Arrange the families in a circle by what they are built from, and position starts to do the remembering for you.
~13 min read
You come out of functional group families holding a dozen families and a dozen sets of behaviour. That is a list, and a list has one weakness: the only way to use it is to recall it. Forget which side of the fence oxides sit on and there is only a gap.
The chemistry wheel is the fix. Arrange those same families around a circle — not alphabetically, not by how often they turn up, but by the properties that cause their behaviour. Once they are in that order, position tells you most of what the list had to. Nobody found molecules arranged in a circle; the wheel is a drawing somebody made, because the families do vary along two properties at once and a circle shows two on one page. The families are real. The arrangement is an arrangement.
A model, not a law
The wheel organises tendencies. It has real exceptions, the families it draws are broader than they look, and different authors draw it differently. A model presented as a model is the stronger tool: you can see its edges and stop trusting it there.
The version drawn here uses the two axes that carry the most information. Nothing from terpenes or functional groups changes — this only puts it somewhere.
A circle on its own is decoration. It becomes an instrument when both directions across it mean something you could in principle measure a molecule for.
The vertical axis is electrical character. Every molecule in an essential oil is built from carbon and hydrogen, and most carry one oxygen somewhere. Carbon and hydrogen pull on shared electrons about equally, so a bond between them is electrically flat. Oxygen does not share evenly — it drags electron density towards itself and leaves a slightly positive region behind, a polar bond in the sense chemical bonding uses. Where that oxygen sits is what this axis measures.
The electropositive pole — the top
Molecules that give. A phenol carries an oxygen–hydrogen group on an aromatic ring, and that hydrogen comes away more readily than any other in oil chemistry: it is genuinely acidic. Alcohols have the same group on a plain carbon chain and hold it far more tightly, so they sit high but not at the extreme. Hydrocarbons sit near the top too, with nothing to hold a negative pole.
The electronegative pole — the bottom
Molecules that take. The oxygen is locked into a group with no hydrogen to give away: a carbonyl, an ester linkage, an oxygen bridged inside a ring. It sits as a concentrated pocket of negative charge, accepting a hydrogen bond and never donating one. That asymmetry is the bottom half of the wheel.
The band between them
Moderate rather than neutral. Alcohols above the line and esters below it are the least extreme electrically, and the mildest to handle. That is not a coincidence.
The horizontal axis is molecular size and volatility. Left is small and quick, right is large and slow, and this one is close to a real measurement. A monoterpene is ten carbons: limonene is C10H16, molecular weight 136, boiling around 176 °C. A sesquiterpene is fifteen: beta-caryophyllene is C15H24, molecular weight 204, boiling somewhere above 260 °C. Heavier molecules attract each other more strongly, need more energy to break free of the liquid, and so evaporate more slowly. That is all volatility is.
Diagram to come
The two axes shown separately, as two straight rulers stacked one above the other, BEFORE they are crossed into a circle. The point of the drawing is that each axis stands on its own and means something on its own. Top ruler: electrical character, running from Electropositive at the left end to Electronegative at the right, with four small molecule sketches sitting along it in order — a phenol ring with its OH, a straight-chain alcohol with its OH, a ketone with its carbonyl in the middle of a chain, an aldehyde with its carbonyl at the end. Bottom ruler: size and volatility, running from Small and volatile to Large and tenacious, with three carbon skeletons along it in order — a ten-carbon monoterpene, a fifteen-carbon sesquiterpene, and a bracket at the far right marked as the range steam distillation no longer reaches. Underneath both, one line of type saying that crossing these two rulers at right angles is what makes the wheel.
A liquid in an open bottle is not still. Its molecules move at a range of speeds, and at any moment a few at the surface are moving fast enough to break away into the air. Vapour pressure measures how many. Boiling point is the temperature at which that escape becomes wholesale rather than occasional, and because it is far easier to measure, it is the number chemists quote as a stand-in for volatility.
Two things decide it. The first is mass, which is why the horizontal axis works at all. The second is how strongly the molecules stick to each other — and that is where the vertical axis leaks into the horizontal one. Compare limonene and carvacrol: both ten-carbon molecules of almost the same size, 136 against 150. Limonene boils at about 176 °C. Carvacrol boils at about 237 °C, and its close relative thymol, at the same molecular weight, is a solid at room temperature that does not melt until around 50 °C. Sixty degrees of difference on molecules of the same size, and the whole of that gap is the oxygen–hydrogen group letting each molecule bond to its neighbours.
