Identification and Quantitation of Nicotine Polacrilex in Nicotine Pouches and Other Oral Nicotine Delivery Products

Nicotine polacrilex was obtained from Siegfried St. Vulbas SS (Saint-Vulbas, France). Nicotine, phenanthrene, methanol, ammonia, silica, and crystalline cellulose were obtained from Sigma Aldrich (Merck Group, Burlington MA, USA). For the quantitation, a set of four standards containing a known concentration of nicotine polacrilex was prepared. An experimental nicotine pouch was initially made that contained 11.45% nicotine polacrilex with other ingredients including water. This material was freeze dried and the final concentration of nicotine polacrilex was 17.67%. This material was taken as the first standard. From this standard, three more standards were obtained by solid-to-solid dilution with microcrystalline cellulose (MCC; Dupont, Endurance VE-090). For this purpose, 200 mg initial standard and 200 mg MCC were thoroughly mixed in an agate mortar to prepare a standard containing 8.83% nicotine polacrilex. The dilution process was repeated to obtain standards containing 4.42% and 2.21% nicotine polacrilex.

The standards or the samples to be pyrolyzed were prepared by mixing solid-to-solid 200 mg material (standard or content of an oral nicotine delivery product) with 200 mg of silica that contains phenanthrene. This was performed in an agate mortar assuring a thorough mixing. From each sample/silica mixture 2.5 mg material was precisely weighed in a pyrolysis autosampler quartz tube and submitted for analysis by Py-GC/MS. Phenanthrene was used as a chromatographic standard and the preparation of silica containing phenanthrene is further described.

The silica-containing phenanthrene was obtained from silica gel (Selecto Scientific, Atlanta, GA, USA) with particle size 100–200 μm, and 60 Å pore size. To 10 g silica was added 5 mL solution in carbon disulfide (CS₂) containing 2.5 mg/mL phenanthrene. Additional 10 mL CS₂ were placed over the silica, and the slurry was well homogenized as the CS₂ was evaporating. A uniform distribution of phenanthrene on silica was achieved.

The pyrolyzer used in the study was a filament type Pyroprobe Model 5200 (with autosampler) (CDS Analytical, Oxford, PA, USA). Pyrolysis was performed in flash mode in helium and the pyrolysis temperature T_eq = 590 °C. The total heating time was THt = 20 s, and heating rate β = 20 °C/ms. The pyrolysate was analyzed on-line using a GC/MS instrument, which was a 7890B/5977B GC/MS with a High Efficiency Source (HES) from Agilent (Wilmington, DE, USA). The GC/MS analysis of the pyrolysates was performed in conditions listed in Table 1 (6).

The DB-1701 type column (Agilent/J&W Scientific, Wilmington, DE, USA) has medium polarity and separates well low molecular weight components of the pyrolysates. The selection of pyrolysis temperature T_eq = 590 °C was made based on literature information (7) showing that polyacrylates are completely decomposed at a temperature of about 480 °C. Similar to the case of other polymers with a saturated backbone carbon chain, random cleavage of the C—C bonds takes place during pyrolysis. Following this cleavage and β-eliminations, small molecules such as methacrylic acid is generated, and from the divinylbenzene moiety of the Amberlite IRP64 are 1,4- and 1,3-ethylvinyl-benzene, 1,4- and 1,3-methylvinylbenzene as well as 1,4- and 1,3-diethylbenzene generated. Part of the methacrylic acid that can be formed by depolymerization can be decarboxylated by CO₂ elimination. The hydrocarbons with aromatic rings and alkyl or alkenyl substituents are stable around 600 °C in particular for a very short residence time in a pyrolysis chamber (8).

The pyrograms (GC/MS chromatograms of pyrolysates) obtained at 590 °C of the material from a nicotine pouch are typically very complex, due to the pyrolysis products of MCC and of other ingredients such as maltitol.

However, the peak for the extracted ion m/z = 117 is characteristic in the mass spectrum of 1,3- and 1,4-ethyl-vinylbenzene and can be utilized for identification and quantitation of nicotine polacrilex. The window 30 min to 40 min for the extracted ion chromatogram for m/z = 117 from the pyrogram of a sample of nicotine pouches that contains nicotine polacrilex is shown in Figure 1.

Figure 1.

Only background noise is generated in the corresponding retention time window for the extracted ion 117 from the pyrograms of nicotine pouches samples that do not contain nicotine polacrilex.

The presence of the peaks for the extracted ion m/z = 117 at the corresponding retention times can be considered characteristic for nicotine polacrilex.

The quantitation of nicotine polacrilex can be made by using the area of peak at 36.31 min from the extracted ion chromatogram of the ion with m/z = 117.

The calibration curve obtained with standards is quadratic, following the equation: 1 Y=2.1178−14X2+5.9667−07X−0.07746 \[Y={{2.1178}^{-14}}{{X}^{2}}+{{5.9667}^{-07}}X-0.07746\] where Y is the % of nicotine polacrilex and X is the peak area of ion 117. The calibration using each standard in five replicates generated a correlation coefficient R² = 0.9995. The use for the calibration of the areas normalized by I.S. generated a slightly lower correlation coefficient R² than the calibration using only the peak areas of the standards. For this reason, the phenanthrene peak areas were used only to verify the integrity of the pyrolytic process. The variability in the peak area of ion m/z = 178 at retention time t_R = 58.33 min corresponding to phenanthrene was about 2.2%, and this variability is graphically shown in Figure 2 for a number of standards and material from nicotine pouches.

