NMR for cosmetics — NorrChemica Analysis Notes, NMR Series Part 16. Navy card with a schematic 1H spectrum: the two glycerol signals at 4.27 and 4.12 ppm of an essential oil with 0.8 % of vegetable oil, above the flat trace of a pure essential oil.

NMR for cosmetics: adulterated essential oils, free formaldehyde, aloe vera, rancid lipsticks and mineral oil in lip products

Analysis Notes · NMR Series · Part 16

NMR for cosmetics: adulterated essential oils, free formaldehyde, aloe vera, rancid lipsticks and mineral oil in lip products

Which questions about a cosmetic ingredient or product an NMR spectrum answers better than chromatography or titration, which official methods prescribe NMR, and which questions it does not answer

Sources J. Agric. Food Chem., Molecules, Anal. Chem., Int. J. Cosmet. Sci., Contact Dermatitis, Spectroscopy Europe, F1000Research, J. Chem., Int. J. Mol. Sci., USP–NF, ASTM, EU and UK legislation, SCCS Samples essential oils, plant oils, aloe vera, lipsticks, petrolatum, lip products, hair straighteners, toothpaste, surfactants, polymers, face serums, perfumes

Most methods that test a cosmetic product first separate its components or convert one of them into a derivative. A nuclear magnetic resonance (NMR) spectrum is recorded on the dissolved sample without either step. Every compound that contains hydrogen and is present in a sufficient amount appears in one 1H spectrum, and the area of each signal is proportional to the number of nuclei that give it.

The spectrum therefore shows compounds that a separation does not detect or that a reagent changes. Gas chromatography does not show a vegetable oil that was added to an essential oil; the 1H NMR spectrum detects it from 0.02 % (w/w), and four of 20 essential oils bought from online shops contained between 10 and 42 % (w/w) of it.[1, 2] 13C NMR measures free formaldehyde beside the preservative that releases it, without shifting the equilibrium between the two.[3] An official control laboratory found aldehydes, a marker of rancid fat, above the detection limit in 26 of 38 lipsticks.[4] The trade association of the aloe industry has the products for its certification seal tested by NMR,[5] and the United States Pharmacopeia prescribes NMR for the composition of poloxamers.[6] Figure 1 shows what six surveys found. The limit of the method is its sensitivity. The detection limits of the 1H methods cited here lie between 0.0001 and 0.1 % (w/w), and polycyclic aromatic hydrocarbons, which are expected at less than 1 mg/kg, are below this range.[7]

SHARE OF THE PRODUCTS EXAMINED Essential oils from online shops that contained vegetable oil 4 of 20 Retail argan oils rejected as adulterated 4 of 28 Hair-straightening products above the formaldehyde limit 7 of 10 Lipsticks with more than 100 mg/kg of aldehydes 15 of 38 Aloe vera powders without a satisfactory content of acemannan 6 of 9 Lip products with at least 0.4 g/100 g of aromatic mineral oil 8 of 57

Figure 1. Findings of six surveys in which cosmetic products or ingredients were examined by NMR. Our drawing of published numbers: essential oils;[1] argan oils;[8] hair-straightening products;[9] lipsticks;[4] aloe vera powders, of which three of nine were satisfactory;[10] lip products.[11] None of the surveys is representative of a market.

Seven terms used in this article. The chemical shift is the position of a signal in the spectrum, given in parts per million (ppm); it depends on the surroundings of the nucleus in the molecule. The integral is the area of a signal. In quantitative NMR (qNMR) the integral is compared with that of a weighed standard, and the amount of a compound follows without a reference sample of it. A triacylglycerol is an ester of glycerol with three fatty acids; vegetable oils consist of such esters.[1] A formaldehyde releaser is a preservative that liberates formaldehyde gradually in the finished product.[12] The detection limit is the smallest amount that a method distinguishes from the noise. % (w/w) means grams per 100 g of sample; 1 mg/kg is 0.0001 % (w/w). Where ppm stands for a concentration, 10 ppm is 0.001 %.

At a glance: the question and the answer from the NMR spectrum

Question What the spectrum shows Limit
Was an essential oil diluted with vegetable oil? The glycerol signals of triacyl­glycerols at 4.12 and 4.27 ppm give the amount; the 13C spectrum identifies the added oil from 3 % (w/w) Detection limit 0.02 % (w/w) by 1H; the volatile composition remains the task of gas chromatography
How much of the formaldehyde in a preserved product is free? Free formaldehyde and the intact releaser as separate 13C signals, measured without a reagent Detection limit about 0.002 to 0.01 % (w/w), which is above the labelling threshold of 0.001 %
How much hydrogen peroxide or formaldehyde does a hair product contain? Hydrogen peroxide as a separate signal at 10.2 ppm; total formaldehyde as methanediol at 4.82 to 4.84 ppm after hydrolysis Detection limit for hydrogen peroxide 0.0001 to 0.01 % (w/w), below each legal maximum
Is a lipstick or an oil rancid? The sum of all aldehydes between 9 and 10 ppm; at 600 MHz their classes appear separately Detection limit about 40 mg/kg in oils; fragrance aldehydes are counted too
Is an aloe vera raw material genuine? Acetylated poly­saccharide, glucose and malic acid; malto­dextrin, degradation products and preservatives In a finished product only when the aloe constituents are concentrated enough
What is the composition of a poloxamer, an ethoxylate or a cyclodextrin ether? Oxyethylene content, average chain length or molar substitution, by the test of the pharmacopoeia Average values, not the distribution of chain lengths or of substituents
How much surfactant does a raw material contain? Its amount against a weighed standard, without a reference sample of the same surfactant Detection limit about 60 ppm for sodium lauryl sulfate; applied to surfactant standards and real samples
Is a plant oil the oil named on the label? Oleic, linoleic and linolenic acid from one 1H spectrum; in one survey, four of 28 retail argan oils were rejected as adulterated Natural variation sets the limit: about 20 % (w/w) of sunflower oil in argan oil in the study cited
How much aromatic mineral oil does a petrolatum or a lip product contain? All protons on aromatic rings between 6.5 and 9.2 ppm, as one value: the mineral oil aromatic hydrocarbons (MOAH) Detection limit below 0.1 g/100 g, and 0.03 g/100 g with the clean-up for lip products; aromatic ingredients must be removed first; single polycyclic compounds are not measured
Does a serum contain the declared amount of an active ingredient? Bakuchiol in oil serums: 0.59 % found where 1 % was declared, and none detected in one that lists it Two emulsions could not be measured; other ingredients covered some signals
Is a perfume the original or a counterfeit? The whole 1H spectrum as a fingerprint, compared with those of the authentic perfume One study, at 700 MHz; several batches of the original are needed for the comparison

What the spectrum of a cosmetic sample contains

The signals used in this article lie between 0.2 and 10.2 ppm of the 1H spectrum (Figure 2).

hydrogen peroxide in DMSO-d6, 10.2 ppm aldehydes of oxidised fat, 9–10 ppm aromatic protons of mineral oil (MOAH), 6.5–9.2 ppm bakuchiol, four signals between 5.59 and 6.88 ppm glycosidic protons of hydroxypropyl betadex, 5.0–5.4 ppm CH2 of methanediol (formaldehyde in water), 4.82–4.84 ppm glycerol CH2 of triacylglycerols, 4.12 and 4.27 ppm OCH2 and OCH of poloxamers, 3.2–3.8 ppm CH2 between two double bonds of fatty acids, 2.75–2.82 ppm acetyl CH3 of the aloe polysaccharide, 2.0–2.2 ppm saturated hydrocarbons of mineral oil (MOSH), 0.2–3.0 ppm CH3 of the oxypropylene units of poloxamers, 1.08 ppm terminal CH3 of fatty acids, 0.84–0.91 ppm 0 1 2 3 4 5 6 7 8 9 10 11 chemical shift of 1H, ppm

Figure 2. Positions of the 1H signals used in the applications of this article. Our drawing of published values: hydrogen peroxide;[13] aldehydes;[4] mineral oil hydrocarbons;[7] bakuchiol;[14] hydroxypropyl betadex;[15] methanediol;[9] glycerol CH2 groups;[1] poloxamers;[6] fatty acids;[16] aloe polysaccharide.[17] The values were measured in different solvents.

