Menthol has long been an important flavorant in tobacco products, and both its historical and present uses are topics of increasing debate. Menthol can exist in eight different stereoisomeric forms (as four enantiomeric pairs) that possess different sensory properties. As regards use in tobacco products, the open scientific literature and available industry documents focus on the D-menthol and l-menthol enantiomeric pair, and in particular on l-menthol, but are ambiguous about the actual importance of D-menthol in tobacco products. This study provides the first openly available measurements regarding the stereoisomeric forms of menthol as found in selected United States sub-brands of smokeless tobacco (SLT), cigarettes, and cigarette smoke.
Gas chromatography/mass spectrometry (GC/MS) was applied using a “chiral” GC column to separate and determine the forms of menthol present in headspace air above various samples of United States sub-brands of SLT, cigarette filler material, and cigarette smoke particulate matter. Additional GC × GC/Time-of-flight mass spectrometry measurements were also made.
The dominant form of menthol by far in any of the samples was l-menthol.
For the selected cigarettes and SLT products tested from the U.S. market, the only form of menthol found was l-menthol. Other forms may be present in products that were not tested. No evidence was found of thermal racemization upon smoking of l-menthol to a d+l mixture.
“Menthol” (i.e., 5-methyl-2-(1-methylethyl)cyclohexanol, C10H20O) is a terpenoid alcohol found in the oil of mint plants, such as Mentha piperita (peppermint) and Mentha arvensis (cornmint). This alcohol possesses three chiral carbons. As such, it can exist in 23 = 8 different stereoisomeric forms as four enantiomeric pairs. For each such “d,1” pair, the two forms are nonsuperimposable mirror images of one another. The d form has a “dextrorotatory” (clockwise) effect on plane-polarized light, and the l form has a “levorotatory” (counterclockwise) effect. Any pair taken from the eight forms that are not enantiomers of one another (e.g., l-menthol and l-neomenthol) are diastereomers, that is, nonsuperimposable compounds that are also not mirror images of one another. Physical properties such as vapor pressure and boiling point will be identical within any given enantiomeric pair but different for diastereomers. Sensory properties can be very different within a given enantiomeric pair and among diastereomers. The fact that human olfactory receptors contain chiral groups and can distinguish between some enantiomers is demonstrated by the distinctively different scents of caraway and spearmint, which are due to the two enantiomeric forms of one compound, carvone.
Natural and synthesized forms of “menthol” have been used as flavoring ingredients in cosmetics, confectionary products, oral health products, over-the-counter pharmaceuticals, and tobacco products (Burdock, 2009; Chen, Isabelle, Pickworth, & Pankow, 2010; Perfetti, 1985). The threshold level for sensation of a “cooling” effect is much lower for l-menthol (aka (−)-menthol or (1R,2S,5R)-menthol) than for D-menthol (aka (+)-menthol or (1S,2R,5S)-menthol) or for any of the other stereoisomers (Eccles, 1994; Emberger & Hopp, 1985).
Emberger and Hopp (1985) report for peppermint oil that (a) menthol, neomenthol, isomenthol, and neoisomenthol forms are all present; (b) of these, the menthol forms comprise >90% of the total “menthols”; and (c) l-menthol greatly predominates (>99%) over D-menthol (see also Manuale, Betti, Marchi, Yori, & Romeo, 2010; Orav & Kann, 2001; Singh et al., 2005). As noted, diastereomers have different physical properties. This allows separation by traditional means of the menthols from the neomethols, isomenthols, and neoisomenthols in natural products. Purification by crystallization from natural mint oils can thereby lead to nearly pure l-menthol (Emberger & Hopp, 1985). Synthetically, routes are available that lead to racemic d+l menthol as well as enantiopure l-menthol (Hopp, 1993). The 2007 worldwide production of “menthol” was 19,000 metric tons of which about 25% was used in tobacco products (Clark, 2007).
The available tobacco industry documents (http://legacy.library.ucsf.edu/) are ambiguous as to whether one or more forms of “menthol” have been used in tobacco products but do focus almost exclusively on the d- and l-menthol pair and in particular on l-menthol. Given the tremendous importance of mentholated tobacco products, we find it astonishing that the scientific literature is heretofore nearly silent on the forms of “menthol” in cigarettes and smokeless tobacco (SLT) products, except that in an article from Lorillard, Heck (2010) states “Both l-menthol and dl-menthol are used in tobacco products.” From RJ Reynolds Tobacco Company, Coleman, Perfetti, and Suber (1998) are similarly unrevealing and report on measurements of the relative amounts of d- and l-menthol in mouthwash, toothpaste, aftershave lotion crème de menthe, and skin cleaning pads but not in any tobacco products: either no measurements on tobacco products were made in the underlying study by these Philip Morris scientists or they were withheld from the publication. Moreover, we are not aware of any reports of the forms of “menthol” found in cigarette smoke, for example, whether the temperatures that are achieved in a burning cigarette can thermally racemize l-menthol to d+1 menthol.
