IN FLUE-CURED TOBACCO
David M. Peele, Marvin G. Riddick, and Mike E. Edwards, R.J. Reynolds Tobacco Company, Avoca Division, Merry Hill, NC, 27957; and Jeff S. Gentry and Timothy B. Nestor, R. J. Reynolds Tobacco Company, Winston-Salem, NC, 27102.
Summary
Nitrosation of naturally occurring tobacco alkaloids is commonly accepted as the mechanism of tobacco specific nitrosamine (TSNA) formation in tobacco. As minimal-to-no TSNA are present in green tobacco, TSNA formation and accumulation are generally considered to occur during curing. Most recent hypotheses focus on microbial reduction of nitrate to nitrite and other oxides of nitrogen (NOx compounds) that react with tobacco alkaloids to form TSNA during curing. This natural microbial process remains the prevalent hypothesis for TSNA formation in cured burley tobacco. However, a second mechanism for the formation of TSNA, independent of microbial activity, is presented in this work on flue-cured tobacco.
It is common practice to flue-cure Virginia or blonde tobacco in bulk barns that incorporate forced air ventilation and temperature control. Modern bulk barns generally use liquid propane gas (LPG) with direct-fired burners that exhaust combustion gases directly into the barn where the tobacco is exposed to those combustion gases. Our studies indicate that LPG combustion by-products, namely NOx (in the exhaust stream of direct-fired, flue-cured barns), react with naturally occurring tobacco alkaloids to form TSNA. Heat exchange curing methods preclude exposure of the tobacco to combustion gases and by-products, thereby eliminating this significant source of TSNA formation, without degrading leaf quality or smoking character. Research findings from 1998 and 1999 will be presented to demonstrate the significance of direct-fired curing as a source of TSNA formation in flue-cured tobacco.
Introduction
For decades, nitrosamines have been a focus of considerable research for which references abound in the scientific literature. The presence of nitrosamines in cigarette smoke and their possible role in tobacco carcinogenesis were first postulated in the scientific literature in 1962. Decades of research have provided no basis for concluding cause and effect relevancy of nitrosamines to chronic disease in smokers,2,3,4,5 despite implied contentions to the contrary.6 Without such relevancy, the effect of reducing levels of nitrosamines in tobacco or tobacco smoke cannot be definitively asserted. However, responsible product stewardship advocates that technologies to reduce nitrosamine levels in tobacco be pursued and implemented as proven effective and commercially practicable.
Extensive research by many investigators since the 1970's reported the presence of as many as thirty-five (35) nitrosamines in cigarette smoke. A subset of those nitrosamines was considered to be specific to tobacco, resulting from nitrosation of endogenous tobacco alkaloids. Today, eight (8) tobacco specific nitrosamines (TSNA) have been identified and seven (7) have been reported as present in cigarette smoke.6,7,8 The breadth of TSNA investigations over the last three (3) decades spans a wide range of topics including formation and accumulation of TSNA in tobacco, 9,10,11 quantitative analyses of TSNA in tobacco and tobacco smoke, 2,3 epidemiology, animal studies, and possible chronic health effects of TSNA in smokers.12,13,14,15
Nitrosation of naturally occurring tobacco alkaloids is commonly accepted as the mechanism of TSNA formation in cured tobacco. N-nitrosonornicotine (NNN), 4-(N-methylnitrosamino)-1-(3-pyridyl)-1-butanone (NNK), N-nitrosoanabasine (NAB), and N-nitrosoanatabine (NAT) are the more prevalent TSNA in tobacco and tobacco smoke and, thus, receive the most attention in scientific literature and discussion. NNK is considered to be derived from nicotine, NNN derived from both nicotine and nornicotine, NAB derived from anabasine, and NAT from anatabine. While all four are generally found in flue-cured and burley tobacco, NNK is typically the most prevalent in flue-cured tobacco while NNN is the most prevalent in burley tobacco. TSNA are generally absent or below quantitative limits in Oriental tobacco.
