Levels of the nicotine analog 6-methyl nicotine as a naturally formed tobacco alkaloid in tobacco and tobacco products

Abstract

S-6-methyl nicotine (S-6MN) has appeared as a nicotine substitute in commercial electronic e-cigarette products and pouches, including with the claim that such use is not regulated under current U.S. law. This work describes an analytical chemistry based search for the natural S/R presence of 6MN and three other MN compounds in additive-free cured leaf tobaccos and in multiple commercial tobacco products. The samples were extracted using 5 N NaOH, then methyl t-butyl ether. The extracts were analyzed using gas chromatography (GC) with mass spectrometric (MS) detection, and liquid chromatography (LC) with high resolution MS/MS detection. GC peaks with the correct retention times and MS patterns were found and confirmed for 6MN. Further confirmation for the presence of 6MN was obtained by LC/MS/MS. The all-sample average level of 6MN was determined to be 0.32 µg per g of tobacco material; the levels were too low to determine the S/R distributions. For 2MN, strong but not fully confirmed (*) evidence was obtained; analytical results are presented for 2MN* at an all-sample average level of 0.10 µg per g of tobacco material. No evidence for either 4MN nor 5MN was found. Because most commercial nicotine is as extracted and purified from tobacco, 6MN can be expected in all such nicotine, and therefore in most nicotine-containing e-cigarettes (ENDS) as well as reagent-grade nicotine. Analyses of GC/MS data from past analyses of nine high-nicotine e-cigarette liquids purchased during the period 2018 to 2022 indicated a mean ± 1 s.d. result for 6MN of 6.3 ± 1.4 µg/mL.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-01392-6.

Introduction

General

Methyl nicotines are defined here as nicotine analog compounds in which an additional methyl group is present on either the pyridine ring or the pyrrolidine ring of the nicotine structure. When the methyl group is on the pyrrolidine ring, a prime is included on the ring-position number in the compound name.

In 2024, Jordt et al.¹ reported the use of 6-methyl nicotine (6MN, Fig. 1) in the SPREE BAR™ vaping system. In 2023, the owner of that trademark claimed:

Fig. 1.

S-forms of nicotine and the four methyl (pyridinyl) nicotines.

‘Although Metatine [6MN] produces the same sensation as nicotine and may also be addictive, Metatine is not made or derived from tobacco or nicotine, and Metatine does not consist of or contain nicotine from any source…. SPREE BAR products containing Metatine (and no tobacco-derived materials or nicotine) do not meet the definition of a “tobacco product” in the Federal Food, Drug and Cosmetic Act and do not require a Pre-Market Tobacco Application (PMTA) to be sold in the USA.’²

Other reports of 6MN in tobacco products have appeared^3,4.

Nicotine and 6MN as well as 2MN, 4MN, and 5MN have two enantiomeric forms each, S and R. The S configurations are depicted in Fig. 1; for nicotine, the S form is far more active at nicotinic acetyl choline receptors (nAChRs)⁵. The seven methyl S-nicotine analogs with the methyl group on a carbon of the pyrrolidine ring were recently discussed in some detail by Xing et al.⁶, including as regards their activities at nAChR) sites; they also discussed the 1’ quarternary amine methyl nicotine in which an additional methyl group resides on the pyrrolidine N.

We first review the discovery origin story of S-6MN as a nicotine analog compound active at nAChR sites. Second, we describe an analytical-chemistry-based search for evidence that S/R-6MN and other MN compounds may be present naturally in tobacco and tobacco products. Indeed, neither 6MN nor any of the other MN compounds in Fig. 1 are among the 9759 compounds listed by Rodgman and Perfetti⁷ as having been found in tobacco; their Table 17.9 does list (their entry 40) the related compound pyridine, 3-(1-methyl-2pyrrolin-2-yl)-6-methyl- (aka 6-methy-2,3-dehydronicotine) as having been reported in internal tobacco industry documents^8,9 as present in a “Turkish tobacco essential oil”. A confirmed natural presence of 6MN might affect considerations regarding governmental regulatory authority over MN compounds in tobacco products.

The discovery of 6MN: the 1960s work of Haglid and collaborators

With the structure of nicotine as the starting point, F. Haglid at Kungliga Tekniska Högskolan (KTH, “Royal Institute of Technology”) and collaborators conducted a broad synthesis-based search for “nicotine analogue” compounds that exhibit significant agonistic activity at nicotinic acetylcholine receptors (nAChRs). Seven publications describing that work are considered here^10–16, ordered chronologically based on the dates of publisher receipt or conference presentation. All but one¹⁴ acknowledge that the financial support for the work was from “Svenska Tobaks AB”, the then Swedish state tobacco enterprise and corporate antecedent of Swedish Match AB. Haglid¹⁶ elaborated that

“This entire work has been generously supported by grants from Svenska Tobaks AB, through its Medical Advisory Board…”

Regarding the motivation for their work, the first Haglid et al. paper states only that¹⁰

“In view of the large amount of work devoted to the physiological and insecticidal action of nicotine, it is surprising that so little systematic work has been done on studying synthetic analogues of nicotine…. With a view to studying relationships between structure and pharmacological action more systematically it was decided to synthesise [various] series of nicotine analogues in [each of] which, as far as possible, only one part of the nicotine molecule at a time is changed.”

