Teaching hydrogen bridges: it is not FON anymore!

1 Introduction

Hydrogen bonds or hydrogen bridges (we use H-bridges in this article) are important in all fields of chemistry and materials science, but the teaching of this concept with antiquated information leaves students at all levels with many misconceptions (Lamoureux, Chaves-Carballo, & Arias-Alvarez, 2021). Perhaps the most unfortunate heuristic that persists in the twenty-first century is that H-bridges only form among the elements of fluorine, oxygen and nitrogen (FON). The power of Pauling to define this concept in his seminal book ‘The Nature of the Chemical Bond’ influenced many generations of chemists.

Referring to the electronegativity scale, we might expect that fluorine, oxygen, nitrogen, and chlorine would possess this ability, to an extent decreasing in this order. It is found empirically that fluorine forms very strong hydrogen bonds, oxygen weaker ones, and nitrogen still weaker ones. Although it has the same electronegativity as nitrogen, chlorine has only a very small hydrogen-bond-forming power; this may be attributed to its large size (relative to nitrogen), which causes its electrostatic interactions to be weaker than those of nitrogen (Pauling, 1960).

The definition of H-bridges from the International Union of Pure and Applied Chemistry (IUPAC) in the twentieth century did not clarify this misconception. “Both electronegative atoms are usually (but not necessarily) from the first row of the Periodic Table, i.e., N, O, or F” (Muller, 1994).

Although this heuristic be incomplete and confusing, at some point the IUPAC definition in some textbooks was transformed to ‘H-bridges only form with the elements F, O, or N’ (and variations thereof). Students who memorize this rule at an early stage might become disappointed when it does not always work.

The researcher asked these students about the relationship between H, F, O, and N atoms and hydrogen bonding. They stated that when the topic of hydrogen bonding was mentioned in chemistry classes, these atoms would commonly be mentioned; they were thus familiar with them, and they had the perception that only H, F, O, and N atoms could form a hydrogen bond. The participants said that this perception is called ‘H, F, O, N rule’ … After analyzing the obtained data, it was determined that most of the students who employed this rule for solving the problem did this unconsciously and thus gave the wrong answer (Karakoyun & Asiltürk, 2020).

To learn more about how students at the Universidad de Costa Rica think about H-bridges, we gave a survey to a class of 100 s-year students in organic chemistry for health sciences (see Supplementary Information). In this survey, 98 % of the students replied that they learned about H-bridges in first-year chemistry at the university level. ‘The only elements that form H-bridges are F,O,N’ was affirmed by 84 % of the students. Interestingly, when asked to represent an H-bridge, most students (75 %) drew a single Lewis structure of water without trying to indicate any interaction between molecules. A question about H-bridges that do not contain F,O,N was answered by the majority (89 %) as a blank reply. When asked if it would be difficult to re-learn the concept of H-bridges, 30 % of the students considered it difficult because it conflicted with the previously learned FON rule.

Many experimental examples in the literature showed that H-bridges are not limited to these elements (Alkorta, Rozas, & Elguero, 1998); it was with a sense of urgency that a new definition should be created. IUPAC codified into the definition of H-bridges in 2011 that the range of elements and functional groups is not as simple as once taught.

The original examples of hydrogen bonding were found to involve the electronegative atoms F, O, or N. The current IUPAC definition given in the ‘Gold Book’ still specifically mentions these atoms, although it adds a caveat suggesting that the phenomenon is not limited to these atoms … Clearly, it has been realized that X may be any element having electronegativity larger than that of H (F, N, O, C, P, S, Cl, Se, Br, and I) and Y could be any of these elements and also π-electrons (Arunan et al., 2011b).

The use of the heuristic (F,O,N) continues in the literature and textbooks to this day (Brown et al., 2022; Chang, Overby, 2019; Petrucci, Herring, Madura, Bissonette, 2017). There are several problems with continued use of this heuristic:

1. Indicating only these three elements oversimplifies H-bridges.

2. If students memorize this heuristic as unconditional in early chemistry classes, they might face cognitive dissonance when they encounter common exceptions or non-conventional cases (Alkorta et al., 1998) in higher-level classes.