A recent review summarized the work of many authors on the effects of fertilization rates, plant maturity at harvest, genotype, alkaloid level, and tobacco storage on the levels of TSNA in burley tobacco.6 However, little work has been reported on TSNA formation in flue-cured tobacco. It is generally agreed that minimal-to-no TSNA are present in green Virginia or burley tobacco. TSNA formation and accumulation are considered to occur post-harvest and most predominantly during curing.6,16.17,18,19 Most recent hypotheses on TSNA formation focus on microbial reduction of nitrate to nitrite and other NOx compounds that react with tobacco alkaloids to form TSNA during the curing process.6 The natural microbial process remains the prevalent hypothesis on TSNA formation in burley tobacco. However, direct-fired curing presents a second (and potentially more important) source of nitrosating agent in the flue-curing process. Precedent for this hypothesis exists in the literature from the milk and beer industries. 20,21The objectives of this research were to demonstrate the importance of LPG combustion by-products (namely, NOx) on TSNA formation in flue-cured tobacco and the relevance of that finding to direct-fired versus heat-exchanged methods of flue-curing tobacco.
Flue-Curing of Tobacco
It is common practice to flue-cure Virginia or blonde tobacco in bulk barns that incorporate forced air ventilation and temperature control. The flue-curing process typically takes from five to seven days and involves three sequential steps of yellowing, leaf drying, and stem drying.22 Figure 1 graphically depicts a typical flue-curing schedule in terms of relative humidity (RH) and wet/dry bulb temperatures. Generally, the yellowing phase requires 36 to 72 hours in which dry bulb temperature and relative humidity are maintained at approximately 35oC and 85%, respectively. During the next 48-hour leaf-drying phase, temperature is slowly ramped from 35oC to 60oC and relative humidity dropped from 85% to approximately 35%. The last 48-hour stem-drying phase is carried out at 75oC (following a gradual ramp from 60oC) and relative humidity allowed to drop to approximately 18%. Finally, the tobacco is allowed to reorder for post-cure processing. The quality of flue-cured tobacco is highly dependent on temperature and humidity control within the barn such that natural biochemical changes are allowed to occur while the process of drying the leaf and stem take place.
In the United States today, flue-cured tobacco is predominantly cured in commercial bulk barns. As a result of energy pressures in the United States during the 1960âs, farmer-built Îstick barnsâ with heat-exchanged flue systems were gradually replaced with more energy efficient bulk barns using direct-fired liquid propane gas (LPG) burners. These LPG direct-fired burner systems exhaust combustion gases and combustion by-products directly into the barn where contact is made with the curing tobacco. The vast majority of commercial direct-fired burners are generally capable of 100,800 to 151,200 kg-cal (400,000 to 600,000 BTU) per hour, resulting in a LPG consumption rate of approximately 570 to 1140 L (150 to 300 gal) per cure. Based on US Environmental Protection Agency emission factors, approximately 1.7 kg of NOx are produced per 1000 L of LPG combusted (14 lb NOx/1000 gal LPG). Consequently, the net effect is that the tobacco in LPG direct-fired bulk barns is exposed to approximately 1.0 to 1.9 kg (2 to 4 lb) of NOx per cure.
In stark contrast to direct-fired curing, heat-exchange burner configurations completely vent combustion gases and combustion by-products to the external atmosphere rather than into the barn. The heat-exchange process precludes exposure of the tobacco to LPG combustion by-products, particularly NOx, thereby eliminating an important source of nitrosating agent for TSNA formation.
