Vitamin E and antioxidants in balms: what they do and do not do
Tocopherol is an antioxidant, not a preservative. Forms, the measured dose of 0.02 to 0.1 percent, the pro-oxidant flip, rosemary extract and chelators in anhydrous balms.
Ask ten people in a soap and balm forum what vitamin E is for and six will say it makes the product last longer. That is half right in a way that causes harm, because the wrong half is the half people rely on when they add water to something.
Vitamin E is an antioxidant, not a preservative. It slows oxygen attacking the fatty acids in your oils and does nothing against bacteria, yeast or mould. Dose mixed tocopherols at 0.02 to 0.1 percent, not the 0.1 to 1 percent usually quoted, because the measured optimum for alpha-tocopherol is about 0.01 percent.
Two completely different jobs
Spoilage in a balm comes from two unrelated processes needing different countermeasures.
Oxidation is a chemical chain reaction. A hydrogen is abstracted from a fatty acid, preferentially from a bis-allylic carbon between two double bonds, and the radical takes up oxygen to form a hydroperoxide that later breaks down into the aldehydes you smell as crayon or old paint. It happens in a sealed sterile jar. Antioxidants interrupt the chain, and the bis-allylic index that says which oils need help is in rancidity and oxidation.
Microbial growth is biological. Bacteria need a water activity around 0.9 or above, moulds and osmotolerant yeasts roughly 0.6 to 0.7, and a true anhydrous balm sits far below that, which is why it needs no preservative. Preservatives interrupt this. Antioxidants do not.
This is not interpretation. US 21 CFR 170.3(o) defines antioxidants as substances used to preserve food "by retarding deterioration, rancidity, or discoloration due to oxidation". EU Regulation 1333/2008 Annex I says they "prolong the shelf-life of foods by protecting them against deterioration caused by oxidation", and reserves protection "against deterioration caused by micro-organisms" for preservatives. Both are written in prospective verbs: retard, prolong, protect against future change, never reverse. The sharpest citation is a negative one: tocopherol does not appear on Annex V of the EU Cosmetics Regulation, the list of permitted cosmetic preservatives. If it worked as one it would be there.
If your balm contains aloe, hydrosol, honey, milk or fresh plant material, or will be used with wet fingers or in the shower, it needs a broad-spectrum preservative system. Vitamin E, rosemary extract and grapefruit seed extract will not do it. See do balms need preservatives and mould in balm.
Which form of vitamin E to buy
Vitamin E is a family of eight molecules: four tocopherols (alpha, beta, gamma, delta) and four tocotrienols. What you buy makes a large practical difference.
| Form | Typical use | Protects oils? | Notes |
|---|---|---|---|
| Mixed tocopherols (T50, T70, T95) | 0.02-0.1% | Yes, best choice | Gamma and delta tolerate higher loading before flipping pro-oxidant. The number is the tocopherol content |
| d-alpha-tocopherol | 0.01-0.02% | Yes, low optimum | Highest vitamin activity, optimum about 100 ppm in bulk oil, pro-oxidant soonest |
| dl-alpha-tocopherol (synthetic) | 0.01-0.02% | Yes | Racemic mixture, cheaper, behaves similarly in an oil |
| Tocopheryl acetate | 0.5-1% | No | Hydroxyl blocked, so no radical scavenging in the jar. A skin claim ingredient |
| Tocopheryl linoleate or succinate | varies | No | Esterified like the acetate, same limitation |
The acetate point defeats a lot of well-intentioned formulating. Tocopherol works because a phenolic hydroxyl donates a hydrogen atom to a lipid radical and stops the chain. Esterifying that hydroxyl makes the molecule shelf stable, which is why the acetate is used in supplements and serums, and also means the group cannot donate anything. If your bottle says tocopheryl acetate and you bought it to stop sunflower oil turning, it will not.
Dose on the active, not on the bottle. Tocopherol is sold diluted in a carrier, usually sunflower oil, at 50, 70 or 95 percent, so a T50 at 0.1 percent of the formula delivers 0.05 percent tocopherol.
The dose almost everyone gets wrong
The 0.1 to 1 percent figure repeated across craft sources is 5 to 100 times above the concentration measured to protect an oil best. The numbers behind that:
Concentration optima measured in soybean oil (Evans, Kodali and Addis, JAOCS 2002;79:47): alpha-tocopherol optimises at about 100 ppm, which is 0.01 percent, and turns pro-oxidant above that. Gamma optimises near 300 ppm, and delta showed no upper limit to 1900 ppm. A review-level safe limit is about 200 ppm. Food law agrees: Codex permits mixed tocopherol concentrate or d-alpha-tocopherol in named vegetable oils at a maximum of 300 mg/kg, 0.03 percent, and no additives at all in virgin or cold pressed oils.
