What is solid fat content, and how much of a butter is still solid at skin temperature?

Solid fat content is the percentage of a fat that is crystalline at a given temperature. Measured curves for cocoa butter, coconut and palm, and why blends go soft.

Solid fat content, or SFC, is the percentage of a fat that is crystalline at one stated temperature, usually measured by pulsed NMR. It is a curve, not a number, and it explains what a melting point hides. Cocoa butter measured by pulsed NMR was 70.5 percent solid at 20 C, 25.6 percent at 30 C, 10.3 percent at 35 C and fully liquid at 40 C. Coconut oil was 34.5 percent solid at 20 C and essentially gone by 30 C, at 0.17 percent. Palm oil was only 24 percent solid at 20 C but still carried 4 percent of crystals at 35 C. Those three shapes, a cliff, an early collapse and a long tail, are why one butter gives a balm that snaps and melts cleanly, another goes runny on a warm windowsill, and a third keeps some structure well above skin temperature. The fact that catches formulators out is that blends do not average. A mix of about 35 percent cocoa butter and 65 percent coconut oil melted at 16.7 C, below either fat on its own.

• Cocoa butter: 70.5% solid at 20 C, 10.3% at 35 C
• Coconut oil: 0.17% solid at 30 C
• Palm oil: 4.05% solid at 35 C
• Cocoa and coconut eutectic: 16.7 C
• NMR vs DSC: about 20 points apart

What solid fat content measures, and what a melting point hides

A natural butter is not one substance. It is a mixture of many triglycerides, each with its own melting behaviour, so it does not melt at a temperature. It melts over a range, losing crystals progressively as it warms. A melting point, a slip point or a clear point picks one moment in that range, usually near the end. Solid fat content describes the whole range: at each temperature, what fraction of the fat is still crystalline and what fraction is liquid oil.

That is the quantity that decides how a balm behaves. The crystals form the network that holds the liquid oil in place, as the oleogels and wax networks page explains for wax, and a butter adds a second network of fat crystals to the wax one. Firmness at any temperature depends on how much of that second network is still there, so two butters with similar melting points can feel quite different on the same day.

Measured SFC curves for common balm fats

The table gathers pulsed NMR results for the fats that turn up most in balms and for which open, fully reported data exist. The palm figures are not a single laboratory's result: they are certified values for reference materials that the Malaysian Palm Oil Board produced through inter-laboratory proficiency programmes, with an uncertainty at each temperature. The two cocoa butter columns are separate studies that used the same type of instrument.

The two cocoa butter studies agree within two percentage points at every temperature both measured, which is about as close as two laboratories get on a natural fat. The MPOB certificate puts the uncertainty for palm oil at between 0.14 and 1.82 points depending on temperature, so differences of a point or two between sources are noise. Differences of ten or twenty points are not, and usually come from the method, covered further down.

Three shapes of curve

Cocoa butter: a plateau and a cliff. It stays above 70 percent solid all the way from 0 to 20 C, then loses about 45 points between 20 and 30 C and the rest by 40 C. Below about 20 C it barely changes, so a cocoa butter balm feels much the same in a cold room and a cool one. Above it, every degree counts. That is the switch the lotion bar relies on: hard on the shelf, transferring under a warm palm.

Coconut oil: an early collapse. It starts high, 81 percent solid at 5 C, but falls steadily and is down to a quarter solid at 25 C. Between 25 and 30 C it loses almost everything that is left, finishing at 0.17 percent. A balm leaning on coconut oil for its body is a summer liability for exactly that reason; the coconut oil in balms page covers where it earns its place and how much to use.

Palm oil and palm stearin: a long tail. Palm oil is softer than cocoa butter at room temperature, about a quarter solid at 20 C, but it never quite lets go. It still holds 4 percent solid at 35 C and 1.25 percent at 40 C. Palm stearin, the hard fraction, holds 13.7 percent at 35 C, nearly 9 percent at 40 C and 5 percent at 45 C. That tail is the reason the palm oil and palm stearin page uses stearin at 5 to 15 percent for structure that survives a hot delivery van. Palm olein, the liquid fraction, is the mirror image: 6.68 percent solid at 20 C and 1.4 percent at 25 C.

The three temperatures to read

Three readings cover most decisions.

