Impact of HPLC Parameters on Chiral Separation of Nornicotine Enantiomers

American Chemical Society (ACS) grade isopropyl amine (IPAm), diethylamine (DEA), triethylamine (TEA), nornicotine, nicotine, and pure nicotine and nornicotine enantiomers were obtained from Sigma-Aldrich (St. Louis, MO, USA) and used as received. Hexane, ethanol (EtOH), and isopropanol (IPA) were HPLC grade and were obtained from Fisher Scientific (Pittsburgh, PA, USA). The tobacco standards were obtained from the Center for Tobacco Reference Products ( CTRP) at the University of Kentucky and used as received. The low alkaloid-tobacco was obtained by purchasing a low alkaloid-cigarette product from a local market.

Packed columns with CHIRALCEL OD-3 (coated cellulose (3,5-dimethylphenylcarbamate)), CHIRALPAK IA-3 (immobilized amylose (3,5-dimethylphenylcarbamate)), IB N-3 (immobilized (3,5-dimethylphenylcarbamate)), IC-3 (immobilized cellulose (3,5-dichlorophenylcarbamate)), IG-3 (immobilized amylose (3-chloro-5-methylphenylcarbamate)), and AY-3 (coated amylose (2-methyl-5-chlorophenylcarbamate)) (250 × 4.6mm, dp = 3 μm) were obtained from Chiral Technologies, Inc. (West Chester, PA, USA). All columns were conditioned as recommended by the manufacturer before use and tested with different modifiers and additives.

An Agilent 1200 Series HPLC equipped with a quaternary pump, variable wavelength detector (VWD-UV), auto liquid sampler (ALS), and oven heater set to ambient was employed with all columns and modifiers. All analyses were performed at 254 nm. Nornicotine and nicotine standards, at approximately 10 ng/μL, consisting of both the (R)- and (S)- enantiomers dissolved in dichloromethane (DCM) or hexane were used to optimize the chromatography conditions. Injection volume was set to 10 μL for all analyses. All separations were obtained under isocratic conditions with hexane as the main component of the mobile phase.

A VLN^© (very low nicotine) brand cigarette was purchased from a local grocery store. The extraction protocol employed herein was validated by CORESTA (Cooperation Centre for Scientific Research Relative to Tobacco, Method No. 62 (39) participants for the quantitative extraction of nicotine from tobacco matrices. The tobacco from two cigarettes or approximately 2 g of the tobacco standards was emptied into a standard coffee grinder and ground for 10–15 s, then used without further modification. First, 0.5 g of the ground tobacco sample was mixed with 10 mL of water, followed by the addition of 5 mL of 8N NaOH. The mixture was shaken for 1 h using a ThermoFisher Scientific shaker (Pittsburgh, PA, USA). The supernatant was then isolated by centrifugation at 4000 rpm for 4 min. Next, 10 mL of hexane was added to the extract supernatant and shaken for 1 h. The hexane layer was separated, and the extraction was repeated with an additional 10 mL of hexane. The combined hexane solutions were dried over a small amount of anhydrous sodium sulfate. Finally, the hexane extract was analyzed using an optimized HPLC method to determine the distribution of nornicotine enantiomers. Nornicotine extractions of the standard tobacco samples from the University of Kentucky were prepared using the same extraction method.

Initially, six different chiral columns (listed in Chiral stationary phases) were screened using a 90/10% hexane/(EtOH + 0.5% DEA) mobile phase to determine which of the column(s) provided the shortest retention time and acceptable peak shapes with the highest resolution. These preliminary results indicated that the IB N-3 and IG-3 offered the best resolution with nornicotine enantiomers with retention times under 15 min (Figure 1). The nornicotine enantiomer retention times and peak shapes obtained with IB N-3 were shorter and sharper, respectively, than those observed with IG-3, while still possessing an R value greater than 2. Systematic variations in flow rate, modifier (EtOH) concentration, and additive (DEA, TEA, and IPA) concentration revealed that the enantiomer retention times with IB N-3 were consistently shorter than those attained with IG-3 (Table 1). Both columns, however, yielded R values greater than 2. The overall method philosophy of making the analysis time as quick as possible, all subsequent optimizations were therefore performed using IB N-3.

Figure 1.

Chiral HPLC/UV/VIS separation of nornicotine enantiomers using IG-3 and IB N-3 columns.