3. The heuristic does not clarify the difference between H-bridge donors (HBD) and H-bridge acceptors (HBA). For instance, X–H⋯Y–Z might differ from Y–H⋯X–Z, in the sense that one interaction might be more significant than the other.

4. The heuristic does not differentiate between intermolecular and intramolecular H-bridges.

5. Most examples used to illustrate this rule of thumb are the jejune dimers of H–F, NH₃ and H₂O, which are not as simple as they are shown (Herschlag & Pinney, 2018).

6. Too many students prefer a simple textual heuristic over structural examples, but learning about H-bridges requires the ability to visualize structures at a symbolic and molecular level to describe and to predict systems at a macroscopic level.

We noted in a previous article the limitations of using only these three elements and have proposed a new system to organize the H-bridge interactions in grids (Lamoureux et al., 2021). Students need practical assistance in visualizing the H-bridge interactions; grids are useful to organize these many interactions using structures of the complexes. Previously published textbooks and articles about H-bridges that rely on textual descriptions might cause misconceptions in the understanding, whereas we consider organized images more helpful from a symbolic perspective. In this article, we propose that textbook authors and lecturers should be cognizant of the 2011 IUPAC definition and the use of grids to represent better these H-bridges. The first step is to communicate that the (F,O,N) heuristic lacks utility, which this article concludes.

2 F in H-bridges

There is only one important HBD connected to F: i.e. H–F. One example containing fluorine is not useful to create a general heuristic. H–F is used in most textbooks and, by extrapolation, most students believe H–F is always an HBD. This rule is not invariably true. In some cases, such as NH₃ as HBA and F–H as HBD in dissolution, it is probable that a proton transfer occurs to form the salt (⁺NH₄ F⁻), as HF can react as a weak acid – pK _a is important for H-bridges or proton transfer (Gilli, Pretto, Bertolasi, & Gilli, 2009). In other cases with weak HBA, the H-bridged complex is insignificant (Figure 1). The focus on only HF devalues the HBD ability of HCl, HBr and HI. It is not pedagogically sound to oversimplify the selection of elements as the X–H group in HBD.

Figure 1:

HF is not invariably a HBD.

Is H–F an HBA only when it interacts with another H–F HBD (F–H⋯F–H)? That example is certainly the most commonly shown, although H–Cl (Figure 2) also acts as an HBD to HF (HBA) in the gaseous phase; Cl–H⋯F–H and F–H⋯Cl–H are almost isoenergetic (Johnson & Tschumper, 2018). There are a few more examples of HF as HBA with experimental crystallographic evidence, although they are neither common nor easily predictable (Wiechert, Mootz, & Dahlems, 1997).

Figure 2:

HF acting as a HBA.

When HF is mixed with H₂O (Figure 3), both the H–F and the H₂O dimers should be shown; the most significant H-bridged complex that forms, however, is the F–H⋯OH₂ complex. It should be indicated that HO–H⋯FH is unlikely to form as the distance between atoms is much larger than the alternative combination (Gardner, Al-Halabi, & Kroes, 2004). Hydrogen fluoride is a better HBD than water and H₂O is a better HBA than HF.

Figure 3:

The H-bridged complexes possible with HF in H₂O.

As there are many more R–F molecules than H–F, what about organofluorines as HBA? Fluorine covalently bound to carbon is an extremely weak HBA (Dunitz & Taylor, 1997). The evidence from X-ray crystallography show that R–F only rarely acts as an HBA (Allen, Cole, & Verdonk, 2012). As we reported (Lamoureux & Chaves-Carballo, 2022), the ability depends on the situation. The literature seems to indicate that intermolecular H-bridges (O–H⋯F–R) are not significant whereas intramolecular cases of 5- and 6-membered complexes between O–H and F–R have experimental and calculational support (Lamoureux & Chaves-Carballo, 2022). These intramolecular examples, however, are the exceptions to the rule and should be taught only in advanced classes because the evidence might vary over time (Cole & Taylor, 2022; Vulpetti & Dalvit, 2021).