The lack of information that exists to date on the forms of “menthol” in tobacco products stands in stark contrast to the enormous importance of mentholated tobacco products. Indeed, menthol is currently the only “characterizing flavor” allowed in cigarettes in the U.S. market (Family Smoking Prevention and Tobacco Control and Federal Retirement Reform, 2009). And, based on cigarettes sold, the market share of menthol cigarettes in the United States in 2001 was estimated to be 26% (Federal Trade Commission, 2003). Furthermore, menthol cigarettes are popular internationally: Giovino et al. (2004) reported that in the Philippines, Cameroon, Hong Kong, and Singapore, the market share of menthol cigarettes (based on numbers of cigarettes sold) has been ∼60%, 40%, 26%, and 22%, respectively. Lastly, menthol is also an important flavor in SLT products. A survey carried out in the state of New Jersey (University of Medicine and Dentistry of New Jersey—School of Public Health, 2006) showed that, among the flavored products, the “mint” category (with “menthol” as the primary flavorant) was second most popular, immediately behind “wintergreen.” (Overall, flavored products as a group were found to account for >60% of all SLT sold between 2003 and 2005.)
The public health effects of mentholated cigarettes have been investigated in numerous epidemiological studies. In a recent menthol supplement to the journal Addiction, data are discussed that indicate (a) disproportionately higher rates of menthol cigarette smoking for Blacks and young adults and that quit attempts among menthol smokers (particularly Blacks) are less successful than among nonmenthol smokers (Stahre, Okuyemi, Joseph, & Fu, 2010; Trinidad, Perez-Stable, Messer, White, & Pierce, 2010) and (b) young adult nondaily smokers who smoke menthol cigarettes show greater nicotine dependence than those smoking nonmenthol cigarettes (Ahijevych & Ford, 2010). Possible causes of greater harm from mentholated cigarettes include (a) the ability of “menthol” to mask the harshness of tobacco smoke allowing easier and deeper inhalation of mentholated tobacco smoke (Wayne & Connolly, 2004; Foley, Payne, & Raskino, 1971), which would make mentholated cigarettes potentially more addictive as well as attractive as “starter” products (Substance Abuse and Mental Health Services Administration, Office of Applied Studies, 2009); (b) an impairment by menthol of metabolic clearance of nicotine (Benowitz, Herrera, & Jacob, 2004); and (c) the inhibition of the detoxification of the tobacco-specific nitrosamine (TSNA) carcinogen 4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanol (NNAL; Muscat et al., 2009). For SLT, Squier, Mantz, and Wertz (2010) have reported that “menthol” can enhance the transdermal and transbuccal penetration of nicotine as well as the TSNA N-nitrosonornicotine (NNN).
The ongoing debate concerning the effects of “menthol” in tobacco products (e.g., Ahijevych & Garrett, 2004; Gardiner & Clark, 2009; Heck, 2010; Wayne & Connolly, 2004; Werley, Coggins, & Lee, 2007) is confounded by the complex sensory and biochemical effects of the “menthol” compounds. For example, low/threshold levels of l-menthol can yield a cooling sensation, whereas a warming sensation and/or skin irritation have been reported for higher levels (Eccles, 1994). Celebucki, Wayne, Connolly, Pankow, and Chang (2005) report “menthol” levels in cigarettes as ranging from 1.6 to 4.3 mg per cigarette, but to our knowledge, no reports exist in the literature on per cigarette deliveries by mentholated cigarettes; Heck (2010) does cite a transfer efficiency (from rod to mainstream smoke) of ∼10% for contemporary cigarettes. For mint-flavored SLT, Chen et al. (2010) report menthol levels ranging from 0.85 to 5.3 mg/g. The purpose of this study is to identify the stereoisomeric forms of menthol in a representative group of sub-brands of U.S. cigarettes, SLT, and in the smoke from selected sub-brands of U.S. cigarettes.
All tobacco products were purchased in the Portland, OR, metropolitan area. These included the following seven cigarette sub-brands (all king-size and all hard-pack): Newport, Marlboro Menthol, Camel Menthol, Basic Menthol Lights, Pall Mall Menthol, Salem, and Kool. The eight sub-brands of SLT products examined included Grizzly Long Cut Mint, Husky Long Cut Mint, Kodiak Long Cut Ice, Kodiak Premium Mint, Longhorn Long Cut Mint, Skoal Bandit Pouches Mint, Skoal Pouches Mint, and Timberwolf Packs Pouches Mint.
The sources of the chemical standard materials were d-menthol and l-menthol from Sigma-Aldrich and l-menthol from Vigon International. Authentic standards of the six other forms of “menthol” were not acquired because preliminary measurements indicated that none of them was present in any of the samples at higher than trace levels.