- Mixed tocopherols for oil protection: 0.02-0.1 percent of the formula. Above that you are making a vitamin E claim, not protecting oil.
- Pure alpha-tocopherol: 0.01-0.02 percent. Its optimum is the lowest of the family.
- Rosemary extract: 0.0015-0.04 percent, the range authorised as the food additive E 392.
- Worked example, 250 g batch: 0.13 g mixed tocopherols (0.05 percent) plus 0.05 g rosemary extract (0.02 percent).
- All percentages are weight percent of the total formula, which must add to 100.
The pro-oxidant effect is the mechanism behind that ceiling. When alpha-tocopherol quenches a lipid radical it becomes a tocopheroxyl radical, which is relatively stable. At low concentrations it is mopped up; at high concentrations there are enough of them to abstract hydrogen from intact fatty acids and start new chains. Gamma and delta tolerate far more loading first, which is why "I add 2 percent to be safe" is not safe.
Two points change the sums. Tocopherols trap peroxyl radicals at 104 to 109 per molar per second against 10 to 60 for the lipid itself, so one molecule protects a great many fatty acids and you do not need much. And your oil does not start at zero: sunflower already carries 400 to 935 mg/kg of alpha-tocopherol, four to nine times the optimum, so adding more alpha to a sunflower base is the least promising move here.
The temperature reversal nobody mentions
The familiar ranking, delta above gamma above beta above alpha, comes from accelerated testing at elevated temperature. Below about 40 C the order reverses to alpha, beta, gamma, delta. A balm on a shelf lives below 40 C all its life, so the usual advice to prefer gamma and delta rests on evidence gathered in conditions your product never sees. That argues for mixed tocopherols on both counts: alpha for the room-temperature case, gamma and delta for headroom against the pro-oxidant flip. Reviewers call this literature abundant, contradictory and strongly system dependent.
Excess costs you cosmetically too: concentrates are dark amber with a seedy odour that shows through a lightly scented balm. Tocopherols degrade in storage and faster in light, and refining strips them, so a refined oil can be less stable than its composition suggests.
Rosemary extract, the other workhorse
Rosemary oleoresin, or ROE, is an extract of Rosmarinus officinalis whose activity sits in two diterpenes, carnosic acid and carnosol. The EU specification requires not less than 5 percent of the sum of the two, worth checking on a certificate. It is authorised as food additive E 392 at 15 to 400 mg/kg (0.0015 to 0.04 percent) across 33 food categories, with a JECFA temporary acceptable daily intake of 0 to 0.3 mg per kg of body weight.
It performs. In a Rancimat comparison at 120 C (Song, Sui and Jiang, Foods 2023;12:2177), rosemary extract at 400 mg/kg was set against a BHA and BHT blend at 200 mg/kg. Induction periods rose from 2.20 to 3.40 hours in soybean, 2.32 to 4.15 in rapeseed, 1.88 to 3.35 in cottonseed, 3.83 to 6.22 in rice bran and 1.54 to 3.39 in camellia. The extract beat the synthetic blend in all five, at twice the dose, so it is not a like-for-like potency comparison.
The constraints are sensory. Standard ROE is a dark green viscous liquid with a pronounced herbal smell, and at 0.1 percent in an unscented lip balm you will smell it. Deodorised grades are worth paying for in pale or delicately scented products, though they carry less carnosic acid. ROE is not a preservative either, despite being sold beside them.
The full toolkit
| Option | Use level | What it does | Limitation |
|---|---|---|---|
| Mixed tocopherols | 0.02-0.1% | Chain-breaking antioxidant, donates hydrogen to lipid radicals | Pro-oxidant above the optimum, adds colour and odour |
| Rosemary extract (E 392) | 0.0015-0.04% | Chain-breaking antioxidant, potent at low levels | Green colour, strong herbal smell in standard grades |
| Ascorbyl palmitate | 0.05-0.2% | Oil-soluble vitamin C ester, can regenerate spent tocopherol | Poorly soluble cold, needs warm dispersion |
| Chelators (EDTA, citric acid) | not recommended | Bind trace iron and copper in a water phase | Poor in bulk oil, no phase to dissolve into, and under-dosing promotes oxidation |
| Synthetic phenolics (BHT, BHA) | 0.01-0.1% | Effective, cheap, well characterised | Consumer resistance, disfavoured in some markets |
| Choosing stable oils | n/a | Removes the substrate | Changes feel and cost |
| Packaging and headspace | n/a | Excludes the oxygen the reaction needs | The highest return on this list |
The last two rows are not a joke. Swapping half a grapeseed fraction for high-oleic sunflower does more for shelf life than any antioxidant dose, and filling the tin to the top does more than either. See carrier oils compared and the oil comparison table.
