Suresh Ralapati, batf/National Laboratory Center

In summary, this section has attempted to provide an overview of analytical procedures for the determination of nicotine in tobacco. Emphasis has been placed on three examples of current analytical practice. However, many other analysis strategies beyond the examples presented are in use today. For example, in a recent CORESTA Routine Analytical Chemistry Sub Group joint experiment coordinated by Hans Thomsen to investigate the determination of nicotine in tobacco by gas chromatography, eighteen laboratories reported eleven separate combinations of sample preparation procedure and chromatographic conditions as described in Table 4 (Thomsen HV, personal communication). Therefore, the choice of tobacco nicotine determination that is practiced in a particular laboratory will often be determined as much by practical considerations (e.g., instrument availability considerations) as it will be by scientific considerations such as the tobacco type of interest, etc.

IV. Determination of Nicotine in Tobacco Processing Environments

Tobacco is used in a variety of products (e.g., chewing tobacco, snuff, pipe tobacco and cigarettes). Growers and manufacturers process all tobacco in some way before or during its preparation into a variety of products for use by consumers. The flavor, taste and aroma, as well as the cost and consumer desires usually determine the choice and level of different tobacco types (Burley, flue-cured, Oriental, Maryland, cigar types, Latakia, Perique and various types of fire-cured and air-cured tobaccos) used in the production of different products. The taste and/or smoke flavor of tobacco leaf may vary depending on the location in which it is grown, the climate, agricultural and post-agricultural practices employed, the location of the leaves on the tobacco stalk and the maturity of the leaf (38). Because of the diversity in leaf characteristics and the inherent flavor in different tobaccos, scientists have explored the chemistry of tobacco in great detail. One of the chemical constituents often examined is nicotine. The nicotine content of tobacco lamina, stem and processed tobacco materials can often be used in conjunction with other chemical, physical and sensory measures to distinguish between different tobacco types, grades of tobacco and different processing means (39).

This section will describe and characterize the main types of tobacco used in the tobacco industry. Typical tobacco nicotine concentrations of those tobacco types will be presented. Tobacco processing techniques will then be discussed in terms of their effect on changes in tobacco nicotine concentrations. A description of the processing that occurs with tobacco from the field to finished products will be discussed. Tobacco processing within the cigarette industry will then be described with emphasis on how those processing steps affect the nicotine concentration in tobacco blends. Analytical techniques for direct and indirect, qualitative and quantitative determination of nicotine, where these determinations have been practiced and documented in the industry, will be discussed

Tobacco Types
Flue-cured tobacco derives its name from the “flues” used to heat the buildings (small barns) where the tobacco is packed to undergo its curing process (40). During the flue-curing process stalk-ripened tobaccos are placed in barns and the temperature of the barns is slowly increased in a particular manner to wilt, yellow, fix and then dry the tobacco. During this process, enzymatic activity converts the majority of the tobacco starch to sugar and various tobacco pigments degrade to produce a golden or yellow color in the leaf (39,41). Flue-cured tobacco is principally used in the cigarette industry. Its smoke is often described as having a sweet, aromatic character (38). Flue-cured tobaccos contain significantly higher levels of reducing sugars than air-cured tobaccos such as Burley or Maryland. Flue-cured tobacco generally has a lower level of tobacco nicotine compared to Burley. Although the average nicotine level of flue-cured (~ two to four percent) is lower than Burley, the nicotine level of particular grades of flue-cured from different growing locations, having different climates, and employing varied agricultural practices can vary greatly (39). Table 5 lists the mean percent tobacco nicotine levels for US flue-cured and Burley tobacco crops grown between 1967 and 1990. The data in Table 5 are archival in nature. Each set of data (flue-cured and Burley) for each year represents an average of multiple analyses conducted on inventoried tobaccos from all tobacco grades purchased during that year. Methods for determination of tobacco nicotine have varied and have been improved continuously over the period of 1967 – 1990. Spectrophotometric analysis of total tobacco alkaloids measured as nicotine was the method used in Table 5 from 1967 - ~1980. Gas chromatographic analysis of tobacco nicotine has been used as the method of choice since ~ 1980. During the period of 1967 – 1990 the data reported have been analyzed in different laboratories in R. J. Reynolds Tobacco Company (RJR) and by numerous analysts. All data were collected on dried and ground tobacco samples. All results are reported as either total tobacco alkaloids (measured as nicotine) or as nicotine on a dry weight basis. The historical data of Table 5 illustrates the year-to-year tobacco nicotine variability in flue-cured and Burley tobacco. Over the 23-year period, flue-cured tobacco nicotine ranged from 2.14 – 3.50%. Burley tobacco nicotine ranged from 2.43 – 4.63%. Although, the percent tobacco nicotine for flue-cured is considered, in general, to be lower than the percent tobacco nicotine of Burley tobacco, there are years when this is not the case. In 1977, 1979 and 1985 the average nicotine level in flue-cured crop was higher than the Burley crop (Figure 5).

Burley tobacco is considered a light, air-cured tobacco (38). It is normally light-brown to reddish brown in color. It possesses a taste, smoke flavor and aroma described as having a “chocolate, nutty, protein, winey, and ammoniacal” character (38). Burley tobacco possesses excellent smoking characteristics for use in blended tobacco cigarettes and in pipe tobaccos. Smokers do not usually prefer smoking Burley tobacco alone because they are considered too “harsh and strong”. Burley tobacco is low in sugar, which is metabolized from starch during the air curing process. The nicotine content of Burley is generally higher than that of flue-cured tobacco (~ two to four percent) depending on the type, leaf position on the stalk, growing season and growing location (39,42). In smoking products, sugar based casings are normally applied to Burley tobaccos to mellow its smoking character. Burley tobacco is also often blended with flue-cured tobacco (that is high in sugar) to help “smooth” the more “basic” smoke of Burley tobacco (38).

Maryland tobacco is air-cured and is considered to have excellent smoking characteristics. It has characteristics similar to Burley tobacco but is much lighter in taste. Straight Maryland cigarettes are popular in some European countries such as Switzerland (38). The tobacco nicotine level of Maryland tobacco is about one to two and one-half percent by weight.

Oriental tobaccos are unique. Their small leaf size (three to six inches) relative to flue-cured and Burley leaf (twelve to sixteen inches) and their aromatic odor and taste characterize the uniqueness of Oriental tobaccos. Oriental tobaccos are usually used in blended cigarettes although straight Oriental cigarettes are produced and are popular in the regions where they are grown (most countries bordering the Aegean, eastern Mediterranean and Black seas). Oriental tobaccos have high levels of volatile flavor oils and provide for a rich smooth flavor effect when smoked. The smoke of Oriental tobaccos range from highly aromatic in character to a resinous-cedar character largely depending on the region where the tobacco is grown (38). The nicotine content of Oriental tobaccos is normally low and range between one to two and one-half percent (39).

