the erm chlorine, thing, ion…and the sodium atom loses erm, one electron is it, yeah one electron, erm, which the chlorine atom gains, and that yeah that completes its outer shell and makes the sodium positively charged and the chlorine negatively charged".

Amy told me that "in ionic bonding it's the electrons that are transferred, I think."

So Amy had acquired a common alternative conception, i.e. that ionic bonding involved electron transfer, and that this occurs to atoms to complete their electron shells.

Ionic bonding refers to the forces between ions that hold the structure of an ionic substance together, rather than a mechanism by which such ions might hypothetically be formed – yet often learners come away form learning about ionic bonding identifying it with a process of electron transfer between atoms instead of interactions between ions which can be used to explain the properties of ionic substances.

Moreover, the hypothetical electron transfer is a fiction. In the case of NaCl such an electron transfer between isolated Na and Cl atoms would be energetically unfavourable, even if reactants containing discrete atoms were available (which is unrealistic).

Whether students are taught that ionic bonding is electron transfer is a moot point, but often introductory teaching of the topic focuses not on the nature of the bonding, but on presenting a (flawed) teaching model of how the ions in the ionic structure could form by electron transfer between atoms. As this mechanism is non-viable, and so not an authentic scientific account, it may seem odd that teachers commonly offer it.

One explanation may simply be custom or tradition has made this an insidious alternative conception. Science teachers and textbooks have 'always' offered the image of electron transfer as representing ionic bonding. So, this is what new teachers had themselves been taught at school, is what they often see in textbooks, and so what they learn to teach.

Another possible explanation is in terms of what what is known as the atomic ontology. This is the idea that the starting pint for thinking about chemistry at the submicroscopic level is atoms. Atoms do not need to be explained (as if in nature matter always starts as atoms – which is not the case) and other entities such as ions and molecules do need to be explained in terms of atoms. So, the atomic ontology is a kind of misleading alternative conceptual framework for thinking about chemistry at the submicroscopic level.

An element needs a certain number of electrons

An element needs a certain amount of electrons in the outer shell

Keith S. Taber

Bert was a participant in the Understanding Science project. In Y10 Bert was talking about how he had been studying electrolysis in class. Bill had described electrolysis as "where different elements are, are taken out from a compound", but it transpired that Bert thought that "a compound is just a lot of different elements put together"*. He seemed to have a tentative understanding that electrolysis could only be used to separate elements in some compounds.

if they're positive and negative then they would be able to be separated into different ones.

So some things are, some things aren't?

Yeah, it matters how many electrons that they have.

Ah. [pause, c.3s] So have you got any examples of things that you know would definitely be positive and negative?

Well I could tell you what happens.

Yeah, go on then.

Well erm, well if a, if an element gives away, electrons, then it becomes positive. But if it gains, then it becomes negative. Because the electrons are negative, so if they gain more, they just go a bit negative.

Yeah. So why would an element give away or gain some electrons? Why would it do that?

Because erm, it needs a certain amount of electrons in the outer shell. It matters on what part of the periodic table they are.

Okay, let me be really awkward. Why does it need a certain number of electrons in the outer shell?

[Pause, c.2 s]

Erm, well, I don't know. It just – 

So Bert thought that an element "needs a certain amount of electrons in the outer shell" depending upon it's position in the periodic table, but he did not seem to recall having been given any reason why this was. The use of the term 'needs' is an example of anthropomorphism, which is commonly used by students talking about atoms and molecules. Often this derives from language used by teachers to help humanise the science, and provide a way for students to make sense of the abstract ideas. If Bert comes to feel this is a sufficient explanation, then talk of what an element needs can come to stand in place of learning a more scientifically acceptable explanation, and so can act as a grounded learning impediment.

References to atoms needing a certain number of electrons is often used as an explanatory principle (the full shells explanatory principle) considered to explain why bonding occurs, why reactions occur and so forth.

Bert's final comment in the short extract above seems to reflect a sense of 'well that's just the way the world is'. It is inevitable that if we keep asking someone a sequence of 'well, why is that' question when they tell us about their understanding of the world, they eventually reach the limits of their understanding. (This tendency has been labelled 'the explanatory gestalt of essence'.) Ultimately, even science has to accept the possibility that eventually we reach answers and can not longer explain further – that's just the way the world is. Research suggests that some students seem to reach the 'it's just natural' or 'well that's just the way it is' point when teachers might hope they would be looking for further levels of explanation. This may link to when phenomena fit well with the learner's intuitive understanding of the world, or tacit knowledge.

Bert's reference to an element needing a certain amount of electrons in the outer shell also seems to confuse description at two different levels: he explicitly refer to substance (element), when he seems to mean a quanticle (atom). Element refers to the substance, at the macroscopic level of materials that can be handled in the laboratory, whilst an atom of the element (which might better be considered to gain or lose electrons) is part of the theoretical model of matter at a submicroscopic level, used by chemists as a basis for explaining much macroscopic, observed behaviour of samples of substances.


A compound is just a lot of different elements put together

Keith S. Taber

Bert was a participant in the Understanding Science project. When interviewed in Y10 he reported that he had been studying electrolysis in chemistry:

"that's where different elements are, are taken out from a compoundthere's a positive anode and a negative cathode. And what it does it attracts the positive part of the compound to the negative cathode, and the negative part goes to the positive , to, you know, so that they can erm get the different elements in the different places, so they can just have one element on its own".

To fully understand what this means from a chemical context the learner needs to appreciate the chemical distinction between elements, compounds and mixtures, so I asked Bert what he thought a compound was:

It's erm, it's er two, er you know, it's just a lot of different elements put together – to create just one.

So if I went and got some elements, let's say I went and got a little file of carbon, a little file of sulphur, a little file of copper, er a little file of magnesium and I were to mix them into a beaker, maybe get a glass rod, give it a good stir, er, give me a compound?

Erm, so it's carbon, erm, carbon, sulphur, magne¬. Carbon, er – What's the fourth one?

Carbon, sulphur, magnesium and copper I think I said.

And copper. All right, erm. Copper, copper sulphate and – and carbon, and I think carbon and magnesium might go just as elements.

Okay, so if I ignored the carbon and magnesium,

Yeah.

if I took some copper and some sulphur,

Yeah.

and mixed them up together,

Yeah.

then I'd get copper sulphate.

Yeah.

And that's a compound now?

Yeah.

In chemistry there is a crucial difference between a mixture and a compound: one which it appeared Bert had not at this point acquired. Presumably his chemistry teacher, in teaching the topic of electrolysis was assuming students in the class would apply prior learning about the difference between elements and compounds, so as to appreciate the significance of electrolysis as a technique which brings about an energetically unfavourable chemical change. This prerequisite knowledge appeared to be lacking for Bert, which provided a deficiency learning impediment when it came to understanding the teaching on electrolysis.

Read about learners' alternative conceptions