1. About 20 C, storage. How firm the product is in a tin on a shelf, and how much it resists a finger. A fat that is 70 percent solid here builds a much firmer balm than one at 25 percent, at the same dose.
2. About 30 C, warm handling. A pocket, a bag in summer, a bathroom shelf above a radiator. This is where coconut oil and palm olein have already gone and cocoa butter is down to about a quarter.
3. About 35 C, skin and heat. What survives here stays as crystals on the skin and in a hot car. A few percent gives cushion and slip; more starts to feel waxy. Zero means everything from this fat is oil once applied.

The butters compared page ranks butters by feel; the SFC curve is the measurement underneath that ranking.

Shea butter: why there is no single curve

Shea is the butter balm makers use most and the one with the least reliable single SFC curve, because shea varies. Goumbri and colleagues measured 37 crude shea butters from Burkina Faso, made by different traditional extraction processes in different localities, by pulsed NMR with both a non-tempered and a tempered method. Stearic acid ran from 34 to 47.2 percent and oleic acid from 40.8 to 51 percent, and "the 37 samples investigated exhibited varying SFC profiles depending on their origin and extraction processes".

The pattern was consistent even where the values were not. Samples with more stearic acid and less unsaturated fatty acid had the highest SFC: "a higher stearic acid content increased the SFC profile, particularly in the range 20 to 35 °C". The authors report that "the complete melting of all samples occurs around 40 °C". By calorimetry, crystallisation began between 16.4 and 22.4 C depending on the sample. Their SFC results are given as curves rather than a table, so this page does not quote a percentage for shea at a given temperature, and no other open source we found tabulates one for whole shea butter.

Two practical conclusions follow. First, a shea balm that came out right with one lot may come out softer or firmer with the next, mostly in the 20 to 35 C band where a balm is used. The fix is a small test pot per lot, as the palm-free balm page recommends for kokum and shea. Second, the shea curve depends on how the shea was cooled and held, which is the subject of the next section, and the source of the grain covered on grainy shea butter.

The method changes the number

Two SFC figures for the same fat can differ by twenty points without either being wrong. The sample's thermal history and the instrument both move them.

Thermal history and tempering

A fat's solid fraction depends on its crystal form, and its crystal form depends on how it was cooled and held, as fat crystal polymorphism describes. So every standard SFC method begins by erasing the sample's history and imposing a fixed one. Kamali and colleagues, for the cocoa butter in the table, melted the sample at 60 C for 5 minutes, chilled it at 0 C for 60 minutes, then measured serially at 5 C steps from 5 to 40 C. That captures the fat in the forms it takes after a quick chill, not the stable forms it reaches after days.

Goumbri's shea work shows what a different history does. Their tempered method held the samples at 26 C, which "had similar effect on the SFC profile of all the samples. Above 25 °C, SFC was shifted to higher values for all samples, indicating that polymorphic transitions occurred during the treatment". A tempered shea is more solid at 30 and 35 C than the same shea measured straight after chilling. A balm poured and left to set for a week is somewhere between the two, which is one reason a batch is judged on day three rather than day one on cooling rate and crystallisation, and one reason a balm can feel firmer after a month, as balm hardens over time explains.

NMR or calorimetry

Pulsed NMR measures the solid fraction of a sample held at a fixed temperature. Calorimetry (DSC) measures heat flow as a sample is heated, and a solid fat index is calculated from the melting curve. Dhaygude and colleagues measured three blends of ordinary and hydrogenated coconut oil both ways. At 15 C, DSC gave 87.55, 88.38 and 95.95 percent solid; pulsed NMR gave 68.05, 68.83 and 72.35 percent for the same blends. That is a gap of 19.5 to 23.6 points. Their explanation is the difference in method: "DSC is dynamic and pNMR is static", and the DSC samples showed a sharp decline between 15 and 25 C where the NMR results showed a gradual slope.

For balm work, use pulsed NMR figures after a stated conditioning, like those in the table, treat a DSC index as describing the shape of the melt rather than the amount of solid, and treat an SFC figure that names no method as an indication only.

Blends do not average: eutectics

The intuitive way to predict a blend is to average the two SFC curves by weight. For some pairs of fats that works. For others the blend is softer than either fat on its own, sometimes dramatically, because the two fats interfere with each other's crystallisation. That is a eutectic, and it is the most common reason a balm made from two hard butters comes out soft.

Cocoa butter and coconut oil

Joshi, Zielbauer and Vilgis at the Max Planck Institute for Polymer Research mapped blends of cocoa butter with coconut oil across the whole range by calorimetry. They found "a phase diagram of eutectic type" with the eutectic "at 65 wt % of CO and 35 wt % CB": about two parts coconut oil to one part cocoa butter. That mixture melted at 16.7 C. Pure coconut oil is still about a third solid at 20 C, and pure cocoa butter over two thirds, yet the blend between them is liquid at a cool room temperature.