The commercial applications for neurologically active nicotine analogs that might have been in mind by the researchers and/or the sponsor may have included uses as: (1) substitutes for nicotine in consumer tobacco products; (2) therapeutic pharmaceuticals; and/or (3) insecticides acting on nAChRs, as with the subsequently developed “neonicotinoids”¹⁷. Regarding the first possibility, while regulation of the tobacco industry was still minimal in the 1960s, in 1966 the U.S. Public Health Service nevertheless recommended a “progressive reduction of the ‘tar’ and nicotine content of cigarette smoke”¹⁸. And, in the same year, the U.S. Federal Trade Commission (FTC) began allowing¹⁹: (1) marketing that featured cigarette tar and nicotine delivery values measured by FTC machine-smoking protocols (rescinded in 2008); and thereby (2) the implied claiming of health benefits for low-tar/nicotine products. (The correlation between the resulting growth of U.S. advertising for low-tar cigarettes and the subsequent increase in low-tar cigarette sales in the U.S. has been analyzed by Reed, Anderson, and Burns²⁰.) Problematically, because deliveries of tar and nicotine by conventional cigarettes are highly correlated²¹, a very low-tar cigarette would, naturally, also be: (1) a very low-nicotine cigarette; so (2) presumably not easily satisfying of nicotine dependence, unless perhaps a potent nicotine analog had been added. That such an option would come to mind within the industry is demonstrated within the UCSF Truth Tobacco Industry Documents collection²² by multiple copies of a summary of an internal Philip Morris discussion from October 2000 of cigarette “harm reduction” which states²³:

“13. Can the ratio of tar to nicotine be altered in a useful way to affect harm reduction?

         …

17. Could nicotine analogs play a role in number 13?”

Regarding the approach chosen by Haglid et al.^10–16, as compared to extraction of tobacco and other alkaloid-containing plant materials, lab synthesis was undoubtedly seen as a more straightforward route to sufficient amounts of an organized variety of nicotine-related compounds at adequate purity for characterization of properties. Whether any of the compounds synthesized and tested might in fact be naturally occurring minor tobacco alkaloids was not discussed. Of the scores of compounds synthesized, Haglid¹⁴ stated in 1965 that only one, 6MN, had been found to exhibit significant nAChR activity. The circumstances of that determination, however, were not described until 1967 when Haglid¹⁵ stated

“The pharmacological investigation, kindly carried out at the Department of Physiology (Prof. U.S. von Euler), Karolinska Institutet, Stockholm, showed that 4-methylnicotine exerted a very low pharmacological activity in all the tests used¹ while that of 6-methylnicotine was fairly strong… The stimulating effect of 6-methylnicotine on the isolated guinea-pig’s ileum was 25% stronger than that of natural nicotine.¹”

We surmise that provision of their reference 1 (herein Erdtman et al.¹⁰) was to identify: (1) the procedural specifics of the activity “tests used”; and (2) the magnitude of the nAChR activity of “natural nicotine”, and not to suggest that the 6MN work had been done as part of the Erdtman et al.¹⁰ study.

U.S. tobacco company interest in nicotine analog compounds

U.S. tobacco companies quickly became aware of the work of Haglid et al. At the time of this writing, searching the UCSF Truth Tobacco Industry Documents collection²² for materials containing the keyword “Haglid” returned 951 results with 403 of these having assigned metadata dates in the 1960–1970s²². Within the Master Settlement Agreement portion of the collection, 733 of these are affiliated as follows: Philip Morris, 537; R.J. Reynolds, 142; Brown and Williamson, 36; and Lorillard, 18. Among the Philip Morris results are numerous copies of an internal scientific study seeking to understand why 6MN exhibits significant nAChR activity while 4MN does not. Each includes a figure showing measured values for nAChR activity with guinea pig ileum for 6MN and 4MN. The copies carry two different titles, e.g. Seeman et al.²⁴ vs. the possibly subsequent Seeman et al.²⁵. For additional relevant discussion of past tobacco industry work on nicotine analogs, the reader is referred to the prescient 2005 review by Vagg and Chapman²⁶.

Overall, even if the product development efforts for SPREE BAR™ as a nicotine analog product did not begin with tobacco-industry based knowledge of the work of Haglid et al., there has likely been wide acquaintance with the UCSF collection within the electronic nicotine delivery system (ENDS, “e-cigarettes”) industry. Anyone searching that collection for information on “nicotine analog” or “nicotine analogue” would have discovered the work of Haglid et al.: at the time of this writing, 291 of the “nicotine analog” document results, and 587 of the “nicotine analogue” document results cite “Haglid”.

Tobacco alkaloids

The term “tobacco alkaloids” has been defined as referring to the group of alkaloid compounds that may be formed in: (1) the living tobacco plant either by directed biosynthesis or accidental cellular reactions; and (2) non-living plant material by biotic (e.g., bacterial) or abiotic (e.g., chemical oxidation) processes acting on compounds formed in the plant (Leete²⁷). Overall, as a practical matter, “tobacco alkaloids” has been taken to include compounds formed during tobacco curing and product manufacturing.