Cigar tobaccos are all air-cured tobaccos (43-46). They are classified as a medium to heavy bodied tobacco and are medium to dark-brown in color. Cigars have a core made of the cigar filler. The filler is wrapped with a tobacco leaf or a reconstituted tobacco referred to as the cigar binder. The outer wrapper of the cigar is known as the cigar wrapper. Depending on the cost of the cigar this outer cigar wrapper may be a high-quality, fine, unblemished single tobacco leaf or a specially prepared reconstituted tobacco sheet prepared from cigar tobacco made to look like a cigar tobacco leaf. The cigar wrappers are usually grown under a cloth to shade them from the direct sunlight, temperature and wind that can damage the leaves. This type of tobacco is known as “shade grown” tobacco. Not only is cigar tobacco air-cured, but it is also fermented under controlled conditions. To ferment the air-cured tobacco, the tobacco is moisturized, bulked and allowed to stand for varying periods of time. During the fermentation process, the tobacco reaches temperatures of 115 to120°F. The fermentation process produces the characteristic color and aroma of cigar tobacco. Numerous chemical and biological changes occur during the fermentation process. Those changes produce a variety of color, flavor and odor differences (via Browning and Maillard Reactions) in the finished fermented tobacco. Much of the nicotine in the cigar tobacco is biologically degraded during the fermentation process. Air-cured cigar tobacco typically has a concentration of three and one-half to four percent tobacco nicotine prior to fermentation. After fermentation as much as 80 percent of its original nicotine is lost (43).

There are other curing and fermenting processes that are used on tobacco (38). Tobaccos can be fire-cured in which they are smoked over low burning fires during the curing process. The fire-cured tobacco is then packed for aging. Fire-cured tobaccos are not fermented. There is another unique tobacco called Latakia, which is an Oriental sun-cured variety grown principally in Syria that is fire-cured. These types of tobaccos are generally used in pipe tobacco blends and give unique smoky notes to the pipe smoke. Perique tobacco is a fermented tobacco type that has a very unique flavor. Perique tobaccos are prepared by fermenting green tobacco leaf under pressure in wooden cases for several years. The finished leaf is black and has a sweet aroma. Perique is also used in pipe tobacco blends. The tobacco nicotine content of fire-cured tobaccos can vary depending on the nicotine content of the original leaf. There is not a substantial loss of nicotine from fire-curing. The nicotine content of Latakia is ~one percent or lower, as it is an Oriental-type tobacco. The nicotine content of Perique tobacco is also very low ( less than ~ one percent) due to the extensive fermentation that it undergoes.

Processed Tobaccos
There are three principal types of processed tobacco; processed stems, reconstituted sheet materials and expanded tobacco.


Processed Stems. After tobacco leaf is purchased - primarily flue-cured and Burley leaf- it is normally taken to a stemming operation where it is conditioned with water to a moisture necessary to separate the lamina from the stem or midrib of the individual leaves. Only about 65 to 75 percent of the weight of the purchased leaf tobacco will end up as tobacco cut filler. The rest of the material is classified as tobacco by-products and is used to prepare processed tobaccos.

As depicted in Figure 6 (42), the stem and dust are separated from the broken leaf (tobacco scrap) and lamina. The lamina is dried to about 12 percent moisture and packed in bales. The bales are taken to storage where they are held for 18 to 36 months on average. The dust from the stemming operation and the drying operation are also dried to a low enough moisture, usually less than10 percent, before they are boxed up and taken to storage. Likewise, the stems and tobacco scrap are dried and boxed for later use. When called for, the boxes of stems are sent to processing areas to prepare useful materials that can be added to or blended with the tobacco lamina. Burley or flue-cured stems can be processed. Processed stem materials are cut or crushed rolled stems (CRS) (42). Cut rolled stems can also be partially expanded. Stems, after being moistened, cut and sized to about one to two inches, are rolled under pressure and cut into disk shaped pieces of a size similar to tobacco cut filler (CRS). The cut stem pieces can also be partially expanded using steam (67). The nicotine level of processed stems is low and is normally about 25 percent of the nicotine level found in the lamina (42).

Stemmery dust is not usually sent to the tobacco reconstitution process as it usually contains large amounts of sand and dirt (less than 10 percent). The sand content of stemmery dust is difficult to remove. As a result, some manufacturers simply dispose of this material and place it into land-fills.

Reconstituted Tobacco. Preparation of reconstituted tobacco sheet is excellent economic way tobacco manufacturers have found to convert tobacco dust, fines, scrap and stem into useful tobacco filler. Tobacco reconstituted sheet materials can be of such good quality and appearance that they can substituted for tobacco lamina in cigarettes and for cigar filler. As previously mentioned, reconstituted sheets are also used as cigar binders and in some cases, cigar and cigarillo outer wrappers (38).

There are at least five processes for making reconstituted tobacco sheet (47,48). Browne (42) has described five different methods, but the two major methods employed today are the cast sheet, or slurry process, and the paper making process. Several reviews and articles have been written on the production and use of reconstituted tobacco which detail the physical, process engineering and smoking and health aspects of reconstituted sheet materials (49,50). The books by Sittig and Halpern (51,52) review much of the patent literature on tobacco reconstituted sheet materials.

For the cast sheet process, tobacco by-products such as tobacco dust, fines, scrap and stems in various proportions are dried and ground to a very fine powder. The powder is mixed into an aqueous solution of binder. Typical binders are methylcellulose, carboxymethylcellulose, guar or locust-bean gum (42). Additional fiber, such as cellulose, can be added. Additionally, combustion and ash modifiers, humectants and flavors can be added. The suspension is usually mixed thoroughly, sheered in a blender or put through a refiner to prepare viscous slurry of fine particle size. The slurry is transferred to a head-box, and a thin film of the slurry is cast onto a stainless-steel belt, dried, reconditioned if necessary, doctored from the belt and either rolled up or cut into pieces the size of lamina. Certain binders can be used to prepare slightly expanded sheet materials. These binders are heat sensitive and expand during the drying process. Some cast sheets are prepared without the addition of a binder. In this type of cast sheet the natural pectin in tobacco is chemically released with the use of ammonia and diammonium phosphate as processing aids. The pectin is cross-linked during the drying process to produce an excellent tobacco reconstituted sheet material (66).

The nicotine content of cast sheet material is totally dependent on the initial tobacco mix of tobacco used to form the sheet. The cast sheet process is over 95 percent efficient in terms of material balance. Typically the nicotine content of cast sheet materials is between one and four percent nicotine (55). But it is also true that about five to twenty percent of the nicotine can be lost in processing during the drying step. Therefore, the overall amount of tobacco nicotine that is found in the cast sheet material is always less than the amount of nicotine that was originally present in the starting material.

Paper-Processed Reconstituted Tobacco Sheet. To prepare reconstituted tobacco sheet by the paper process, a mix of tobacco by-products are used as the feedstock material (Figure 7)(42). Tobacco stems (flue-cured and Burley stems) are the largest component of the mix, usually at least 50 percent by weight. The rest of the mix is made of tobacco dust, tobacco scrap, by-products from the primary tobacco blending area, cigarette tobacco making by-products, and broken sheet from sheet operation (broke). The starting materials are called from tobacco storage. Often the materials are ground (scrap) or cut (stems) to reduce their size and to improve the extraction efficiency. The tobacco mix is then placed in extraction vessels with warm water and agitated to remove as much water-soluble material as possible. In the extraction process about 45 to 55 percent of the weight of the initial tobacco is soluble and can be separated from the tobacco pulp. The extract is then separated from the pulp. The pulp is washed with water and the combined dilute extracts are combined and concentrated in an evaporator. The “concentrated” extract (20 percent concentration of tobacco solids and solubles) is still quite dilute. The washed pulp is then digested, beaten mechanically and / or refined to fibrillate the cellulose and reduce the size of the fibers comprising the biomass composition of the pulp.

