Students’ ideas about chemical bonding

Almost all molecules have bonds falling between the two extremes of “covalent” and “ionic” bonding. The behaviour of a substance is influenced by intermolecular bonds, which, if extensive, influence boiling and melting points, structure and potential use. Students are introduced to intermolecular bonds during post-16 chemistry courses. Relatively little work has been carried out on students’ ideas about chemical bonding prior to the age of 16.

Covalent bonds

The simplest idea associated with the formation of a single covalent bond is that a pair of

electrons is shared between two atoms, and for a double bond two electron pairs are

shared. In either case the sharing confers additional stability on both atoms involved and a

fixed amount of energy is required to break the bond.

The development of basic ideas

Barker (1994) reports the changes in students’ basic ideas about covalent bonds and molecular structure over a two-year period. About 18% of 16 year olds could distinguish between single and double covalent bonds in methane, ethene and water molecules in terms of the numbers of electrons involved. About 66% of the student population could do this about fifteen months later. A further 25% at this stage distinguished between single and double bonds, but did not specify the numbers of electrons involved. About 7% of students at the end of the study thought the bonds had 1 or 2 electrons.

In a companion question, Barker explored students’ ideas about the energetics involved in bond formation by asking students why a methane molecule has the formula CH4. Very few students at any point in the survey responded in energetics terms, but about 6% at the start and 16% at the end said, “C and H are more stable as CH₄.” A very popular response, given by 56% of 16 year olds and 61% of 18 year olds was “C needs four bonds”. This answer ignores the hydrogen in the molecule and attributes anthropomorphic behaviour to the carbon atom. Taber and Watts (1996) found this language to be extensive and not only used by students but also by teachers in their drive to promote understanding of science issues.

Progression in understanding

Taber (1997a) carried out case studies exploring post-16 chemistry students’ developing understanding of chemical bonding. An early report (Taber 1993a and b) describes “Annie’s” three interviews about chemical bonding and indicates progression in her understanding. In the first interview, she recognised that a covalent bond exists in diatomic molecules in which the two atoms are identical. She did not explain covalent bond formation in terms of sharing electrons. Instead, Annie said that the atoms “pull together”. To decide if a bond was covalent Annie looked at the chemical elements involved to establish if both were non-metals. If this was so, then a covalent bond would form between them. After several months on an A level chemistry course, Annie described covalent bonds in terms of electrons being shared and realised that one result of electron sharing was that atoms acquire “full shells” of electrons. Towards the end of her course Annie was interviewed again. She could describe the electrostatic attractions between atomic nuclei and the electrons, which indicates she had moved towards an accepted view of a covalent bond.

Annie’s progress is reflected in the increasing sophistication of her ideas. Taber developed a model for progression in understanding chemical bonding ideas among post-16 chemistry students. He argues that students begin these courses with a range of conceptual tools gained from earlier study of “curriculum science” and that these are developed into first an “Octet rule framework” towards a “minimum energy explanatory principle” using ideas based on simple quantum theory using atomic orbitals. A key point is that his evidence supports students finding it easier to acquire or add new conceptual tools to the old set, rather than to dismantle existing models. Barker’s study supports this - although students will have been taught “new” ideas based on atomic orbitals, in answering her question about molecular structure existing models for explaining molecular structure were used in preference. Even if students had “learned” the new material, they still retained their existing models. Thus, there seems to be an issue here in encouraging students to assimilate and apply new information.

Associated difficulties

In learning about covalent bonds students also find out about the shapes of molecules and that almost all covalent bonds are polarised. In addition students are presented with “rules” of combination, for example, the “Octet rule” which predicts, in a limited way, the maximum number of electrons permitted in any atomic orbital. Thus, besides learning the basic chemical idea about electrons being shared, students are also expected to assimilate many other associated concepts. In their work with Australian 17 year olds, Peterson and Treagust (1989) found that students’ ideas developed during an advanced chemistry course, but their progress was often accompanied by misconceptions about these associated areas. For example, they found that 23% of 17 year olds thought that electrons were equally shared in all covalent bonds, while about one-quarter attributed the shape of molecules to repulsion between the bonding pairs of electrons, or to bond polarity. Only about 60% of students knew the correct position of the electron pair in a bond between hydrogen and fluorine. The same question asked of first year university students studying chemistry (Peterson, 1993) yielded a 55% correct response, implying that most students who learn about bond polarity retain their knowledge.