In inorganic chemistry, one learns that the fluoride ion (F⁻) is an excellent HBA. There are also many examples of metal-bound fluoride (F-M) as significant HBA (Figure 4) (Desiraju & Steiner, 2001). When F is connected to an inorganic anion complex (Grabowski, 2020), for example, ⁻BF₄ or ⁻PF₆, the H-bridge ability diminishes although they are much stronger HBA than neutral organic F–C or F–B, in which the interaction is unlikely (Figure 4).

Figure 4:

Inorganic examples of F as a HBA.

Both our experimental results and the abundant information available in the CSD concur to indicate that the inorganic fluorine atom in ^–PF₆ forms weak hydrogen bonds with a variety of donors. With respect to organic fluorine, the X-H⋯F(δ-)(P) interaction is ‘charge assisted’, i.e., the weak interaction is seemingly reinforced by the different ionic charge carried by anion and cations (Grepioni et al., 1998).

What should one teach about F in H-bridges? The situation is multitudinously complicated; the species that contain F are manifold and various, and each type might have a separate HBA significance. H–F is a significant HBD, for example, and a poor HBA. It is better for the students to identify HBD and HBA when they are presented with a known H-bridged complex, and to predict whether the H-bridge is significant. The blind use of the heuristic that F forms H-bridges without considering its participation as HBD or HBA leads to incorrect predictions in most possible cases, and as such fails as a useful tool to teach chemistry.

3 O in H-bridges

Oxygen seems to be a natural choice as an element to illustrate the formation of H-bridges and water the perfect example of a molecule that forms intermolecular OH⋯O interactions. These examples are undoubtedly an important part of the story of H-bridges, but not the only part. In advanced cases, oxygen has both advantages and drawbacks in forming H-bridges (Cramer, Sager, & Ernst, 2019) and the presence of oxygen in an organic solvent does not ensure that an H-bridge is formed (Tessensohn, Lee, Hirao, & Webster, 2015).

It is important to use the current definitions from IUPAC when teaching about organic solvents with oxygen. In this sense, the terms ‘protic’ and ‘aprotic’ are deprecated and should be replaced by HBA or HBD classification.

It is recommended to classify solvents according to their capability to donate or not donate, as well as to accept or not accept, hydrogen bonds to or from the solute, as follows:

Hydrogen-bond donating solvents (short: HBD solvents), formerly protic solvents, now ‘protogenic solvents.’

Non-hydrogen-bond donating solvents (short: non-HBD solvents), formerly aprotic solvents.

Hydrogen-bond accepting solvents (short: HBA solvents), now ‘protophilic solvents.’

Non-hydrogen-bond accepting solvents (short: non-HBA solvents)” (Perrin et al., 2022).

The diverse interactions of oxygen in H-bridges are organized in Figure 5. In the first column, one finds no polar solvent (‘polar’ defined as relative permittivity, ε, greater than ∼15) (Perrin et al., 2022) that contains oxygen and that is neither an HBD nor an HBA. Among non-polar solvents (ε less than 15), furan (ε ∼ 2.9) is not an HBD and insignificantly accepts H-bridges. In a search of the Cambridge Structural Database in which 118 instances of a furan ring are present, for example, the oxygen is an HBA in only three cases (Allen, et al., 2012). Hexachloropropanone (ε ∼ 3.9) contains no hydrogen and, as an HBA, it has about the same acceptor properties as benzene (i.e., unlikely) (Laurence, Mansour, Vuluga, Planchat, & Legros, 2021). The classification of furan and hexachloropropanone as non-polar, non-HBD and non-HBA solvents is hence apt.

Figure 5:

Grid dividing solvents containing oxygen into polar or non-polar and with various HBD or HBA properties.