Equipment, Materials, and Methods
A series of curing experiments were conducted in four small, 8-rack R&D barns capable of holding approximately 450 kg of green tobacco (or 65 kg of cured leaf). Two of these 8-rack R&D barns were fitted with 20 kilowatt, electrical resistance-heating units of 17,140 kg-cal (68,000 BTU) per hour capability. These electric barns serve to represent heat-exchange curing structures. Two similar barns were equipped with LPG direct-fired burner units of 18,900 kg-cal (75,000 BTU) per hour capability. Fans in each barn were sized to provide volume-adjusted air exchanges comparable to commercial bulk barns. Curing schedules were generally in accordance with that described in Figure 1.
Nitric oxide gas was purchased from Praxair Distribution, Inc. (Bethlehem, PA) as 99.5% nitric oxide. The gas was described as Product Number 2.5-K, Batch Number BP101697NX with certificate of analysis dated September 11, 1998.
The tobacco used for all controlled studies was mid- to upper-stalk Virginia tobacco (cv. K326). Samples were collected from curing structures to appropriately represent the entire barn in terms of box or rack location. All TSNA analyses in tobacco were performed in accordance with Risner, et al and reported in this work as the sum of NNN, NNK, NAT, and NAB.
Results and Discussion
In an initial experiment, primed tobacco
was loaded into an 8-rack electrical R&D barn, an 8-rack LPG direct-fired
R&D barn, and a direct-fired commercial barn. At the end of cure, tobacco
from each barn was sampled and analyzed for TSNA levels. The results are
summarized in Table 1. The electrical barn, absent an external nitrosating
agent from LPG combustion, demonstrated TSNA levels below detection. Both
the R&D and commercial direct-fired barns showed measurable levels
of TSNA. Approximate emissions of NOx from the direct-fired
barns were calculated from recorded LPG consumption and are also given
in Table 1.
| Table 1. Effect of Burner Configuration on TSNA Level. | ||||
| Barn type | Burner
Configuration |
TSNA
in
tobacco (ppm) |
NOx from combustion (kg) | |
| R&D Electrical | Heat-exchange | 0 | 0 | |
| R&D LPG | Direct-fired | 2 | 0.18 | |
| Commercial LPG | Direct-fired | 13 | 1.15 | |
Following the observation above on heat-exchange
versus direct-fired configurations, the effect of exogenous nitrosating
agents (expected to be present in the combustion stream of direct-fired
curing barns), was investigated. Tobacco was loaded into each of the four
8-rack R&D barns. Two of the barns were electrically heated while the
other two were LPG direct-fired. The tobacco was yellowed for 48 hours
per a typical flue-curing schedule. Into one electrical and one LPG direct-fired
barn, 1.8 kg of NOx gas were introduced during the last 24 hours
of the yellowing period. The remaining electrical and LPG direct-fired
barns were operated as control barns not receiving any NOx treatment.
The tobacco from each barn was sampled at the end of cure and analyzed
for TSNA levels. The data, summarized in Table 2, show that the TSNA levels
in tobacco from the LPG direct-fired control barn are significantly higher
than levels from the electric control barn. Furthermore, and more importantly,
the introduction of exogenous NOx substantially increased TSNA
levels found in the tobacco, regardless of barn type.
| Table 2. Effect of exogenous NOx on TSNA levels. | ||
| Barn Type | NOx Treatment (kg) | TSNA in Tobacco (ppm) |
| Electric (control) | 0 | 1 |
| Electric | 1.8 | 174 |
| LPG Direct-fired (control) | 0 | 5 |
| LPG Direct-fired | 1.8 | 107 |
Given the significant increase in TSNA levels observed upon the introduction of exogenous NOx, it was reasonable to expect a dose-related response of TSNA to NOx exposure. Tobacco was loaded into each of two electric R&D barns where no combustion source of NOx was present. As in the previous experiment, the tobacco was yellowed for 48 hours and a quantity of NOx was introduced into each barn during the last 24 hours of the yellowing period. One barn received a 0.45 kg (1 lb) NOx treatment while the other barn received a 1.8 kg (4 lb) NOx treatment. The tobacco was sampled just prior to any NOx treatment (time zero) and subsequently on a 12-hour schedule for the next 48 hours. Finally, the tobacco was sampled at the end of cure. TSNA levels in the tobacco were determined for each sampling interval (corrected for moisture) and are illustrated in Figure 2. As expected, a significant dose response was observed. The 1.8 kg exogenous NOx treatment resulted in TSNA levels at least an order of magnitude higher than those observed for the 0.45 kg NOx treatment. In addition to demonstrating a significant dose response, these data also show TSNA levels increased rapidly during the 24 hours of NOx treatment and continued to increase, albeit at a slower rate, over the life of the cure. This observation suggests that NOx is absorbed or adsorbed by the leaf and can continue to react with natural alkaloids to form TSNA.