Much of the published work comes from one institution: the Chemisches und Veterinäruntersuchungsamt (CVUA) Karlsruhe, an official control laboratory of the German state of Baden-Württemberg, which measures cosmetics and their raw materials on a 400 MHz spectrometer.[7]

When NMR is the preferable method

Free formaldehyde beside a releaser

In water, formaldehyde is present as its hydrate, methanediol. A product preserved with a releaser contains bound and free formaldehyde in equilibrium (Figure 3).

ONOHONHO + 2 H2O ONHONH + 2 HO–CH2–OH H2C=O + H2O DMDM hydantoin formaldehyde bound as N–CH2–OH 5,5-dimethylhydantoin methanediol free formaldehyde in water formaldehyde carbon atom of the formaldehyde unit, observed directly by 13C NMR in each of its forms

Figure 3. Release of formaldehyde from DMDM hydantoin, a member of the amide class of Figure 4. The abstract of [3] does not name the compounds that were studied; the one drawn is our example. Textbook chemistry, our drawing.

A reagent that reacts with free formaldehyde removes it from this equilibrium, and the releaser supplies more. According to an article in a trade journal on the stability of these preservatives, almost all methods and many reagents disturb this equilibrium; its authors used 13C NMR for that reason.[18]

In 2007 Emeis, Anker and Wittern, of a skin-care manufacturer and a chemical company in Hamburg, used 13C NMR to observe the formaldehyde carbon in each of its forms. They compared several releasers of two kinds, O-formals and N-formals, that are used in industrial and consumer products, and they followed the structures and the liberation of formaldehyde at different pH values and dilutions (Figure 4). Only the N-formals derived from amides were a reservoir of formaldehyde at the concentrations usual in cosmetic products; the O-formals and the N-formals of amines decomposed completely. In finished products, free formaldehyde was quantified down to about 0.002 % (w/w) in an acceptable measuring time. The authors regard NMR as the only method that determines free formaldehyde unequivocally.[3]

R1OOH O-formal decomposes completely R1NR2OH N-formal derived from an amine decomposes completely R1ONR2HO N-formal derived from an amide reservoir of formaldehyde CH2 group that is released as formaldehyde; R1 and R2 are organic groups

Figure 4. The three classes of formaldehyde releasers compared in [3] and the behaviour reported for each at the concentrations usual in cosmetic products. General formulas, our drawing.

In 2010 Emeis, de Groot and Brinkmann applied the method to the preparations with which dermatologists test patients for contact allergy to these preservatives. The five preparations in petrolatum contained no free formaldehyde. All nine commercial preparations in water contained it, and so did nine of twelve aqueous solutions made by the authors, at 0.019 to 0.37 %; the detection limit in this study was 0.01 %. The petrolatum preparations released formaldehyde when they came into contact with water.[19]

Under Regulation (EU) 2022/1181, a product that contains a releaser carries the warning “releases formaldehyde” when the total concentration released exceeds 0.001 % (10 ppm); products that complied with the earlier rule could be placed on the EU market until 31 July 2024 and sold until 31 July 2026.[12] Great Britain has applied the same threshold since 15 July 2026, and products placed on the market before that date may be sold until 14 January 2027.[20] Our article on the cosmetics rules of 2021 to 2026 lists these dates with the other changes. The lower of the two detection limits reported for 13C NMR, about 0.002 %, is twice this threshold. An NMR measurement therefore does not show that a product is below it; it gives the free and the bound share of the formaldehyde.

Small regulated molecules: formaldehyde, hydrogen peroxide and fluorine compounds

The CVUA Karlsruhe published a 1H NMR method for formaldehyde in hair-straightening products in 2013, when their illegal use had become popular in Europe. The sample is diluted and hydrolysed with alkali, and the total formaldehyde is read from the CH2 signal of methanediol at 4.82 to 4.84 ppm. Seven of ten products contained more than 0.2 % of formaldehyde, the maximum allowed at that time: between 0.42 and 5.83 %, on average 1.46 %. The photometric reference method of the EU confirmed the NMR results.[9] Formaldehyde as such is now prohibited in cosmetic products.[12]

In 2016 Monakhova, a co-author of the Karlsruhe methods, and Diehl, both of a commercial NMR laboratory in Cologne, described the determination of hydrogen peroxide in hair sprays, nail treatments, disinfectant solutions and chemical reagents. The sample is diluted with deuterated dimethyl sulfoxide (DMSO-d6), which prevents the OH signals from merging. Hydrogen peroxide then gives a separate signal at 10.2 ppm, which is measured against that of the methyl groups of dimethylformamide at 2.9 ppm. The detection limit was 0.0001 % (w/w) for reagents and disinfectants and 0.01 % (w/w) for hair sprays, and the recovery was 103 to 111 %.[13] The Cosmetics Regulation allows at most 12 % of hydrogen peroxide for hair, 4 % for skin, 2 % in nail hardeners and 0.1 % in oral products sold to consumers;[21] the detection limits are below all four.

Oral products may contain at most 0.15 % of fluorine, calculated as F, from all of its compounds together.[21] Sodium fluoride and sodium monofluorophosphate are often used in oral care products to prevent caries. In 2009 Deng, Hou and co-workers determined the two compounds simultaneously in toothpaste by quantitative 19F NMR, with sodium trifluoroacetate as the internal standard. They describe the method as simple, fast and free of interference.[22]

Rancidity: aldehydes in lipsticks and oils

Unsaturated fatty acids react with oxygen to give hydroperoxides, and these break down to aldehydes, ketones, acids and hydrocarbons; the off odours and flavours of these products are called rancidity (Figure 5). The classical measure of oxidation is the peroxide value, a titration; it has no clear relation to the smell. For this reason, headspace gas chromatography of single carbonyl compounds, with hexanal as the indicator, is often preferred.[4] A review from the CVUA Karlsruhe in 2023 states that the gas chromatographic methods designed for oils did not prove effective for cosmetic matrices.[23]

OO O2 OOHOO cleavage O linoleic acid residue of a triacylglycerol hydroperoxide first product of oxidation hexanal one of the aldehydes formed O nonanal, the calibration compound CHO proton at 9.76 ppm carbon of the CHO group

Figure 5. Oxidation of a linoleic acid residue to a hydroperoxide and cleavage to hexanal, one of several aldehydes that form. Textbook chemistry, our drawing. Nonanal is the calibration compound of [4].