Ionic bonds

The basic ideas associated with ionic bond formation involve the transfer of electron(s) between two electrically neutral atoms to make ions with overall positive and negative charges. The number of electrons transferred or accepted by an atom is related to the valency of the element. The positive and negative charges are “all over” the ions, so depending on the packing arrangements ions form ionic bonds with more than one ion of opposite charge at a time, forming a giant structure we call a crystal.

Students find ionic bonding hard to learn, describe and explain

Emerging evidence suggests the topic is problematic for students and that these difficulties could present significant obstacles to understanding. Barker’s (1995) study provides preliminary evidence for students’ difficulties from a rather broadly phrased question probing the formation of ionic bonds between sodium and chlorine atoms. The question comprised a diagram of a gas jar containing chlorine into which a piece of hot sodium metal was lowered together with a description of the reaction. Students were asked to explain what was happening in the jar. At the beginning of the study, about 20% gave answers suggesting they knew about ionic bonds, including the response “an electron is transferred from sodium to chlorine and a stable compound forms”. A further 54% at this stage suggested simply that sodium and chlorine are “reacting” or “forming a compound”. By the end of the survey, despite receiving teaching during the intervening fifteen months, these figures were only 34% and 48% respectively, compared to much higher figures (reported above) for covalent bonding.

At a more specific level, Taber’s interview work (1993a and b) with Annie also indicates problems. Annie began her post-16 chemistry course by recognising a class of bonds found between metals and non-metals she called “ionic”. Annie could not recognise the bond type present in a diagrammatic representation of a sodium chloride crystal, describing this as “just sodium and chlorine atoms” arranged “in rows” (p 18). Taber summarises her view of sodium chloride:-

“… the structure is held together, but without any bonding; there are charges on the neutral atoms; atoms are combining without overlapping; and the atoms are exchanging not just electrons but force pulls related to the electronic configuration.” (p 19)

In her second interview, Annie identified the ions in sodium chloride, but used the term “molecule” to describe ionic substances, as though the elements combine to form discrete particles just as carbon and hydrogen atoms combine to form a methane molecule. Annie knew that when ions combine, the overall effect produces something neutral. In her final interview, Annie recognised that electron transfer is involved in ionic bonding, but she remained confused about whether any sort of bonding existed in sodium chloride, explaining:-

“… it’s almost like they’re mixed but they haven’t combined. I think they’re held together just by the attraction of their forces in effect.” (p 23)

Annie knew that positive and negative charges implied attraction, but could not describe accurately their role in the sodium chloride structure. Barker’s responses suggest that 16-17 year old chemists cannot describe ionic bonding accurately, while Taber’s work provides detailed evidence explaining why this could be. Further details of students’ problems are discussed.

Ionic compounds form discrete molecules

Butts and Smith (1987) report the results of twenty-eight interviews with 17 year old Australian students who had studied chemical bonding. These students were asked to draw and explain the structure of sodium chloride. While most associated the compound with ionic bonding, many did not appreciate that ionic bonds are three-dimensional. Butts and Smith also report that some students consider sodium chloride to be molecular, suggesting that covalent bonds were present between sodium and chlorine, but that ionic bonds between molecules were needed to create the full structure. Taber (1994) suggests that students acquire this idea because they do not “share the framework of electrostatics knowledge” of the teacher, and also because they are taught about the formation of ionic bonds in a way which promotes the molecular model.

Students in the Australian study were asked to describe what would happen when sodium chloride was dissolved in water. All students responded that the particles would be dispersed, although some thought that sodium and chloride ions would still attract one another so there would be a “residual” structure in the water. Two students suggested that the salt would react with the water, forming sodium, chloride, hydrogen and hydroxide ions. Barker (1994) reports similar findings. She found that about 28% of students beginning post-16 courses and 40% of the same group completing their course intuitively visualised hydrochloric acid as hydrogen chloride molecules in solution. Students used the idea that the elements “swapped partners” with chlorine to explain hydrogen gas displacement on addition of magnesium metal. Extrapolating these responses suggests that magnesium chloride molecules in solution would be the product.