You can watch this without equipment. One drop each of three oils on three paper strips, the time written on each. The strip bare in ten minutes and the strip still faintly scented next morning are the two ends of the horizontal axis, measured on your kitchen table.
Diagram to come
The central diagram of the Science course. Drawn in code, sized to fill the width of a phone screen and stay legible there. STRUCTURE: one circle drawn as a wide ring rather than a filled disc, with two axes crossing at the centre and running out beyond the ring to their end labels. The vertical axis is titled Electrical character, Electropositive at the top, Electronegative at the bottom. The horizontal axis is titled Size and volatility, Small and volatile at the left, Large and tenacious at the right. Neither axis carries numbers or tick marks: these are directions, not scales, and the drawing must not imply a measurement it cannot support. FAMILIES: ten family names sit inside the ring at clock positions, each occupying a short wedge rather than a point, because a family covers a region. Clockwise: 12 o’clock Phenols; 1 Sesquiterpenes; 2 Sesquiterpenols; 4 Lactones and coumarins; 5 Esters; 6 Aldehydes; 7 Ketones; 8 Oxides and ethers; 9 Monoterpenols; 10 to 11 Monoterpenes. Under each family name, in smaller type, one example constituent, in the same order: carvacrol, beta-caryophyllene, patchoulol, bergapten, linalyl acetate, citral, menthone, 1,8-cineole, linalool, limonene. THE EDGE: 3 o’clock carries no family. The ring fades outward there into a shaded margin labelled Past here, too heavy to distil — resins, waxes and the heaviest plant acids lie beyond the drawing, and the wheel should be seen to have an edge rather than running on forever. THE ONE ADMISSION: phenols at 12 o’clock carry a short arrow pointing a little clockwise, labelled Pulled right by hydrogen bonding, because a phenol is as small as a monoterpene and boils far higher. It is the single place where the drawing admits its own two axes are not independent, and it must not be tidied away. RIM WORDS: outside the ring, at the eight compass points, the property words, set in a lighter weight so they read as consequences rather than labels — reading clockwise from the top: Donates a hydrogen, Most reactive with protein, Aroma arrives late, Stays for hours or days, Sweet and soft-edged, Oxygen fully exposed, Handle with the most care, Penetrating and camphorous, Aroma arrives first, Evaporates first, Oxidises in the bottle. COLOUR: design tokens only — the electropositive half warming towards terracotta at the top, the electronegative half cooling towards slate at the bottom, the ring itself parchment, all type in soil. TWO STATES: the plain wheel, and a version in which named wedges can be shaded, because the four-oils diagram further down reuses this same drawing with regions highlighted.
Start at the top and go clockwise.
Twelve o’clock — phenols
The electropositive pole. Carvacrol, thymol, eugenol: an aromatic ring with an oxygen–hydrogen group on it, acidic enough to matter. The most reactive family on the wheel, and how phenols are handled follows from that.
One and two o’clock — sesquiterpenes, then sesquiterpenols
Fifteen carbons instead of ten. Beta-caryophyllene at 204, patchoulol around 220. Still electropositive, heavy, slow to leave — the woods and the resins.
Three o’clock — nothing
The edge of the drawing. Go far enough right and a molecule is too heavy for steam to carry — the triterpene acids, the waxes, the gums stay in the plant material, which is why distillation gives something chemically different from the resin or peel.
Four and five o’clock — lactones and coumarins, then esters
Below the line, electronegative. A lactone is an ester closed into a ring; a furocoumarin such as bergapten is heavier again at 216. Plain esters — linalyl acetate at 196 — sit further round. Sweet, soft-edged, among the gentlest families.
Six and seven o’clock — aldehydes, then ketones
Both carry a carbon–oxygen double bond; only the position differs. In an aldehyde it is at the end of the chain, exposed on one side; in a ketone it is in the middle, shielded by carbon. Small to look at, large in consequence — which is why aldehydes sit at the pole and ketones just off it.
