Qualitative and Quantitative Analyses of the Enantiomers of Nicotine and Related Alkaloids Employing Chiral Supercritical Fluid Chromatography in Commercial Nicotine Samples and in E-Cigarette Products

Diethylamine (DEA) was obtained from Sigma-Aldrich (St. Louis, MO, USA). Methanol (MeOH) and isopropanol (IPA) were HPLC grade and were obtained from Fisher Scientific (Pittsburgh, PA, USA). Packed chiral columns (Chiralpak IG-3) with amylose derivative (250 × 4.6mm, dp = 5μm) were obtained from Chiral Technologies (West Chester, PA, USA). S-nicotine, R/S-nicotine, and R/S-nornicotine were obtained from Millipore Sigma (Darmstadt, Germany), while R-nicotine and R-nornicotine were purchased from Toronto Research Chemicals (North York, ON, Canada). (Table 1). Nicotine (single production lots) were donated from several companies that sell nicotine to the public (see Table 2 for sample descriptions and supplier/manufacturers).

E-cigarette samples were purchased on-line via the internet during September–October 2022. E-cigarette samples were received “as is” from suppliers/dealers (Table 3). The e-cigarettes were deconstructed carefully, and nicotine was extracted using the method described by Cheetham et al. (5). After extraction, the e-liquid was concentrated for chiral supercritical fluid/diode array ultraviolet (DAD-UV) analysis (8).

Since knowledge about the relative amount of nornicotine, a minor alkaloid common to Nicotiana tabacum L. as well as a possible reagent intermediate in synthetic nicotine preparation schemes, in the samples under investigation as well as its enantiomer distribution can play an important role in describing the nature of the nicotine-rich samples, clearly establishing the detection limit for nornicotine was of paramount importance. Hence, having established an optimized protocol for chiral supercritical fluid chromatographic nornicotine enantiomer separation, a series of sequential dilution standards (R-nornicotine) were prepared ranging in concentration from a high of 1 μg/μL to a low of 100 pg/μL.

Employing an arbitrary benchmark of a nornicotine enantiomer signal to noise ratio of at least 2/1, the UV/VIS detector was capable of easily detecting between 100 and 200 pg/μL. Hence, this relatively low nornicotine enantiomer detection limit allowed for the more thorough evaluation of the nature of the nicotine rich samples under investigation (see Figure 2).

Figure 2.

The analysis of nicotine in an e-liquid formulation requires that the nicotine is first isolated from all other components present, e.g., propylene glycol (PG), glycerol (G), flavorings (F), etc. To be confident in the result obtained, therefore, an extraction method was developed based on the work of Cheetham et al. (5) so that only the nicotine concentration was elevated relative to the other components of the e-liquid formulation.

The most obvious strategy to isolate the nicotine in an e-liquid is to take advantage of its acid-base properties and use liquid-liquid extraction techniques.

In our studies, it was found that the nicotine in unflavored e-liquid formulations could be simply isolated by dissolving the e-liquid in basic water (pH > 10) and extracting with hexanes or dichloromethane, with no apparent significant extraction of either PG or VG (vegetable glycerin) into the organic phase.

The addition of flavorings, however, introduced the need for extraction under acidic conditions in order to remove these ingredients. The flavored extracts were subsequently dissolved in 1 M hydrochloric acid and washed with dichloromethane, followed by pH adjustment and extraction into dichloromethane.

Using this methodology, it was quite readily feasible to extract/enrich nicotine from e-liquid formulations and determine its enantiomeric distribution and aid in determining if the nicotine were tobacco-derived (TDN) or synthetic (SyN). Unfortunately, at times there was a significant loss of sample. The average yield of nicotine from the e-cigarette liquids was approximately 40% and was much lower for the dilute e-cigarette formulations (3 mg/mL). These observations consequently required a relatively large sample volume to obtain adequate nicotine for enantiomeric distribution analysis.

Adaptation of this technique to a more high-throughput environment, however, will require further optimization to both streamline the extraction workflow and to improve the nicotine recovery yield.

Chiral chromatography involves the resolution of enantiomeric mixtures through their differing interactions with a chiral stationary phase, thus allowing quantitation of the relative (or absolute) amounts of each enantiomer present. As mentioned in the introduction of Cheetham et al. (2), “the use of chiral chromatography to distinguish between TDN and SyN is not expected to be conclusive, since both types can be produced containing ≥ 98% S-nicotine.” The authors of this paper do not wholly believe this comment by Cheetham et al. (5). This is due in part to the notion that several minor alkaloids often co-distill with nicotine. The presences of these trace amounts of minor alkaloids are not present in synthetic nicotine. Therefore, it is possible that TDN and SyN at > 98% S-nicotine could be conclusively differentiated.