The clean refined pulp is very dilute at this point in the process, usually less than one percent suspension in water. The dilute refined pulp is pumped to a head-box where it is formed into a web on a wire screen of a standard papermaking machine, dried by suction and heated drums or hot air to a moisture level of 40 to 60 percent. In a parallel operation the extract, was “concentrated” as previously mentioned. Additional ingredients can be added to the extract at this point. Humectants, combustion and ash conditioners, sugars and flavors are often added to the extract to improve its material performance characteristics (friability, moisture retention, flexibility, etc.), burning properties and tobacco taste and smoke flavor. The “concentrated” extract is then applied to the wet web, dried in ovens, conditioned to a moisture of about 10 to 12 percent and either rolled up or cut into pieces about the same size as tobacco lamina (3 inch by 5 inch rectangular shapes).

The paper process operation for the production of reconstituted tobacco sheet is reasonably efficient. Operating efficiencies range from about 80 to 95 percent at conversion costs in the range of $0.40 to about $0.90 depending on the converter, the source of tobacco materials, the additives / processing aids used to prepare the sheet and the physical dimensions and properties of the reconstituted tobacco sheet produced. To be economical, the paper process requires a high throughput. Initial capital investments are high. As a result only large tobacco manufacturers and national monopolies can justify their own facilities and equipment. Many smaller tobacco manufactures send their tobacco by-products to other converters for processing (42).

The nicotine content of reconstituted tobacco sheet material is totally dependent on the initial tobacco mix of tobacco used to form the sheet. The efficient of the papermaking operation in terms of material balance is 80 to 95 percent. Typically the nicotine content of reconstituted tobacco sheet materials is between one and four percent nicotine (55). But it is also true that about five to twenty percent of the original nicotine present in the feedstock can be lost in processing, especially during the drying step. Therefore, the overall amount of tobacco nicotine that is found in the reconstituted sheet material is always less than the amount of nicotine that was originally present in the starting material.

Expanded Tobacco. Numerous processes have been developed over the years to expand tobacco and thus increase its volume and reduce the weight of the space it occupies (42,67,68). The technique typically involves the impregnation of tobacco with an expansion agent which is a liquid or gas form and then releasing that agent from the tobacco to increase the volume of the tobacco. Expansion agents have included water, steam, a wide variety of organic fluids and inorganic fluids. Volume expansions vary based on the mechanical / engineering reduction-to-practice and operating conditions, the impregnating liquid or gas employed, the tobacco type (e.g., flue-cured, Burley, Oriental, etc.) and tobacco form (lamina, stems). A schematic diagram for a typical expansion process is shown in Figure 8. In this process, tobacco lamina is shredded into cut-filler form (lamina cut to 25 to 35 cuts per inch wide). The tobacco is moistened to 25 to 35 percent and then saturated (impregnated) with an expansion agent. The impregnated tobacco is then contacted with a stream of steam to expand the tobacco. The expansion agent is most often recovered and reused. The expanded tobacco is then dried, cooled, conditioned to a set moisture and either used immediately or packaged for storage. Today, the preferred expansion agent is carbon dioxide. Use of expanded tobacco is cost effective and is widely used in the cigarette industry. The normal inclusion rate ranges from about five to fifty percent, with the normal level of incorporation in the range of ten to twenty percent.

Expansion processing does not, in most cases, dramatically change the nicotine content of expanded tobacco compared to the infeed tobacco. Nicotine losses on the order of five to ten percent are typical for most expansion processes used today. Factors that may effect these typical nicotine losses are the type of tobacco impregnated, any additives that might be applied to the tobacco, temperatures used to dry and condition the tobacco after expansion, and other design parameters that may be unique to each tobacco expansion facility.

Tobacco Processing
Tobacco is a natural product and as such a certain amount of variability in the physical and chemical properties of tobacco is expected to occur for each type grown, for each growing season, and for each location where it is grown in the world. Year-to-year, crop-to-crop and location-to-location variations are due to a variety of circumstances; some controllable (e.g., certain kinds of agronomic practices), and some uncontrollable (e.g., weather-related variables, political environment in tobacco growing areas of the world). To compensate for tobacco variability, blends of different types of tobaccos are usually used in smoking (pipe, cigar and cigarette tobacco blends) and smokeless (wet and dry snuff, chewing and plug tobacco) tobacco products to insure consistency in the final product mixture. For each different tobacco product a variety of different process steps take place due to manufacturing costs, tobacco availability, the different forms of products manufactured (e.g. dry snuff versus moist chewing tobacco) and the desired end use and preferences of the consumers (e.g. pipe tobacco versus cigarettes). Depending on the particular finished tobacco product, various processing steps (e.g., conditioning, blending, casing, cutting, bulking, flavoring, etc.) can be used to compensate for certain variations in the chemical composition of the tobaccos used (relative amount of sugars, acids, volatile oils and alkaloids). In this way it is possible for a manufacturer to provide consistent products to consumers year after year. The following section will describe the processing steps that tobacco undergoes from when it is harvested to when it is prepared into finished products. Although several chemical constituents are known to change during the processing of tobacco, special emphasis will be placed on changes in tobacco nicotine.


Effects of Harvesting, Curing Methods and Bulking on Tobacco Nicotine Content. The reported range of nicotine in different tobacco types used in the manufacture of tobacco products varies greatly (Table 6). Data on Table 6 are a compilation of literature data from numerous sources (39,43,45-47,54) and should not a considered a single data set. The data are presented to show relative change in tobacco nicotine that has been reported to occur at various stages in tobacco processing. The data under each tobacco types give examples of data in the literature and are from different studies. The specific analysis methods used to report tobacco nicotine changes for the numerous studies cited within the table are not known. The particular varieties of each tobacco reported in Table 6 and the specific agronomic practices under which the tobaccos were grown and are also unknown. The reported ranges of nicotine for each tobacco type in Table 6 encompass normal cultivars and hybrid varieties of tobaccos as seen in Table 7. Table 8 illustrates the typical ranges of nicotine by stalk position that is normally seen in different tobacco types grown in the United States. At the time of harvest green tobacco contains a level of nicotine that is governed in large part by the climatic, geographic and agronomic conditions that the tobacco experienced during its growing cycle. Data from the first part of Table 6 show the change in nicotine content those different tobacco types exhibit depending on the curing method employed by the grower. It should be emphasized that curing tobacco is an art. Different growers cure their tobaccos slightly differently depending on the maturity of the tobacco, the weather conditions during the time of curing, the overall climatic condition that the crop experienced during the season, etc. Typical nicotine losses that occur from curing are from 5 to 15 percent of the nicotine present in the green leaf. Data from this section of Table 6 illustrate that Oriental tobacco can lose nearly 40 percent of the nicotine in green leaf during sun-curing, although the nominal amount of nicotine lose (~0.5 percent) is similar to barn-cured burley tobacco. Stalk-cut, fire-cured burley tobacco can loose relatively little nicotine (five to fifteen percent) or nearly 30 percent (56) depending on how much heat the tobacco experiences during the curing process. Once the grower cures the tobacco it is made ready to be sold at auction. The cured, dried tobacco is moistened to prevent breakage during transport and handling. The conditioned tobacco is bulked before sale and then is either sold at auction or stored for a period of time before it is sold. The time that the tobacco is bulked often depends on the amount of tobacco the grower produced that year, the amount of stored tobacco the grower has from last year, the price of the leaf at market and other circumstances known to the grower or the market itself. Regardless, nicotine losses generally on the order of less than 0.1 percent are estimated to occur during the time that tobacco is bulked prior to sale (39).