Eight o’clock — oxides and ethers
The oxygen tucked inside a ring, no hydrogen of its own, carbon on both sides. Electronegative, but the mildest kind. 1,8-cineole is the famous one, at 154 and boiling around 176 °C — polar and unusually quick to leave. Penetrating, camphorous, gone fast.
Nine, ten and eleven o’clock — monoterpenols, then monoterpenes
Back above the line and back to ten carbons. Linalool, geraniol, terpinen-4-ol, menthol — all at 154 or 156, each with one donatable hydrogen. Then the hydrocarbons, the lightest and quickest on the wheel.
Some neighbours are neighbours for a real reason. Monoterpenes sit beside phenols because in oregano and thyme gamma-terpinene converts to para-cymene, and para-cymene to carvacrol and thymol — one pathway, three stops, all of them in the same bottle. Others are adjacent only because a circle has to close, and some close relatives end up far apart: linalool and linalyl acetate are one esterification apart, yet the electrical axis puts alcohols nearly opposite esters.
Here is the payoff. Know roughly where a constituent sits and you can say, before looking anything up, how quickly it evaporates, how it smells, how carefully it needs handling, and how long it stays detectable.
| Where on the wheel | Volatility | Handling | Aroma | Lingers for |
|---|---|---|---|---|
| Upper left — monoterpenes | Highest. From about 155 °C | Well tolerated, still diluted. Oxidises with age, harsher once oxidised | Fresh, sharp, citrus or pine | Minutes to an hour |
| Left — monoterpenols | High. Around 200–230 °C | Among the mildest. Dilute as normal | Soft, green, floral, woody | One to a few hours |
| Top — phenols | Lower than their size suggests. 230–250 °C | The most care of any family. Heavier dilution, never near eyes or ears | Pungent, hot, herbaceous | Several hours |
| Lower left — oxides and ketones | High. Cineole around 176 °C | Moderate. Dilute as normal; labels carry their own cautions | Penetrating, camphorous | Under an hour to a few |
| Bottom — aldehydes | Moderate, lower than their size suggests | Care. Reactive with protein, and a common reason a label asks for more dilution | Intense, lemony or spicy | Several hours |
| Lower right — esters and lactones | Moderate to low | Esters among the gentlest. Furocoumarins bring the sunlight caution | Sweet, fruity, soft-edged | Hours; furocoumarins never leave |
| Right — sesquiterpenes | Lowest on the wheel. Above 260 °C | Gentlest in most cases. Dilute as normal | Deep, woody, resinous | Many hours to days |
The poles are the reactive ends
Reactivity does not increase steadily down the wheel. It peaks at both electrical extremes and is lowest in the band between. Phenols at the top and aldehydes at the bottom need the most care; alcohols and esters, either side of the middle, need the least. If you take one rule off this diagram, take that.
The horizontal axis is one fact stated four ways: why monoterpenes are the first thing you smell and the first gone, why a wood is still faintly there next day, why a citrus is a top note and a frankincense a base note, why the heaviest plant compounds never reach a bottle at all.
Diagram to come
One molecule, three fates, arranged as three panels sharing a single molecule sketch at the top so the reader sees that one property is producing all three outcomes. The molecule is drawn once with its polar oxygen region shaded and its nonpolar carbon skeleton plain. Panel one, evaporation: the same molecule at a liquid surface, with dotted hydrogen bonds tethering it to its neighbours — polarity is what holds it in the liquid. Panel two, absorption: a simplified cross-section of skin as a brick-and-mortar wall, flattened cells as bricks and lipid layers as mortar, with a strongly nonpolar molecule lodged in the lipid and going no further, a strongly polar molecule sitting on the surface unable to enter the lipid at all, and a moderately polar molecule passing through. Panel three, reactivity: the polarised bond drawn with its partial charges marked, and a protein surface approaching the positive end. No arrows implying an effect in the body — panel three stops at the chemistry.
A polar molecule is one where the electrons are unevenly shared, leaving one region slightly negative and another slightly positive. It is a small imbalance — nothing like a full charge — but enough to decide what a molecule will sit next to. Water is strongly polar, and polar things dissolve in it. The membranes around every cell in your body are the opposite: two layers of fatty tails facing each other, an oily sheet with water on both sides. A molecule has to get on with both to go anywhere interesting, and how well it does is set almost entirely by the vertical axis.