Chelators: the counterintuitive part
Start with the fact that reverses the usual intuition. Transition metals are most catalytically active in the dry state. Water deactivates iron and copper by hydrating their coordination shells, so an anhydrous balm is the worst case for trace-metal catalysis, not the safest. Bulk oil is not truly water-free either: the trace water it holds sits in reverse micelles where hydroperoxides aggregate and metals concentrate. Codex limits iron to 1.5 mg/kg refined and 5.0 virgin, copper 0.1 and 0.4, the numbers to ask a supplier for.
Chelation is nonetheless a poor answer in a balm. Chelators work much less well in bulk oil than in emulsions, and EDTA disodium is a water-soluble salt with no phase to dissolve into, so it sits as grit. Refineries add citric acid at about 50 ppm or edetates at about 30 ppm into hot oil during final refining, which you cannot reproduce by stirring powder into a cooled balm. Worse, under-dosing a chelator is worse than omitting it: at a 1 to 1 molar ratio with iron, EDTA promoted oxidation, while 2 to 1 and 4 to 1 inhibited it. A pinch you cannot dissolve or dose falls on the wrong side of that line.
The advice for a balm is exclusion, not chelation. Buy oils low in iron and copper, avoid unlined metal and prolonged metal contact, and watch herbal infused oils and unrefined material carrying their own metal load. Glass or stainless equipment does the job. No evidence either way was found for phytic acid in anhydrous systems, so this page makes no claim about it.
Synergy with tocopherol is disputed, not settled
The claim that tocopherol and rosemary extract are synergistic is repeated everywhere, and the evidence does not support stating it as fact. One study reports an optimum near a 1.5 to 1 ratio of tocopherol to carnosic acid plus carnosol. Frankel and colleagues, in bulk corn oil, found alpha-tocopherol had no effect on rosemary's activity or a slightly negative one. Given the pro-oxidant flip above, a negative interaction is plausible: a second hydrogen donor past the optimum can make things worse.
A defensible default for a balm on moderately unsaturated oils is 0.05 percent mixed tocopherols plus 0.02 percent deodorised rosemary extract: for a 100 g batch, 0.05 g and 0.02 g on a scale reading to 0.01 g. If your scale reads to 1 g, make a larger batch rather than dosing bigger.
When to add them
- As early in the oil's life as possible. An antioxidant delays the onset of oxidation, so dose the bulk oil when you open it or at the start of an infusion, not after months in a warm cupboard.
- Into the oil phase, stirred properly. A blob of tocopherol in one corner of the pot protects one corner of the pot.
- Below 50 to 60 C. Tocopherol degrades with prolonged heat, the same window used for flavour, described under pour temperatures.
- Not to rescue. If the oil smells of crayons the end products already exist and cannot be unmade: see rancid balm.
How to tell whether it is working
Most makers add tocopherol on faith. You can do better without a laboratory.
- Run a control. Split a batch, dose half, label both. It is the only way to attribute a difference to the antioxidant rather than to the oil or the storage.
- Accelerate carefully. Matched samples at 40 to 45 C are indicative, not a substitute for real-time data, and the tocopherol ranking itself changes above 40 C.
- Smell blind. Code the jars, have someone else present them, and record the first date you can pick the undosed one out.
- Measure a series, not a point. If you test peroxide value commercially, repeat it over time and pair it with an anisidine or acid value: a single reading can fall as an oil gets worse. See shelf life testing.
Vitamin E and scars: the claim with the weakest evidence
None of the above is why most people buy vitamin E. Ask a shopper what it is for and the answer is scars. The best known controlled test of that idea went the other way. Baumann and Spencer, Dermatologic Surgery 1999;25(4):311-315 (PMID 10417589), applied topical vitamin E after skin surgery and reported that in 90 percent of cases it had no effect on the cosmetic appearance of the scars or actively worsened it, and that 33 percent of subjects developed a contact dermatitis.
Frame that carefully. It is one controlled study of one application, it is old, and its endpoint was scar appearance rather than the antioxidant chemistry the rest of this page is about. It is not evidence that vitamin E is useless. It is evidence of an awkward asymmetry: the job tocopherol demonstrably does, protecting the oils in the jar, is the one nobody markets it for, while the claim it is most often sold on has the weakest support behind it.
