More than ‘pour-and-mix’ – Extending Content Knowledge at the college level through an analysis of coumarin in cinnamon

The several faces of Content Knowledge

The course’s layout has to some degree been motivated by students’ unsatisfactory retention of knowledge from their preceding courses, which might be interpreted from an extended view on professional knowledge for teachers. Professional knowledge has traditionally been suggested to comprise Content Knowledge, Pedagogical Content Knowledge and Pedagogical Knowledge (Shulman, 1987). Content Knowledge undoubtedly is essential to developing Pedagogical Content Knowledge (e.g., Abell, 2007) so that lack of the former will negatively affect the latter.

Therefore, it is essential for future teachers to be well educated in terms of Content Knowledge to lay the foundation for their Pedagogical Content Knowledge. This is the distinctive knowledge type of subject teachers, which supports them in selecting appropriate Content Knowledge for their students and which advises them how to convey it to students, i.e., by considering typical student difficulties and by employing suitable methods and scaffolds. At the same time, there has been a growing feeling that Content Knowledge as a concept might be too broad to precisely capture the depth of knowledge needed by teachers: Some suggest distinguishing Academic Content Knowledge (e.g., Gess-Newsome et al., 2019), which typically is referred to in universities, from School-related Content Knowledge (e.g., Dreher, Lindmeier, Heinze, & Niemand, 2018; Hermanns & Keller, 2021) outlining that portion of Content Knowledge typically taught in schools or needed by the teacher to comprehensively understand the compulsory teaching content in the curriculum.

Nixon, Toerin and Luft (2019) introduce three facets of Content Knowledge that do not suggest a hierarchy of knowledges but that progressively widen the scope of knowledge. Core Content Knowledge refers to fundamental concepts of a discipline that everybody in it should know (e.g., nature of the chemical reaction, acid-base-theories, models of the atom, etc.). Specialized Content Knowledge extends beyond this, describing knowledge that is only needed in select instances of a discipline – laboratory techniques might arguably be subsumed under this heading (e.g., theoretical chemists need not be proficient titrators to do their jobs). Lastly, Linked Content Knowledge provides knowledge that allows to relate knowledge between sub-disciplines (e.g., from inorganic and physical chemistry: explaining the precipitation of silver-I-ions with chloride as a function of AgCl’s solubility product) as well as beyond the discipline’s boundaries. This latter kind of knowledge enables teachers to relate the ‘purely chemical’ to their students’ lives – staying with the example of silver halides, this would have entailed (in a pre-DigiCam era) to bring the processes of black-and-white-photography to the classroom.

The proposed college course aims at (1) recapitulating Core Content Knowledge from preceding teaching at college from introductory lectures, (2) at introducing Specialized Content Knowledge in the form of analytical procedures, and (3) at equipping students with Linked Content Knowledge. This combination of knowledges is intended to result in their synthesis into one coherent application, i.e., students understand the relevance of why and how an analytical approach works. This is decidedly different from the ubiquitous pour-and-mix-recipes frequently found in books on experimentation, where the primary focus is on producing an effect but not explaining what happens (as mentioned above, this kind of black-box-experience cannot be satisfactory when educating teachers). Moreover, Linked Content Knowledge provides extra-disciplinary links to help seeing an every-day relevance of the chemical content. This, then, could support students in their later chemistry classes to relate chemistry knowledge to their own students’ everyday lives and experiences, which is arguably beneficial to learning (e.g., Childs, Hayes, & O’Dwyer, 2015). Table 1 provides a sketch of how the three types of knowledge complement each other in the course before illustrating the idea in a concrete lab-investigation – entries in the table can only be exemplary and no exhaustive syllabus should be expected.

Core Content Knowledge – fluorescence of coumarin

Coumarin (Figure 2) is the δ-lacton of o-coumaric acid – which itself is the 2-hydroxy derivative of cinnamic acid – and as such belongs to the benzopyranone-family. It crystallizes in long white needles that dissolve easily in ethanol but only little in water (Abernethy, 1969). Chemical analyses for coumarin in food most often use HPLC (Ballin & Sørensen, 2014; Sproll, Ruge, Andlauer, Godelmann, & Lachenmeier, 2008). Liquid-phase- (Miller, Poole, & Chichila, 1995), thin layer- and gas chromatography-mass spectrometry are also frequently employed (Woehrlin, Fry, Abraham, & Preiss-Weigert, 2010); UV-VIS-spectrometry or fluorometry might be used after adjusting to pH > 9 (Haskins & Gorz, 1957; Rose et al., 2015). Coumarin is known to react with alkalines by opening of the lacton-ring yielding a cis-configured anion of o-coumaric acid (Haskins & Gorz, 1957). The cis-isomer can be brought to its trans-configuration by irradiation with UV-A (λ = 365 nm) (Abernethy, 1969; Haskins & Gorz, 1957; Marbach, Harel, & Mayer, 1983; Schwarze & Hoffschmidt, 1959). Returning to the cis-configuration gives blueish-green fluorescence with a maximum intensity at λ = 490 nm (Rose et al., 2015 – see Figure 3 for the overall reaction).