Chemists put a number on this by shaking a compound up with oil and water and measuring how it divides. The result is a partition coefficient, and most oil constituents land in a band where all of them prefer oil to water, some far more strongly than others. That number predicts absorption through skin better than almost anything else, and with a middle-preferring shape rather than a straight line. Skin is flattened cells set in lipid, like bricks in mortar. A compound that barely likes oil never gets into the mortar; one that likes oil enormously gets in and stays there, going nowhere. The ones that travel are in between — which is why the moderate families are both the mildest to handle and the most reliably absorbed. Those measurements come from laboratory work: isolated skin samples and synthetic membranes in a dish, which is not a living person with circulation under the skin. Size gates the same door, and every constituent on this wheel is small enough to pass it.
The third consequence is reactivity. A polarised bond is an open bond: it has an electron-poor end that something electron-rich can attack, and proteins are covered in electron-rich groups. In laboratory chemistry, aldehydes react readily with the amine groups on proteins and phenols disrupt protein structure directly — measured on isolated proteins and on cells in a dish, not in people. That laboratory reactivity is the chemical reason those two families sit at the top of every dilution table, and it stops there. It is a statement about molecules meeting protein on a bench, not about what anything does in a body. See safety for what it means at the shelf.
One property, then. It holds the molecule in the liquid, decides whether it crosses a membrane, and decides whether it reacts with what it meets. Evaporation, absorption and irritation are three readings off the same dial, which is why one axis can carry all three.
A wheel is only worth drawing if you can put something on it. Take two oils from opposite ends of it and see what position gets right.
Lemon — 60–75% limonene, and limonene is a monoterpene hydrocarbon. One family, one place: ten o’clock, hard left, high on the electropositive side. Almost the entire bottle sits in a single wedge.
And one thing the wheel predicts that the list does not mention. Lemon is cold pressed from the peel, not steam distilled, so nothing was filtered out by weight: the heavy furocoumarins at four o’clock come along in the bottle, where steam would have left them behind. That is the chemical reason expressed citrus carries a sunlight caution. After applying lemon to skin, avoid direct sunlight and UV for up to 12 hours, and dilute in a carrier oil as with any topical use.
Oregano — carvacrol and thymol, with the para-cymene and gamma-terpinene they came from still in the bottle. Twelve o’clock with a tail running back to eleven: the electropositive pole itself.
Lavender — 25–45% linalyl acetate, 20–47% linalool, 0.3–10% ocimene. An ester, an alcohol and a trace hydrocarbon. This oil does not sit in one place; it sits in two, and they are nearly opposite. A middle note: linalool boils around 198 °C and linalyl acetate around 220 °C, both slower than lemon and both quicker than a wood. Both families sit in the moderate band either side of the middle, away from both electrical poles, which makes it undemanding — dilute in a carrier oil for topical use, as always. Sweet and soft-edged rather than sharp, which is the ester fraction doing that. What the wheel cannot predict is the range: linalool is recorded anywhere from 20% to 47%, and where a batch falls depends on where and when the plant was grown.
Frankincense — alpha-pinene and limonene alongside sesquiterpenes, with boswellic acids named among its constituents, distilled from resin rather than leaf or peel. Two places again, and this time the two ends of the horizontal axis at once. A base note that opens bright: the monoterpenes at ten o’clock reach you first and are gone within the hour, the sesquiterpenes at one o’clock arrive underneath and are still there at the end of the evening. No phenols, no aldehydes, nothing at either electrical pole. Still dilute in a carrier oil. The boswellic acids are the heaviest thing on that list, out at the right-hand edge where volatility runs out — the heavier a compound, the less of it steam carries over and the more stays behind in the resin. An oil and the material it was distilled from are related, not identical.