Crooks²⁸ lists 15 tobacco alkaloid compounds plus two groups of derivatives. Without being specific here about chirality, these are: (1) nicotine; (2) nornicotine; (3) myosmine; (4) β-nicotyrine; (5) N’-isopropylnornicotine; (6) norcotinine; (7) cotinine; (8) nicotine-N-1’-oxide; (9) anabasine; (10) N-methylanabasine; 11) anabaseine; 12) anatabine; 13) N-methylanatabine; 14) nicotelline; 15) 2,3’-bipyridyl; 16) the amide derivatives of nornicotine with acetic acid through dodecanoic acid; and 17) the amide derivatives of anabasine and anatabine with formic acid. Benowitz et al.²⁹ adds: 18) β-nornicotyrine; 19) metanicotine; 20) pseudooxynicotine; and 21) the ionic quaternary species N-methylmyosmine; Jacobs et al.³⁰ adds: 22) anatalline. The chiral differentiation of 40 tobacco-related alkaloids (including 6MN) is discussed by Hellinghausen et al.³¹. Rodgman and Perfetti⁷ enumerate additional other minor tobacco alkaloids. The structures of myosmine and of the S-forms of nicotine, nornicotine, and cotinine are given in Fig. 2.

Fig. 2.

S-forms of nicotine, nornicotine, and cotinine with the single form of myosmine (a non-chiral compound).

In tobacco, nicotine is believed to be formed from nicotinic acid and N-methylpyrrolinium as mediated by a nicotine synthase enzyme³². Involvement of substrate analogs in the pathway could also lead to nicotine analogs as with 2- and 6-methylnicotinic acid in place of nicotinic acid leading to 2MN and 6MN, respectively. Methylation of nicotine might also occur with mediation by methyltransferases³³, including as with radical involved methyltransferases³⁴. Overall, while it cannot currently be said that any methyl nicotines are clearly expected to be present in either tobacco or tobacco products, pathways for their formation can be suggested. The goal of the laboratory portion of this work was to investigate the possible presence of 2MN, 4MN, 5MN, and 6MN in tobacco and tobacco products.

Materials and methods

Tobacco samples

Eight groups of tobacco materials were analyzed (Table 1) to determine their MN levels. These included additive-free leaf tobacco obtained in 2023, smokeless tobaccos, domestic cigarettes, and foreign cigarettes. Tobacco products archived in our laboratory from as long ago as 2004 were included, even though alkaloid compounds are reactive and subject to oxidation and other conversion reactions: for this work, inclusion of historical products was advantageous even if the results might only be interpretable as qualitative.

Table 1.

Tobacco and tobacco product types analyzed in this work.

Type	Details
Leaf, additive free (4 samples, purchased 2023)	Kentucky burley, dark air cured, Virginia bright flue cured, Virginia red cured (from Whole Leaf Tobacco LLC, Akron, OH)
Smokeless, “CRP” reference products (4 samples, purchased 2009)	CRP1 (snus pouch), CRP2 (moist snuff), CRP3 (dry snuff powder), and CRP4 (looseleaf chewing) / (from CORESTA https://www.coresta.org/coresta-smokeless-tobacco-reference-products
Smokeless (U.S.) (7 samples, purchased 2007)	Skoal cherry tobacco long cut, Kayak peach long cut moist snuff, Camel Stripes fresh dissolvable tobacco, Skoal X-tra Crisp Blend long cut, Camel Snus Original, Camel Spice snus, Taboka
Cigarettes (U.S.) (5 samples, purchased 2004)	Marlboro 100’s, Basic 100’s, Gold Pack, American Spirit Turquoise, Camel 99’s, Basic Flip Top Box
Cigarettes (U.S.) (2 samples, purchased 2022)	American Spirit Turquoise, Marlboro 100’s
Cigarettes (U.S.) (2 samples, purchased 2024)	American Spirit Turquoise, Marlboro Red
Cigarettes (various, foreign) (10 samples, purchased between 2017 and 2022)	SinSinergi Mint (Inda), Sampoerna Splash Tropical (Ind), Benson & Hedges Crystal Violet (Mexa), Pall Mall (Mex), Dunhill Switch (Paka), Milano X-merge (Pak), MEVIUS Wind Blue (Phila), Winston Less (Phil), Raison Ice Café (Vieta), Camel Caster (Viet).
Cigarettes (Nigeria) (13 samples, purchased 2024)	Marlboro Red, Marlboro Gold, Marlboro Silver, Omni, Omni Lights, B&H Switch, B&H Special Filter, Camel Menthol, Camel Menthol Silver, Camel Turkish Royal, Eclipse, Advance Premium Lights, Merit Ultralight

^aInd = Indonesia; Mex = Mexico; Pak = Pakistan; Phil = Philippines; Viet = Vietnam.

Chemicals and standard reference materials

Sigma-Aldrich (St Louis, MO) provided the HPLC-grade solvents, the authentic standard reference materials for nicotine, nornicotine, cotinine, myosmine, and for the surrogate standard (SS) chemical (quinoline) and the internal standard (IS) chemical (1,2,3-trichlorobenzene). For the study of the MN compounds, racemic authentic standard reference materials for 2MN, 5MN, and 6MN were from Santa Cruz Biotechnology (Dallas, TX). Examination of those MN standards by gas chromatography/mass spectrometry (GC/MS) (see details below) indicated > 98% purity, and provided the GC retention times on three columns, and EI low-resolution MS spectra needed during the identification/quantitation work with the samples.