In 2010 Lachenmeier and co-workers at the CVUA Karlsruhe, which sometimes receives lipsticks that taste rancid, tested whether the aldehyde protons could serve for screening. They describe their work as the first use of these signals for the quality control of oils. Aldehyde protons appear between 9 and 10 ppm. The laboratory expressed the sum of all aldehydes as nonanal, whose signal is at 9.76 ppm. A sample of 200 mg, scraped from the surface of the lipstick, was dissolved in 1.4 mL of deuterated chloroform at 50 °C, extracted with water to remove hydrophilic compounds, centrifuged and filtered. The measurement on a 400 MHz spectrometer took about 25 minutes per sample and ran automatically, more than 50 per day. The detection limit, stated for oils, was about 40 mg/kg.[4]

The authorities had submitted 72 vegetable oils and 38 lipsticks for control. Of the oils, 22 contained detectable aldehydes, two of them more than 100 mg/kg; both of these tasted rancid. Of the lipsticks, 26 contained aldehydes above the detection limit, 15 of them more than 100 mg/kg. One lipstick with a rancid taste contained more than 2,000 mg/kg. The authors name the causes: production temperatures above 75 °C, pigments such as iron oxides that promote oxidation, unsaturated oils such as almond oil in the formula, and frequent opening during use. They propose the method for product development and for accelerated shelf-life tests in industry.[4]

The method counts every aldehyde with a signal near that of nonanal, and the authors name the fragrance compound 7-hydroxycitronellal among them; a positive result therefore needs the ingredient list. It does not separate the individual aldehydes, and the authors recommend gas chromatography to confirm a positive result.[4]

In 2021 Wann, Grootveld and co-workers at De Montfort University in Leicester heated culinary oils in a laboratory simulation of shallow frying and assigned the aldehyde signals at 600 MHz, where they separate by class. The saturated n-alkanals give a triplet at 9.74 to 9.77 ppm and the unsaturated aldehydes five doublets between 9.48 and 9.62 ppm; Figure 6 shows three of them.[16]

O n-alkanals 9.74–9.77 ppm, triplet O (E)-2-alkenals 9.48–9.51 ppm, doublet O (E,E)-2,4-alkadienals 9.52–9.54 ppm, doublet OHO 4-hydroxy-(E)-2-alkenals 9.57–9.60 ppm, doublet SCHEMATIC SPECTRUM n-alkanals 4-hydroxy-2-alkenals 2,4-alkadienals 2-alkenals 9.9 9.8 9.7 9.6 9.5 9.4 chemical shift of 1H, ppm

Figure 6. Four of the classes of aldehydes that [16] assigns in oxidised oils, with the chemical shifts of their CHO protons at 600 MHz. The range of 9.57 to 9.60 ppm is shared by the 4-hydroxy- and the 4-hydroperoxy-(E)-2-alkenals. One member of each class is drawn as our example. The spectrum is schematic. Our drawing.

In 2012 Skiera and co-workers, in work for a doctoral thesis at the University of Würzburg with the CVUA Karlsruhe and a commercial NMR laboratory in Cologne, showed that the p-anisidine value, the classical measure of aldehydes, can be modelled as a linear combination of their NMR integrals.[24] They measured the hydroperoxides in 290 edible oils by 1H NMR; the performance was similar to that of the peroxide value, although the two methods disagreed considerably for some kinds of oil.[25]

Aloe vera: the test behind the certification seal

The International Aloe Science Council, the trade association of the industry, certifies raw materials and products with a seal. Its standard states that only products that contain acemannan, the acetylated polysaccharide of the gel, may be labelled as aloe vera.[26] The programme requires at least 5 % of acemannan in the dry matter of a raw material.[27] In 2010 the council selected a commercial laboratory in the United States as its primary provider of the NMR measurements for the seal.[5]

OOOOOHOOH OOHOHOHOHHO OHOOHOOH acetylated mannose unit of the aloe polysaccharide glucose constituent of the gel malic acid constituent of the gel OHOOHOOHOOH OHOOH OOHOOHOOH isocitric acid marker of whole-leaf material lactic acid formed when the material degrades glucose unit of maltodextrin an adulterant

Figure 7. Compounds that the 1H spectrum of an aloe material shows.[17, 28] Top row: constituents of the gel. The acetyl CH3 group of the polysaccharide, marked in gold, gives signals between 2.0 and 2.2 ppm.[17] Bottom row: compounds that indicate whole-leaf material, degradation or adulteration. Our drawing; the position of the acetyl group on the mannose unit is an example.

A quantitative 1H NMR method for aloe products was published in 2010 by Jiao and co-workers, of a supplier of plant ingredients in the United States. One spectrum, with nicotinamide as the internal standard, gives the acetylated polysaccharide, glucose and malic, lactic and acetic acid. The authors validated the method on two standard materials of the council, one from the inner leaf and one from the decolourised whole leaf, and applied it to 18 commercial products.[29] The laboratory that measures for the seal describes its own method as similar to this one.[17]

The first three of these compounds are constituents of the gel (Figure 7). Maltodextrin, an adulterant, is detected and quantified. Isocitric acid and its lactone, which form in the green part, mark material made from the whole leaf. Lactic, acetic, succinic, pyruvic, fumaric and formic acid form as the material degrades, and the preservatives sodium benzoate and potassium sorbate give their own signals.[17, 28]

In a survey published in 2007, Bozzi and co-workers examined nine aloe vera gel powders from leading international suppliers by 1H NMR, compared them with fresh gel and assessed authenticity, adulterants, impurities and preservatives.[30] According to a later review, only three of the nine contained satisfactory amounts of acemannan.[10]

In finished products the method applies only when the aloe constituents are concentrated enough to be observed and the signals of other ingredients do not cover them.[17] The raw material is therefore the suitable sample, not a cream that contains a small proportion of it.

Polymers and surfactants: composition without a reference substance

For several polymers and oligomers used in cosmetics, the official test of composition is an NMR measurement. Poloxamers are block copolymers of ethylene oxide and propylene oxide. The United States Pharmacopeia–National Formulary (USP–NF) determines their oxyethylene content from two areas of the 1H spectrum: the methyl doublet of the oxypropylene units at about 1.08 ppm, and the band of all OCH2 and OCH groups between 3.2 and 3.8 ppm (Figure 8). The monograph requires 81.8 ± 1.9 % of oxyethylene for poloxamer 188 and 73.2 ± 1.7 % for poloxamer 407.[6]

HO O CH3 O O H a b a oxyethylene unit oxypropylene unit oxyethylene unit area A1: doublet of the CH3 groups at about 1.08 ppm area A2: band of all OCH2 and OCH groups at 3.2–3.8 ppm CALCULATION OF THE MONOGRAPH α = A2 / A1 − 1 oxyethylene, % (w/w) = 3300 α / (33 α + 58) EXAMPLE: POLOXAMER 188, a = 80, b = 27 α = 8a / 3b = 7.90 oxyethylene = 81.8 % (w/w); requirement of the monograph: 81.8 ± 1.9 %

Figure 8. A poloxamer and the test of the USP–NF monograph. Structure, signal areas and formula from [6]. The relation α = 8a/3b and the example are our calculation from the values of a and b given in the monograph.

The same pharmacopoeia tests polyoxyl 20 cetostearyl ether, an ethoxylated fatty alcohol known to formulators as ceteareth-20, by the ratio of two areas of its 1H spectrum, from 2.35 to 4.9 ppm and from 0.4 to 2.35 ppm. The ratio gives the average number of oxyethylene units per molecule, which must lie between 17.2 and 25.0.[31] Hydroxypropyl betadex is a hydroxypropyl ether of β-cyclodextrin. The methyl doublet of its hydroxypropyl groups at 1.2 ppm is compared with the signals of the glycosidic protons, those on C-1 of the glucose units, between 5.0 and 5.4 ppm (Figure 9). The result is the molar substitution, the number of hydroxypropyl groups per glucose unit, which must lie between 0.40 and 1.50 and within 10 % of the value on the label.[15]

OOOOHOHOOH OOH2NOHOOH OOHNOOHOOH glucose unit of hydroxypropyl betadex with one hydroxypropyl group deacetylated unit of chitosan acetylated unit of chitosan CH3 group of the substituent

Figure 9. Repeat units of two further materials whose composition is measured by 1H NMR: hydroxypropyl betadex[15] and chitosan.[32] The position of the hydroxypropyl group on the glucose unit is an example; the CH3 group of each substituent is marked. Our drawing.

The standard ASTM F2260, written for chitosan in biomedical and pharmaceutical use, determines the degree of deacetylation by 1H NMR; it is valid for the chloride and the glutamate salt between 50 and 99 %. The chitosan is first depolymerised with nitrous acid in deuterium oxide to chains of about 15 to 30 units, which lowers the viscosity and gives well-resolved signals. The standard notes that the measurement needs no calibration curve, because the spectrum reports the relative numbers of chemically distinct protons directly.[32] All of these tests give the mean composition of the material, not the distribution of chain lengths or of substituents.