In the second column in Figure 5, both polar and non-polar oxygenated solvents can be only HBA. Among propanone (ε ∼ 21), dimethylsulfoxide (DMSO, ε ∼ 47), diethyl ether (ε ∼ 4) and tetrahydrofuran (THF, ε ∼ 7.5), each lack an HBD and can form no H-bridge among the same molecules but each can form H-bridges with, for instance, water as an HBD.

The alcohols in column 3 in Figure 5 – 1,1,1,3,3,3-hexafluoropropan-2-ol (hexafluoroisopropanol, HFIP) and 1,1,1,3,3,3-hexafluoro-2-(trifluoromethyl)propan-2-ol (nonafluoro-t-butanol, NFTB) – are examples of solvents that can donate H-bridges but are poor acceptors. The value of ε for the polar HFIP is ∼17 whereas the non-polar NFTB is expected to have an ε value below 15 (Motiwala et al., 2022); both solvents are considered poor acceptors (Laurence et al., 2021). The presence of trifluoromethyl groups obviously affects the acceptor ability; once again the use of the heuristic ‘O forms H-bridges’ does not tell the entire story.

The most common examples of HBD/HBA solvents in organic chemistry are shown in the fourth column in Figure 5. Both methanol (ε ∼ 33) and ethanol (ε ∼ 25) are polar and can easily H-bridge with themselves – or with other alcohols – and these solvents can be used for most general uses in which an HBD/HBA solvent is required. The alcohols 2-methyl-2-propanol (ε ∼ 12) and 1-octanol (ε ∼ 10) should, however, be portrayed as non-polar solvents even though they are HBD/HBA solvents.

The presence of H-bridging ability (HBD or HBA, or lack thereof) is neither necessary nor sufficient to predict the physical properties of these solvents. In Figure 5, there is no obvious correlation between the H-bridge ability and boiling or melting point. Water solubility is also related to several factors. All the ‘polar’ solvents are miscible with H₂O, but some of the ‘non-polar’ solvents – THF, NFTB and 2-methyl-2-propanol – are also miscible. Furan, hexachloropropanone, and diethyl ether are only slightly miscible in water at near 300 K, whereas 1-octanol is considered immiscible.

As a further example, two oxygen atomic nuclei in the same molecule are compared (Figure 6). Oxygen is only an HBA in carboxylate anions. Carboxylic acids have one oxygen that is an HBA but non-HBD yet the other oxygen is HBD and non-HBA. Esters are known to H-bridge only on the carbonyl oxygen as the other oxygen shows no experimental evidence of being an HBA (Allen, et al., 2012). To an untrained eye, all these molecules look alike and should have the same H-bridging properties. The prediction by students that all oxygens act as HBA is due to the oversimplified heuristic ‘O forms H-bridges’.

Figure 6:

Carboxylate anions, carboxylic acids and esters have varied HBA/HBD oxygens.

4 N in H-bridges

With the assertion of (F,O,N) H-bridges, some textbooks still use NH₃ as a prime example of H-bridging. An article in the journal Science titled “Does Ammonia Hydrogen Bond?” (Nelson, Fraser, & Klemperer, 1987) directly investigates this case. In this article they indicate data contrary to the rule that ammonia forms H-bridges.

Given that HF and H₂O hydrogen bond, their crystal structures as well as their gas-phase dimer structures can be predicted; hence the hydrogen-bonding concept is valuable in these systems. In contrast, the concept is misleading as a predictor of the NH₃ dimer orientation and separation distance and not very illuminating as a rationalization of the NH₃ crystal structure. For these reasons it does not seem useful to view NH₃ as a hydrogen-bond donor, particularly in its gas phase or pair interactions. If NH₃ is to be classified as a hydrogen-bond donor, it must be considered a very poor donor, forming weaker, longer, and less linear hydrogen bonds than even HCCH, CF₃H and H₂S.