Near that ratio the blends held at 22 C for 24 hours did not crystallise at all: no melting signal could be recorded afterwards because nothing had solidified. The authors also report solid fat content measurements on the blends at 22 C after 24 hours showing the same eutectic behaviour. A cocoa and coconut body balm blended near that ratio can refuse to set in a cool kitchen.

The authors explain it by molecular shape: the kink of the oleic chain shortens cocoa butter's triglycerides to a length similar to coconut oil's short saturated ones, so mixed pairs form. A fully saturated long-chain fat behaves differently. The same study mixed cocoa butter with tristearin and found a monotectic system in which the two "were completely immiscible in the solid phase": no softening dip, the hard fat simply crystallising on its own.

Cocoa butter and shea butter

Shea is a natural source of the same symmetrical, stearic-rich triglycerides that give cocoa butter its melt, and Rodríguez-Negrette and colleagues concluded that shea was a good cocoa butter extender in chocolate. Even so, studying the pair with isosolid diagrams, they report that "softening occurred due to both dilution effects and a slight eutectic formation". The effect is small compared with coconut oil, but it runs in the same direction: a cocoa and shea blend tends to be a little softer than the weighted average of the two.

From a butter's SFC to a balm's firmness

A balm is wax, butter and liquid oil. A rough estimate of how much of it is solid at a temperature is the wax percentage (beeswax stays solid at every temperature in this range) plus each butter's percentage multiplied by its SFC at that temperature. This is the arithmetic the lotion bar page uses. It is an upper bound for the butter's share, because part of the butter dissolves in the liquid oil, the dilution effect Rodríguez-Negrette describe, and because any eutectic pair falls below it.

Set against a typical 7 to 12 percent of wax in a tin, the table shows why the choice of butter matters more at 20 C than at 35 C. At 20 C, 20 percent cocoa butter adds about fourteen points of solid, more than the wax itself, so the balm feels very much harder than the same formula with coconut oil, which adds about seven. At 35 C the wax is doing nearly all the work whichever butter is present, and the butter adds two points or less. That is why balm melting in the heat is fixed with wax rather than with more butter, and why balm too hard or too soft adjusts room-temperature firmness through the butter and heat resistance through the wax. The wax ratio calculator handles the wax side.

What SFC explains at the bench

Several common complaints are SFC problems in disguise.

Matching a butter you cannot get is the same exercise. The blended and faux butters page frames it as matching solid fat at 20 and 30 C, and the curves above are the figures to match against. For the cocoa butter end of the range specifically, the cocoa butter page covers grades, bloom and the deodorised versus natural choice, and mango and kokum butter covers the other stearic-rich butters.

Getting SFC figures for your own materials

In practice SFC is something you obtain rather than measure.

• Ask the supplier. Confectionery and food fats are specified by SFC at several temperatures, so a food-grade supplier may have the figures even when a cosmetic sheet does not. Ask for the values with the method and conditioning stated, for example pulsed NMR to AOCS Cd 16b-93 or ISO 8292-1, and whether the sample was stabilised before measurement.
• Use published curves. For cocoa butter, coconut oil and palm fractions the table above is a sound working basis. Shea needs a test pot per lot.
• Compare lots at the bench. Set 20 g of each lot in identical tins, hold them together at a fixed temperature for 48 hours and compare firmness. You will not get a percentage, but you will see whether a new lot is firmer or softer before it reaches a batch.

Where these numbers stop applying

The curves here are for fats on their own. In a balm, part of each butter dissolves in the liquid oil, so the butter contributes less solid than its own curve suggests; how much less depends on the oil and the proportions, and no source we found measures it for a balm. The cocoa butter values are for samples chilled and measured promptly, not for cocoa butter that has spent weeks converting to its stable forms, which may melt over a narrower range. The coconut figures are for one ordinary coconut oil, the palm figures are reference materials from Malaysian production, and the shea results come from 37 crude West African butters that varied widely among themselves. Refined, fractionated or East African shea, mango, kokum and tallow are not covered by any curve on this page, and the eutectic results apply to the pairs tested, not to every combination of butters.

What the numbers settle is the principle: a balm's firmness at any temperature follows how much of its fat is still solid there, that amount differs sharply between butters, and two butters mixed can be softer than either. Read the curve at 20, 30 and 35 C, check the method, and test any blend.

Reviewed and updated 21 October 2026. Spotted an error? Tell us and we will fix and log it.