In 2004 Koike and co-workers, at the analytical research centre of a Japanese manufacturer of consumer products, determined anionic and cationic surfactants by 1H NMR against an internal standard, potassium hydrogen phthalate or the sodium salt of deuterated trimethylsilylpropionic acid. They give the reason: in many cases no reference surfactant corresponds completely to the analyte. One spectrum took less than 5 minutes, and the detection limit for sodium lauryl sulfate was about 60 ppm. The results for surfactant standards and real samples agreed with the titration method of the Japanese industrial standard, with relative standard deviations of 0.12 to 0.44 %.[33]

Adulterated essential oils

Vegetable oil: the adulterant that gas chromatography does not detect

An essential oil that is diluted with a vegetable oil keeps its colour, refractive index and density.[1] Gas chromatography, the reference technique for essential oils, does not detect the addition: triacylglycerols are not volatile and remain in the injector (Figure 10).[2]

(R)-limonene volatile: detected by gas chromatography R1OOOOR2OOR3 triacylglycerol of a vegetable oil not volatile: not detected by gas chromatography essential oil with 0.8 % (w/w) of vegetable oil essential oil without vegetable oil 4.27 4.12 4.4 4.3 4.2 4.1 4.0 chemical shift of 1H, ppm SCHEMATIC SPECTRUM glycerol CH2 groups, the source of the two signals

Figure 10. Left: (R)-limonene, the main constituent of orange oil, and the general structure of a triacylglycerol; R1, R2 and R3 are the chains of three fatty acids. Right: schematic 1H spectrum between 4.45 and 3.95 ppm, with the two shifts of [1]. Our drawing.

In 2021 Truzzi, Bertelli and co-workers at the University of Modena and Reggio Emilia mixed lavender, citronella, rosemary and orange oil with seven vegetable oils at 0.8 to 50 % (w/w) and recorded 1H and 13C spectra of the mixtures in deuterated chloroform at 600 MHz. The CH2 groups of glycerol give two 1H signals, at 4.12 and 4.27 ppm, and the pure essential oils showed none at these positions. At 0.8 % (w/w), the lowest concentration tested, the signal-to-noise ratio of the glycerol signals was 130 in the 1H spectrum (about 5 minutes) and 5 in the 13C spectrum (47 minutes).[1]

In the 13C spectrum, fourteen signals of the palmitic, oleic and linoleic acid residues have different relative areas in each vegetable oil. The comparison of these ratios with those of the pure oils identified the added oil in all mixtures from 3 to 50 % (w/w). Corn, sunflower and soybean oil, which have similar fatty acid compositions, were the most difficult to distinguish.[1]

The authors then bought 20 essential oils from online shops. Four contained triacylglycerols: sunflower oil in a thyme oil, soybean oil in a mint oil, and corn oil in a palmarosa and an oregano oil. A diffusion-ordered spectrum (DOSY), which separates large molecules from small ones, confirmed that the signals came from triacylglycerols and not from free fatty acids or other esters.[1]

In a second paper of the same year the group measured the amount of the added oil from the glycerol signals, against an external reference compound. For mixtures of rosemary oil with 0.8 to 50 % of corn oil, the mean accuracy was 91.81 % from the 1H and 89.77 % from the 13C spectra. The detection limit of the 1H measurement was 0.02 % (w/w) and the limit of quantification 0.07 % (w/w); the detection limits for the two glycerol signals of the 13C spectrum were 0.64 and 0.82 % (w/w). By the 1H measurement, the four adulterated oils of the survey contained between 10 and 42 % (w/w) of vegetable oil (Figure 11).[2]

0 10 20 30 40 50 vegetable oil found, % (w/w) 32.8 33.1 palmarosa oil with corn oil 41.7 43.3 thyme oil with sunflower oil 10.3 10.3 mint oil with soybean oil 17.8 17.5 oregano oil with corn oil from the 1H spectrum from the 13C spectrum

Figure 11. Vegetable oil found in the four adulterated essential oils of the survey, measured from the glycerol signals of the 1H and of the 13C spectrum. The oils are palmarosa (Cymbopogon martinii), thyme (Thymus capitatus), mint (Mentha arvensis) and oregano (Origanum vulgare). Our drawing of the values of Table 5 of [2].

Synthetic linalool in lavender oil: the configuration and the deuterium content

A second fraud adds nothing foreign: synthetic linalool or linalyl acetate in lavender oil, of which the two compounds are the main constituents, is a well-known adulteration.[34] An ordinary gas chromatogram shows the same two compounds as in a genuine oil. (R)-Linalool and (R)-linalyl acetate of high enantiomeric purity are characteristic of genuine lavender oil (Figure 12), and synthetic material is partly or fully racemic. Enantioselective gas chromatography, which separates the (R) and the (S) form on a chiral column, is the established test for this adulteration.[34]

OH OO (R)-linalool (R)-linalyl acetate stereocentre; the configuration shown is the one typical of genuine lavender oil

Figure 12. Linalool and linalyl acetate in the (R) configuration.[34] Our drawing.

The natural deuterium content of a molecule depends on the way in which it was made. In 1992 Hanneguelle, Thibault, Naulet and Martin measured the deuterium at each position of the molecule by 2H NMR, a method called SNIF-NMR (site-specific natural isotope fractionation), for linalool and linalyl acetate from chemical synthesis and from essential oils of known botanical and geographical origin. The overall 13C or 2H content of the molecule, measured by isotope ratio mass spectrometry, was in general not an efficient criterion for distinguishing natural from synthetic samples. The distribution of deuterium varied widely with the origin. A discriminant analysis of ten site-specific hydrogen isotope ratios and the overall carbon value distinguished natural from synthetic material unambiguously.[35]

In 2002 Bilke and Mosandl at the University of Frankfurt measured the 2H/1H ratio of the whole molecule by gas chromatography coupled to isotope ratio mass spectrometry. Natural linalool and linalyl acetate were clearly differentiated from synthetic samples, and the authors describe the method as an additional tool beside the enantioselective analysis.[34] A site-specific 2H measurement at natural abundance takes a long time and is not routine.

Plant oils and butters: composition, authenticity and quality indices

In 2004 Knothe and Kenar, at a laboratory of the United States Department of Agriculture in Peoria, showed that the 1H spectrum at 400 MHz gives the amounts of the common unsaturated fatty acids of a vegetable oil. Gas chromatography requires the conversion of the oil into derivatives; NMR measures the oil itself. The methyl signal of linolenic acid lies downfield of those of the other fatty acids and is integrated separately (Figure 13). The signals of the CH2 groups next to one double bond and between two then give oleic and linoleic acid. The results agreed well with those of gas chromatography.[36]

OHO oleic acid one double bond OHO linoleic acid two double bonds OHO α-linolenic acid three double bonds CH2 group between two double bonds SCHEMATIC SPECTRUM OF A VEGETABLE OIL CH=CH 5.30–5.40 glycerol CH2 4.11–4.32 =CH–CH2–CH= 2.75–2.82 OCO–CH2 2.26–2.36 =CH–CH2 1.95–2.11 (CH2)n 1.20–1.40 CH3 0.84–0.91 ω-3 CH3 0.95–0.99 1 2 3 4 5 chemical shift of 1H, ppm

Figure 13. Top: the three common unsaturated fatty acids of vegetable oils. Bottom: schematic 1H spectrum of a vegetable oil with the shift ranges assigned at 600 MHz in [16]. Our drawing.