Effects Stemming Operation and Aging on Tobacco Nicotine Content. Once the tobacco is sold to the tobacco dealer or tobacco product manufacturer, it is quickly removed from the auction house and taken to a stemmery. At the stemmery the tobacco is brought to a moisture level (~ 20 percent) where the midrib of the leaves can be removed without destroying the lamina. Cigar leaves are stemmed individually as these half-leaves can be used as cigar wrappers. Filler tobaccos and tobacco intended for cigarettes are mechanically stemmed or threshed to remove the maximum amount of lamina from each leaf. The stemmed tobacco lamina is separated from the midribs mechanically. The stems are dried to less than 10 percent moisture and packaged for storage. The tobacco lamina is sized to remove sand, dust and small particles (scrap) and then dried to ~12 percent moisture and packaged for storage. The stems and scrap can be used in tobacco reconstitution. The dust is usually not used due to the amount of sand that it contains (usually greater than ten percent). The tobacco lamina will later be used as cigar or cigarette cut-filler. Typical nicotine losses in the tobacco lamina during stemming and redrying are three to five percent, (middle section of Table 6)(39,40).


The tobacco lamina used in the cigarette industry is usually stored in warehouses before they are used. Modern warehouses have vapor-proof concrete floors and are designed for the use of forklifts, tow motors and other types of lifts to be able to place and remove packaged tobacco container with ease. Most of the modern warehouses have doors, louvers or vents that prevent the tobacco from extreme exposure to the environment. Temperature and moisture variations can be controlled up to a point in most modern warehouses. Storage times can be from eighteen months to three years. Tobaccos used in production of chewing tobacco, snuff (wet and dry), pipe tobacco and cigars are aged various periods of time (sometimes up to several years) and are also fermented in some cases depending on the product application. Tobacco is aged for numerous reasons, but two important reasons are economics and taste considerations. Economically, it is important for tobacco manufacturers to maintain substantial inventories of different types of domestic and imported tobacco for blending purposes. These inventories can ensure that the industry can maintain production of consistent products when crop failures occur or in years where the crop productions are low (due to drought or disease). Additionally, there have been times when wars, political unrest, natural disasters and even man-made disasters (Chernobyl incident) have limited the amount of tobacco that the tobacco industry can purchase. The second reason tobacco is aged is to improve the flavor characteristics of the tobacco. Some tobacconists believe that cured, dried, unaged tobacco is less desirable than aged tobacco that has undergone several cycles of moisture and temperature loss and gain during the aging period. Stored tobacco gains moisture (one to two percent) in the spring through late summer and loses moisture during winter (one to two percent). The gains in moisture and subsequent increases in temperature during the late spring and summer may produce conditions in the packaged tobacco where some sort of mild fermentation occurs. These conditions may also increase the chances for tobacco volatiles to migrate in the stored tobacco, homogenizing the taste qualities of the tobacco. Regardless, changes in the chemical constituents have been observed to occur during aging of tobacco in modern warehouses. For flue-cured tobacco there is an increase in moisture and a decrease in sugars, total nitrogen, water-soluble nitrogen, amino nitrogen, nicotine, total acids and pH during storage. For air-cured burley tobacco there is an increase in calcium and ammonia and a decrease in total dry weight of tobacco, protein nitrogen, soluble nitrogen, amide nitrogen, amino nitrogen and nicotine. Table 6 (lower portion) contains data on the loss of nicotine that occurred for flue-cured, Burley and Oriental tobacco stored in warehouses. Over a 30 month aging period flue-cured tobacco lost about 11 percent of its original nicotine. Oriental tobacco lost about 14 percent of its original nicotine during two years of aging and burley tobacco lost over 33 percent of its original nicotine content during 2.3 years of aging (39,40).

During cigar tobacco fermentation a tremendous amount of change occurs in the chemical constituents of tobacco. Considerable losses of dry tobacco weight, soluble carbohydrates, organic acids, pectins, pentosans, hemicelluloses, essential oils, resins, waxes, pigments and little or no change in cellulose or lignin occurred during fermentation. There is an evolution of carbon dioxide, ammonia, volatile nitrogenous and methyl alcohol during fermentation of cigar tobacco. The pH of cigar tobacco increases after fermentation. The water retention of cigar tobacco increases and its fire-holding capacity improves after fermentation. As shown on Table 6, the tobacco nicotine content of fermented cigar tobacco decreased substantially. Losses of between 50 and 80 percent are common. Nicotine is oxidized during fermentation to oxynicotine, water soluble pyridine derivatives, nicotinic acid, 2,3-dipyridyl- and methyl-3-pyridyl ketone and several other oxidized compounds (39,40). After cigar tobacco is fermented it will continue to age with minimal further losses of chemical constituents.

When tobaccos are called from inventory and before the various strips are combined or made into a blend of tobacco, the strips of tobacco must be sufficiently pliable to be handled without undue breakage during blend processing. To accomplish this remoisturizing of the tobacco, two distinct types of conditioning systems have been used. The first type employs a batch processing of packaged bales or blocks of tobacco. In this batch process, tobacco in bales (or other packaging forms) is placed in sealed vessels. The bales of tobacco are steamed to heat the tobacco. The steaming also adds water to the tobacco. The heated and moisturized tobacco delaminates easier and is less subject to breakage during manufacturing. There are several types of equipment available today to accomplish this task. The classical equipment used for many years was the Thermo-Vacuum process or the Vacu-Dyne process. In this process tobacco was placed into a vessel, the vessel was evacuated, and steam was allowed to escape into the evacuated chamber via probes. The probes were pushed into the tobacco inside the vessel to furnish both heat and water to the tobacco. Several cycles were normally used to bring the dry tobacco to a pliable form that minimized tobacco breakage during processing of the leaf. The second type of process used today to remoisturize dry packaged tobacco from storage is an in-line process. Tobacco from inventory is brought to the processing area. The tobacco is removed from the bales employing “kickers” that partially break up the tobacco bales, or in the case of the smaller burlap-wrapped Oriental tobacco, the Oriental bales are sliced into sections. The tobacco types are fed into direct conditioning cylinders and steamed at about 150 to 170F°. When the tobacco exits the direct conditioning cylinders, it is very pliable, moisturized and delaminated. The in-line process is used extensively today in large manufacturing areas, while the batch processing technique and equipment still finds a lot of application in smaller pilot plant operations. By employing either the batch process technique / equipment or the in-line conditioning process technique / equipment only a minimal amount of nicotine is lost. Typically less than one percent of the leaf nicotine is lost during this type of tobacco conditioning (Table 6).