To obtain the same information for 4MN, since a vendor for that MN could not be found, it was necessary to run a synthesis reaction to at least obtain a product mixture that contained 4MN at sufficiently high concentration to allow clear characterization of its GC retention and MS spectral properties. The homolytic radical conditions used were created with S-nicotine + t-butyl hydrogen peroxide + FeSO₄ as described by Itokawa et al.³⁵ 250 mg (1.54 mmol) of nicotine (free-base) was dissolved in 15 mL of deionized water followed by 0.5 mL of 95–98% H₂SO₄ and 0.86 g (5.1 mmol) of FeSO₄·H₂O_(s). The mixture was stirred while adding 560 mg (3.1 mmol) of t-butyl hydrogen peroxide over 20 min. After 20 min of additional stirring, the reaction solution was transferred to a separatory funnel. The solution was made alkaline by addition of 5 mL of 3.7 N NaOH, then extracted with 20 mL of chloroform. The extract was analyzed without further purification by temperature programmed GC/MS with a 30 m DB-5MS (bonded, 95% methyl, 5% phenyl polysiloxane) capillary column, EI ionization, and other column and instrumental details as specified below. The observed GC retention times and EI MS patterns (as compared to the standards) allowed confirmed assignments of peaks for 2MN, 5MN, and 6MN. Many additional GC peaks were also present, for which the MS patterns indicated nicotine derivatives with from one to at least three added methyl groups, including with methylation having occurred on both rings. One of the major GC peaks exhibited an EI mass spectrum very similar to that shared by 2MN, 5MN, and 6MN (see Table 2; for detailed mass spectra at ≥0.5% of the base peak see Table S.1 in the Supplementary Materials), each having a clear molecular ion (M⁺) at m/z = 176 amu along with the N-methyl pyrrolidine fragment ¹²C₅¹H₁₀¹⁴N⁺ as the base peak at m/z = 84 amu, as is shared also with nicotine (but not with the C-bonded methyl pyrrolidine nicotine analogs for which a major peak is for a C-bonded methyl pyrrolidine fragment ¹²C₆¹H₁₂¹⁴N⁺ at m/z = nominal mass 98 amu). We assigned the additional apparently monomethyl nicotine peak to be 4MN based on: (1) this MS evidence; (2) consistent capillary GC relative retention time data found in the 1983 work of Seeman et al.³⁶ for 2MN, 4MN, and 6MN under isothermal 175 ^oC conditions on the GC stationary phase SE-54 (non-bonded, 94% methyl, 5% phenyl, 1% vinyl polysiloxane) as given in Table 3 along with those measured here using the above reaction mixture and isothermal 175 ^oC conditions with the very similar DB-5MS phase; and (3) the fact that we verified that the EI mass spectra of the methyl pyrrolidine nicotine analogs as obtained from KemPharma (Gainesville, FL) are distinctly different from those of the Fig. 1 methyl pyridine nicotine analogs each of which has a major peak at m/z = nominal mass 98 amu, as explained above.

Table 2.

Electron impact (EI) ionization (70 eV) mass spectra of nicotine and four methyl nicotine compounds.

Nominal m/z (amu)	Nicotine	2-methyl nicotine (2MN)	4-methyl nicotine (4MN)	5-methyl nicotine (5MN)	6-methyl nicotine (6MN)
42	14.2	11.2	9.4	11.8	13.9
77					5.3
82	5.2
84	100	100	100	100	100
85	6.1	5.9	6.0	5.8	5.5
92	6.1
106					6.2
118	5.5	5.4
119	8.3
132				4.6	5.9
133	32.7	19.9	16.9	9.3	12.7
147		14.0	25.0	26.5	43.2
148					6.0
161	18.7
162	19.1
175		8.9	12.5	17	28.5
176		14.5	14.5	19.4	28.3

Intensity values included are those for m/z values with intensities of ≥5% of the base peak at m/z = 84 amu. MS source temperature at 225 °C. (Table S.1 in the supplementary materials gives intensity values for all m/z values with intensities of ≥0.5% of the base peak.)

Table 3.

Isothermal 175 oC gas chromatographic (GC) retention time (RT) values and relative retention time (RRT) values on the stationary phase SE-54 (94% methyl, 5% phenyl, 1% vinyl polysiloxane) as reported by Seeman et al.³⁶ and on DB-5 (95% methyl, 5% phenyl polysiloxane) using the S-nicotine methylation reaction mixture.

Compound	Seeman et al.36 175 °C isothermal, 30 m, 0.25 mm i.d., SE-54 (film thickness and column flow rate not given)		This work 175 °C isothermal, 30 m, 0.25 mm i.d., DB-5MS, 0.25 μm film thickness, 1 mL/min column flow rate
	RT (min)	RRT	RT (min)	RRT
Nicotine	NAa	1.00	2.718	1.000
2MN	NA	1.12	2.980	1.096
6MN	NA	1.14	3.066	1.128
5MN	NA	NA	3.409	1.254
4MN	NA	1.31	3.538	1.302

^aNA = not available.

For characterization of the GC Kovats Indices of the analyte compounds on the columns used, an n-alkanes standard mixture (C₁₀ to C₂₅ at 1 mg/mL per component in dichloromethane) was obtained from Restek, Inc. (Bellefonte, PA); the C₂₆ to C₂₉ n-alkanes were obtained at > 98% purity from Sigma-Aldrich, Inc.

Tobacco and tobacco product sample processing

The main experimental question at hand pertained to whether the subject MNs were present in the tobacco samples, though not the exact levels. The work therefore emphasized an abundance of samples with single “replicates” vs. fewer samples with multiple replicates. The 2024 domestic cigarettes and smokeless tobacco products were analyzed in duplicate.