The fatty acid profile is a test of authenticity. Argan oil is expensive and is used in cosmetics, and sunflower oil is thought to be one of its most common adulterants. In 2020 Gunning, Kemsley and co-workers in Norwich built a model of authentic argan oil from the spectra of oils of assured origin. They applied it to 28 oils bought mostly from online retailers, all labelled as 100 % pure argan oil. Four were rejected. Their spectra were consistent with more polyunsaturated fatty acids than expected, and two showed significant amounts of α-linolenic acid, which authentic argan oil contains only in traces. The authors estimate the rate of fraud in the sector at 14 %, with a confidence interval of 7 to 23 %. In test mixtures with argan oil, sunflower oil was detected from about 20 % (w/w). The spectra of this study were recorded at 60 MHz on the neat oil; at that field the proportions of mono- and polyunsaturated fatty acids are estimated by regression against gas chromatographic data.[8] At 400 MHz the signals of the fatty acids are integrated separately.[36]

Suppliers specify oils by classical indices: the saponification, acid and iodine values, each determined by a separate titration, and the ester value, which is the difference between the first two. In a paper published online in 2022, Triyasmono, Holzgrabe and co-workers at the University of Würzburg obtained all four from one 1H spectrum at 400 MHz with dimethyl sulfone as the internal standard. The CH2 signals next to the carboxyl groups between 2.20 and 2.37 ppm gave the saponification, acid and ester values, and those of CH=CH between 5.27 and 5.37 ppm the iodine value. For red fruit oil, the results agreed with those of the standard methods.[37]

Mineral oil in petrolatum and lip products

Petrolatum, paraffin oils and paraffin waxes are mixtures of hydrocarbons from mineral oil. Two sum values describe them: mineral oil saturated hydrocarbons (MOSH) and mineral oil aromatic hydrocarbons (MOAH). The aromatic fraction is examined closely, because it may contain carcinogenic polycyclic aromatic hydrocarbons as contaminants. The usual method couples liquid chromatography to gas chromatography with flame ionisation detection (LC-GC-FID). It assigns the compounds to the two fractions by their retention and gives no information on their chemical nature.[7]

decalin saturated: reference for MOSH naphthalene aromatic: reference for MOAH OH BHT antioxidant that passes the clean-up benzo[a]pyrene expected below 1 mg/kg: not measurable MOAH: 6.5–9.2 ppm MOSH: 0.2–3.0 ppm protons on aromatic rings protons on saturated carbon 0 1 2 3 4 5 6 7 8 9 10 chemical shift of 1H, ppm

Figure 14. Top: decalin and naphthalene, the compounds in which the two fractions are expressed;[7] butylated hydroxytoluene (BHT), which passes the clean-up for lip products;[11] and benzo[a]pyrene, a polycyclic aromatic hydrocarbon. Bottom: the two regions of the 1H spectrum that are integrated.[7] Our drawing.

In 2017 Lachenmeier, Kuballa and co-workers at the CVUA Karlsruhe measured the two fractions by the kind of proton: those on aromatic rings between 6.5 and 9.2 ppm and those on saturated carbon between 0.2 and 3.0 ppm, in both cases without the signals of the solvent (Figure 14). The sample was diluted in deuterated chloroform and measured at 400 MHz without further preparation, and the detection limit was below 0.1 g/100 g. The 27 samples, which were cosmetics, raw materials and technical mineral oil products, contained up to 1.10 g/100 g of MOAH. A German federal institute had reported about 1.7 to 5.0 % of MOAH for cosmetic products based on petrolatum, and the authors conclude that earlier methods may have overestimated the aromatic fraction.[7]

Aromatic ingredients of a compounded product, such as UV filters, preservatives, perfumes and colourants, give signals in the same region and would be counted as MOAH. Individual polycyclic aromatic hydrocarbons, which are expected at considerably less than 1 mg/kg, are beyond the sensitivity of NMR; the MOAH value is an upper limit for them.[7]

For lip products the laboratory added a clean-up step in 2019. The sample is dissolved in cyclohexane and passed through silica gel, which retains the polar aromatic ingredients. Of the aromatic ingredients tested, BHT, an antioxidant, passed through at 90 % and must be taken into account; for the others that dissolved in cyclohexane the figure was 0.5 % or less. The detection limit was 0.03 g/100 g, and the recovery was sufficient above 0.2 g/100 g. Of 57 lipsticks, lip glosses and lip balms from retail, 36 contained at most 0.1 g/100 g of MOAH, 13 between 0.1 and 0.4 g/100 g, and 8 at least 0.4 g/100 g, with a maximum of 1.6 g/100 g. Of the products in the last group, 88 % were liquid. The authors write that products with such contents should be examined further for compliance with the purity criteria of the Cosmetics Regulation. They describe the NMR method as a good complement to LC-GC-FID: the migration of saturated hydrocarbons into the aromatic fraction, a problem of the chromatographic separation, does not occur.[11]

Finished products: bakuchiol in serums and counterfeit perfume

Bakuchiol in face serums

Bakuchiol is a plant compound that is used in cosmetics as an alternative to retinoids (Figure 15). In 2025 Grzelecka, Paradowska and co-workers at the Medical University of Warsaw, with colleagues at the University of Würzburg, compared methods for its content in six products from the Polish market: 1H qNMR at 300 MHz with nicotinamide as the internal standard, high-performance liquid chromatography (HPLC) and UV spectrophotometry.[14]

HO bakuchiol stereocentre MEASUREMENT IN FACE SERUMS Signals integrated 5.59–5.71, 5.80–5.94, 6.36–6.46 and 6.78–6.88 ppm Internal standard nicotinamide Solvent CDCl3 and DMSO, 5 : 1 by volume Sample 250 mg of serum in 1 mL

Figure 15. Bakuchiol and the conditions of the qNMR measurement in [14]. Our drawing.

Bakuchiol in six cosmetic products

Product Label qNMR HPLC
1, oil serum 1 % declared 0.59 0.51
2, oil serum listed, no content not detected not detected
3, oil serum 1 % declared 0.83 1.06
4, oil serum listed, no content 3.57 3.62
5, emulsion 1 % declared not measured not measured
6, emulsion listed, no content not measured not measured

Values of [14], in % (w/w). “Listed, no content”: bakuchiol is in the list of ingredients, and the label does not state how much.

UV spectrophotometry gave 0.17 % for product 2, in which neither of the other methods detected bakuchiol. The two emulsions did not dissolve completely, and bakuchiol could not be extracted properly from them. In two of the oil serums the signals of other ingredients overlapped with some of those of bakuchiol. An NMR measurement took about 4 minutes, an HPLC analysis about 40.[14]

Counterfeit perfume: a fingerprint of the whole product

In 2021 Pacholczyk-Sienicka, Ciepielowski and Albrecht at Lodz University of Technology recorded 1H spectra at 700 MHz of eight authentic perfumes of six fragrances, of six counterfeits bought over the internet and of 14 imitations sold as “inspired” versions. A principal component analysis of the spectra separated the authentic perfumes from the counterfeit and the inspired ones without the assignment of a single signal. For one fragrance, NMR identified 23 compounds, and gas chromatography with mass spectrometry (GC-MS) detected 34. The authors name the limits of the study: few authentic samples were available, and two of the same perfume differed slightly, so that several batches of the original are needed for a comparison.[38]

What the safety report asks for

A cosmetic product needs a safety report before it is placed on the market in the EU or in Great Britain. Part A of the report must contain the physical and chemical characteristics of the substances and mixtures and their purity. Where traces of a prohibited substance are present, it must contain evidence that they are technically unavoidable.[39]

For the dossier of an ingredient, the Scientific Committee on Consumer Safety (SCCS) asks for the spectra and chromatograms that identify the substance and names NMR among the techniques. Impurities are to be characterised and quantified, for example by HPLC with diode array detection, by liquid or gas chromatography with mass spectrometry, or by NMR spectroscopy.[40] An assigned NMR spectrum of a raw material is one document that meets these requirements: it identifies the compound, and a quantitative measurement gives the content. Part 4 of this series describes how such a purity value is obtained.