Taber (1998) found evidence indicating a possible explanation for this thinking. His detailed work led to the suggestion that students perceive ionic bond formation in terms of the electrovalency of the atoms involved. In this model, sodium chloride exists as molecules of “NaCl” because sodium and chlorine both have electrovalencies of one; a sodium atom loses one electron which is gained by a partner chlorine atom and the two ions form a discrete pair. Similarly, magnesium chloride exists as MgCl₂, because chlorine (valency one) combines with magnesium (valency two), allowing each magnesium atom to lose two electrons, one to each partner chlorine atom. The model means that students view ionic bond formation in the same way as covalent bond formation, with the key factor being the generation of “full electron shells”. Shells can be filled by sharing or transfer of electrons - either results in a discrete molecule, the formula being determined by the valencies of the elements. Taber reports one consequence of this - one student argued that a sodium ion could not form six ionic bonds unless the ion had a 6⁺ charge.

A “molecular framework” for ionic compounds

Taber continued his work on ionic bonding with a survey instrument administered to 370 students (1997b). These data led him to formulate a “molecular framework” which students use to describe ionic bonds. The framework comprises three conjectures called “valency”, “history” and “just forces”. The valency conjecture states that the number of ionic bonds an ion can form is determined by the electronic configuration. The history conjecture states that bonds can only form between atoms that have donated or accepted electrons. The “just forces” conjecture states that ions interact with other ions, but an ionic bond can only be formed between one sodium ion and one chloride ion (p 101), so these extra interactions are “just forces” not bonds. These imply belief that ionic compounds adopt a molecular structure like covalent molecules, but with ionic bonds between ions rather than covalent bonds between atoms.

Intermolecular bonds

Intermolecular bonds do not normally feature in pre-16 chemistry courses in the UK. Ideas about hydrogen bonding, other types of dipole-dipole bonds including those frequently termed “van der Waals’ forces” are taught in post-16 courses. The topic has received relatively little attention from chemical education researchers.

Hydrogen bonds

Hydrogen bonds arise when hydrogen is bonded to the highly electronegative elements fluorine, oxygen and nitrogen. For example, in hydrogen fluoride, the electrons in the covalent bond between hydrogen and fluorine are distributed towards the electronegative element, distorting the electron cloud and creating permanent positive and negative charges on the molecule, referred to as a “dipole”. The hydrogen nucleus contributes the positive charge and the distorted electron cloud around the fluorine atom takes a negative charge. The positive charge from one molecule may align with the negative charge on another, resulting in a specific type of electrostatic attraction called a “hydrogen bond”.

Progression in the development of basic ideas

Barker (1995) and Taber (1993a) have explored students’ thinking about hydrogen bonds. In Barker’s survey, 250 students beginning post-16 chemistry study were asked to identify the bonds between water molecules and to explain what distinguished these from covalent bonds. At the start, about 18% identified these as hydrogen bonds, increasing to about 69% fifteen months later. About 20% began by suggesting the bonds were “liquid” bonds or “ weak” bonds between molecules, possibly because a lack of formal teaching led to guessing from the diagrams provided. About 8% at the first stage described hydrogen bonds as “an attraction force, not a bond”. Fifteen months later, few students gave the “liquid/weak” bond response, but 24% gave the “attraction” description. This suggests that students learn to distinguish between intermolecular bonds and other types of bond, and ascribe these different properties. This is neither chemically accurate or necessary. Taber’s work with Annie (1993a) gives a more specific view of progression in understanding of hydrogen bonds. Annie was presented with a diagram representing a chain of hydrogen fluoride molecules. The molecules were shown with the appropriate distorted electron cloud, and were drawn touching one another. Annie did not think any bonding was present between the molecules. Taber suggests this may have been because the shapes did not overlap one another. In her second, post-teaching interview, Annie could describe the difference between the O-H bond within a water molecule and the bond between two water molecules:-

“You’ve got the two hydrogens added to an oxygen. And then the hydrogen brings a small bonding between like another oxygen, to hold the structure together but it’s not like, it is a bond, but it’s not as strong, as like, the ionic bond would be” (p 42).

In her third interview, Annie talked about hydrogen bonds involving lone pairs of electrons and demonstrated much clearer understanding of the intermolecular role of hydrogen bonding.

Other intermolecular bonds

Other, temporary dipoles arise because electrons continually move around within molecules. Temporary positive charges bond with temporary negative charges. This type of interaction can be called a “van der Waals’ force”. Each electrostatic attraction is small in energy terms, but when thousands or millions are being made and broken their effect on the structure and function of a substance is significant.