Varying the mobile phase flow rate between 1.0 and 0.8 mL/min resulted in changes to both the retention times and R values of the nornicotine enantiomers (Table 1), with the (R)- nornicotine enantiomer eluting from the column first. Several observations on the influences of changes in flow rate on enantiomer retention times and R values (relative internally to each set of conditions) can be made from the data in Table 1:

For IB N-3

• At 0.2% DEA and a 90/10 mobile phase composition, enantiomer retention times decreased, ~1 min, at the higher flow rate with no notable change in R

• At 0.5% DEA and 90/10 mobile phase composition, enantiomer retention times decreased, ~1 min, at the higher flow rate with a ~0.1 decrease in R

• At 0.2% DEA and 85/15 mobile phase composition, enantiomer retention times decreased, ~2.3 min at the higher flow rate with a ~0.1 decrease in R

• At 0.5% DEA and 85/15 mobile phase composition, enantiomer retention times decreased, ~2 min, at the higher flow rate with a ~0.1 decrease in R

For IG-3

• At 0.2% DEA and 85/15 mobile phase composition, enantiomer retention times decreased, ~5 min, at the higher flow rate with an ~0.3 decrease in R

• At 0.5% DEA and 85/15 mobile phase composition, enantiomer retention times decreased, ~6 min, at the higher flow rate with an ~0.5 decrease in R

• At 0.2% DEA and 80/20 mobile phase composition, enantiomer retention times decreased, ~5 min, at the higher flow rate with an ~0.2 decrease in R

• At 0.5% DEA and 80/20 mobile phase composition, enantiomer retention times decreased, ~5 min, at the higher flow rate with an ~0.4 decrease in R

Varying the additive percentage in the mobile phase resulted in observable changes to both the retention times and R values of the nornicotine enantiomers (Table 1).

For IB N-3

• At 1 mL/min and 90/10 mobile phase composition, enantiomer retention times decreased, ~1 min, at the higher additive percent with no notable change in R

• At 0.8 mL/min and 90/10 mobile phase composition, enantiomer retention times decreased, ~1 min, at the higher additive percent with no notable change in R

• At 1 mL/min and 85/15 mobile phase composition, enantiomer retention times decreased, ~0.3 min, at the higher additive percent with no notable change in R

• At 0.8 mL/min and 85/15 mobile phase composition, enantiomer retention times decreased, ~0.5 min, at the higher additive percent with no notable change in R

For IG-3

• At 1 mL/min and 85/15 mobile phase composition, enantiomer retention times decreased, ~2 min, at the higher additive percent with a decrease, ~0.5 in R

• At 0.8 mL/min and 85/15 mobile phase composition, enantiomer retention times decreased, ~1 min, at the higher additive percent with a decrease, ~0.4 in R

• At 1 mL/min and 80/20 mobile phase composition, enantiomer retention times decreased, ~1.3 min, at the higher additive percent with a decrease, ~0.4 in R

• At 0.8 mL/min and 80/20 mobile phase composition, enantiomer retention times decreased, ~2 min, at the higher additive percentage with a decrease, ~0.3 in R

Varying the modifier percentage in the mobile phase resulted in observable changes to both the retention times and R values of the nornicotine enantiomers (Table 1).

For IBN-3

• At 0.2% DEA and 1 mL/min mobile phase flow rate, enantiomer retention times decreased, ~5 min, at the higher modifier percentage with a decrease, ~0.7 in R

• At 0.2% DEA and 0.8 mL/min mobile phase flow rate, enantiomer retention times decreased, ~3 min, at the higher modifier percentage with a decrease, ~0.5 in R

• At 0.5% DEA and 1 mL/min mobile phase flow rate, enantiomer retention times decreased, ~4 min, at the higher modifier percentage with a decrease, ~0.7 in R

• At 0.5% DEA and 0.8 mL/min mobile phase flow rate, enantiomer retention times decreased, ~3 min, at the higher modifier percentage with a decrease, ~0.6 in R

For IG-3

• At 0.2% DEA and 1 mL/min mobile phase flow rate, enantiomer retention times decreased, ~6 min, at the higher modifier percentage with a decrease, ~0.4 in R

• At 0.2% DEA and 0.8 mL/min mobile phase flow rate, enantiomer retention times decreased, ~6 min, at the higher modifier percentage with a decrease, ~0.5 in R

• At 0.5% DEA and 1 mL/min mobile phase flow rate, enantiomer retention times decreased, ~5 min, at the higher modifier percentage with a decrease, ~0.3 in R

• At 0.5% DEA and 0.8 mL/min mobile phase flow rate, enantiomer retention times decreased, ~7 min, at the higher modifier percentage with a decrease, ~0.4 in R

When similar percent changes in modifier were made using isopropanol, nornicotine enantiomer retention times of less than 20 min could not be attained with either column even with percentages of isopropanol approaching 25%. Thus, in keeping with the philosophy of building methods with quick analysis times, isopropanol was no longer investigated as a mobile phase modifier.