The extraction method was adapted from the CORESTA³⁷ approach for determination of nicotine in tobacco materials. For each extraction, 1 to 2 g of tobacco material was placed in a clean 40 mL “VOA” vial fitted with a Teflon-lined septum cap. In the case of cigarette samples, the filterless rods from two cigarettes (with the paper wraps) were cut into pieces into the vial using a solvent-cleaned scissors. In the case of oral nicotine pouch samples, two pouches were cut open inside the vial. For each sample vial, 100 µL of the quinoline SS spike solution at 8 mg/mL in methyl t-butyl ether (MTBE) was added onto the tobacco material. For the tobacco leaf samples, this was followed by 12.5 mL of 5 N NaOH (to place the alkaloids in their neutral, free-base forms); for the other samples, 10.0 mL of 5 N NaOH was used. Each sample vial was then capped, swirled to wet the sample material, and allowed to stand for 25 min. 10 mL of MTBE was then added to each sample vial. All sample vials were then placed on a shaker and agitated for 2 h (to draw the free-base alkaloids into the mostly MTBE phase), then held overnight at 4 °C. After being returned the next day to room temperature, the MTBE phase of each sample was analyzed in duplicate. For each replicate: (1) for determination of nicotine, 0.1 mL of the MTBE phase was placed in a ~ 2 mL brown glass autosampler vial with a septum cap then diluted using 0.9 mL MTBE; (2) for determination of minor alkaloids, 1.0 mL of the MTBE phase was placed in ~ 2 mL brown glass autosampler vial with a septum cap. Every autosampler vial then received 20 µL of a 1,2,3-trichlorobenzene IS spike solution at 2 mg/mL in isopropyl alcohol.

Gas chromatography/mass spectrometry (GC/MS)

The analyses of the sample extracts proceeded by capillary column GC/MS using four different GC columns and a separate temperature program for each (Table 4). Kovats Index measurements proceeded using at slower temperature program rates. All GC/MS runs proceeded using an Agilent (Santa Clara, CA) 7693 autosampler, Agilent 7890 A GC, and Agilent 5975 C MS. The injection volume was 1.0 µL, the injector temperature was 235 °C, the “split” ratio was 10:1. The MS was operated in the electron impact (EI) ionization mode at 70 eV, using a source temperature of 225 °C, and scanning from 34 to 400 atomic mass units. All quantitations corrected for variations in the instrument response using IS-normalized relative response factors defined as [analyte response]/[analyte mass injected]/([IS response]/[IS mass injected]).

Table 4.

Temperature programs used with four gas chromatography (GC) columns used in this work for sample analysis and for Kovats index measurements. All runs carried out with a constant 1 mL/min column flow rate.

GC column	Sample analysis	Kovats index measurements
1. Medium polarity - Restek Rxi-624Sil MS, 30 m long, 0.25 mm i.d., 1.4 μm film thickness.	40 °C for 2 min; 10 °C/min to 100 °C; 12 °C/min to 280 °C; 280 °C for 8 min. (Prep. for next run: controlled oven cooling at 20 °C/min to 220 °C, then maximal cooling to 40 °C.)	80 °C for 1 min; 10 °C/min to 280 °C; 280 °C for 8 min, then 20 °C/min to 220 °C and hold at 220 °C for 2 min.
2. Polar - Agilent DB-HeavyWax, 30 m long, 0.25 mm i.d., 0.25 μm film thickness.	80 °C for 2 min; 10 °C/min to 200 °C; 20 °C/min to 280 °C; 280 °C for 4 min. (Prep. for next run: controlled oven cooling at 20 °C/min to 220 °C, then maximal cooling to 80 °C.)	50 °C for 1 min; 10 °C/min to 280 °C; 280 °C for 5 min, then 20 °C/min to 220 °C and hold at 220 °C for 1 min.
3. Non-polar - Agilent DB-5MS, 30 m long, 0.25 mm i.d., 0.25 μm film thickness.	80 °C for 1 min; 10 °C/min to 310 °C; 310 °C for 7 min. (Prep. for next run: maximal cooling to 80 °C.)	50 °C for 1 min; 10 °C/min to 300 °C; 300 °C for 5 min; then 20 °C/min to 200 °C and hold at 200 °C for 1 min.
4. Chiral: Agilent J&W CP-Cyclodextrin-β-2,3,6-M-19, 50 m long, 0.25 mm i.d., 25 μm film thickness.	120 °C for 90 min; 20 °C/min to 220 °C; 220 °C for 2 min.  (Prep. for next run: maximal cooling to 120 °C.)	NA

Liquid chromatography/mass spectrometry/mass spectrometry (LC/MS/MS)

LC/MS/MS Instrument: Vanquish UHPLC system with Thermo Scientific (Waltham, MA) Q Exactive (for MS1/MS2 measurements).

LC Conditions: (1) guard column: Agilent InfinityLab Proshell 120 EC-C18, 3.0 mm i.d. × 5 mm long, 2.7 μm particle size; (2) analytical column: Agilent InfinityLab Proshell 120 EC-C18 3.0 mm i.d. × 150 mm long, 2.7 μm particle size; (3) mobile phase flow rate: 0.50 mL/min; (4) mobile phase component A: 10% acetonitrile, 90% 10 mM ammonium formate adjusted to pH = 8.0 with NH₄OH; (5) mobile phase component B: 100% acetonitrile; (6) mobile phase gradient elution: 0 to 10.0 min, 0 to 30% B; 10.0 to 12.0 min, 30 to 90% B; and 12.0 to 14.0 min, hold at 90% B. Conditioning the column for the next run proceeded according to: 14.0 to 14.1 min, 90 to 0% B; and 14.1 to 18.0 min, hold at 0% B.

MS1 conditions: scanning from 50 to 200 amu at 140,000 resolution.