Sample, time and detection limit, and what NMR does not measure

Sample, measurement and detection limit in the cited work

Analysis Sample and method Detection limit
Vege­table oil in an essen­tial oil[1, 2] About 50 µL in 550 µL of CDCl3. 1H: 32 scans in 5 min 20 s; 13C: 256 scans in 47 min; 600 MHz 0.02 % (w/w) by 1H; 0.64 to 0.82 % (w/w) by 13C
Free form­aldehyde beside a releaser[3, 19] Finished product or patch-test preparation. 13C, in an “acceptable measuring time” About 0.002 % (w/w) in finished products; 0.01 % in patch-test preparations
Hydrogen per­oxide[13] 10 to 100 µL, diluted with DMSO-d6. 1H 0.0001 to 0.01 % (w/w)
Alde­hydes in lip­stick[4] 200 mg in 1.4 mL of CDCl3, extracted with water. 1H: 128 scans, about 25 min; 400 MHz About 40 mg/kg, stated for oils
Surfac­tants[33] Solution with an internal standard. 1H: less than 5 min About 60 ppm for sodium lauryl sulfate
MOAH in petro­latum[7, 11] 50 mg in 1.5 mL of CDCl3. 1H; 400 MHz Below 0.1 g/100 g
MOAH in lip products[11] 50 mg in 1.5 mL of cyclo­hexane, passed through silica gel. 1H with suppression of the solvent signal: 32 scans; 400 MHz 0.03 g/100 g
Baku­chiol in an oil serum[14] 250 mg in 1 mL. 1H: about 4 min; 300 MHz Not stated
Perfume, finger­print[38] 50 µL in 600 µL of CDCl3. 1H: 32 scans; 700 MHz Not appli­cable
Polox­amer, USP–NF[6] 0.1 to 0.2 g in 1 mL of D2O or CDCl3. 1H, 0 to 5 ppm Not appli­cable
Hydroxy­propyl betadex, USP–NF[15] At least 10 mg in 0.75 mL of D2O. 1H: at least 8 scans, field of at least 6 tesla Not appli­cable

CDCl3: deuterated chloroform; D2O: deuterium oxide. The pharmacopoeial texts are those of USP 35–NF 30. 40 mg/kg is 0.004 % (w/w). A field of 6 tesla corresponds to a 1H frequency of about 255 MHz.

For the amounts to be correct, the laboratory measures against a known quantity of a standard and allows every signal to recover fully between two pulses. Part 4 and Part 9 of this series describe both conditions. Figure 16 places the detection limits beside the legal ones.

LEGAL LIMITS AND EXPECTED LEVELS Single polycyclic aromatic hydrocarbons, expected level Warning “releases formaldehyde” Fragrance allergens in leave-on products Fragrance allergens in rinse-off products Hydrogen peroxide in oral products Fluorine in oral products Hydrogen peroxide in nail, skin and hair products DETECTION LIMITS OF THE NMR METHODS Hydrogen peroxide Free formaldehyde beside a releaser Aldehydes in oils Vegetable oil in an essential oil Aromatic mineral oil hydrocarbons (MOAH) 0.00001 0.0001 0.001 0.01 0.1 1 10 100 content, % (w/w), logarithmic scale

Figure 16. Legal limits for cosmetic products and detection limits of the NMR methods of this article on one logarithmic scale. Legal limits: [12] for the warning on formaldehyde, [21] for the others; expected level of polycyclic aromatic hydrocarbons from [7]. Detection limits: hydrogen peroxide;[13] formaldehyde by 13C NMR;[3, 19] aldehydes;[4] vegetable oil by 1H NMR;[2] MOAH without and with the clean-up.[7, 11] Our drawing of published values.

Questions that NMR answers less well

Question Reason, and the method that answers it
Is the formaldehyde released below the labelling threshold of 0.001 %? The detection limit of the 13C method, about 0.002 to 0.01 % (w/w), is above the threshold[3, 12, 19]
Does a fragrance allergen exceed its labelling threshold? The thresholds are 0.001 % in leave-on and 0.01 % in rinse-off products,[21] at or below the detection limits of the NMR methods for organic compounds in this article, which begin at 0.002 %; fragrance allergens are measured by GC-MS
How much of a single polycyclic aromatic hydrocarbon does a mineral oil product contain? These compounds are expected at considerably less than 1 mg/kg, beyond the sensitivity of NMR[7]
Was synthetic linalool added to a lavender oil? Enantioselective gas chromatography is the established test;[34] the 2H NMR method is a specialised measurement[35]
Which terpenes does an essential oil contain, and in what proportions? Gas chromatography is the reference technique for the analysis of essential oils[2]
Which aldehyde makes a lipstick rancid? The screening at 400 MHz gives the sum; gas chromatography separates the individual aldehydes[4]
How much of an active ingredient does a cream or lotion contain? The emulsions of the bakuchiol study did not dissolve completely and were not measured[14]
Does a product meet the limits for heavy metals and micro-organisms? These are measured by element analysis and by microbiological methods; see our article on cosmetics testing

Summary

  • An NMR spectrum is recorded on the dissolved sample without a separation or a reagent, and the area of each signal gives the amount of the compound.
  • 13C NMR measures free formaldehyde beside its releaser without shifting the equilibrium, down to about 0.002 % (w/w) in finished products.
  • 1H NMR measured formaldehyde in hair-straightening products, seven of ten of which exceeded the limit of that time, and hydrogen peroxide down to 0.0001 % (w/w); the fluorine compounds of toothpaste are determined by 19F NMR.
  • The aldehydes of rancid fat appear between 9 and 10 ppm. An official laboratory found them in 26 of 38 lipsticks, and in 15 at more than 100 mg/kg. At 600 MHz the classes of aldehydes give separate signals.
  • The aloe industry certifies its products by NMR: the spectrum shows the acetylated polysaccharide, maltodextrin and the acids that form on degradation. In one survey, three of nine aloe powders contained satisfactory amounts of acemannan.
  • The United States Pharmacopeia prescribes NMR for the oxyethylene content of poloxamers, the chain length of an ethoxylated fatty alcohol and the molar substitution of hydroxypropyl betadex; an ASTM standard does so for the degree of deacetylation of chitosan. Surfactants are measured against an internal standard where no identical reference substance exists.
  • Gas chromatography does not show vegetable oil in an essential oil. The glycerol signals at 4.12 and 4.27 ppm detect it from 0.02 % (w/w); four of 20 essential oils from online shops contained 10 to 42 % (w/w) of it.
  • One 1H spectrum gives oleic, linoleic and linolenic acid in a plant oil. In one survey, four of 28 retail oils labelled as pure argan oil were rejected as adulterated.
  • Aromatic mineral oil hydrocarbons are measured as one value between 6.5 and 9.2 ppm, with a detection limit below 0.1 g/100 g, and of 0.03 g/100 g after a clean-up; eight of 57 lip products contained at least 0.4 g/100 g.
  • In oil serums, qNMR and HPLC gave similar bakuchiol contents: about half of the declared amount in one product, and none detectable in another that lists the compound. Emulsions could not be measured.
  • In one study, the comparison of whole 1H spectra separated authentic perfumes from counterfeit and “inspired” ones; several batches of the original are needed for it.
  • For organic compounds in cosmetic products, the NMR methods cited here do not reach thresholds of 10 mg/kg and below, such as that for the labelling of released formaldehyde; NMR does not replace gas chromatography for the volatile composition of an essential oil or for the detection of synthetic linalool in lavender oil.

NMR analysis of cosmetic ingredients from NorrChemica

Our NMR service measures raw materials and anhydrous products in solution at 400 MHz, priced per sample: a 1H spectrum is Package I, 1H and 13C spectra are Package III, and a quantitative measurement against a calibrated standard is Package V. Measurements at 600 MHz and above are quoted on request. A chemist assigns each spectrum, and the signed report includes the native and processed data files. Suitable samples are plant oils and butters, essential oils, mineral and plant waxes, surfactants, polymers, single ingredients and oil-based serums. The service does not accept biohazardous samples or material of animal origin. Please describe the sample and the question on the request form. Research use only.