The structures of the complexes of NH₃ suggest that NH₃ is an active hydrogen-bond acceptor but is quite reluctant to donate hydrogen bonds. This gas-phase view of NH₃ interactions is in sharp contrast with the traditional view of NH₃ interactions described in most freshman chemistry textbooks (Nelson et al., 1987).

In a subsequent publication, the same author (Klemperer, 1995) as of the Science paper writes:

Every chemist is taught that the hydrides of the first row-HF, H₂O, and NH₃-are hydrogen bonded. This has been the classic explanation of the unobvious observation that their boiling points are higher than the corresponding second-row hydrides- HCl, H₂S, PH₃. The spectroscopy of (HF)₂ and (H₂O)₂ immediately established the dimeric structures as hydrogen bonded. The hydrogen bond in both of these species is quite close to an X–H–X angle of 180°. From the outset the ammonia dimer was quite different than either chemical intuition or electron structure calculations predicted. There is little evidence for a strong, highly directional, nearly linear hydrogen bond (Klemperer, 1995).

It seems even recent papers have begun to rectify what was once thought as an obvious model. Liquid ammonia, both experimentally (Luehrs, Brown, & Godbole, 1989) and according to simulations using quantum chemistry (Tongraar, Kerdcharoen, & Hannongbua, 2006), has been shown to donate H-bridges only weakly. In the gaseous phase, the ammonia dimer is not the same as a water dimer.

Guided by analogy or chemical intuition, one might guess that the ammonia dimer possesses a ‘classical’ quasi-linear hydrogen bond similar to other dimers such as those of water or hydrogen fluoride … Thanks to sophisticated experiments and extensive computations there is now consensus that the ammonia dimer [in the gas phase] is a fluxional molecular complex with an equilibrium structure that is characterized by a bent hydrogen bond … Contrary to earlier conceptions (as presented in many standard textbooks on that subject) the spatial arrangement of nitrogen atoms showed that no extended hydrogen bonded network exists in liquid ammonia. Nevertheless, some degree of hydrogen bonding was inferred from the temperature dependence of the N – H and H – H radial distribution functions. However, the hydrogen bond interaction in liquid ammonia proved to be much weaker than that in water and no clear hydrogen bond peak was observed in either N – H or H – H correlations, unlike the case of water (Boese, Chandra, Martin, & Marx, 2003).

Most recent data indicate that ammonia as HBD forms only less significant H-bridges: “Does ammonia hydrogen bond? According to the criteria established here, among the complexes investigated only the NH₃ dimer is [weakly] hydrogen bonding.” (Grein, 2021).

In hindsight, one might predict this result (Figure 7) by the large separation in pK _a values between ammonia (∼34) and the ammonium ion (∼9) (Herschlag & Pinney, 2018). This example is a case in which textbooks and chemical intuition provide a simple, but erroneous, result. Textbook authors and lecturers should learn from this error and correct this portrayal of ammonia H-bridges. The first step is to reject the use of a general heuristic and show structures of concrete examples.

Figure 7:

NH₃ is a poor HBD.

The inadequate ability of N–H to act as an HBD is also evident in hydrazine (Ju & Xiao, 2002), organic amines (Marten et al., 1996), and even intramolecular cases of aminoalcohols and aminoethers (Siewert, Zherikova, & Verevkin, 2022). For ethanolamine (Figure 8), the most stable conformation has the N as an HBA to the O–H HBD; a minor conformer has a weak N–H⋯OR H-bridge. As the chain length between groups increases, the N–H⋯O interaction becomes unlikely (Siewert et al., 2022).

Figure 8:

Intramolecular H-bridges in aminoalcohols.

All these data for NH₃ do not mean that N–H is never an HBD nor that N is always an HBA. One must take into consideration, in an examination case-by-case, whether the species containing nitrogen is an HBD or HBA. Shown in Figure 9, a primary amide has a nitrogen that is a non-HBA but an HBD, a counterpoint to that of NH₃. In imidazole, with two distinct nitrogens, one is a non-HBA whereas the other is an HBA (Figure 9). These examples are two common species encountered in biochemistry; their H-bridging ability should not be undermined by what is taught in general chemistry.