The chiral chromatography in the report by Cheetham et al. (5) employed a modified version of a method by Hellinghausen et al. (14). Employing the AZYP Nicoshell SPP chiral column, Cheetham et al. (5) showed excellent baseline resolution between the S- and R-nicotine enantiomers.

Nornicotine in the tobacco plant is predominantly formed by enzymatic demethylation of S-nicotine (15, 16), a process that appears biased toward the S-nicotine enantiomer and leads to an observed enantiomeric % ratio (R/S ratio) of nornicotine in tobacco of ~30%/70% (17,18,19,20,21) which does not match the enantiomeric % R/S ratio of nicotine from the same tobacco plant (~2%/98%) (6, 7, 21, 22) (see Table 4.)

If the nornicotine enantiomeric % ratio (R/S ratio) are mismatched (e.g., 30%/70% (R/S)) to their respective nicotine R/S ratios (e.g., 2%/98% (R/S)) this would indicate that the samples are TDN. If the nicotine enantiomeric % ratio (R/S ratio) the nornicotine enantiomeric % ratio (R/S ratio) are well-matched (e.g., ~equal) then the samples would be expected to be SyN.

Cheetham et al. (5) attempted to test this hypothesis. They analyzed the nicotine samples using LC-MS/MS and found that they could achieve excellent separation of R- and S-nornicotine on their standards but poorer separations on the commercial nicotine samples (TDN and SyN samples). The method of Ashraf-Khorassini et al. (8) does not appear to suffer from this disparity.

Due to the different production pathways for tobacco-derived and synthetic nicotine, it might be expected that the impurity profile of each might offer a potential means to distinguish the two sources (TDN or SyN). To this end, all the alkaloid standards were screened, as well as the commercial nicotine samples and the e-liquids for common impurities that can be found in tobacco products (nicotine degradants, e.g., nornicotine, anabasine and anatabine) via gas chromatography/selected ion monitoring/mass spectrometry (GC/SIM/MS).

All GC/MS and GC/SIM/MS analyses were performed using a 7890 GC equipped with a 5975 Mass Selective Detector (MSD) from Agilent (Wilmington, DE, USA). Separations were obtained using an Agilent J&W DB-WAXetr capillary column (30 m long × 250 μm I.D. with a film thickness of 0.25 μm) from J&W (Wilmington, DE, USA). The following operating parameters were used for each analysis (see Table 5).

Due to the relatively low concentration of anabasine, anatabine, and nornicotine in the samples, GC/MS/SIM was used. Table 6 summarizes the ions that were used for the detection of each compound, including their retention time and approximate detection limits.

The U.S. Pharmacopeia (USP) monograph for nicotine lists seven nicotine-related compounds that must be analyzed and found to be ≤ 0.3 wt. % individually and ≤ 0.8 wt. % collectively in order to be considered acceptable for use (22). (Please note that both TDN and SyN can be USP grade and used as analytical standards.)

These seven compounds, also known as nicotine degradants, are anabasine, anatabine, cotinine, nicotine-N-oxide, β-nicotyrine, nornicotine, and myosmine. Of these seven, anabasine and anatabine would be expected to be found exclusively in tobacco-derived nicotine, since the synthetic pathway would exclude their formation. However, Cheetham et al. (5) did not find anabasine and anatabine in two of the four samples of TDN. This may have been due to the sensitivity of their detector. Myosmine and nornicotine are both common intermediates in the chemical synthesis of nicotine and as such could potentially be more abundant in synthetic nicotine. Cheetham et al. (5) found that it was equally probable to find myosmine at variable levels in both TDN and SyN samples. Nicotine-N-oxide, β-nicotyrine, and cotinine are oxidation products of nicotine, and their presence would not necessarily be indicative of either production route. Cheetham et al. (5) conjectured that the presence of nornicotine (when it was observed in either the TDN or SyN samples) was indicative that the sample was TDN. None of the SyN samples contained appreciable levels of nornicotine (5), although they could have, as nornicotine is a common degradation product of both TDN and SyN.

The exact same nicotine and nornicotine analytical standards (Table 1) used in this study were tested by Ashraf-Khorassini et al. (8) who used chiral-SFC with DAD-UV on a Chiralpak IG-3 column to resolve the R- and S-isomers of nicotine and the R- and S-isomers of nornicotine. As previously noted, the DAD-UV detector is less sensitive than a SIM detector. Table 7 shows those results.

Employing the chiral-SFC method described by Ashraf-Khorassini et al. (8), baseline separations of R- and S-nicotine and R- and S-nornicotine were achieved.