Most “menthol” measurements were carried out with headspace gas sampling (25 μl) followed by gas chromatography/mass spectrometry (GC/MS) with an Agilent 7890 GC/5975C MSD (Agilent Technologies, Santa Clara, CA). For a given enantiomeric pair, the MS response factors will be identical under any given conditions. The relative amounts of D- and L-menthol in a given sample were thus determinable without quantitative analysis. The chiral capillary GC column used with the Agilent 7890 GC/5975C MSD was obtained from Supelco, Inc., Bellefonte, PA: β–DEX 120 phase, 0.25-μm film thickness, 30 m long, and 0.25 mm internal diameter (ID). Helium at 1 ml/min was the carrier gas. After headspace injection, the GC temperature program was hold at 40°C for 2 min, then 40°C–220°C at 4°C/min. All injections were in “splitless” mode at an injector temperature of 180°C.
Headspace sample vials were prepared by placing materials of interest into individual 2-ml GC autosampler vials (Agilent Technologies). Three types of samples were investigated: (a) cigarette filler material (∼50 mg), (b) SLT (∼100 mg), and (c) cigarette smoke particulate matter as collected on a section of the cigarette filter (∼30 mg). Five sub-brands of cigarettes were smoked using eight 50 ml puffs and a puff interval of 30 s. The sub-brands included Newport, Marlboro Menthol, Camel Menthol, Basic Menthol Lights, and Pall Mall Menthol.
To protect the chiral column from relatively large loadings of nicotine, for each sample, (a) a 1.0-cm disk of Millipore Cellulosic Absorbent Pad (Millipore, Bedford, MA) was lodged in the vial above sample material, care being taken to avoid contact between the disk and sample material and (b) 10 μl of an oxalic acid–saturated water solution was injected onto the disk. Each sealed vial was equilibrated >30 min prior to analysis. Individual standard vials containing the following were also prepared: ∼5 mg d-menthol, ∼5 mg l-menthol, ∼5 mg d+1-menthol, ∼50 mg cigarette filler (Newport) + ∼5 mg d-menthol, ∼100 mg SLT (Kodiak Premium Mint) + ∼5 mg d-menthol , and ∼100 mg SLT (Husky Long Cut Mint) + ∼5 mg L-menthol.
A small number of supporting measurements to assist in the investigation of the possible presence of any of the neomenthols, isomenthols, or neoisomenthols were also carried out by sampling the headspace vials using the relatively more sensitive solid phase microextraction (SPME) method as followed by analysis using a Pegasus 4D GC×GC/Time-of-flight mass spectrometry (TOFMS) (Leco Corporation, St. Joseph, MI). In this 2D GC instrument, the primary GC column was nonchiral and obtained from Agilent: DB-VRX; 45 m long; 0.25 mm ID; 1.4-μm film thickness. The secondary GC column was from Restek: Stabilwax; 1.5 m long and 0.25 mm ID; 0.5-μm film thickness.
Various GC/MS chromatograms obtained using headspace sampling, the chiral column, and the Agilent 7890 GC/5975C MSD are provided in Figure 1. The chromatogram for the d+1-menthol standard mixture is given in Figure 1a, and chromatograms for various tobacco-related samples are provided in Figure 1b–f. Near baseline separation of d-menthol and l-menthol was achieved. The retention times and separation for the tobacco-related samples were essentially the same as in Figure 1a, and l-menthol was the only form of menthol positively detected in any of the samples, whether cigarette filler material, SLT, or cigarette smoke particulate matter. For the smoke particulate matter samples collected using cigarettes known to contain l-menthol (Figure 1f), the absence of any detectable d-menthol demonstrates that significant smoking-related thermal racemization of l-menthol to a d+1 mixture does not occur in cigarettes. A few of the analyses by SPME on the GC×GC/TOFMS indicated the presence of trace levels of two menthone isomers and other forms of menthol in some of the SLT samples, possibly (+)-neomenthol and (−)-isomenthol, as based on the observed presence of these compounds in peppermint oil (Manuale et al., 2010).
The three conclusions from this work are (a) no other form of menthol besides l-menthol was found at significant levels in a range of sub-brands of cigarettes and SLT products from the U.S. market. (For those tested products, the use of l-menthol over other forms is likely due to its predominance in nature and its strong sensory effects.) (b) Other forms of menthol may well be present in brands and forms of tobacco that were not tested here, whether from inside or outside of the U.S. market. (c) l-menthol in burning cigarettes does not racemize to a d+1 mixture.
Declaration of Interests
This work was supported by the Cooley Family Fund for Critical Research of the Oregon Community Foundation and through the support of Regina M. Dowd, Michael J. Dowd, Patrick J. Coughlin, Keith F. Park, and Steven T. Huff.