How far neat oils go into skin, and what the studies did not test

What in vivo Raman and microscopy studies of neat oils found about how deep plant oils, paraffin oil and petrolatum go into skin, and what the 500 dalton rule says.

Balm labels make two kinds of statement about where the product goes. "Absorbs quickly" describes something the user feels. "Penetrates deep into the skin" is a claim about depth, and depth has been measured, but only for neat oils applied to skin. No study cited here tested a balm or any wax-structured product.

The layer the oils were measured in

Every study on this page measures depth inside the stratum corneum, the cornified outer layer of the epidermis, so its thickness is the ruler for reading the numbers. This site's page on moisturisers and the skin barrier gives it as "typically 10 to 20 micrometres thick over most of the body and considerably thicker on palms and soles".

What the in vivo depth studies measured

Four studies measured how far applied oils go in living human skin. Three used confocal Raman microscopy, which the 2017 paper describes as a tool to analyse depth profiles "non-invasively". The fourth, Patzelt and colleagues in 2012, used laser scanning microscopy and measured occlusion separately by transepidermal water loss. The table gives each study as its published abstract reports it.

The two depth figures are not the same figure

The 2015 and 2017 papers come from the same group and both give a depth, but they describe different things. The 2015 figure is a maximum: the oils "do not penetrate deeper than 11 µm", a result obtained with three of the four analysis methods and stated in the abstract for human and porcine skin together. The abstract does not give a separate figure for human skin alone. Set against the 10 to 20 micrometre thickness above, 11 micrometres on its own does not place the oil in the upper part of the layer. What does is the authors' conclusion that the oils are "predominantly remaining in the uppermost corneocyte layers". Predominantly is their word, and it is not the same as entirely.

The 2017 figure is a band, in human skin in vivo: "all oils remain in the upper layers of the SC (0-20% of SC thickness)". What both support is that the tested oils were found mainly near the top of the stratum corneum, and that the 2015 authors concluded they do not reach the living cells beneath it.

Paraffin oil, plant oils and jojoba

None of the three studies that compared mineral and plant oils found the plant oils going deeper. Stamatas and colleagues conclude that "there is no statistical difference between the paraffin oil and vegetable oils in terms of skin penetration and skin occlusion". Patzelt and colleagues found that "paraffin oil, as well as the vegetable oils, penetrated only into the first upper layers of the stratum corneum". The 2017 paper, with two mineral-derived and two plant-derived oils, found that "both oil types remain in the superficial layers of the SC".

Jojoba, whose composition is set out in jojoba oil, is a partial exception. On depth it went where the other vegetable oils went in the 2012 study, and it sat in the same 0 to 20 percent band in 2017. On occlusion it did not match them. The 2012 abstract says that "the application of the vegetable oils (except jojoba oil) as well as paraffin oil, led to a similar occlusion of the skin surface". It does not say whether jojoba occluded more or less than the others.

What the oils do to the lipids they reach

The 2017 paper looked at more than where the oils were. It measured how ordered the barrier's own intercellular lipids were at each depth, and it found an effect from both kinds of oil, not only the plant ones. Its conclusion: "Skin treated with mineral- and plant-derived oils shows significantly higher disordered lateral and lamellar packing order of ICL in these layers of the SC compared to intact skin", where "these layers" are the top 0 to 20 percent in which the oils remained.

Deeper down, only the plant oils showed an effect. The conclusion states that "plant-derived oils significantly changed the ICL ordering in the depths of 30% and 70-90% of the SC thickness". The results section is more cautious about the deepest band: at 70 to 90 percent it reports "slight changes with more disorder states" at p<0.1. The authors explain the deeper effect as "likely due to the penetration of free fatty acids in the deeper layers of the SC". That is their explanation, and "likely" is their word; the abstract does not report measuring free fatty acids at those depths.

A caveat about how Raman spectra are read

The 2015 paper is also a methods comparison. It analysed the same spectra four ways and reports that with one of them, "analyzing the lipid-to-keratin peak ratio using the perpendicular drop-down cutoff procedure", "the penetration profile of oils, and especially of petrolatum, into the skin was interpreted incorrectly". The authors recommend a Gaussian function-based deconvolution procedure "to exclude any possible mistakes".

The 2008 abstract says only that its lipid and water profiles "were calculated from the Raman spectra", not by which procedure. Its swelling figures disagree with 2015: "modest" swelling of 10 to 20 percent for its three oils, against "no significant swelling effect" for oils other than petrolatum in 2015. The two do agree that petrolatum swelled the stratum corneum more than the liquid oils did, 40 to 60 percent in 2008 and 32 percent in vivo in 2015, which fits the 2012 finding that petrolatum gave the most effective occlusion by transepidermal water loss.