Table 8 shows the results of the percent R- and S-nicotine and percent R- and S-nornicotine for the donated commercial TDN and SyN samples. TDN-1 through TDN-3 are the TDN samples from Siegfried and Nicotine River. TDN-1 and TDN-2 are TDN freebase samples in neat form and in propylene glycol, respectively. TDN-3 is a TDN sample as a Polacrilex (20%) nicotine salt.

The conclusion drawn from the data in Table 8 is that the results on TDN-1 and TDN-2 are equivocal. The data do not support the mismatch hypothesis of Cheetham et al. (5). Cheetham (5) who also tested TDN-1 and TDN-2 showed results similar to what we report here. Cheetham et al. (5) noted that the Nicotine River samples were highly distilled and contained low to no nornicotine. We also found these results. Although the samples from Siegfried and Nicotine River (TDN-1 and TDN-2) were explicitly labeled as tobacco-derived nicotine, unequivocal data proving this was not evident. The Siegfried nicotine salt sample was also deemed equivocal as the mismatch hypothesis could not be established. TDN-1 through TDN-3 contained very low levels to no of S-nornicotine. As a result, establishing the mismatch hypothesis was not possible. There were eight commercial samples of SyN from five suppliers. Different manufacturers employ varied methods to prepare their SyN products. Next Generation Lab's (NGL's) synthetic pathway starts from ethyl nicotinate. Ethyl nicotinate is reacted with N-vinyl-2-pyrrolidinone to form myosmine. Myosmine is reduced to nornicotine, followed by a methylation step resulting in a racemic (50/50) mixture of the nicotine stereoisomers, S-nicotine and R-nicotine (23). S-nicotine and R-nicotine are then prepared via stereoselective recrystallization. Zanoprima Lifesciences (London, UK), as a supplier of synthetic S-nicotine uses myosmine as a starting material (23). Myosmine is first stereoselectively converted to S-nornicotine using a commercially available recombinant enzyme, then S-nornicotine is converted to S-nicotine through methylation. Contraf-Nicotex-Tobacco uses ethyl nicotinate to produce S-nicotine and then uses stereoselective recrystallisation to prepare very pure products. N-Joy starts with racemic (R/S) nicotine and then uses stereoselective recrystallisation to prepare very pure S-nicotine (23).

The presence of small levels of nornicotine in the commercial SyN samples is not unexpected. For example, Zanoprima Lifesciences (London, UK), as a supplier of synthetic S-nicotine uses myosmine as a starting material (23). In other methods to prepare SyN (e.g., stereoselective recrystallization) it is very possible that small levels of nornicotine could be produced via oxidation or reduction reactions (23).

In the present study Millipore Sigma R/S-nicotine and S-nicotine were used as analytical nicotine standards. Additionally, R-nicotine from Toronto Research Chemicals was used as an analytical standard (Table 1). All of these analytical standards are prepared from tobacco nicotine. As a result, finding nornicotine and other secondary nicotine alkaloids in the nicotine analytical standard is not unexpected. All nicotine standards must meet USP standard of at least 98% nicotine, but trace amounts of nornicotine and other secondary nicotine alkaloids are possible (see data in Table 7).

SyN-1 through SyN-3 are sold by Next Generation Lab (NGL) and are S-Synthetic nicotine, R-Synthetic nicotine and R/S-Synthetic nicotine, respectively. The match hypothesis for SyN states that if the nicotine enantiomeric % ratio (R/S ratio) and the nornicotine enantiomeric % ratio (R/S ratio) are well-matched (e.g., ~equal) then the samples would be expected to be SyN. This appears to be true for samples SyN-1. For SyN-2, which is NGL's SyN R-nicotine, the vast majority of the nicotine is R-nicotine (96.9%) with only a small amount of S-nicotine carried over from the selective recrystallization. As no nornicotine was found, the match hypothesis could not be tested, and the sample is characterized as equivocal. For SyN-3 which is NGL's R/S Synthetic nicotine, the R/S nicotine ratio is nearly 50/50, as one might expect. However, the nornicotine enantiomeric % ratio (R/S ratio) was not well-matched (e.g., not equal). As a result, the sample would not be expected to be SyN and is characterized as equivocal, based on the match hypothesis.

Sample SyN-4 is S-nicotine from Nicobrand. This sample conforms to the match hypothesis and as a result is characterized as SyN S-nicotine.

Sample SyN-5 is S-nicotine in propylene glycol from Tobacco Technology, Inc. It is almost pure S-nicotine (99.8%). However, the nornicotine enantiomeric % ratio (R/S ratio) was not well-matched (e.g., not equal). As a result, the sample would not be expected to be SyN and is characterized as equivocal, based on the match hypothesis.

Samples SyN-6 and SyN-7 are samples of S-nicotine in propylene glycol and S-nicotine salt in propylene glycol, respectively, from Nicotine River. Both of these samples conform to the match hypothesis and as a result are characterized as SyN S-nicotine and SyN S-nicotine salt, respectively.