Figure 9:

Nitrogens in an amide (left) and imidazole (right).

Once again, the situation with nitrogen is complicated; ‘N forms H-bridges’ does not reflect the nuanced chemistry. When a heuristic fails to reflect a general tendency, even for the simplest examples, removal of the heuristic is timely.

5 No FON

At this point, it is pertinent to distinguish not only the significance of H-bridges but also their prevalence. Water has both significant and common H-bridges HOH⋯OH₂. The HF dimer, although significant, is uncommon in a practical sense. Ammonia H-bridges, although perhaps common in general chemistry, have been shown to be less significant (vide supra). Instructors should strive to present both the most significant and the most prevalent H-bridges, but the most appropriate examples depend on the level of the student. In introductory classes, common possibilities in a range should be explored so that students can relate non-covalent forces to their daily lives.

The most relevant cases are not limited to (F,O,N). HBD could be C–H, P–H, S–H, Cl–H, Se–H, Br–H, or I–H. HBA could be C, P, S, Cl, Se, Br, I or also alkenes, alkynes or aromatic rings. The hydrogen sulfide dimer (HS–H⋯SH₂) contains no (F,O,N); it might be considered less significant even though pertinent during teaching sulfur chemistry. In biological or medicinal chemistry systems, there are many examples of C–H⋯SR interactions (Fargher, Sherbow, Haley, Johnson, & Pluth, 2022; Suzuki, Matsubara, Nakata, Ito, & Noguchi, 2022). Unconventional (i.e. beyond F,O,N) examples of C–H⋯P interactions can be introduced during teaching the chemistry of phosphorus (Hansen, Du, & Kjaergaard, 2014).

As shown above, proposing an H-bridge interaction among molecules requires expertise, critical thinking, and previous knowledge about the HBD and HBA character of its participants. In an introductory chemistry course, the student’s abilities are limited to recognize, identify, define, or remember a concept, according to Bloom’s taxonomy of cognitive levels. In this sense, it is illogical to expect a first-year student to draw correctly all H-bridge interactions without more in-depth tutelage: he/she has been asked to evaluate and apply criteria that he/she does not yet have. When students lack a solid background, the FON rule becomes a law that leaves no chance for future questions, and might conflict with the understanding of topics in higher level courses. This is the reason why we recommend that, in introductory courses, H-bridges should be taught so the students can recognize and identify them in a series of examples, as well as the HBA and HBD participation of the involved atoms.

6 Should H-bridges be ‘Explained’?

There is a tendency to try to explain the phenomenon of H-bridges in terms of deeper knowledge at the same time as presenting the concept of H-bridges. We contend that explanations of H-bridge are not necessary in introductory chemistry and cause an overload of information. First, teach the description of H-bridges and allow students to predict their presence and significance. The theory behind the concept can be postponed to advanced courses.

The complications of the theory of H-bridges can be seen from the published IUPAC description. Can an instructor legitimately expect an introductory student, especially a student who does not take chemistry as a major subject, to understand these subtleties?

The forces involved in the formation of a hydrogen bond include those of an electrostatic origin, those arising from charge transfer between the donor and acceptor leading to partial covalent bond formation between H and Y, and those originating from dispersion (Arunan et al., 2011b).

There are several other possible explanations in the literature for the ‘nature’ of H-bridges (Oliverira, 2015; van der Lubbe & Fonseca Guerra, 2019). Some incompatible theories appear below.

1. Electrostatic/Coulombic: Pauling wrote that the hydrogen bond is electrostatic (ionic) in nature: “hydrogen-bond formation must be due largely to ionic forces.” (Pauling, 1960) This statement was repeated in a recent physical-organic chemistry textbook. “Since the hydrogen bond is a simple Coulombic interaction, any partial negative charge can accept a hydrogen bond, not just electronegative atoms, but even π systems.” (Anslyn & Dougherty, 2006).