Figure 2:

Structural formula of coumarin.

Figure 3:

Overall reaction of the analytical test for coumarin.

Linked Content Knowledge – coumarin and the Nature of Science

The Nature of Science denotes a concept derived from the philosophy and epistemology of science. It is concerned with describing what gives science a self-contained and idiosyncratic view of the world. It, therefore, tries to capture the defining aspects that distinguish scientific endeavors from non-scientific ones, be it research in the humanities, claims from esotericism and mysticism, or deliberation in theology.

Lederman (2007) suggests that school should attend to select characteristics of scientific knowledge as a comprehensive picture of the Nature of Science lies outside the scope of school teaching (see also McComas, 2020). These aspects should be discussed with students to prevent them from developing a positivist misconception of science: science is tentative and never definite, i.e., it is subject to change; science is creative in nature, i.e., at some point in time someone must have thought it up (also: subjective Nature of Science) as it cannot be read directly from nature; science is empirical, i.e., it derives from observation and investigation; science is socio-culturally embedded, i.e., it always is influenced by its generative context (e.g., banning of certain strings of research on ethical grounds, extensive funding of certain research for economic reasons). Lederman (2007), moreover, emphasizes that students need to be made aware of the difference between theories and laws in science as students frequently misjudge the latter to be a logical product of the former.

Other researchers choose to augment the Nature of Science (McComas & Olson, 1998; Osborne, Collins, Ratcliffe, Millar, & Duschl, 2003) by, e.g., emphasizing the role of hypotheses and predictions in science, or its gradual development through scrutiny by a peer community. The bulk of alternative suggestions can be paralleled with Lederman’s canon (Neumann & Kremer, 2013). In his earlier approaches to Nature of Science, Lederman focuses the ontology of science by addressing Nature of Scientific Knowledge, which only later is complemented by introducing the Nature of Scientific Inquiry which devotes itself much more to the epistemological aspects of science (Lederman & Lederman, 2012) – it is the former of these two perspectives that is concentrated on in the following paragraphs.

The college course is not primarily concerned with fostering abilities of scientific inquiry, but some aspects from Lederman’s list dealing with the Nature of Scientific Knowledge can be addressed in the analysis of coumarin in cinnamon (i.e., tentativeness, subjectiveness/creativeness, socio-cultural embeddedness). The analytical procedure and its chemistry, thereby, gain an immediate relevance for the students (college and school alike) who might otherwise be less interested in ‘some smelly heterocycle’ that occurs in cinnamon, which they are not too keen on to begin with (“Can’t stand the stench.”) … learning that there are coumarin-lean cinnamon alternatives and why they are less common might, however, kindle their interest in the odd cookie.

Coumarin in nature and trade – touching the socio-cultural Natures of Science

Sweet clover (Melilotus alba), woodruff (Galium odoratum) and tonka bean (Dipteryx odorata) exhibit considerably high coumarin contents which contribute much to their characteristic odor (the full aroma of, e.g., cinnamon species results from a more complex interplay of at least 20–40 substances (Jayaprakasha, Jagan Mohan Rao, & Sakariah, 2000, 2003; Marongiu et al., 2007; Miller et al., 1995; Senanayake, Lee, & Wills, 1978; Ter Heide, 1972)). Moreover, coumarin occurs in some cinnamon species of the cassia variety.

‘True’ or Ceylon cinnamon (Cinnamomum verum) is nearly void of coumarin (Miller et al., 1995) and weak in cinnamaldehyde, which is responsible for the pungent aroma defining cinnamon’s quality (Woehrlin et al., 2010). Cassia cinnamon is the head term for barks from several cinnamon trees: Chinese cinnamon (Cinnamomum cassia), Indonesian cinnamon (Cinnamomum burmanni), and Vietnamese cinnamon (Cinnamomum loureioi). Cassia cinnamon shows high contents of cinnamaldehyde and coumarin (Table 2). Content of components, however, may vary greatly between plant species, between plants of one species and even within the same plant according to age and/or locus (Woehrlin et al., 2010). Earlier claims of eugenol being absent from cassia or coumarin being found in true cinnamon in traces only (Miller et al., 1995) cannot be maintained according to more recent analyses (Woehrlin et al., 2010).