Sample SyN-8 is S-nicotine from AlChem. This sample conforms to the match hypothesis and as a result is characterized as SyN S-nicotine.

It should be noted that the results of this study mirror some of the results obtained by Cheetham et al. (5). For example, Cheetham et al. (5) tested Next Generation Lab's (NGL's) R/S-SyN. In this paper we also tested NGL's R/S-SyN. In both cases, similar results were found. Figure 3 is a plot of concentrations of R- and S-nicotine in the commercial nicotine samples. Figure 4 is a plot of concentrations of R- and S-nor-nicotine in the commercial nicotine samples.

Figure 3.

Figure 4.

Table 9 shows the results of the percent R- and S-nicotine and percent R- and S-nornicotine for the e-liquids removed from 11 e-cigarettes purchased from the internet. The e-cigarettes were differentiated based on how the products were advertised on the internet and/or on how the product packaging labelling stated the products as either to be containing tobacco-derived nicotine (TDN) or synthetic nicotine (SyN). Figure 5 is a plot of concentrations of R- and S-nicotine in the e-cigarette samples. Figure 6 is a plot of concentrations of R- and S-nornicotine in the e-cigarette samples.

Figure 5.

Figure 6.

All of the e-cigarettes marketed as having tobacco-derived nicotine, TDN Ecig-1 through TDN Ecig-3 (N-Joy, Puff Bar, and blu, respectively) appear to contain tobacco-derived nicotine. This is evident by the very high level (> 99%) of S-nicotine and the level of S-nornicotine in the e-liquid. The levels of S-nicotine and level of S-nornicotine in the e-liquid are remarkably like the levels of S-nicotine and S-nornicotine in commercial tobacco-derived nicotine samples. However, one could argue that the nicotine used could be synthetic as similar results would be expected. Because of this ambiguity, the results are considered equivocal.

There were eight e-cigarettes marketed as having synthetic nicotine in their e-liquids. SyN Ecig-1, SyN Ecig-2, and SyN Ecig-3 (Mr. Fog Max, Helix Bar Max, and Vapor Lax) appear to have levels of S-nicotine and S-nornicotine in the e-liquid consistent with the levels of S-nicotine and S-nornicotine in commercial synthetic nicotine samples. The levels of S-nicotine and S-nornicotine in the e-liquid are remarkably like the levels of S-nicotine and S-nornicotine in commercial tobacco-derived nicotine samples. Therefore, one could argue that the nicotine used could be tobacco-derived as similar results would be expected. As a result, the results are equivocal.

SyN Ecig-4 (Blue Razz Ice) represents a dilemma. It contains ~74% S-nicotine and ~26% R-nicotine. Additionally, it has all S-nornicotine and no R-nornicotine. The data does not provide sufficient information to discern the type of nicotine used in the e-liquid. Therefore, the product description is uncertain.

SyN Ecig-5 (Fuze Disposable Banana) also represents a dilemma. It contains ~99% S-nicotine and ~0.1% R-nicotine. Additionally, it has ~70% S-nornicotine and ~30% R-nornicotine. The data does not provide sufficient information to discern the type of nicotine used in the e-liquid. Therefore, the product description is uncertain.

SyN Ecig-6 (Dinner Lady - Citrus Ice) is advertised as a slim e-cigarette that uses synthetic nicotine. It contains ~52% S-nicotine and 48% R-nicotine. Additionally, both products have essentially no S-nornicotine or R-nornicotine. This nicotine data is very similar to the commercial R/S-Synthetic nicotine sample from Next Generation Labs (~ 52% S-nicotine and ~48% R-nicotine). However, the commercial R/S-Synthetic nicotine sample from Next Generation Labs contains small amounts of R- and S-nornicotine. The data does not provide sufficient information to discern the type of nicotine used in the e-liquid. Therefore, the product description is equivocal.

SyN Ecig-7 and SyN Ecig-8 are two products from a company called GoBoosted.com (Santé, Ontario, Canada).

GoBoosted.com. advertises that their products contain synthetic nicotine. Their products (Peach Ice and Blackcurrant Lychee) contain ~100% S-nicotine and no R-nicotine. Additionally, it has ~100% S-nornicotine and no R-nornicotine. This nicotine data is very similar to the commercial S-Synthetic nicotine sample from Next Generation Labs (~ 98% S-nicotine and ~2% R-nicotine).

However, the commercial S-Synthetic nicotine sample from Next Generation Labs contains small amounts of S-nornicotine (~ 100%) and essentially no R-nornicotine. The data does not provide sufficient information to discern the type of nicotine used in the e-liquid. Therefore, the product description is equivocal.