Effects of Processing Tobaccos / Blends in the Cigarette Industry on Tobacco Nicotine Content. It was mentioned above that the first step in tobacco processing is to obtain tobaccos from a tobacco dealer or from a company’s inventory. Questions such as what types of tobaccos are needed, how much of each type is needed and what criteria should be used to determine quality and usability of the tobacco, are very important and need to be addressed. Before a blend can be put together, the blend specialist or cigarette designer must have a supply of different types of quality tobaccos and know information on the physical, chemical and organoleptic or sensory characteristics of those tobaccos. To obtain the necessary tobaccos, the tobacco manufacturers can contract grow the tobaccos needed based on specifications set by the manufacturers, buy tobacco at auction, or they can contract with one or more tobacco dealers to purchase specified types and grades of tobacco. Some tobacco dealers will stem and sometimes even store a manufacturer’s tobacco until needed. A Company’s buying program and practices are usually based on a five-year or longer strategic plan. The buying plan is based on the needs that the company has in terms of its recent and projected market performance, its present tobacco blends, its present inventory of tobaccos, projections of future tobacco availability (domestic and off-shore) and new product development. The tobacco manufacturer sets minimum quality and uniformity standards for each type and grade of tobacco based on stalk position, tobacco maturity and consistency. Typical criteria for purchasing are tobacco maturity, color, odor and sensory attributes (when available). Once the lots are purchased and stemmed the typical criteria of concern are particle size, nicotine level and total sugar content. Nicotine and sugar determinations can be obtained employing in-line or off-line near infrared (NIR) techniques (57-59) or by analysis of composite tobacco samples by colorimetric, gas chromatographic, ultraviolet spectroscopy or any number of other techniques known and available to the tobacco industry. The measurements for tobacco nicotine and sugars are done as a chemical check for the uniformity / consistency of each tobacco lot. Historically, the standard deviation expected in tobacco nicotine and sugar per lot of tobacco is 0.2 percent and 1.5 percent, respectively.

The cigarette industry in the United States (U. S.) normally uses what has been termed the “American Blend” which is comprised of varying ratios of flue-cured, Burley, Oriental, reconstituted sheet(s), expanded tobacco and a variety of tobacco by-products (42,48). The majority of flue-cured and Burley tobaccos used in U. S. cigarettes are grown and purchased domestically, although some manufacturers incorporate off-shore flue-cured and burley tobaccos in their products. Oriental leaf is 100 percent off-shore tobacco. The majority of domestic flue-cured tobaccos are grown along the eastern coast of the United States. Flue-cured tobacco is grown and sold in four different tobacco belts (growing regions). These are Florida/Georgia, South Carolina, eastern North Carolina and the so-called “Old Belt” that encompasses western North Carolina and Virginia (39). Domestic Burley tobacco is grown and sold primarily in two different belts, Kentucky and eastern Tennessee. Both flue-cured and Burley tobaccos have similar growing seasons, but are harvested and cured differently, and are sold over several months. For example, the Florida/Georgia belt flue-cured tobacco matures, is harvested, cured and sold first (the tobacco market opens in early summer), while the Burley tobacco market opens early in the fall and often continues through early winter. At the beginning of the flue-cured market in the Florida/Georgia belt, lower stalk tobaccos are sold, stemmed, analyzed both chemically and physically for quality and placed in inventory. At the auction, prior to the sale, agents from the U. S. Department of Agriculture (USDA) grade each lot of tobacco sold. For example, the USDA grade B3K indicates that the lot of tobacco is a flue-cured type, that it is a sample of leaf (B), that it is considered to have a uniform quality of 3^rd (3) and that the leaf was ripe (K). A lot of flue-cured tobacco can range from ~100-200 pounds. At the auction, the lots of tobacco are sold quickly, are identified as purchased by the different tobacco companies or dealers, and often a company designation is placed on the different lots of tobacco to specify a particular company grade. The tobacco is shipped to a stemmery.

At the stemmery hundreds of lots of tobacco are processed together in a sequential manner based largely on stalk position. Tobaccos from each belt are processed separately. Composite samples (five to ten) of each grade of tobacco from each belt are collected and analyzed for tobacco nicotine and sugar. An average tobacco nicotine and sugar value is assigned to each grade of tobacco for characterization of that grade and belt for a particular crop year. Additionally, other physical and often sensory characteristics are measured on the composite tobacco samples. As the sale of tobacco continues, the tobacco sold and processed at the stemmery come from stalk positions higher up the stalk. Flue-cured and Burley tobaccos from different stalk positions and belts have different physical and chemical properties. For example, lower stalk tobaccos are lower in nicotine compared to upper stalk tobaccos. Similarly, flue-cured tobaccos of similar stalk positions from the Florida/Georgia belt can be different in tobacco nicotine and sugar levels depending on the climatic and soil conditions compared to tobaccos grown in the Old Belt. During the tobacco market season, millions of pounds of tobacco will be graded, purchased, stemmed, analyzed (by composite samples) and placed in inventory. Each container of tobacco is designated by type, grade, belt and crop year. After the last lot of burley tobacco is placed into inventory, a complete picture of the results of the domestic tobacco-buying program for that year can be assessed and compared to the present inventory in light of the present and future needs of the company. The ultimate goal for the company’s buying program is to obtain an inventory of sufficient quantities of high quality tobaccos that can be used to maintain product quality and consistency.

A company’s tobacco utilization program and inventory management program are very important to the effective and efficient use the millions of pounds of tobacco in inventory, while maintaining product quality and consistency. The objectives of these programs are to improve long-term blend consistency, minimize blend variability caused by disruptions in tobacco flow (e.g., periodic fumigation of tobacco sheds, unusual environmental conditions, catastrophic events, (fire, water damage, etc), political unrest in various regions of the world), and allow manufacturing operations to make tobacco grade substitutions without making dramatic changes in product specifications. One type of plan used in the tobacco industry to accomplish this objective involves the grouping and blending of different types of tobaccos (e.g., flue-cured upper stalk or lower stalk tobaccos) having similar chemical, physical, sensory properties to create common grouped blends of tobaccos. This approach can be accomplished by the use of computer simulation modeling techniques and statistical cluster analysis. Ideally, grouped lots of tobaccos, evaluated based on similarities (i.e., grade [quantified by stalk position, quality / uniformity and maturity], sensory ratings, chemical parameters and physical parameters) from the entire tobacco inventory are analyzed and placed into common group clusters. Each common group cluster of tobaccos contains thousands of containers of tobaccos, which are located in inventory and designated by a specific grade, belt and year. A typical common group cluster, for example, upper stalk flue-cured tobacco, would contain tobacco from several grades, belts and years. Based on the company’s current blends, new product development needs, and future tobacco needs, the entire inventory can be clustered into several common group blends. Examples of common group blends might be: upper stalk flue-cured tobacco, lower stalk flue-cured tobacco, upper stalk Burley tobacco, lower stalk Burley tobacco, off-shore purchased flue-cured tobacco, off-shore purchased Burley tobacco, common blends of Oriental leaf, common blends of tobaccos used in the preparation of processed tobaccos, such as expanded or reconstituted tobacco sheet materials, or tobacco blends used in lower cost value cigarette brands. As the clusters are developed based on a set of criteria (grade [quantified by stalk position, quality / uniformity and maturity], sensory ratings, chemical parameters and physical parameters), it is possible to adjust the size (total available tobacco weight) and character of the common tobacco groups by weighting the different criteria. This is often important when tobacco crops differ greatly in criteria due to climatic and/or environmental changes that can occur in different growing seasons. Once adjustments to each common group cluster (in terms of size and common group character) are made to meet the needs of the company’s production schedule, tobacco blends containing set percentages of different common group tobaccos can be developed to meet the product specifications. A company to effectively and to efficiently use its tobacco inventories and to provide consistency in the physical, chemical and sensory parameters of its products can use computer simulation models, such as the one described. Additionally, it is hoped that the use of such models may allow for better product production and reduced waste.