MS2 Conditions: scanning from 50 to 200 amu at 70,000 resolution with: (1) monitoring for 163.1230 amu (nicotine-H⁺) and higher-energy collisional dissociation (HCD) and normalized collision energy (NCE) at 33; and (2) monitoring for 177.1386 amu (methyl nicotine-H⁺) with HCD and NCE at 33.

Results

GC retention times, GC Kovats index values, EI mass spectra from GC/MS measurements, LC retention times

Absolute GC retention times on the four GC columns are given in Table 5. The measured Kovats Index values on GC columns 1, 2, and 3 are given in Table 6. 2MN, 6MN, 5MN, and 4MN were well separated and eluted in that order on all of the columns. On columns 1, 2, and 3, nicotine and 2MN were baseline resolved in the standard runs. For tobacco-related samples however, with a large nicotine peak, 2MN eluted on the tail of nicotine, which will always create a problem for low-resolution-MS-based confirmation of the presence of a low level of 2MN because with m/z = 84 amu as the MS base peak for both nicotine and 2MN, obtaining a mass spectrum for a small candidate 2MN peak using subtraction of the background signal from a large nicotine tail will not result in a properly-sized m/z = 84 amu signal for proper MS pattern concordance. Absolute LC retention times for the instrumental conditions used are given in Table 7.

Table 5.

Gas chromatographic (GC) retention time (RT, minutes) values and relative retention time (RRT) values under temperature programming conditions as given in Table 4.

	GC column 1 (medium polarity) Rxi-624Sil MS		GC column 2 (polar) DB-HeavyWax		GC column 3 (non-polar) DB-5MS		GC column 4 (chiral) CP-Cyclodextrin-β-2,3,6-M-19
	RT (min)	RRT	RT (min)	RRT	RT (min)	RRT	RT (min)	RRT
Compound
Nicotine	18.02	1.000	11.62	1.000	8.69	1.000	S 35.51	S 1.000
Nornicotine	19.08	1.059	14.62	1.258	9.63	1.108	S 78.95a, R 79.22a	S 2.223a, R 2.231a
Myosmine	19.15	1.063	14.88	1.281	9.71	1.117	68.82	1.938
Cotinine	22.63	1.256	18.62	1.602	13.00	1.496	NAb	NA
Methyl nicotines
2MN	18.51	1.027	11.72	1.009	9.34	1.075	S 37.41a, R 38.19a	S 1.054a, R 1.075a
6MN	18.61	1.033	11.97	1.030	9.53	1.097	S 45.52a, R 45.67a	S 1.282a, R 1.286a
5MN	19.27	1.069	12.90	1.110	10.11	1.163	S 56.17a, R 57.68a	S 1.582a, R 1.624a
4MN	19.49	1.082	13.19	1.135	10.31	1.186	S 61.61c, R NA	S 1.735c

^aAssumes the same S first/R second order as observed with nicotine

^bNA = not available.

^cAssumes that the methylation of S-nicotine reaction used here to obtain 4MN yielded S-4MN.

Table 6.

Kovats index values on three types of gas chromatography (GC) columns under temperature programming conditions as given in Table 4.

Compound	Medium polarity		Polar		Non-polar
	This work Rxi-624Sil MS	NIST 17 MS Database	This work DB-HeavyWax	NIST 17 MS Database	This work DB-5MS	NIST 17 MS Database
Nicotine	1428.3	NAa	1862.1	1863 (DB-Wax)	1354.1	1360.5 (VF-5MS)
Nornicotine	1524.4	NA	2161.8	NA	1426.3	1435.4 (VF-5MS)
Myosmine	1530.7	NA	2191.7	NA	1433.0	1427.3 (HP-5)
Cotinine	1893.0	NA	2844.9	NA	1709.1	1706.3 (Rxi-5Sil MS)
2MN	1471.6	NA	1870.4	NA	1404.9	NA
6MN	1481.2	NA	1894.0	NA	1418.8	NA
5MN	1542.2	NA	1983.1	NA	1464.7	NA
4MN	1567.9	NA	2020.9	NA	1478.3	NA

^aNA = not available.

Table 7.

Liquid chromatographic (LC) retention time (RT, min) values and relative retention time (RRT) values.

	RT (min)	RRT
Nicotine	7.67	1.000
6MN	8.80	1.147
2MN	9.30	1.213
5MN	9.56	1.246
4MN	10.50	1.369

Column: agilent InfinityLab proshell 120 EC-C18 3.0 mm i.d. × 150 mm long, 2.7 μm particle size. Gradient elution conditions as described in the text.

Tobacco alkaloid levels in the samples

Nicotine, nornicotine, cotinine, and myosmine all gave significant GC/MS peaks for every sample: the GC retention times and MS patterns on GC column 1 were unequivocal, so the identifications and quantitations of these tobacco alkaloids were unambiguous and reliable. Table 8 gives the averages by sample type for the measured levels of nicotine as (g nicotine)/(g tobacco material). For the other compounds, to place their levels in context as minor tobacco alkaloids, the values are given as log[(g alkaloid)/(g nicotine)]; candidate peaks were observed for almost all of the samples for 6MN and approximately half of the samples for 2MN. Candidate peaks were not observed for 4MN or 5MN in any of the samples.

Table 8.

Results of quantitative determinations of nicotine and other tobacco alkaloids in tobacco samples.