Barker explored students’ thinking about intermolecular bonds other than hydrogen bonds by asking students to explain why the vapour at 1000 ^oC above a mixture of titanium(IV) and magnesium chlorides comprised titanium(IV) chloride only, given that titanium(IV) chloride is “covalent” and magnesium chloride “ionic” in nature. At the start, only 1% of respondents suggested that intermolecular bonds between titanium(IV) chloride molecules would break, a figure which increased to 16% fifteen months later. Initially, students starting post-16 chemistry study divided into four groups. Those who thought that covalent substances have lower boiling points, so more heat was needed to vaporise the magnesium chloride numbered 22%. About 13% thought that ionic bonds can’t be broken by heating. Almost one-quarter (24%) suggested that covalent bonds are weaker than ionic bonds so break.

About one-third (33%) gave no response or an uncodeable response. By the end of the study these responses were still prevalent; the figures giving these answers were 14%, 15% and 31%, with 11% giving an uncodeable or no response. These data point to the widespread use of qualitative and vague ideas focusing on the behaviour of substances, despite the fact that the course followed by these students presented all intermolecular bonds in a chemically correct, context-led way.

At her first interview, Annie (Taber, 1993a) was asked about the structure of iodine. She explained that iodine molecules were held together by “forces of pressure”, not chemical bonds. After teaching, she was aware of the existence of van der Waals’ forces, and correctly placed these between iodine molecules, but thought that they would also occur in compounds like sodium chloride, as though she was applying them to any structure which she could not otherwise explain. Annie knew at this second stage that van der Waals’ forces would be affected by heat, but could not explain this in an accepted way. In her final interview, Annie retained the idea that van der Waals’ forces existed in sodium chloride, and realised that these bonds would break before covalent bonds when a substance was heated. Annie’s views support those reported in the large scale study.

Associated difficulties

In learning about intermolecular bonds some students develop misconceptions. One common error touched on by Annie and reported more formally by Peterson and Treagust (1987) is misunderstanding of the different locations of inter- and intramolecular bonds. About 23% of students thought that intermolecular bonds were within a covalent molecule. In his later study, Peterson (1993) found that 36% of first year university chemists thought that silicon carbide had a high melting point because of “strong intermolecular forces”.

Students also misunderstand the relative strengths of inter- and intramolecular bonds. Peterson and Treagust report that one-third of their sample of Australian sixth formers thought that “strong intermolecular forces exist in a continuous covalent network” (p 460).

Summary of key difficulties

1. Compounds with ionic bonds behave as simple molecules

Students see the formulae of ionic bonds written as “NaCl” or “MgCl₂”. There is no distinction between these formulae and “CH₄” or “H₂O”, which are mainly covalent compounds. The three-dimensional structure of compounds with mainly ionic bonding is ignored. Although this is chemical convention, students learning chemistry need help to realise that compounds with mainly ionic bonds behave differently from those with covalent bonds. For example, understanding that ions separate when a mainly ionic compound dissolves in a solvent, rather than the “molecule” staying together.

2. The central (first) element in a formula is responsible for bond formation

The convention for writing formulae contributes to the misconception that the first element in the formula is the more “powerful”. In methane, for example, carbon is perceived as “needing” four bonds, while hydrogen is the weaker partner with each “needing” only one bond.

3. Atoms “want” to form bonds

An extension of the idea that atoms “need” to form bonds is that atoms make decisions about making bonds. This reasoning may come from analogies such as “holding hands” or “finding a partner” used in teaching. This strategy causes problems later when students attempt to learn the role of energetics in bond formation.

4. There are only two types of bond – covalent and ionic

Pre-16 teaching focuses almost entirely on covalent and ionic bonds to the extent that students think that all bonds must be “ionic” or “covalent”. As the vast majority of chemical bonds fall between these two extremes or are intermolecular, this is unhelpful.

5. Covalent bonds are weaker than ionic bonds

Teaching presents differences between “covalent” and “ionic” compounds in terms of melting points, boiling points and physical states. Simplistic explanations ignore the role of intermolecular bonds, leading students towards, for example, poor models for explaining changes of state.