The impact on nornicotine enantiomer resolution and retention times of a group of amine additives including TEA, DEA, and IPAm revealed that using TEA produced undesirable peak tailing (Figure 3). The nornicotine enantiomer retention times for the DEA containing mobile phase were shorter than those observed when IPAm and TEA were employed as mobile phase modifiers, with other conditions remaining constant. The R values for the different additives were 2.14, 1.99, and 1.11 for DEA, IPAm, and TEA, respectively, with all other conditions held constant.

Figure 2.

Chiral HPLC/UV/VIS separation of nornicotine enantiomers as a function of changes in additive percentage using IB N-3 column.

Figure 3.

Chiral HPLC/UV/VIS separation of nornicotine enantiomers using selected additives using IB N-3 column.

Authentic gravimetric standards of nornicotine enantiomers were prepared at selected unit ng/μL levels and optimized chromatographic conditions were employed for their separations with UV detection at 254 nm. The results indicate that a nornicotine enantiomers detection limit of approximately 1 ng on column, with a signal to noise ratio of greater than 2/1, for the enantiomers, Figure 4.

Figure 4.

Chiral HPLC/UV/VIS separation of nornicotine enantiomers at their detection limit.

In an investigation into the possible simultaneous separation of nicotine and nornicotine enantiomers in a single injection, the optimized conditions for nicotine enantiomer separation were found to be not conducive to the separation of nornicotine enantiomers (18, 21). The nornicotine enantiomer retention times were >40 min, which was accompanied by a broad peak shape, even though an R value of 3.87 was noted (Figure 5). Under the conditions, that is, optimized nicotine enantiomer separation, this one particular set of chiral HPLC method parameters, did not possess the capability to optimally and efficiently separate both sets of enantiomers in one injection. However, through a systematic adjustment of the mobile phase modifier percentage, the simultaneous separation of both nornicotine and nicotine enantiomers was attained with:

• 1)

  R values greater than 2, and

• 2)

  retention times less than 10 min in one injection, or alternatively, R values greater than 2 and retention times less than 13 min (Figure 5).

Figure 5.

Chiral HPLC/UV/VIS simultaneous separation of nornicotine and nicotine enantiomers as a function of mobile phase composition (IB N-3 column).

This discovery most likely represents one of the first disclosures of a single chiral HPLC method for the simultaneous separation of both nornicotine and nicotine enantiomers that possesses these characteristics.

Systematic variations in normal phase HPLC mobile phase flow rates (between 1.0 and 0.8 mL/min), mobile phase modifier percentage, and mobile phase additive percentage, produced consistent trends in shifts of enantiomer retention times and R values. The R value and retention time %RSD values for the endpoints were less than 3%. The nornicotine enantiomers were well resolved with R values ranging between ~2 and ~5 (Figures 1–5) with the R value results from CHIRALPAK IG-3 consistently greater than those from CHIRALPAK IB N-3, regardless of mobile phase conditions. The nornicotine enantiomer pair retention times were significantly longer, ~29/~25 and ~17/~15 min for CHIRALPAK IG-3 compared to ~14/~8 and ~13/~8 min for CHIRALPAK IB N-3.

For both columns:

• Increases in mobile phase flow rate always resulted in notable reductions in enantiomer retention times, sometimes as great as 5 min, with accompanying meaningful somewhat unattractive yet having no significant impact on overall separation performance (0.1 to 0.5) shifts in R values

• Increases in additive percentages always resulted in observable reductions in enantiomer retention times, from 0.2 to 2 min, with accompanying notable shifts somewhat unattractive yet having no significant impact on overall separation performance (0.3 to 0.5) in R values

• Increases in modifier percentages always resulted in significant reduction in enantiomer retention times, from 1 to 7 min, with accompanying notable shifts somewhat unattractive yet having no significant impact on overall separation performance (0.3 to 0.6) in R values

Thus, from the “quick” analysis time perspective, the following conditions would provide for the fastest nornicotine enantiomer analysis time and more than adequate nornicotine enantiomer resolution: column, IB N-3; mobile phase flow rate, 1 mL/min, mobile phase composition, 15% EtOH modifier and 0.5% DEA additive. Depending on the complexity of the tobacco extract matrix, an alternative mobile phase composition of 10% EtOH and 0.5% DEA could be employed to improve enantiomer resolution and eliminate co-eluting extract components.