	Leaf (2023)	Smokeless CRP (2009)	Smokeless U.S. (archived)	Cigarettes U.S. (archived)	Cigarettes U.S. (2022)	Cigarettes U.S. (2024)	Cigarettes foreign (archived)	Cigarettes Nigerian (2024)	All-types- average	All-types- average
	(g nicotine)/(g tobacco material) ± 1 s.d.									(g alkaloid)/(g tobacco material)
Nicotine	0.0230 ±0.0087	0.0118 ±0.0042	0.0076 ±0.0043	0.0096 ±0.0010	0.0104 ±0.0040	0.0095 ±0.0024	0.0129 ±0.0039	0.0102 ±0.0028	0.0118	0.0118
	log [(g alkaloid)/(g nicotine)] ± 1 s.d.									(µg alkaloid)/(g tobacco material)
Nornicotine	-1.38 ±0.18	-1.70 ±0.12	-1.76 ±0.26	-1.43 ±0.12	-1.57 ±0.15	-1.52 ±0.19	-1.47 ±0.23	-1.35 ±0.07	-1.52	360
Cotinine	-2.59 ±0.78	-2.36 ±0.29	-2.29 ±0.33	-2.62 ±0.25	-2.73 ±0.44	-2.76 ±0.37	-2.71 ±0.24	-2.59 ±0.16	-2.58	31
Myosmine	-3.28 ±0.39	-2.64 ±0.68	-2.76 ±0.34	-3.15 ±0.15	-3.26 ±0.09	-3.23 ±0.34	-3.24 ±0.16	-3.09 ±0.19	-3.08	10
6MN	-5.03 ±0.32	-4.21 ±0.17	-4.33 ±0.19	-4.61 ±0.15	-4.57 ±0.03	-4.53 ±0.13	-4.44 ±0.15	-4.84 ±0.48	-4.57	0.32
2MN*	NDa	-5.32 ±0.27	-5.01 ±0.28	-5.07 ±0.17	-5.11 ±0.03	-4.97 ±0.14	-4.99 ±0.19	-5.07 ±0.13	-5.08	0.10
4MN	ND	ND	ND	ND	ND	ND	ND	ND	ND	ND
5MN	ND	ND	ND	ND	ND	ND	ND	ND	ND	ND

^aND = not detected

For 6MN, on both GC columns 1 and 2, small but quantifiable candidate peaks for the m/z = 176 amu confirmation ion (with m/z = 84 amu as the quantitation ion) were observed in nearly every chromatogram for the Table 1 samples within ±1.5 s of the expected retention time, and with the proper relative intensities for m/z = 84, 147, 175, and 176 amu in background-subtracted mass spectra. Additional confirmatory evidence for 6MN in Table 1 samples was obtained by LC/MS/MS with high resolution MS1/MS2 using high resolution MS1 scanning for M+H⁺ at the 6MN LC retention time (see Table 7 below) at 177.1386 amu (= ¹²C₁₁¹H₁₇¹⁴N₂⁺ = ¹²C₁₁¹H₁₇¹⁴N₂ minus m_electron) then examination of the MS2 spectrum against the authentic standard. Therefore, all of the 6MN candidate peaks for the quantitation ion on GC column 1 were quantitated as actual 6MN (Table 8). While the levels were too low to allow differential S/R quantitation by chiral GC, especially in the presence of the relatively massive levels of S-nicotine, it is considered likely that the S enantiomer was at least dominant because of original formation by reaction of 6-methyl nicotinic acid in the nicotine formation pathway, or subsequent methylation of S-nicotine.

For 2MN, candidate peaks for the m/z = 176 amu confirmation ion were present in many of the chromatograms for the Table 1 samples using GC column 1 within ±2.2s of the expected retention time (though for each the confirmation ion peak size was ~ 3× smaller than for the corresponding candidate peak for 6MN). Further, for the 2MN candidate peaks on GC column 1, while the relative intensities at m/z = 147, 175, and 176 amu were approximately correct in the background-subtracted mass spectra, for reasons discussed above with 2MN on a nicotine tail, the overall MS patterns including m/z = 84 amu were not as clear as for the 6MN candidate peaks. Because of the particular difficulty created by significant nicotine for 2MN on GC column 2 as a confirmatory column, for ~ 1/3 of the samples, confirmations of the 2MN candidate peaks were sought on a third column (GC column 3). In each of those column 3 chromatograms, a peak for the confirmation ion was present within ±1.4 s of the expected retention time for 2MN and with the proper relative intensities for m/z = 147, 175, and 176 amu. And, on that column, a 2MN candidate signal (if present) was always observable in the total ion chromatogram, but the background-subtracted MS patterns were not as clear as for the 6MN candidate peaks. For samples run on both GC columns 1 and 3, for each sample in which a quantifiable candidate peak for 2MN was present in the GC column 1 run, a quantifiable candidate confirmation ion peak for 2MN was also present in the GC column 3 run. (In addition, all of GC/MS chromatograms for the analyses related to Table 9 (discussed below) gave strong evidence for the presence of 2MN in those samples, qualified only by the MS pattern being affected by the overlap with nicotine for m/z = 84 amu.) As for 6MN, additional confirmatory evidence for 2MN in the Table 1 samples was sought by LC/MS/MS, but as in the GC portion of the search for 2MN in the Table 1 samples, the low candidate compound levels and high nicotine levels prevented clear identifications. Overall, the quantitation values in Table 8 are considered tentative and accordingly qualified with an asterisk as values for 2MN*.

Table 9.

Results of quantitative determinations of nicotine and 6MN as based on archived GC/MS results of this laboratory plus a new GC/MS analysis of reagent grade nicotine, all using the GC column 1 conditions.