Suggested activities^[1]

1. Explore students’ understanding of simple events

Water boiling, sodium chloride and sugar dissolving, ice melting, iodine subliming and propanone evaporating can all be used to investigate students’ thinking about chemical bonding. Make the events explicit by carrying them out in the students’ presence and using molecular models to probe thinking about which bonds break and form.

2. Use cognitive conflict to show why elements form different types of bond

Show students that bonding depends on atoms forming compounds in the most energetically favourable way. Make three grids with 2, 8, and 8 boxes in each, aligned as if to represent electrons. Make the boxes big enough so that a mini-chocolate bar will sit inside. To start, use eleven chocolates in one grid arranged in a 2.8.1 formation, representing the electrons in one atom of sodium and seventeen chocolates in a second grid arranged in a 2.8.7 formation to represent the electrons in one atom of chlorine. Say that when compounds are made, all the electron spaces are usually filled. How might sodium and chlorine atoms arrange electrons so all the spaces are filled? Invite a student to move the lone chocolate from the “sodium atom” to the “chlorine atom”. To earn a chocolate, a student must then reason what would happen if magnesium replaced sodium? This time, make a “magnesium atom” by placing twelve chocolates in a 2.8.2 arrangement in the grid used for sodium, but keep chlorine the same. Ask the same question as before – how might magnesium and chlorine atoms arrange electrons so all the spaces are filled? Students will avoid the chemist’s answer, saying for example that an extra space is created or that the extra electron is “lost”. Eventually someone will realise that an extra atom of chlorine is needed. At this point bring out the third grid, showing that the unit formula for magnesium chloride is MgCl₂. The discussion can then be extended to show how different ways of filling the electron spaces can be used, depending on the most energetically favourable situation. Electron transfer may be preferred in some reactions, while electron sharing is used to form other compounds. Extension further allows discussion of physical states of common compounds such as methane, water and sodium chloride. These three can be used to introduce covalent, ionic and polar covalent bonds, with the additional idea that in practice most compounds fall between the extremes of ionic and covalent bond types.

3. Use electrostatics to explain bond formation

Describe the particles involved in chemical bond formation, stating always which is negatively and which positively charged. For example, electrons in a covalent bond are negatively charged and the bond forms between electrons and positively charged nuclei. An ionic bond comprises positively and negatively charged ions attracting each other in a three-dimensional arrangement. An intermolecular bond forms between the positive charge region on one molecule and negative charge region on another. Using this type of language helps students to focus on the particles involved in chemical bonding, rather than the bulk properties of the compounds. This approach is helpful to students making a transition towards electron orbital interactions which are used at university level to explain these ideas.

4. Be consistent in using “bonding” terminology

A range of different terms is used to describe bonds in English, particularly intermolecular bonds. These include “van der Waals forces”, “London forces”, “attractive forces” and “attractions”. There is no need for this. Such language over-complicates the picture. Terms such as “induced dipole-dipole bonds” and “permanent dipole- permanent dipole bonds” are much more descriptive, as these explain clearly the kind of interaction involved. Students can then work out (or use the term because they have already worked out) the particles involved in making these bonds, rather than be left with a memory exercise. Also, this language means that consistency can be introduced in discussing relative bond energies – chemists do not talk about “relative attractive force” energies! Using language which supports energetics teaching is helpful to students developing correct thinking.

5. And two things to avoid

Teachers working with pre-16 year olds tend to over-rely on the “octet rule” as an infallible tool for students to use in determining formulae and bonding. This contributes to students’ problems with ionic bonding, because they use this (or maybe are taught to) as a technique to determine the formulae of all compounds. In teaching ionic bonds, the rule is applied to show that some atoms “can fill their shells” by electron transfer, instead of electron sharing. The implication is that an ionic bond forms between oppositely charged ions combining to make a molecule, such as “NaCl”. This formula satisfies the octet rule, and teaching may end there, leaving students with Taber’s “molecular framework”. As a direct result students cannot fully understand how crystalline lattices form, the behaviour of acidic solutions and the influence ionic bonds have on melting point. This is even before the inert gas compounds are considered.

Use of anthropomorphic analogies to explain how bonds form should also be avoided. These only give students false ideas about atoms “wanting” to form bonds, or “needing” a certain number of bonds, or “finding a partner”. The analogy is patronising – far better to make students think about the chemical elements as chemicals than “living” organisms.