Figure 2.

The back calculation of the level of nicotine polacrilex in the standards showed an average error of 1.87%, with an error of about 0.2% for the highest standard, and about 4.3% for the lowest standard. The accuracy of the method can be also considered very good, the back calculation of the lowest standard showing a difference less than 2% from the expected level.

In addition to the evaluation of the nicotine polacrilex level based on the level of peak area in the pyrogram of ion with m/z = 117, the generated nicotine level can also be estimated. The window 40 min to 50 min for the extracted ion chromatogram for m/z =162 from the pyrogram of a sample of nicotine pouches that contains nicotine polacrilex is shown in Figure 3.

Figure 3.

The calibration curve for the nicotine obtained from the same series of standards used for nicotine polacrilex generated a quadratic curve. The level of nicotine in the standards were 3.53%, 1.77%, 0.88%, and 0.44%.The calibration follows the equation: 2 Y=2.60902−14X2−3.07948−07X+1.373973 \[Y={{2.60902}^{-14}}{{X}^{2}}-{{3.07948}^{-07}}X+1.373973\] where Y is the % of nicotine and X is the peak area of ion 162. The normalization with the area of phenanthrene peak was not performed for the areas of ion with m/z = 162. The back calculation of the level of nicotine in the standards showed an average error of 5.63%.

However, for some standard samples the error was as low as 0.43% and for other standard samples as high as 11.86%. It was not possible to determine the cause of the larger errors.

The back calculated level of nicotine in the standards showed an average of 20.16% (very close to the expected theoretical 20% nicotine in nicotine polacrilex), but the lowest back calculated nicotine level was 16.74% and the highest 23.9%. This indicated a lower reliability in the measurement of nicotine level by the pyrolysis technique. An additional aspect related to nicotine measurement using pyrolysis data is the potential decomposition of part of the nicotine during pyrolysis. By pyrolysis at 900 °C, during flash pyrolysis (at heating rate = 20 °C/ms), only about 7% of nicotine suffer decomposition and the main decomposition products are myosmine, β-nicotyrine, and cotinine (9). Based on this information, the presence of these three compounds was verified in the pyrogram of nicotine polacrilex, although pyrolysis of this material was performed at a lower temperature (590 °C). Only β-nicotyrine at about 1–2% of the level of nicotine was detected in the pyrogram. This level is lower than the average error for measuring nicotine and it cannot be determined if β-nicotyrine was generated during pyrolysis or was present as an impurity in the nicotine used to make the nicotine polacrilex. The maximum ratio of nicotine to nicotine polacrilex is about 20%. The results of the measurement of the level of nicotine polacrilex and the results of the measurement of nicotine can provide information whether besides nicotine polacrilex other forms of nicotine were added to the analyzed material. If the ratio (nicotine)/(nicotine polacrilex) is higher than 20–23% by weight it is likely that other forms of nicotine were added.

Because the measurement of nicotine level from Py-GC/MS data showed a relatively large variability, an HPLC procedure based on a method recommended by U.S. Pharmacopeia (3) has been developed for nicotine measurement in oral nicotine delivery products. For this analysis, about 200 mg sample was weighed in a 20 mL scintillation vial, and the precise weight was recorded. To the sample were added 200 μL solution of concentrated NH₄OH and 4.8 mL methanol, and the samples were extracted for 60 min on a wrist action shaker (Burrell Co., Pittsburgh, PA, USA). The extract was filtered through a 0.45 mm PVDF (polyvinylidene fluoride) filter (Whatman Autovial, Cytiva, Maidstone, UK) and analyzed by an HPLC procedure. The analysis was performed on a 1200 Ser. HPLC from Agilent (Wilmington, DE, USA), consisting of a solvent delivery system, quaternary pump, autosampler, column temperature controller, and variable wavelength detector.

The separation was performed in gradient conditions on a XTerra^® RP 18,5 μm column, 150 × 4.6 mm (Waters, Milford, MA, USA). For the gradient, aqueous component A contained 0.024 M ammonium carbonate and was brought to pH = 9.7 with ammonium hydroxide, and organic component B was acetonitrile. The gradient started with 100% A for 2 min, was brought to 59% A at 10 min, kept at 59% A for 2 min, and returned to initial conditions at 14 min with a total run time of 15 min. The flow rate was 1 mL/min, detection was performed at 254 nm, and the injection volume was 5 μL. The temperature of the column was not controlled. For the quantitation, a set of 7 standards with concentrations between 8 μg/mL and 512 μg/mL nicotine was used. The calibration curve was linear with the equation: 3 Y=0.244243X−2.38494 \[Y=0.244243X-2.38494\] where Y is μg/mL of nicotine and X is the nicotine peak area in the chromatogram. The calibration had a correlation coefficient R² = 1.0000. Because the extraction of nicotine from nicotine polacrilex was performed as described in Azzopardi et al. (3), the nicotine extraction efficiency was not verified. The precision of the measurement based on the back calculation of the nicotine level on standards was 2.5%. The limit of quantitation (LOQ) of the procedure could be taken as the value of the lowest calibration standard and LOQ = 8 μg/mL, although the sensitivity of the method is better. Based on signal to noise ratio for the peak of the lowest standard the LOQ could be taken as low as 0.5 μg/mL. The accuracy of the method was also very good, with a difference less than 0.2% for the highest standard and less than 5% for the lowest standard (triplicate measurements).