Brand	Product	Year of GC/MS analysis		Nicotine (µg/mL)	6MN (µg/mL)	6MN as mass % of nicotine	log [(g 6MN)/(g nicotine)]
e-liquids
JUUL™	“Mango”	2018		63,100	8.2	0.013	-3.89
JUUL™	“Fruit Medley”	2018		61,900	6.9	0.011	-3.95
JUUL™	“Crème Brulee”	2018		57,300	7.2	0.013	-3.90
JUUL™	“Cool Menthol”	2018		60,200	7.6	0.013	-3.90
JUUL™	“Virginia Tobacco”	2018		58,900	6.8	0.012	-3.94
JUUL™	“Menthol”	2021		53,600	5.9	0.011	-3.96
Unknown	“Fruit Fusion”	2022		40,900	5.3	0.013	-3.89
Unknown	“Strawberry”	2022		51,700	4.4	0.009	-4.07
Unknown	“Mango”	2022		50,700	4.3	0.008	-4.07
			Average ± 1 s.d		6.3 ± 1.4	0.011 ± 0.0018	-3.95 ± 0.007
Reagent chemical
Alfa Aesar	nicotine (≥99%)	2025		1.01 E + 06	121	0.012	-3.94

The log[(g 6MN)/(g nicotine)] and log[(g 2MN*)/g nicotine)] values were computed for each sample. For example, for a 6MN level of 10⁻⁶ g/(g tobacco material) and a nicotine level of 10⁻² g/(g tobacco material), then log[(6MN) /(g nicotine)] = -4.0. The results for the different tobacco material types for 6MN and 2MN* are given in Table 8 as material type averages for the log[(g alkaloid)/(g nicotine))] values. For those few samples where a candidate peak for 6MN was not observed, the log [(g 6MN)/(g nicotine)] value was not computed, and was excluded without significant effect when computing the average. The same approach was used for the 2MN* results even though candidate peaks for 2MN were absent in a larger fraction of the samples because: (1) a markedly lower abundance for 2MN would be expected to cause quantitation problems for non-zero levels; and (2) the goal here is to obtain a qualitative understanding of absence/presence/levels. Table 8 gives the all-types-average log[(g alkaloid)/(g nicotine)] values ± 1 s.d.; the values by type are plotted in Fig. 3.

Fig. 3.

log[(g alkaloid)]/(g nicotine)] values ± 1s by tobacco type. 6MN = 6-methyl nicotine; 2MN* = 2-methyl nicotine (tentative).

Table 8 provides estimates for all the minor alkaloids of the all-types average levels in units of µg/(g tobacco material). These were computed as: (all-types-average g alkaloid)/(g nicotine) × (all-types-average g nicotine)/(g tobacco material) × (10⁶ µg/g). For 6MN, the calculation gave 0.32 µg/g (= 10^−4.57 × 0.0118 × (10⁶ µg/g)). For 2MN*, the calculation gave 0.10 µg/g (= 10^−5.08 × 0.0118 × (10⁶ µg/g)).

Discussion

The commercial appearance of e-cigarettes (ENDS) in 2003 marked the end of the era of tobacco control and the beginning of an era of tobacco and nicotine control (Fig. 4). Then, the appearance in 2023 of 6MN in commercial vaping systems began the current era, that of tobacco, nicotine, and nicotine analog control. While 6MN may be the first analog requiring regulatory consideration, the near term use of other nicotinoid compounds seems certain. Candidate compounds will obviously include those with nAChR activity, with some surely from beyond the MN class, but may also include compounds without nAChR activity. For the latter case, the appearance of vaping products and even cigarettes with a non-nAChR active compound such as 4MN seems possible, offering as they might the chemosensory rewards of inhaled free-base nicotine with lesser or no addiction potential. Bringing nicotine analog products under U.S. FDA control will in any case require amendments to FSPTCA-2009³⁸, by analogy with the passage of increasingly flexible regulation of potent synthetic opiods³⁹. As suggested above, the natural presence of 6MN in tobacco and tobacco products as demonstrated here may facilitate such an extension. In that context, we make the following additional observations:

Fig. 4.

Public health eras related to tobacco.

• Most nicotine currently used in e-cigarettes (ENDS) is derived from tobacco⁴⁰ and similarly so for commercial forms of nicotine in general;

• Complete purification of nicotine (or anything else) is never possible, with separation of nicotine from 6MN inherently difficult because of their very similar chemical characteristics (e.g., liquid-phase volatilities, viz. vapor pressures during distillation).

There have therefore been, and presently are, low levels of 6MN in virtually all non-synthetic nicotine products, whether e-cigarette (ENDS) liquids nicotine pouches, nicotine gums, nicotine-based pesticides, or analytical grade nicotine. (See Table 9 for results obtained by our examination of archival GC/MS data for e-liquids purchased in the period 2018 to 2022 as well as chemical analysis here of the reagent grade nicotine standard material used in this work.)

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

J.F.P., W.L., and K.J.McW. chose the samples and designed the analytical measurement methods. W.L. and K.J.McW. carried out the analytical work and data processing. M.S. and R.M.S. carried out the methylation of nicotine reaction. J.F.P. and W.L. carried out the statistical evaluations of the data. J.F.P. wrote the paper.

Data availability

The datasets containing final analytical results generated during the current study are available from the corresponding author on reasonable request.

Declarations

Competing interests

The authors declare no competing interests.