Diagram to come
The same wheel drawn four times in a two-by-two grid, small, with axis labels dropped to a minimum so the shapes carry the comparison. On each copy the wedges that oil occupies are shaded, and the shading is weighted — a darker tone where most of the bottle sits, lighter where a minor fraction does. Lemon: one dark wedge at ten o’clock, plus a small, deliberately visible mark out at four o’clock for the furocoumarins only a cold-pressed oil carries. Lavender: two wedges of similar weight at nine and at five, nearly opposite each other, with a faint mark at ten for the ocimene. Oregano: one dark wedge at twelve with a lighter tail back to eleven, drawn so the tail reads as a pathway rather than a second ingredient. Frankincense: two wedges of similar weight at ten and at one, plus a mark in the shaded margin beyond three o’clock for the resin acids. Under each, the oil’s name and its aromatic note. A reader should see at a glance that one oil is a point and three are spreads.
Which oils are on sale differs from market to market, so these are a method rather than a shopping list. It transfers to your own shelf: read the constituent list, name the families, find the positions, and read the shape.
A model earns trust by being specific about where it fails. It fails in three places, and knowing them is the difference between using the wheel and being used by it.
Small constituents can run the whole bottle
The wheel weights a family by how much of it there is, which is often the wrong weighting. Detection thresholds differ between compounds by orders of magnitude, so a constituent well under 1% can decide what an oil smells like. Safety works the same way: the furocoumarins that put a sunlight caution on a whole bottle of expressed citrus are a trace fraction of it. A big wedge is not an important wedge.
Most oils sit in two places
Of the oils above, only lemon is close to a point. Lavender straddles the electrical axis, frankincense the volatility axis, and both are ordinary in that. Circles suggest points; real oils are shapes. A family is broader than its wedge, too — alcohols covers menthol at 156 and patchoulol at 222. The family sets a tendency; the oil sets the rule; the label settles it.
The wheel stops at the skin
Everything on it is about a molecule getting somewhere — evaporating, dissolving, crossing, reacting on contact — and nothing about what happens next. How a compound is changed by the body, where it ends up, how quickly it is cleared: none of that is readable off a position. It predicts arrival, not fate.
There is also no single chemistry wheel. The best-known ancestor comes from the French aromatherapy tradition and uses a different horizontal axis — water-loving against fat-loving rather than small against large. Other versions swap families in and out, put alcohols next to esters, or add a third dimension. If you meet a wheel unlike this one, it is probably not wrong. Read its axes first.
You already accept this elsewhere without noticing. The periodic table is a model, and a famously leaky one — hydrogen sits above a column it has almost nothing in common with, the lanthanides are parked underneath because they would not fit. Nobody proposes abandoning it. It is the most productive diagram in chemistry precisely because it makes a good first guess about an element you have never met.
That is the right standard: not whether a model is true, but whether it makes you less wrong more often than having nothing. A wheel that gets you to the right handling caution four times out of five is far more useful than no wheel, provided you check the fifth. The danger is never that a model is approximate. It is that the approximation becomes invisible — that you stop seeing a drawing somebody made and start seeing how molecules are.
So use it in the order it is built for. The wheel generates the first guess. The constituent profile refines it. The label on the bottle settles it. Three steps, and the wheel is only ever the first — which is what makes reading a constituent profile the lesson that has to come next.
Most people meet the wheel while trying to build a blend, and this is where it is most practical.
Top note, middle note and base note are a volatility statement. A top note is a molecule at the left edge of the wheel: light, low-boiling, reaching you first because it leaves the blend first, and gone first for the same reason. A base note sits at the right edge: heavy, high-boiling, arriving late and still there when everything else has gone. You can tell which an unfamiliar oil will be from its constituent list alone — which is why three citrus oils arrive together, leave together, and leave nothing behind them.
Three drops of lemon, two of lavender, one of frankincense is that shape exactly: ten o’clock, the middle band, one o’clock. Bright at the start, soft through the middle, resinous at the end, and far longer-lasting in a diffuser than lemon alone. For a topical version, the same ratio diluted in a carrier oil — and because there is expressed citrus in it, avoid direct sunlight and UV for up to 12 hours after applying it to skin. Build your own in blends.
Keep oils away from eyes and inner ears, and out of reach of children. Dilute in a carrier oil for topical use, and dilute hot oils such as oregano far more heavily. If you are pregnant, nursing, under medical care or taking medication, speak to your healthcare practitioner before use. Only take an oil internally if the label on your bottle says it is for internal use. Nothing here is intended to diagnose, treat, cure or prevent any disease.
Lesson 9 of 18 · Science