Coumarin content for woodruff and tonka bean, respectively, may far surpass that of cinnamon (1% of dried mass, 2–3% respectively; Abernethy, 1969). For both these spices, health adverse effects have long been known even to a wider lay public – aromatization of alcoholic beverages with woodruff (e.g., German Maiwein) might frequently result in headaches largely irrespective of the imbibed amount, recipes for preparing dishes with tonka beans usually draw attention to moderate addition of the spice.

Trade statistics for 2017 (Table 3) show that the bulk of traded cinnamon (by mass) stems from the cassia species, while the more expensive Ceylon cinnamon more than outweighs cassia in trade value. The reduced spiciness of Ceylon cinnamon (Blahová & Svobodová, 2012) – due to reduced content of cinnamaldehyde – makes it of lesser interest for the food industry who prefer processing cassia (Vitenskapskomiteen for mattrygghet [VKM], 2010).

This overview shows that, although there have been heated discussions on the tolerance of cassia (see below), it continues to be the dominant species in international trade. This is, on the one hand, due to its relative ‘cheapness’ which allows production of inexpensive consumer goods. On the other hand, cassia gives an improved spice-effect due to its being rich in cinnamaldehyde; i.e., industry needs to add less cassia to achieve a certain spiciness as compared to Ceylon cinnamon. Judging from Table 2, cassia cinnamon should spice equivalently at approximately 1/3 the mass of Ceylon cinnamon. Referring to import values from Table 3, this would come in at an estimated 17% of the cost incurred when adding Ceylon cinnamon. So, there is an indisputable socio-economic nature to how scientific knowledge develops. In this instance, it is economy that suggests reading the health-risk-assessments (see below) liberally instead of adopting their production to Ceylon cinnamon or encouraging to culture a new cinnamon species that were rich in cinnamaldehyde and, at the same time, lean in coumarin.

Bioactivity of coumarin – touching the tentative and creative Natures of Science

Coumarin has been discussed as being hepatotoxic (Abraham et al., 2010; Dinesh et al., 2015; Lake, 1999; Ratanasavanh et al., 1996; Woehrlin et al., 2010) and carcinogenic in humans (Abraham et al., 2010; Dinesh et al., 2015; Felter, Vassallo, Carlton, & Daston, 2006; Lake, 1999). The former issue draws on animal studies from which a TDI dose (tolerable daily intake) was derived. Based on the NOAEL dose (no observed adverse effects level) with beagle dogs, adjusted for interspecies variation (NOAEL x 0.1; since dogs react differently from humans) and intraspecies variation (again x 0.1; since no two humans react completely alike), the advice is to consume no more than 0.1 mg coumarin per kg body weight per day (Anton et al., 2004). The same TDI is given when incorporating human data from clinical case reports on sensitive humans for whom the NOAEL lies at 5 mg/kg/day (Abraham et al., 2010). The Norwegian Scientific Committee for Food Safety advises an even lower TDI of 0.07 mg/kg/day, drawing on hepatoxicity studies with rats; this TDI is rarely overstepped in various worst case intake scenarios (VKM, 2010). Other authors (e.g., Lake, 1999) dispute such doses as being irrationally restrictive. They criticize the data basis as irrelevant since it either comes from an intransparent sample of dogs (Felter et al., 2006), or resorts to studies on rats, which show an increased susceptibility to coumarin (Dinesh et al., 2015; Felter et al., 2006) due to a metabolism pathway different from the preferred 7-Hydroxycoumarin-pathway in humans (Felter et al., 2006; VKM, 2010). They, moreover, claim that evidence on human coumarin toxicity is inconclusive but, still, they concede that there is a sensitive subpopulation (Dinesh et al., 2015; Felter et al., 2006). The less susceptible majority, however, could tolerate – according to a conservative estimation relying on rat studies – daily doses of 0.64 mg/kg and probably even exceeded this (Felter et al., 2006).

Observed carcinogenicity by coumarin is not genotoxic in nature, i.e. DNA-reactive behavior has been ruled out for humans (Abraham et al., 2010; Anton et al., 2004; Dinesh et al., 2015; Felter et al., 2006). It is now assumed that carcinogenicity results from organ-specific toxicity (Abraham et al., 2010; Felter et al., 2006). Coumarin or coumarin-containing solutions have even been suggested as remedies in cancer treatment (Lacy & O’Kennedy, 2004) as well as for diabetes mellitus type II (Chen et al., 2012). Regarding the latter, experimental data are inconclusive and intake of these preparations is discouraged (Dinesh et al., 2015); coumarin has been administered for improving blood-flow in small veins despite not showing anti-coagulant activity in contrast to some of its derivatives (Felter et al., 2006; Jain & Joshi, 2012).