In the 12 months from June 1997 to June 1998, the U. S. cigarette industry has produced 475.35 billion cigarettes (Maxwell Report, (64)). About 98 percent of the cigarettes prepared and sold were filtered and only about two percent were unfiltered. Cigarettes are classified by the Federal Trade Commission (FTC) based on “tar” yield per cigarette. The smoke nicotine and carbon monoxide (CO) yield per cigarette are also determined and reported by the FTC for each cigarette brand made and sold in the United States. Based on each cigarette brand’s FTC yield of “tar”, it is classified as full flavored (FF, greater than15 mg “tar”), fuller flavored low “tar” (FFLT, 7 to 15mg “tar”) or ultra low “tar” (ULT, 0 to 7mg “tar”). Within each brand category there are several hundred brands monitored by the FTC. Each company also monitors its brands for FTC “tar” and CO yield to be sure that it falls within the values advertised. The cigarette market is highly fragmented and competition between brands is fiercely competitive. Each company tries to gain a “competitive edge,” by both developing an understanding of current and emerging consumer wants and by translating those consumer wants into product variations. To gain an understanding of the success of lead products in each cigarette category, companies conduct competitive brand analyses in efforts to better understand their competition, the competitor’s product and reasons why smokers chose the competitor’s product. Interpretation of the entire area of competitive brand analysis is outside the scope of this chapter and will not be discussed in detail. Suffice it to say that one of the studies normally conducted in competitive brand analysis is a breakdown of a competitor’s product. An analysis of a competitor’s product can reveal differences in its blend or cigarette construction parameters (cigarette length and circumference, cigarette papers, filter types, adhesives, filter ventilation type and level, cigarette draft, etc) compared to other products. Cigarette blend analyses are routinely conducted. A typical set of blend analyses includes an evaluation of the blend composition by tobacco type (percentage of each type of tobacco) and an evaluation of each blend component’s level of tobacco nicotine and sugar. This information may provide clues to competitor’s use of high versus low stalk tobaccos in the blend. In formulating a blend to compete with the competitor’s successful product difficulties can be encountered. There are numerous ways to blend tobaccos to match a particular tobacco blend’s nicotine and sugar level. Not every blend of tobacco at a particular nicotine and sugar level will yield an acceptable smoke taste. The choice of tobaccos used in a blend is often directed by a company ‘s tobacco inventory. It should be noted that in any particular growing season that the vast majority of the total U. S. tobacco grown is purchased by one of the major cigarette producers. The amount of tobacco that each manufacturer buys is proportional to the market share of each manufacturer. Therefore, each tobacco manufacturer is most likely to buy similar types of grades of tobaccos to ensure that its products remain consistent. If a company can not prepare a sustainable blend that is consumer acceptable and that has a relatively constant tobacco nicotine and sugar level over long periods of time, its finished, marketed products will not be consistent to the consumer over time. The results of this inconsistency may well lead to the eventual rejection of product. Additionally, an analysis of the competitive product’s FTC “tar,” smoke nicotine and CO is determined. From these types of data it is often possible to piece together valuable information as to whether the success of a competitor’s product is product-based. It is possible to have a competitive product advantage based on other criteria such as name recognition or a particularly novel advertising or merchandising technique. If a competitor’s competitive advantage is product-based, it should be understood that the product was designed based on a specific FTC “tar” category that has met the wants and acceptance criteria of smokers in that particular category. A company will often produce a prototype product for consumer evaluation that is based on product specifications that are similar to a competitor’s product while trying to maintain the company’s original product taste signature or product performance attributes. If successful, the company will try to market the new or improved product and will set product specifications for its manufacture. Typical product specifications for the new or improved product include: the type and percentage of each blend component based on common group clusters, flavoring and/or casings to be applied to the blend, the cigarette length, circumference, weight, level and type of filter ventilation, the cigarette paper and filter specification, cigarette tipping specifications, and packaging specifications. The total set of product specifications is specified in order to compete with the taste and cigarette performance characteristics of the consumer-successful cigarette prototype and not necessarily to match the product characteristics. Neither tobacco nicotine content nor smoke nicotine yields are product specifications. Tobacco nicotine is a response to the blend ratios of common group blends of tobacco used together to produce a consumer acceptable smoke flavor. Smoke nicotine is a response to a consumer-acceptable combination of various levels of common group blended tobaccos placed into a specific cigarette configuration designed to generate a specific FTC “tar” and CO yield per cigarette.

A typical cigarette manufacturing process based on a predetermined set of product specifications is illustrated in Figure 9 (42). The process can be divided into several steps. Based on product blend specifications exact levels of different common group blended flue-cured, different common group blended Burley, Oriental group blends, reconstituted tobacco materials and possibly group blends of Maryland tobaccos, are metered from common group blended tobacco bulkers and conditioned separately to moisten and delaminate the tobaccos. Each tobacco type is fed to a separate feeder line. Some of the flue-cured and/or Burley tobacco can be shunted to a tobacco expansion process for future inclusion in the blend or can be added via a separate tobacco expansion process that continuously produces expanded tobacco for later inclusion in different blends. The Burley tobacco itself or any part of the tobacco strip blend can be cased with sugars and humectants. Flavors can be added to the strip blend or components at this point. After the various specified tobaccos have been conditioned, delaminated and possibly cased and/or flavored, they are placed into blending silos or bulkers. From the blending silos or bulkers the tobacco strip blend can have additional casing materials applied, if specified. If casing is applied at this phase in the processing, the tobacco is run through a dryer and then to a bulker to reequilibrate the blend to a specified moisture for cutting (usually 18 to 24 percent moisture). The blended tobacco strip is then cut to a specified cut width, usually ~ 25 to 35 cuts per inch). The tobacco cut-filler is then delaminated and dried. Specified levels of processed tobacco stems and expanded tobacco filler are then added in a mixing drum. If specified, the final blend, which contains specified levels of cut tobacco strip materials, expanded tobaccos and stems can be top-dressed with flavors. The finished cut-filler tobacco blend is prepared at specified moisture and is sent to one or more bulkers for use in the preparation of cigarettes.

Product Consistency in Terms of Tobacco Nicotine. The ability for a company to produce a consistent product day after day is often dependent on the quality, quantity and consistency of a company’s raw material and the company’s ability to effectively and efficiently process the raw materials into finished products. For the tobacco industry, a strategically directed tobacco purchasing plan and an effective tactically operated blending program are necessary.

R. J. Reynolds Tobacco Company (RJR) employs a tobacco blending system based on computer simulation modeling techniques and statistical clustering analysis. Its methods for effectively and efficiently utilizing their tobacco inventory are similar to those previously described. Estimates of blend nicotine variability have been calculated based on composite samples of tobacco collected and analyzed for tobacco nicotine during the stemming operation prior to placing the tobacco in inventory.