2. Covalent: “Coulson, Del Bene and Pople, Dannenberg, Gilli et al., Weinhold and Landis, and many others have highlighted the importance of a partial covalent nature in the hydrogen bond.” (Arunan et al., 2011a).

3. Dipole-Dipole: “a special type of dipole-dipole interaction.” (Brown et al., 2022).

4. Resonance covalency using charge transfer based on Natural Bond Orbitals (NBO) (Weinhold & Klein, 2014; Weinhold, 2023).

5. A combination of electrostatic, exchange-repulsion, induction, and dispersion forces determined by symmetry-adapted perturbation theory (SAPT) calculations (Stone, 2017; Tafipolsky, 2016), in which charge transfer + polarization = induction (Scheiner, 2018).

6. Molecular electrostatic potential with polarization forms a Coulombic σ-hole interaction (Murray & Politzer, 2019).

7. Intermediate between unprotonated and completely protonated HBA (Arunan et al., 2011a).

8. H-bridge acceptors do not depend on polarizability, but rather “charge capacity” (Murray, Seybold, & Politzer, 2021). Fluorine has small polarizability but large electronegativity; its charge capacity to form H-bridges depends on a small distance, such as intermolecular H-bridges. The other halogens (with the possible exception of chlorine) do not have this ability, even though they are more polarizable.

How can one teach these conflicting views to students? Should only one viewpoint be taught and the rest ignored, or should all possibilities be taught? The best way, as indicated by IUPAC, is to focus on the phenomenon, not the explanation.

Thus, defining a hydrogen bond as ‘no more than a particularly strong type of directional dipole-dipole interaction’ is certainly incomplete. As Buckingham wrote in a book ‘The hydrogen bond results from inter-atomic forces that probably should not be divided into components, although no doubt electrostatic and overlap interactions are the principal ingredients’ … Clearly, no single physical force can be attributed to hydrogen bonding. There have been numerous attempts to decompose the hydrogen-bonding interaction into electrostatic, polarization, charge transfer, dispersion, and exchange repulsion (Arunan et al., 2011a).

We do not recommend teaching any of these theories to explain H-bridges to introductory students, so as to avoid trying to define the ‘origin’ or ‘nature’ of H-bridges. The best policy for textbook authors is omission over oversimplification (Bell, 2015). The nature of H-bridges is complicated and still under discussion (van der Lubbe & Fonseca Guerra, 2019). Trying to explain this complication to a student in an introductory chemistry course without relying on oversimplified or misleading ideas is perhaps too challenging and so the ‘explanation’ of H-bridges should be removed from the curricula.

7 Conclusions

The traditional approach to teaching H-bridges is imperfect, as shown by the misconceptions and alternative models that students in the twenty-first century still expound (Henderleiter, Smart, Anderson, & Elian, 2001; Schmidt, Kaufmann, & Treagust, 2009; Widarti, Marfuaf, & Retosari, 2019). Textbooks should be modernized to resolve these known problems (Gültepe, 2021). How can one improve this pedagogy? By removing all points of confusion as early as possible with the latest information, by not oversimplifying the topic with misleading heuristics, and by avoiding convoluted theories. At an introductory level, it is more important to describe correctly the phenomenon than to explain it.

The introduction to the topic of H-bridges should begin with an indication that they are temporary ‘bridges’ and not ‘covalent bonds’. The IUPAC definition of HBD and HBA, with examples, should then be provided to encourage the idea that there are at least two parts to H-bridges. Simple examples (both symbolic and molecular models, not just a rule of thumb) of dimers should be provided, linked to the experimental evidence (macroscopic models) of their existence. Grids are convenient to show whether an HBD⋯HBA interaction is significant (Lamoureux et al., 2021). It appears that even though the IUPAC definition of H-bridges changed in 2011, modern students and chemistry educators still rely on old information; it would be best pedagogically to start university chemistry classes without the FON heuristic – and to use more visual examples – to avoid confusion in later courses.