These discussions serve to illustrate that data from scientific investigations can be interpreted conflictingly. It always takes the personal perspective of a scientist to make sense of data or to refute the claims by others. There is a creative nature to science that cannot be completely separated from the ‘final bit of information’ that is communicated. Moreover, there is no such thing, as a ‘final’ bit – the above sketch shows that health risk assessment of coumarin has changed over time. It is liable to change again if new insights on health-adverse effects of coumarin need to be considered. Thus, scientific knowledge is tentative by nature.

Specialized Content Knowledge – the analytic procedure

Samples from cassia are compared with samples from Ceylon cinnamon. A serial dilution of pure coumarin in ethanol-water (4:1) allows for semiquantitative analysis. The analytical approach deviates from the more frequently used thin layer or HPL chromatographies (Sproll et al., 2008) and uses a simplified form of fluorometry (see Haskins & Gorz, 1957).

First, chipped cinnamon sticks (0.5 g) are ground in a mortar with sea sand so that relatively high content of coumarin from unhurt lignin structures (before grinding) might be expected. The grind is suspended in 50 mL of a 4:1 mixture of pure ethanol (w = 0.96) with water – do not use denatured alcohol as a solvent since additives, to discourage oral consumption, might show a fluorescence of their own. The suspension is stirred for 30 min (Sproll et al., 2008) and separated by filtration. The filtrate has a light-yellow, clear appearance (Figure 4). Adding one flake of KOH (approximately 0.2 g) to the solution produces precipitation and darkening of the solution (Figure 5). The alkaline reaction is left for 15 min. The forming precipitate is non-definable in composition, but most likely lignin-analogous structures are formed (McFadden & Ross-Morre, 2002). After centrifugation of the suspension, the supernatant is observed under a UV-A-lamp (λ = 365 nm).

Figure 4:

Ethanolic solutions of 0.5 g cassia (left) and 0.5 g Ceylon cinnamon (right).

Figure 5:

Precipitation after adding KOH; cassia (left), Ceylon cinnamon (right).

A serial dilution of coumarin in ethanol-water (50, 25, 10, 5 mg L⁻¹ – to 50 mL of each solution one flake of KOH is added) allows an estimate of coumarin content in the samples. Typically, cassia samples yield an intensive fluorescence, while Ceylon cinnamon does hardly/not fluoresce (Figure 6). The intensity of fluorescence with cassia is expected to lie between the 25 and 50 mg-preparations (Figure 7), as the 0.5 g of ‘average’ cassia extracted in 50 mL contain approximately 1.8 mg of coumarin (i.e., 36 mg L⁻¹; see Table 2 – caution: one must not directly compare fluorescence from Figure 6 with that in Figure 7 as both the photographs were taken with different apertures).

Figure 6:

Fluorescence under UV-A (λ = 365 nm), cassia (left), Ceylon cinnamon (right); aperture: 4.5, closure: 1 s.

Figure 7:

Fluorescence of dilution series of coumarin in ethanol under UV-A (λ = 365 nm): 50; 25; 10; 5 mg L⁻¹ (left to right); aperture: 6.3, closure: 1 s.

Fluorescence of the samples increases over time under UV-irradiation so that time of exposition to UV needs to be controlled (see Figure 8).

Figure 8:

Comparison of fluorescence of coumarin as a function of irradiation time (t_irr) under UV-A (λ = 365 nm) samples: 0.5 g cassia (left), Ceylon cinnamon (right) in 50 mL methanol: Water (4:1) (aperture: 6.3, closure: 1 s).

Limitations of the analytical procedure

As indicated above, fluorescence cannot be completely ruled out for Ceylon cinnamon. Analogous investigations with extracts of tonka beans may observe a weak blueish fluorescence already with the ethanolic extract which is likely attributable to Umbelliferone (7-Hydroxycoumarin) (Sullivan, 1982). This does not interfere with the dominant fluorescence from o-coumaric acid. Neither samples of eugenol (tested with oil of clove diluted to reflect eugenol contents in Cinnamomum verum), nor cinnamaldehyde (diluted to reflect average content in cassia, see Table 2) show fluorescence under the given conditions. It is, therefore, assumed that the procedure is specific for coumarin – at least, specific enough for the school science classroom.

Hazards

On principle, the procedure is poor in potential risks. Improved extraction is achieved with methanol (w = 0.8) (Sproll et al., 2008) but ethanol has been suggested here to reduce hazards. Ethanol should not be exposed to open flames as it is easily flammable. Flakes of KOH should be transferred with spoons. Students – in college and in schools – should wear protective gloves, safety goggles and a lab coat for the whole procedure. The centrifuge should be loaded equally as to prevent any imbalance in the rotor. UV-A-lamps should be positioned in a manner that make it unlikely for students to look directly into the source.