Two consequences follow. A one in three rate of contact dermatitis at treatment-level exposure is another argument for the fractions of a percent recommended here rather than squeezing capsules into a pot. And a balm is not the place to chase a skin-directed vitamin E claim: scar, healing and repair language is drug language in the US whatever the ingredient, as set out in cosmetic versus drug claims.
What to do with this
Add 0.02 to 0.1 percent mixed tocopherols to any balm carrying significant linoleic or linolenic oils, put it into fresh oil, keep it under 60 C, and stop thinking of it as protection against anything living. For a two year balm, lower the bis-allylic index and fill the container properly, then use the antioxidant to defend the margin rather than to buy it. A first batch on these lines is in how to make lip balm.
Frequently asked questions
Is vitamin E a preservative?
No. Tocopherol is an antioxidant. It slows the reaction between oxygen and the unsaturated fatty acids in your oils. It has no meaningful effect on bacteria, yeast or mould, so it cannot preserve anything that contains water. Confusing the two is the most common and most consequential mistake in home skincare.
How much vitamin E should I add to a lip balm?
Far less than most sources say. Measured optima in bulk oil are about 100 ppm for alpha-tocopherol and about 300 ppm for gamma, so 0.02 to 0.1 percent of a mixed tocopherol concentrate is the evidence-based window. For a 100 g batch that is 0.02 to 0.1 g. Check the strength: a T50 is only half tocopherol.
Does adding more vitamin E make a balm last longer?
No, and past a point it makes things worse. In soybean oil the optimum for alpha-tocopherol was about 100 ppm, above which it acted as a pro-oxidant, because the tocopheroxyl radical it forms can start new chains. Gamma optimised near 300 ppm and delta showed no upper limit to 1900 ppm. High levels also add colour and a seedy smell.
What is the difference between tocopherol and tocopheryl acetate?
Tocopheryl acetate has the reactive hydroxyl group blocked by an acetate ester, which is what makes it shelf stable. That same blocking means it cannot donate a hydrogen to a lipid radical, so it does nothing for the oxidative stability of your oils. It is a skincare active with a vitamin E claim, not an oil protectant.
Is rosemary extract a preservative?
No, for the same reason vitamin E is not. Rosemary extract is a potent antioxidant, authorised in food as E 392 at 15 to 400 mg/kg, and it has some weak antimicrobial activity in the laboratory, but nowhere near the spectrum needed to preserve a product containing water. In an anhydrous balm that does not matter, because such a balm does not need preserving.
Can antioxidants rescue an oil that has already gone rancid?
No, and the regulatory definitions say so in prospective verbs: 21 CFR 170.3(o) and EU Regulation 1333/2008 Annex I both describe antioxidants as retarding or prolonging, never reversing. Once hydroperoxides have decomposed into aldehydes and ketones, those are stable end products with nothing for a radical scavenger to act on. Adding tocopherol dilutes them and nothing more.
When in the process do I add the antioxidant?
Into the oil phase, below 50 to 60 C, as late in the heating as practical but as early in the oil's life as possible. Best of all, dose the antioxidant into the bulk oil when you open the bottle or when you start an infusion, so it is protecting the oil during the weeks it sits around, not just after the balm is poured.
Sources and further reading
- Cosmetic Ingredient Review, Safety assessment of tocopherols and tocotrienols as used in cosmetics, CIR Expert Panel.
- Evans, Kodali and Addis, Optimal tocopherol concentrations to inhibit soybean oil oxidation, Journal of the American Oil Chemists' Society, 2002;79:47.
- Song, Sui and Jiang, Comparison of rosemary extract with synthetic antioxidants in vegetable oils, Foods, 2023;12:2177.
- Codex Alimentarius Commission, Standard for Named Vegetable Oils (CXS 210-1999), FAO/WHO (permitted tocopherol levels and metal limits).
- Kamal-Eldin A and Appelqvist LA, The chemistry and antioxidant properties of tocopherols and tocotrienols, Lipids, 1996.
- Huang SW, Frankel EN and German JB, Effect of individual tocopherols and tocopherol mixtures on the oxidative stability of corn oil triglycerides, Journal of Agricultural and Food Chemistry, 1995.
- American Oil Chemists' Society, Official Methods: Peroxide Value (Cd 8b-90) and Oil Stability Index (Cd 12b-92).
- EFSA Panel on Food Additives, Scientific opinions on extracts of rosemary (E 392), EFSA Journal.
Reviewed and updated 6 September 2026. Spotted an error? Tell us and we will fix and log it.