Target recipes for cigarette blends are composed of different percentages of common grouped RJR grades of tobacco strips. Within each common group grade of tobacco (for example, upper stalk flue-cured or lower stalk Burley strip tobaccos) there are various lots of tobacco differentiated by tobacco belt and crop year. While the lot percentages in a common group tobacco strip recipe can and do change due to crop availability, the common group grade percentages in a cigarette blend recipe remain almost constant. Each lot of tobacco is characterized by an averaged tobacco nicotine value, among other physical, chemical and sensory parameters. Thus, the nicotine variability within a common group of tobacco strips can be estimated. First, by determining the variance in nicotine of each grade across lots (weighting lots by the number of pounds in the lot) and then by combining the grades according to their target percentages to estimate the tobacco nicotine variability.

Similarly, tobacco blend cut-filler nicotine variability can be estimated by determining the weighted average percent tobacco nicotine variance of each common group of tobacco used in the final blend recipe. In the following example, the nicotine variability from sources other than flue-cured and Burley strip has been ignored.

Three major assumptions are included in the estimates that have significant effects: (1) Only one average nicotine determination is used for each lot of tobacco, thus ignoring within lot variability. This underestimates the common group of tobacco strip nicotine variability. (2) The weighted average variance of lots within a common group of tobacco strip implies that tobacco from a common group of tobacco strips is randomly selected for inclusion in the bulked final blend recipe. As random selection of tobacco containers is difficult and not practical, the random selection criteria is not true and as a result this tends to overestimate the bulked final blend recipe nicotine variability, particularly in the short run. (3) Nicotine variability from other cut filler components is considered non-appreciable, thus underestimating cut filler variability. The following tables illustrate how the nicotine variability is calculated.

Table 9 shows an example of how the weighted variance and standard deviation in percent tobacco nicotine is calculated for one hypothetical common group of flue-cured tobaccos (FC1). For this example five grades of flue-cured strip tobacco were selected. Each group is included into a common group of flue-cured tobaccos called FC1 at a particular percentage of the total weight of the common group of flue-cured strip. Each grade of flue-cured strip was previously analyzed for percent tobacco nicotine (along with other analyses) based on five to ten composite samples of tobacco collected during the stemming operation. A standard deviation in percent tobacco nicotine was determined taking into account the weight and size of each tobacco lot. A variance was calculated for each grade of flue-cured strip incorporated into the common group flue-cured blend of tobaccos. Finally, a weighted standard deviation of percent tobacco nicotine was calculated based on the sum of weighted variances from each grade used in the FC1 common group blend. In a similar way, Table 10 gives the calculated weighted standard deviation in percent tobacco nicotine for several other common group grades of tobaccos used in this example of how a blend recipe could be developed by RJR. Table 11 shows examples of two calculations of the total weighted standard deviation in percent tobacco nicotine for hypothetical blends of cut-filler tobacco based on the weighted standard deviation in percent tobacco nicotine of each common tobacco group used in the two hypothetical blends. From the calculations, the total weighted standard deviation in percent tobacco nicotine for blends of cut-filler tobacco are in the range of about 0.22 percent tobacco nicotine. If a tobacco company could maintain such a low variance in percent tobacco nicotine day after day their product consistency (at least in terms of tobacco nicotine) would be quite good.

Janjigian, et al. (60) and Gordin, et al. (61) have conducted studies on the level of tobacco nicotine necessary to elicit a change in sensory response by a certain percentage of a smoker population. These studies employed the classic sensory methodology called “just noticeable difference” (JND). The most conservative estimate of the JND (JND₁₀) for tobacco nicotine was a -0.27/+0.22 percent change in percent tobacco nicotine. They also found that there was a high correlation between tobacco nicotine and smoke nicotine for a single cigarette configuration (specific cigarette wrapper, blend, packing density, filter type, filter ventilation, etc.). Based on the calculated total weighted standard deviation in percent tobacco nicotine for blends of cut-filler tobacco it would appear that if a company could maintain such a low variance in percent tobacco nicotine, then its smoke nicotine within a single cigarette configuration would be expected to be consistent over time.

R. J. Reynolds Tobacco Company, like most companies, has a quality-monitoring program. Data in Table 12 shows the percent tobacco nicotine from four products made numerous times during 1996. The percent tobacco nicotine data of Table 12 represents only one of numerous analytes evaluated in their quality-monitoring program. The percent tobacco nicotine data are from a non-filtered 70mm cigarette, a 85mm filtered full flavor product, a 85mm filtered low “tar” product and a 85mm filtered ultra low “tar” product. The average percent tobacco nicotine and standard deviation for the four products were 2.37, 0.065; 1.92,0.078; 2.03,0.068 and 2.07,0.065, respectively. For each cigarette type there was a relatively narrow range in percent tobacco nicotine. The method for the analysis of percent tobacco nicotine for the data of Table 12 was a high volume colorimetric analysis and was used for all the data of Table 12. The blend recipes for these cigarette brands were all calculated using RJR’s tobacco blending model. The model worked well in selecting tobaccos for each blend over one year. The blend nicotine for the four tobacco blends were statistically different (at the 95 percent confidence level) among all comparisons other than the blends for the 85mm filtered low “tar” product and the 85mm filtered ultra low “tar” product which were not statistically significantly different. Figure 10 is a graphical representation of the data.

Over an extended period of time, cigarette brand styles and configurations tend to change to meet changes in the market (changes in consumer wants, competition, changes in advertising, etc.)(69). Additionally, the tobaccos employed in those brands vary based on environmental changes (hot-dry versus cool-wet growing seasons). Table 13 gives data on the percent tobacco nicotine and smoke nicotine for a single RJR cigarette brand from 1978-1994. Each data entry is a yearly average of at least three or more data points for both percent tobacco nicotine and smoke nicotine obtained from RJR’s quality-monitoring program. The method for the determination of percent tobacco nicotine was a high volume colorimetric analysis. Over the years, laboratories and personnel at RJR have changed and improvements have been made to both the equipment and methodologies for the determination of percent tobacco nicotine. The data of Table 13 although precise may have some bias due to the changes and improvements that were made over time. Smoke nicotine was determined by the official FTC methodology. The average and standard deviation for percent tobacco nicotine and smoke nicotine from data on Table 13 are 1.88, 0.19 and 0.77,0.09, respectively. The range in percent tobacco nicotine was 1.60 to 2.17percent. The range in smoke nicotine was 0.92 to 0.64 mg/cigarette. Figures 11 - 13 show the relationships of percent tobacco nicotine over time, smoke nicotine over time and smoke nicotine to percent tobacco nicotine. Numerous blend and cigarette configurational changes were made to this brand over the 16-year period of time evaluated. As expected, the percent tobacco nicotine and the smoke nicotine varied. These changes are evident from the low linear correlation (R² values) seen in the data of Figures 11 - 13. Yet it is interesting that on average percent tobacco nicotine remained relatively flat, albeit highly variable. On average smoke nicotine was reduced from 1978-1994. There is no linear relationship of percent tobacco nicotine and smoke nicotine over the period from 1978-1994.