Request NMR analysis → Cosmetics testing

NMR tubes made in Germany by Deutero and shipped from Finland. See the NMR tube guide →

References

  1. Truzzi E, Marchetti L, Benvenuti S, Ferroni A, Rossi MC, Bertelli D. Novel strategy for the recognition of adulterant vegetable oils in essential oils commonly used in food industries by applying 13C NMR spectroscopy. J. Agric. Food Chem. 2021, 69, 8276–8286. DOI
  2. Truzzi E, Marchetti L, Benvenuti S, Righi V, Rossi MC, Gallo V, Bertelli D. A novel qNMR application for the quantification of vegetable oils used as adulterants in essential oils. Molecules 2021, 26, 5439. DOI
  3. Emeis D, Anker W, Wittern K-P. Quantitative 13C NMR spectroscopic studies on the equilibrium of formaldehyde with its releasing cosmetic preservatives. Anal. Chem. 2007, 79, 2096–2100 (abstract). DOI
  4. Lachenmeier DW, Gary M, Monakhova YB, Kuballa T, Mildau G. Rapid NMR screening of total aldehydes to detect oxidative rancidity in vegetable oils and decorative cosmetics. Spectroscopy Europe 2010, 22 (6), published 1 December 2010. Link
  5. Process NMR Associates. IASC chooses Process NMR Associates to provide analytical services for certification program. Press release of a commercial laboratory, 24 November 2010. Link
  6. United States Pharmacopeia 35–National Formulary 30, monograph Poloxamer, test “Weight percent oxyethylene” (copy on a third-party site). Link
  7. Lachenmeier DW, Mildau G, Rullmann A, Marx G, Walch SG, Hartwig A, Kuballa T. Evaluation of mineral oil saturated hydrocarbons (MOSH) and mineral oil aromatic hydrocarbons (MOAH) in pure mineral hydrocarbon-based cosmetics and cosmetic raw materials using 1H NMR spectroscopy. F1000Research 2017, 6, 682 (version 2). DOI
  8. Gunning Y, Jackson AJ, Colmer J, Taous F, Philo M, Brignall RM, El Ghali T, Defernez M, Kemsley EK. High-throughput screening of argan oil composition and authenticity using benchtop 1H NMR. Magn. Reson. Chem. 2020, 58, 1177–1186. DOI
  9. Monakhova YB, Kuballa T, Mildau G, Kratz E, Keck-Wilhelm A, Tschiersch C, Lachenmeier DW. Formaldehyde in hair straightening products: rapid 1H NMR determination and risk assessment. Int. J. Cosmet. Sci. 2013, 35, 201–206 (abstract). DOI
  10. Foster M, Hunter D, Samman S. Evaluation of the nutritional and metabolic effects of Aloe vera. Chapter 3 of Herbal Medicine: Biomolecular and Clinical Aspects, 2nd edition (NCBI Bookshelf NBK92765). Link
  11. Weber S, Schmidt T, Schumacher P, Kuballa T, Mildau G, Walch SG, Hartwig A, Lachenmeier DW. Quantification of mineral oil aromatic hydrocarbons (MOAH) in anhydrous cosmetics using 1H NMR. J. Chem. 2019, 1680269. DOI
  12. Commission Regulation (EU) 2022/1181 of 8 July 2022 amending the preamble of Annex V to Regulation (EC) No 1223/2009 on cosmetic products. OJ L 184, 11.7.2022, p. 3. Link
  13. Monakhova YB, Diehl BWK. Rapid 1H NMR determination of hydrogen peroxide in cosmetic products and chemical reagents. Anal. Methods 2016, 8, 4632–4639 (abstract). DOI
  14. Grzelecka M, Siudem P, Tyburc N, Triyasmono L, Holzgrabe U, Paradowska K. Comparison of quantification using UV-Vis, NMR, and HPLC methods of retinol-like bakuchiol present in cosmetic products. Int. J. Mol. Sci. 2025, 26, 6638. DOI
  15. United States Pharmacopeia 35–National Formulary 30, monograph Hydroxypropyl Betadex, test “Molar substitution” (copy on a third-party site). Link
  16. Wann AI, Percival BC, Woodason K, Gibson M, Vincent S, Grootveld M. Comparative 1H NMR-based chemometric evaluations of the time-dependent generation of aldehydic lipid oxidation products in culinary oils exposed to laboratory-simulated shallow frying episodes. Foods 2021, 10, 2481. DOI
  17. Edwards JC. Quantitative proton nuclear magnetic resonance spectrometry (1H-NMR) for determination of acetylated polysaccharides, glucose, maltodextrin, and isocitrate in aloe vera leaf juice. Conference abstract, 2012 (Link), and the page on whole-leaf markers of the same commercial laboratory (Link).
  18. Free formaldehyde preservative stability in personal care, Part I: water. Cosmetics & Toiletries (trade journal), online article (introduction). Link
  19. Emeis D, de Groot AC, Brinkmann J. Determination of formaldehyde in formaldehyde-releaser patch test preparations. Contact Dermatitis 2010, 63, 57–62 (abstract). DOI
  20. The Cosmetic Products Regulation (EC) No 1223/2009 (Restriction of Chemical Substances) (Amendment and Transitional Provisions) Regulations 2026, SI 2026/23. Link
  21. Regulation (EC) No 1223/2009 on cosmetic products, Annex III, entries for hydrogen peroxide, for fluorine compounds and for fragrance allergens (version in force in Great Britain). Link
  22. Deng D, Deng P, Wang X, Hou X. Direct determination of sodium fluoride and sodium monofluorophosphate in toothpaste by quantitative 19F-NMR: a green analytical method. Spectrosc. Lett. 2009, 42, 334–340 (abstract). DOI
  23. Kuballa T, Kaltenbach KH, Teipel J, Lachenmeier DW. Liquid nuclear magnetic resonance (NMR) spectroscopy in transition: from structure elucidation to multi-analysis method. Separations 2023, 10, 572. DOI
  24. Skiera C, Steliopoulos P, Kuballa T, Holzgrabe U, Diehl B. 1H NMR approach as an alternative to the classical p-anisidine value method. Eur. Food Res. Technol. 2012, 235, 1101–1105 (abstract). DOI
  25. Skiera C, Steliopoulos P, Kuballa T, Holzgrabe U, Diehl B. 1H-NMR spectroscopy as a new tool in the assessment of the oxidative state in edible oils. J. Am. Oil Chem. Soc. 2012, 89, 1383–1391 (abstract). DOI
  26. International Aloe Science Council. Aloe FAQs (web page). Link
  27. Aloe vera products: the importance of IASC certification. WholeFoods Magazine (trade magazine), online article. Link
  28. Markus MA, Colson KL, Jiao P, Gafner S. Quantitative analysis of Aloe vera by NMR, automated using the Assure-RMS software. Planta Med. 2012, 78, P_65 (conference abstract). DOI
  29. Jiao P, Jia Q, Randel G, Diehl B, Weaver S, Milligan G. Quantitative 1H-NMR spectrometry method for quality control of Aloe vera products. J. AOAC Int. 2010, 93, 842–848 (abstract). PubMed
  30. Bozzi A, Perrin C, Austin S, Arce Vera F. Quality and authenticity of commercial aloe vera gel powders. Food Chem. 2007, 103, 22–30 (abstract). DOI
  31. United States Pharmacopeia 35–National Formulary 30, monograph Polyoxyl 20 Cetostearyl Ether, test “Average polymer length” (copy on a third-party site). Link
  32. ASTM F2260-18 (2025). Standard test method for determining degree of deacetylation in chitosan salts by proton nuclear magnetic resonance (1H NMR) spectroscopy (scope and significance). Link
  33. Koike R, Jo S, Azuma M, Wakisaka T. Precise and rapid determination of anionic and cationic surfactants by 1H nuclear magnetic resonance using an internal standard. Bunseki Kagaku 2004, 53, 1125–1131 (abstract). DOI
  34. Bilke S, Mosandl A. Authenticity assessment of lavender oils using GC-P-IRMS: 2H/1H isotope ratios of linalool and linalyl acetate. Eur. Food Res. Technol. 2002, 214, 532–535 (abstract and introduction). DOI
  35. Hanneguelle S, Thibault JN, Naulet N, Martin GJ. Authentication of essential oils containing linalool and linalyl acetate by isotopic methods. J. Agric. Food Chem. 1992, 40, 81–87 (abstract). DOI
  36. Knothe G, Kenar JA. Determination of the fatty acid profile by 1H-NMR spectroscopy. Eur. J. Lipid Sci. Technol. 2004, 106, 88–96 (abstract). DOI
  37. Triyasmono L, Schollmayer C, Schmitz J, Hovah E, Lombo C, Schmidt S, Holzgrabe U. Simultaneous determination of the saponification value, acid value, ester value, and iodine value in commercially available red fruit oil (Pandanus conoideus Lam.) using 1H qNMR spectroscopy. Food Anal. Methods 2023, 16, 155–167. DOI
  38. Pacholczyk-Sienicka B, Ciepielowski G, Albrecht Ł. The first application of 1H NMR spectroscopy for the assessment of the authenticity of perfumes. Molecules 2021, 26, 3098. DOI
  39. Regulation (EC) No 1223/2009 on cosmetic products, Annex I, Cosmetic product safety report (version in force in Great Britain). Link
  40. Scientific Committee on Consumer Safety. The SCCS Notes of Guidance for the testing of cosmetic ingredients and their safety evaluation, 12th revision, SCCS/1647/22, sections 3-2.4 and 3-2.5. Link