These finding clearly illustrate the extremely critical impact of column selection, arguably minor changes in modifier percentages, additive percentages, and mobile phase flow rates, on the nornicotine enantiomer R values and retention times. Judicious selection of the values for modifier and additive type and percentage can thus lead to meaningful positive shifts in both retention time reduction and enantiomer R values (>2), improving the performance of the analytical method and addressing several of the issues delineated in the QuEChERS protocol requirements. Among the accepted analytical protocol requirements, a method must possess effective and rugged characteristics. These criteria can be met when the method analytical data supports analyte specificity/accuracy and precision. From the data contained herein, the presence of baseline enantiomer resolution coupled with detection at 254 nm provides strong supporting evidence for specific analyte detection. The overwhelming majority of the % RSD values for both analyte retention times and analyte enantiomer R values, at less than 3%, support the rugged character of the approach. A critical and important aspect of successful chromatographic separations is the nature of the peak shape of the analytes of interest. The influence of additive percentage on the shape of the nornicotine enantiomer peaks is evident (Figure 2), with the greater amount of additive (0.5% DEA) providing for a greater R value.

Some significant observations on peak shape as well as other chromatographic characteristics can be garnered from Figure 3. The nornicotine enantiomer peak shape when using TEA as the additive compared with DEA and IPA was significantly distorted with unacceptable peak tailing. The underlying cause of this behavior is likely not related to the basicity of the amine additives themselves, as the pK_a values are 10.75, 10.98, and 10.63 for TEA, DEA, and IPA, respectively. It has been well established that different additives can affect the performance of polysaccharide-based chiral stationary phases (CSPs) by masking silanol interactions with the silica gel support material (29), the ability of the additive to outcompete the analyte for space in the CSP binding groove (30, 31), or the ability of the additive to mask certain hydrogen bonding interactions within the CSP binding groove (32). While the specific cause was not investigated in this study, it is possible that one or a combination of these possibilities accounts for the additive effects observed as related to peak shape, selectivity, and R values.

In a very desirable/perfect scenario, a method using a single chiral HPLC method possesses the characteristics and the capability of providing information on the enantiomer distributions of both nicotine and nornicotine (arguably the vast majority of Nicotiana tabacum L. alkaloids) would be advantageous. The investigations described herein clearly indicate that the technologies to accomplish this feat have now been established (Figure 5). The unanticipated magnitude of the shifts in nicotine and nornicotine enantiomer retention times to much shorter times accompanied by more than acceptable R values was demonstrated. Reduction in nornicotine retention times by a factor of 4 and by a factor of 2 for nicotine are clearly demonstrated (Table 3). Equally important was the retention of very acceptable R values for both sets of nicotine and enantiomers, 2.60 and 3.09 for nicotine and nornicotine enantiomers, respectively (Figure 5).

Overall, this result highlights the importance of making small incremental adjustments to additive and modifier concentrations when assessing their effects on enantiomeric resolution (R values) and retention times, particularly when optimizing sample analysis times. Obviously, there is little to no reason for creating mobile phase compositions that result in enantiomeric R values greater than one. However, it is critically important to fully understand how the mobile phase compositions influence R values and retention times, one of the main foci of this manuscript, and how subtle changes in mobile phase compositions can actually result in unacceptable enantiomer elution characteristics.

An additional important criterion for an acceptable method is the ability of the approach to detect the analyte(s) of interest at concentrations closely linked to levels of the analyte(s) appearing in samples of interest. Figure 4 presents the chiral HPLC chromatogram of nornicotine enantiomers at a nornicotine concentration of 1 ng on column for each enantiomer with a signal to noise ratio of greater than 2:1, these conditions affording an acceptable detection limit for nornicotine concentrations in a variety of tobacco cultivars.