Data on a 70mm non-filtered RJR product was collected from numerous sources over a period of 44 years. Over that period of time, numerous cigarette design technologies were introduced into cigarette products (69). In the early 1950s filter tips and reconstituted tobacco materials were introduced. In the late 1950s and early 1960s paper additives and porous paper were introduced. In the late 1960s and early 1970s expanded tobacco and filter ventilation were developed and used to varying degrees in cigarette products. All technologies used in the cigarette are interactive. The cigarette is a system of components that interact to produce changes in smoke chemistry. Inclusion of, or even changes in the level or type of technology, may require adjustments in the cigarette design to maintain certain smoke characteristics acceptable to consumers (62). During this period of time numerous improvements were also made in the analytical determination of nicotine in tobacco and smoke (63). Since the data of Table 14 is on a non-filtered cigarette, filter-related technologies obviously have not impacted smoke deliveries, but the other paper and processed tobacco technologies have been implemented into the 70mm product over time. Table 14 lists the “tar” and smoke nicotine yields and methods of analysis used to determine both. References are cited that contained the data in Table 14. Prior to 1966, numerous analytical methods were used to determine the yields of “tar” and nicotine in cigarette smoke. Some of the cited data are not actually “tar,” as we consider it today, but are measures of wet or dry particulate matter. After 1966 the data on Table 14 lists “tar” and smoke nicotine values obtained employing the FTC method. Figure 14 is a plot of smoke nicotine versus time. As can be seen there has been a large decrease in smoke nicotine over time. This initial decrease was due to the implementation of reconstituted tobacco, use of porous papers and small changes in the size (circumference) of the cigarette rods. Additionally, there was competition among cigarette manufacturers to reduce “tar” and smoke nicotine to satisfy the wants of their consumers’ (69). Since about 1960, the “tar” and smoke nicotine for this 70mm non-filtered cigarette has changed very little, comparatively. One obvious change that did occur from evaluation of the figure was an increase in smoke nicotine during the later half of the 1970s. This increase in smoke nicotine was due to the very high nicotine levels in both the flue-cured and Burley crops beginning in 1975 through 1979. In 1977 it is interesting to note that the mean nicotine for the flue-cured crop was higher than the Burley crop, 3.46 versus 3.33 percent, respectively, Table 5. Figure 15 shows the FTC smoke nicotine data for the 70mm non-filtered cigarette during the time period between 1966 to 1997. On average, the smoke nicotine has been relatively constant although wide swings in the data set have been noted (1977-1983).

Summarizing, it is critical for tobacco manufacturers to be able to produce consistent products for consumers. The raw material, tobacco, is a natural product and as such can vary considerable for year-to-year. To maintain consistent product quality, the manufacturers have limited means available to process or remove material inconsistencies. For the cigarette industry, the use of a strategic buying program coupled with inventory modeling programs has been an effective and efficient way to use tobacco inventories and maintain consistent tobacco nicotine levels in their products from year to year. Additionally, cigarette design technologies have been used to provide a means for cigarette manufacturers to provide consistent yields of “tar” and smoke nicotine. With the use of effective and efficient tobacco buying practices, tobacco use modeling efforts and use of advances in cigarette design technology, it is possible for a cigarette manufacturer to reproduce reasonable consistent levels of tobacco nicotine and smoke nicotine in their products year after year.

V. Regulatory Aspects

Tobacco is widely regulated both in the United States and abroad. Historically in the United States, tobacco has been regulated by the federal government to classify products for the purpose of taxation (Bureau of Alcohol, Tobacco and Firearms (70)), to provide cigarette smokers with information about cigarette smoke yields in advertising (Federal Trade Commission (71)), and to promote health education (Department of Health and Human Services (72,73)). Recently, regulations have been introduced in some states that require tobacco companies to conduct new testing methods and to report smoke yield classification schemes for cigarettes and other tobacco products (74-76). State regulations are based on the premise that smoke yields for “an average smoker” can be determined. Therefore state regulations do not focus on issues of taxation or product advertising, but rather focus on issues within the broad umbrella of the smoking and health controversy. Specifically, current state regulations derive from nicotine addiction arguments and, tangentially, human smoking behavior research.

The determination of tobacco nicotine content has been a key element in regulatory schemes applied to tobacco products. Nicotine analyses conducted for this purpose can be divided into two general categories: (1) analyses conducted in government laboratories and (2) analyses conducted in industry laboratories. Of note is the fact that the former are conducted at the expense of the taxpayer and the latter are incurred as a corporate expense. This section of the chapter will describe tobacco nicotine determinations used in the classification of tobacco products for federal taxation purposes. The analytical methods applied to classify products according to state guidelines and to meet state regulatory compliance will also be described and discussed.

Analyses Conducted in Government Laboratories: Classification of Tobacco Products for Tax Purposes – Bureau of Alcohol, Tobacco and Firearms Methods

Determination of Nicotine in Alcohol Tobacco and Firearms (ATF) Regulated Tobacco Products. The Bureau of Alcohol, Tobacco and Firearms (ATF) was established as an independent agency within the U.S. Department of the Treasury on June 6, 1972. On July 1, 1972, under Treasury Department Order 221, the functions, powers and duties of the Internal Revenue Service arising under laws relating to alcohol, tobacco and firearms were transferred to the newly created bureau where they presently reside. The primary responsibility of ATF is to regulate the alcohol, tobacco, firearms and explosives industries. ATF’s mission in connection with the tobacco industry is as follows:

1. 
   To secure voluntary compliance with law and regulations.
   

2. To exercise adequate controls to assure full and timely payment of the taxes imposed on tobacco products.
3. The revenue derived from these taxes is the primary consideration in regulating the tobacco industry.

The Federal taxes on manufactured tobacco products are among the oldest enacted legislation in the United States and are still in use today (77). Historically, there has been nearly two hundred years of federal tobacco legislation, regulation and taxation. Table 15 illustrates key dates from 1791 to 1988 in the regulation and taxation of tobacco. Despite the current decrease in the quantity of cigarettes sold, federal excise taxes from cigarettes rose to a total of $4.7 billion for 1990. The Federal revenue collected from the sale of all tobacco products amounted to $6 billion in 1997.

The mission of the ATF laboratories (ATF National Laboratory Center, Rockville, MD) is to produce the accurate and authoritative scientific information needed by law enforcement and compliance operations in protecting the public and collecting the revenue. The laboratory services provide scientific expertise to the bureau. Tobacco laboratory analysts perform a wide variety of tests to assist in consumer protection and proper tax classification of tobacco products.

Commercial tobacco products are defined (Table 16) and classified at different tax rates (Table 17) according to the Code of Federal Regulations (70,78). ATF regulated tobacco products (Figure 16) can be classified into two groups: Smoking tobacco products and Smokeless tobacco products. Smoking tobacco products include cigarettes, cigars, pipe tobacco and roll-your-own (RYO) tobacco. Smokeless tobacco products include chewing tobacco and snuff. Product classification is performed on a comparative basis: cigars vs. cigarettes, pipe tobacco vs. roll-your-own tobacco, and chewing tobacco vs. snuff. For example, the tax on cigarettes is approximately ten-fold greater than the tax on cigars. The current tax rate is $12.00 per thousand for small cigarettes compared to $1.125 per thousand for small cigars (Table 17). Identification of a tobacco product(s) relies on a wide variety of tests including a determination of nicotine.

The first step in defining and properly classifying an ATF regulated tobacco product is the determination of nicotine (Figure 16). The rationale for analyzing nicotine in ATF regulated tobacco products, regardless of the type of tobacco product, is two-fold:

1.