About this article. All sources were read on 17 November 2026, in the browser unless stated otherwise. Read in full: sources 1 and 4, the two legal texts 12 and 20 and Part A of source 39. Read in part: the passages and tables of the open full text of sources 2, 7, 8, 11, 14, 16, 37 and 38 that carry the cited statements and numbers; the passage of source 10 on commercial products; the three groups of entries of source 21 named in the reference; the scope and the section on significance of source 32; and the cited tests of the three monograph texts 6, 15 and 31, which are those of USP 35–NF 30 on a third-party site and were not compared with the current edition. Read as abstracts, on PubMed or on the sites of the publishers: sources 3, 9, 19, 22, 24, 25, 28, 29, 33, 34 and 36, and for source 22 the notes to the tables shown with the abstract. Sources 5, 17, 18, 26 and 27 are web pages. Sources 23 and 40 were read through a retrieval tool that returned extracts of the first part of each document, which includes the cited passages; the introduction of source 34 was read through the same tool in a copy held by a university repository. Sources 13, 30 and 35 are known to us from abstracts shown by a search engine or read through that tool; the pages of the publishers could not be opened. The bibliographic data of the journal articles were compared with the Crossref record of each article. The place of work of the authors of source 36 is taken from our research notes, and those of the authors of sources 3, 8, 13, 25 and 33 from the Crossref records; for source 24, which has the same authors as source 25, we assume the same places. That the two papers belong to a doctoral thesis is taken from the record of the thesis in the repository of the University of Würzburg. Two sources contain internal differences: the abstract and Table 10 of source 11 give 57 samples and its text lists 56, and the abstract of source 7 gives a lowest MOAH content of 0.02 g/100 g while its table lists 0.01 g/100 g for one paraffin oil; we cite 57 samples and the highest content only. Source 4 gives its detection limit for oil samples and states none for lipsticks. Source 33 gives its detection limit in ppm without a statement of the basis, and we have not converted it. Our own work: all structure drawings and stereodescriptors; the general formulas of Figure 4; the schemes of Figures 3 and 5, which are textbook chemistry, and the choice of DMDM hydantoin as the example; the statements that formaldehyde is present in water as methanediol and that a reagent which consumes free formaldehyde shifts the equilibrium; the examples of Figure 6 and the schematic spectra of Figures 6, 10 and 13; the relation α = 8a/3b and the example of Figure 8; the statements that limonene is the main constituent of orange oil and that linalool and linalyl acetate are the main constituents of lavender oil, which are textbook knowledge; the name ceteareth-20 and the description of hydroxypropyl betadex; the remark that the tests of composition give mean values; the conclusion on the labelling threshold; the remark that a positive aldehyde result needs the ingredient list; the remark on the suitable aloe sample; the statement that an ordinary gas chromatogram shows the same two compounds in a lavender oil with synthetic linalool and the remark on 2H spectra; the remark that an assigned spectrum of a raw material meets the requirements of the safety report; the statement that fragrance allergens are measured by GC-MS; the common names of the four essential oils of Figure 11; the number of six aloe powders in Figure 1, which is nine less three; the conversions of units; the statement that the ester value is the difference between the saponification and the acid value, which is textbook knowledge; the range of detection limits given in the introduction; the comparison of detection limits with legal limits in Figure 16 and in the text; and the last row of the table of questions that NMR answers less well.

NMR Series

  1. Part 1 — Reading 1H and 13C NMR
  2. Part 2 — Reading 19F, 31P and 11B NMR
  3. Part 3 — Reading 2D NMR: COSY, HSQC and HMBC
  4. Part 4 — Quantitative NMR: how a purity figure is made
  5. Part 5 — 15N NMR: which nitrogen is alkylated, protonated or coordinated?
  6. Part 6 — NMR tubes by experiment: which tube for which measurement
  7. Part 7 — Quadrupolar nuclei in NMR: broad lines and missing carbons
  8. Part 8 — Two sets of NMR signals: rotamers, diastereomers or an impurity
  9. Part 9 — NMR yield and isomer ratios: relaxation and the delay between pulses
  10. Part 10 — Silicon NMR: the glass signal, the negative NOE and M, D, T, Q units
  11. Part 11 — Solid-state NMR: magic-angle spinning, cross-polarisation and what a CP-MAS spectrum counts
  12. Part 12 — Enantiomeric excess by NMR: when NMR is the best method, the chiral auxiliaries and the detection limit for the minor enantiomer
  13. Part 13 — Predicted NMR shifts: how accurate ChemDraw, MestReNova and other predictors are
  14. Part 14 — Phosphorus NMR in biology: phospholipids, phosphorylated proteins, oligonucleotide drugs and phosphates in food
  15. Part 15 — Multiplets in 1H NMR: coupling trees, second-order patterns and two signals from one CH2
  16. Part 16 — NMR for cosmetics: adulterated essential oils, free formaldehyde, aloe vera, rancid lipsticks and mineral oil in lip products (this article)
  17. Part 17 — Polymers by NMR: composition, sequence, chain ends and branches, and what a spectrum of many chains can and cannot measure
  18. Part 18 — Benchtop NMR: what a permanent-magnet spectrometer does well, and when a sample needs high field
  19. Part 19 — NOESY, ROESY and TOCSY: regioisomers, E and Z isomers, anomers, and the signals of one compound in a mixture
  20. Part 20 — Rapid-injection NMR: starting a reaction inside the spectrometer, how fast the spectra can follow it, and the intermediates it has revealed
  21. Part 21 — Fluorine NMR for PFAS: total organofluorine, what targeted LC-MS/MS does not detect, and the detection limit
  22. Part 22 — Polysiloxanes by proton NMR: silicone oil in medicines, food and cosmetics, the reactive groups before curing, and the rubber network after it

© 2026 NorrChemica™ · Content may not be reproduced without written permission

Back to blog

Leave a comment

Please note, comments need to be approved before they are published.