In order to augment the controlled experiments
described above, a number of commercially producing farms, representing
varied agronomic, harvesting, and curing practices, were sampled to compare
direct-fired barns to heat-exchange barns. Table 3 summarizes the TSNA
levels found for the various barns sampled. The data demonstrate that the
TSNA levels for each heat-exchange barn configuration are significantly
lower than observed for LPG direct-fired barns.
| Table 3. TSNA Sampling from Various Commercially Producing Barns | |||
| Barn Type/Fuel | Burner Configuration | Number of Samples | TSNA in Tobacco (ppm) |
| Stick/Wood | Flues | 6 | 0.3 |
| Bulk/Diesel | Heat-exchange | 27 | 1.0 |
| Bulk/LPG (Turkey) | Heat-exchange | 23 | ND |
| Stick/LPG | Direct-fired | 1 | 5.9 |
| Bulk/LPG | Direct-fired | 43 | 11.0 |
During the '98/'99 off-season, 27 direct-fired
commercial bulk barns were converted to heat-exchange barns. To facilitate
the conversion, the direct-fired burner and forced air ventilation systems
were removed from the barns and new 100,800 kg-cal (400,000 BTU) per hour
heat-exchange burners and accompanying forced air ventilation systems were
retro-fitted in each of the commercial barns. All farms were using either
LPG or natural gas fuels. To date, cured primings and cutters have been
sampled from each of the heat-exchange barns and compared to similar direct-fired
barn curings from the same farms. The data are summarized in Table 4 and
show approximately 80% reduction of TSNA levels in tobacco from the heat-exchange
barns compared to tobacco from direct-fired barns.
|
|
|
| Burner Configuration | TSNA in Tobacco (ppm) |
| Heat-Exchange | 1.3 |
| Direct-Fired | 7.1 |
Conclusion
Data presented in this paper strongly support the contention that direct-fired curing provides combustion by-products (i.e., oxides of nitrogen) that contribute significantly to the formation of TSNA during the flue-curing of tobacco. Direct-fired curing was shown to be the primary source of TSNA formation in flue-cured tobacco, significantly overshadowing microbial-mediated mechanisms. Heat-exchange curing practices were shown to significantly reduce TSNA levels in flue-cured tobacco. Also important, conversion to heat-exchange curing does not alter the integrity, quality, and smoking characteristics of flue-cured tobacco since the basis of the flue-curing process is not altered by the change in burner configuration.
Furthermore, work at RJR has shown no significant
TSNA formation or accumulation during proper storage or aging of flue-cured
tobacco used for cigarette manufacture. Similar works by Andersen
and Burton on air-cured tobacco support this finding. 25,26
Consequently, TSNA reductions accomplished through heat-exchange curing
of Virginia tobacco are not compromised by proper storage and aging prior
to cigarette manufacture.
1 Druckerey, H. and Preussmann, R. The formation of carcinogenic nitrosamines in tobacco. Die Natur., 49: 488-499 (1962).
2 Witschi, H., Espiritu, I., Maronpot, R.R., Pinkerton, K.E., and Jones, A.D. The carcinogenic potential of the gas phase of environmental tobacco smoke. Carcinogenesis, 18: 2035-20