Does petrolatum go into the intercellular spaces?

Ghadially, Halkier-Sorensen and Elias, in the Journal of the American Academy of Dermatology in 1992, "reexamined the assumption that Vaseline Petroleum Jelly (VPJ) is occlusive". The abstract describes two parts. In one, "barrier recovery was measured in VPJ-treated versus untreated sites after acetone-induced barrier disruption in human volunteers". In the other, introduced with "Moreover", petrolatum "was localized within the stratum corneum (SC) with tracers and ruthenium tetroxide staining", which "allowed visualization of the depth of VPJ penetration and its relation to intercellular membrane structures". The results: petrolatum "accelerated, rather than impeded, barrier recovery", and it "was present within the interstices at all levels of the SC, where it replaced intercellular bilayers". The conclusion: it "permeates throughout the SC interstices, allowing normal barrier recovery despite its occlusive properties".

The abstract ties the acetone treatment to the barrier recovery measurement. It does not say whether the skin used to locate the petrolatum had been treated with acetone.

In 2017 Raman placed petrolatum with the other oils in the top 0 to 20 percent of stratum corneum thickness, on skin the abstract does not describe as disrupted. The 1992 staining located petrolatum relative to the intercellular membranes; Raman gives how much oil is at each depth. "At all levels" and "upper layers" are different statements from different methods. What they share is that petrolatum enters the stratum corneum rather than sitting entirely on top of it.

Does the 500 dalton observation apply to balm lipids?

The 500 dalton figure comes from a short 2000 paper in Experimental Dermatology by Bos and Meinardi, who wrote: "We argue that the molecular weight (MW) of a compound must be under 500 Dalton to allow skin absorption. Larger molecules cannot pass the corneal layer."

Their arguments were drawn from contact allergens and topical and transdermal drugs, not from lipids.

Long-chain triglycerides and jojoba's esters are over the line; medium-chain triglycerides straddle it. A triglyceride of C16 and C18 chains weighs roughly 800 to 900 daltons, a lauric one about 640, and the dominant jojoba esters described in jojoba oil around 590 to 620. Caprylic/capric triglyceride, a common light balm oil, runs from about 471 daltons for the all-C8 molecule to about 555 for the all-C10 one, so part of it is under 500. By Bos and Meinardi's argument, the molecules over 500 would not pass the corneal layer. The depth studies are consistent with that, since the oils they tested stayed in the stratum corneum. They also show that molecules this large do enter the stratum corneum. The rule is about passing it.

Free fatty acids are under it. Oleic acid, at 282 daltons, is a free fatty acid, the class the 2017 authors suggested as the likely reason plant oils changed lipid ordering deeper in the stratum corneum. The rule is phrased as a requirement, "must be under 500 Dalton to allow skin absorption", so being under 500, as oleic acid and tricaprylin are, does not by itself say that a molecule is absorbed.

Paraffin oil and petrolatum are not placed on the table. They are mixtures, and none of the cited sources gives their molecular weights.

What the studies did not test

Subjects and skin. The groups were small: 9 volunteers and 7 infants in 2008, 6 healthy volunteers in 2012. The 2015 and 2017 abstracts do not give numbers. Only the 2008 abstract names the site, the forearm. None of the four describes broken, inflamed or eczematous skin, and none describes lips. The one study on this page that used deliberately damaged skin is the 1992 petrolatum paper, and there the damage is tied to the barrier recovery measurement.

Time and dose. The 2008 study measured at 30 and 90 minutes after a single application. The other abstracts do not give times, and none describes repeated application over days or weeks. None of the abstracts reports the amount applied, so the studies say where the oil went, not what share of a dose went there. How much of an application stays on the surface, and how much transfers to hands and fabric, was not measured by any of them. Dose itself is covered in how much balm to apply.

Materials. The oils tested were paraffin oil, petrolatum, almond oil, jojoba oil and vegetable oils the abstracts do not all name, each applied neat. No study here tested a finished balm, shea butter or other solid butters, beeswax, squalane, or the light esters in cosmetic esters, so this page gives no depth for any of them.

Feel. None of the studies measured how an oil feels, so how fast an oil seems to disappear is not evidence of how deep it went. Texture is its own subject, covered in balm texture science.

What this means for wording on a label

On the last row: the depth evidence does not give plant oils an absorption advantage, so any case for choosing between them rests on other grounds, discussed in natural versus petroleum. If you are reading someone else's label rather than writing your own, the same table works in reverse, and reading a balm label covers the ingredient list side of the exercise.

Reviewed and updated 27 September 2026. Spotted an error? Tell us and we will fix and log it.