The U.S. Food and Drug Administration (FDA) published an Advanced Notice of Proposed Rulemaking (ANPRM) in March of 2018 titled “Tobacco Product Standard of Nicotine Level of Certain Tobacco Products”. The ANPRM addressed reducing nicotine levels in tobacco fillers used for cigarette manufacture by 95 to 98%. (16). A standard cigarette filler contains 5 mg to 15 mg of nicotine per gram of tobacco. In a similar vein, the World Health Organization recommends a 35-fold reduction, which corresponds to 0.4 mg/g of nicotine (17). Application of this nornicotine enantiomer analysis protocol to the analysis of the tobacco contained within a commercially available low alkaloid tobacco cigarette as well as tobacco standards revealed that the method was appropriate for the qualitative and quantitative analysis of the nicotine enantiomeric distribution within samples of this type (Figures 6 and 7). The distribution of nornicotine enantiomers disclosed herein was well within the ranges found for the extracts of selected tobacco standard cultivars and smokeless tobacco (18). The protocol was able to detect the presence of both nornicotine enantiomers with generally acceptable signal to noise ratios (>2/1) and consistent enantiomer retention times, clearly indicating that this protocol for nornicotine enantiomer distribution would be applicable to most all Nicotiana tabacum L. samples.

Figure 6.

Chiral HPLC/UV/VIS separation of nornicotine enantiomers extracted from flue cured, Burley and Oriental standard tobaccos and a low nicotine commercially available tobacco (90/10 Hexane/EtOH, 0.5% DEA, 1 mL/min, IB N-3 column).

Figure 7.

Chiral HPLC/UV/VIS separation of nornicotine enantiomers extracted from commercial VLN cigarette tobacco. Conditions (90/10 Hexane/EtOH, 0.5% DEA, 1 mL/min, IB N-3 column).

A somewhat subtle but arguably very important observation was detected in the nornicotine enantiomer R values for the tobacco standards extracts (Figure 6). Of note was the marked difference in the R value for the flue-cured tobacco extract, 2.26 versus the R values for the Burley and Oriental tobacco extracts at ~2.8. A possible reason for this marked difference could reside with the complexity of the sample matrix, that is, the nature and concentration of the other tobacco components and optical isomers known to be present in tobacco extracts (33, 34). Hence, minor but necessary adjustments to mobile phase composition maybe required to maintain acceptable enantiomer resolution values and resolution of tobacco matrix components from the enantiomers themselves. To illustrate this concern, an important observation was made when a set of two different mobile phase compositions was employed to determine the nornicotine enantiomer distribution in the tobacco extracts. Resolution of the nornicotine enantiomers themselves as well as the other extract components was possible with the following mobile phase composition (90/10 hexane/EtOH, 0.5% DEA, 1 mL/min, CHIRALPAK IB N-3). However, when the mobile phase was changed slightly to 85/15 hexane/EtOH. 0.5% DEA, 1 mL/min, CHIRALPAK IB N-3, in an attempt to shorten retention times, an unacceptable co-elution of other tobacco extract components with the nornicotine enantiomers occurred. Thus, it is of note that optimized mobile phase conditions for the separation of nornicotine and quite possibly nicotine enantiomers as well, including results from separations of enantiomer standards, may have to undergo some subtle modification(s) to ensure complete resolution of the enantiomers present in a tobacco extract having multiple additional components. The addition of known amounts of each pure enantiomer to the tobacco extract could be employed to confirm the shifts in retention times of the enantiomers due to the tobacco extract complexity.

In a marked contrast to the nornicotine enantiomer distribution characteristics of the Burley, flue-cured, and Oriental tobacco extracts, the chiral HPLC analysis of the extract of the tobacco from the VLN product was dominated by the presence of the (S)-nornicotine enantiomer with the (R)-nornicotine enantiomer not being detected (Figure 7). In a similar fashion, the (S)-nicotine enantiomer was the only nicotine enantiomer detected in the VLN extract, as well (Table 2). The impacts of flow rate, additive percentages, and modifier percentages on the nornicotine enantiomer retention times and resolution were very similar to those observed in the chiral HPLC separation of nicotine enantiomers (27).

Several recent publications have addressed directly or indirectly the simultaneous chromatographic separation of tobacco alkaloids (35,36,37,38). Most often adequate separation of both enantiomers was obtained (R greater than 1) but with retention times greater than 20 min and employing at times two chiral columns in series. In one case (38), however, through a virtual combination of the retention times contained within two displayed separate chromatograms of alkaloid standards, it was possible to artificially construct a chromatogram wherein the retention times of nicotine enantiomers (~0.7 and 0.9 min) and nornicotine enantiomers (~3.9 and 4.5 min) would possibly have been produced in a single injection. Thus, the current disclosure contained herein when compared with known published results, provides:

• 1)

  an understanding of the subtle influences of mobile phase, additive, and modifier structures as well as their concentrations on the resolution and retention times of nicotine and nornicotine enantiomers and

• 2)

  for a new relatively fast, easy, inexpensive, accurate, and precise method for the simultaneous determination of nicotine and nornicotine enantiomer distributions in tobacco extracts.