Reading Nutrients by Their Light

The oldest nutrition advice — “eat the rainbow” — turns out to be a spectroscopy instruction. A plant pigment’s colour is a direct read-out of its molecular structure, and that structure is its phytochemical class. This paper separates what the light actually tells us Verified from the organizing metaphor built on top of it Model and the health claims that are not yet earned Open.

1Colour is chemistry Verified — textbook

A pigment’s colour is what it does not absorb. Molecules with extended chains of alternating double bonds (conjugated π-systems) absorb visible light; the longer the conjugation, the longer the absorbed wavelength (λ_max), and the redder the pigment appears. This is Beer–Lambert absorption applied to real food molecules — the same physics in every spectrophotometer.

So the visible-to-UV spectrum is not a metaphor for a nutrient classification; it is one. The major dietary pigment classes sort cleanly by λ_max:

Absorbed wavelength (nm), ultraviolet → infrared. A pigment appears the complementary colour to the band it absorbs.

Pigment class	Absorbs (λmax, approx.)	Appears	Structure
Flavonoids / flavonols (e.g. quercetin)	~330–370 nm (UV)	colourless–pale	polyphenol
Chlorophylls a/b	~430 & ~660 nm (blue + red)	green	Mg-porphyrin
Carotenoids (β-carotene, lutein)	~450–470 nm (blue)	yellow–orange	conjugated polyene
Lycopene	~470–505 nm (longer conjugation)	red	conjugated polyene (11 C=C)
Anthocyanins (flavonoid)	~510–540 nm, pH-dependent	red–purple–blue	flavylium polyphenol
Betalains (beet)	~480 & ~535 nm	magenta	betalamic-acid conjugate

λ_max values are solvent- and pH-dependent and given here as textbook approximations; pin exact figures to a reference (or measure them) before any are quoted as firm. The ordering — more conjugation → redder — is robust.

2“Eat the rainbow” — a real basis, an honest limit

Verified The mapping colour band → pigment class → phytochemical family is genuine chemistry: a purple food is telling you it carries anthocyanins; a deep-orange one, carotenoids. Colour diversity is a fair proxy for phytochemical diversity.

Open What colour diversity does not establish is a causal health benefit per pigment. The evidence that eating a variety of plants is good for you is largely epidemiological and associational — strong as a dietary heuristic, weak as “this purple compound treats that condition.” This paper makes no therapeutic claim; a colourful diet is food, not medicine, and any specific health effect stays Open until a trial closes it.

3Beyond colour: the functional-group fingerprint Verified — analytical chemistry

UV-Vis reads pigments. But most nutrients — fats, proteins, sugars, fibre, moisture — are colourless, and they are read further down the spectrum. Infrared (FTIR, and DRIFTS — diffuse-reflectance IR, the same method used here to characterise fuels) and near-infrared (NIR) spectroscopy fingerprint molecular bonds: C–H, O–H, N–H, C=O. From those bands, composition — fat / protein / moisture / carbohydrate — can be quantified non-destructively, in seconds, without reagents.

NIR compositional analysis is standard, validated practice across the food and agriculture industries. The point that matters for this programme: the spectroscopy G6 already runs on biomass-to-fuel is the same toolset that screens a formulated food for its nutrient and moisture profile — one instrument, two tracks.

4Crystal — food is a crystallization problem Verified — food science

Below the pigments, much of a food’s texture, gloss, snap and shelf-life is set by which crystals form, and in which polymorph. The clearest case sits right in the bar: cocoa butter is polymorphic — it can solidify into six distinct crystal forms, and only Form V gives chocolate its gloss and clean snap. Tempering is the controlled seeding of that one crystal form; skip it and you get the dull, streaky “bloom” of the wrong polymorph. Sugar (grainy vs. smooth), honey’s glucose crystallizing out (why raw honey granulates), starch’s semi-crystalline granules (gelatinization and the retrogradation that stales bread), and ice-crystal size in anything frozen are all crystallography. For a honey–nut–cocoa bar, the crystal state of the fat and the sugar is the texture and much of the stability. It resonates with the series’ crystal-lattice chapters — except here the lattice is edible and the claim is textbook.

5Phyllotaxis — the plant’s geometry Verified — geometry Model — nutrition link

Verified Many edible plants arrange their parts by the golden angle (~137.5°): the Fibonacci spirals of a sunflower head, a pinecone, a pineapple, the florets of Romanesco. That is real botany and the real mathematics of the series’ φ chapter — the golden angle packs seeds and leaves to minimise overlap and maximise light capture.

Model The nutritional link is indirect, and I tag it as such. The geometry optimises the plant’s photosynthesis and seed-packing — its capacity to make and store nutrients densely — not a property of the food once it’s on your plate. Eating a spiral doesn’t transfer the spiral’s virtue. The honest statement: the same golden-angle packing that fills a sunflower head is why it is a dense store of oil and protein — geometry as the reason for the density, not a nutrient in itself.

Other examples, same rule. The Fibonacci spiral is all over the produce aisle — a pineapple's hexagonal eyes tile in 8-and-13 spirals; an artichoke's bracts spiral; Romanesco does it as a fractal (§6). Each is gorgeous botany, and each carries its nutrition elsewhere — the pineapple's vitamin C and bromelain, the artichoke's fibre and cynarin, are biochemistry the spiral never touches. The clean line worth keeping: colour is chemistry you can see (§1); shape is packing you shouldn't overread; and the colourless middle — fats, protein, vitamin C, bromelain — needs the instrument (§3).

6Fractality — surface area is the nutrient interface Verified — structure Model — nutrition link

Verified Self-similar branching recurs wherever biology must pack maximal surface area into finite volume: Romanesco broccoli is a near-perfect natural fractal; roots, leaf venation, lungs, and the intestinal villi and microvilli that absorb your food are all fractal-like.

Model The link to nutrition is real but specific: fractal branching maximises the surface area of exchange. In the gut, villi multiply absorptive area by orders of magnitude; in food, surface area sets digestion and dissolution rate (a powder digests faster than a block — the same mass, more interface). What is not supported is any claim that a fractal-shaped food is “more nutritious” by virtue of its self-similarity. The defensible statement is about surface area and kinetics, not a mystical property of the shape.

7The classification layer Model

Organising nutrients along a single spectral axis — a “periodic table of nutrients” read off their light — is an organizing model, not a physical law. It is useful for teaching and for structuring a screening pipeline, and it rests on the verified spectroscopy of §§1–3. The dm³ contact-geometric framing sits here, as the modelling layer, and is tagged Model throughout — the absorption chemistry underneath it is what carries the weight.

8What this is — and isn’t

Verified The physics (Beer–Lambert, conjugation → λ_max), the pigment assignments, the crystallization and polymorphism of fats and sugars, the golden-angle geometry of phyllotaxis, and the DRIFTS/NIR compositional methods are all established science. Model The single-spectral-axis “nutrient spectrum,” and the nutritional readings of phyllotaxis and fractality, are organizing models built on that verified base. Open Health-outcome claims from colour, shape, or diversity are associational and not settled. No claim here is therapeutic, and none is stated as proven beyond what a spectrophotometer, or a citation, actually supports.