RJ Reynolds
The Tobacco Chemists' Research Conference: a Half Century Forum Advances in Analytical Methodology of Tobacco and Its Products.
Fields
- Type
- REPORT
- Alias
- R 11117
- Site
- R&D
- Biochem Biobehavioral
- Rodgman Ar
- Sr Staff Scientist
- Rodgman Post '89 Docs
- Named Organization
- American Health Foundation
- List of Philip Morris Awardees
- Assn of Official Analytical Chemist
- Univ of Ky Tobacco & Health Researc
- List of Smoking & Health Authors
- Intl Ag for Research on Cancer
- Stanford Research Institute
- American Machine & Foundry
- List of Organizations
- Dept of Agronomy
- Univ of Ky
- Natl Cancer Institute
- Eastern Regional Research Laborator
- Tobacco Science
- Sloan Kettering Institute for Cance
- Medical College of Va
- List of Researchers
- List of Affiliations
- Japan Tobacco Salt
- Natl Institute of Occupational Safe
- Registry of Toxic Effects of Chemic
- Plenum Press
- Academic Press
- Request
- Mitchell Int27
- Burton by Agreement
- US Priority Request 2
- Wallace 1rfp14
- Wallace 1rfp12
- Wallace 1rfp10
- Wallace 1rfp11
- Wallace 1rfp9
- Wallace 1rfp7
- US Comprehensive Request 447
- US Comprehensive Request 130
- US Comprehensive Request 38
- US Comprehensive Request 103
- US Comprehensive Request 174
- US Comprehensive Request 80
- US Comprehensive Request 46
- US Comprehensive Request 326
- US Comprehensive Request 37
- US Comprehensive Request 47
- US Comprehensive Request 264
- US Comprehensive Request 86
- US Comprehensive Request 425
- US Comprehensive Request 251
- US Comprehensive Request 263
- US Comprehensive Request 42
- US Comprehensive Request 241
- US Comprehensive Request 249
- US Comprehensive Request 201
- US Comprehensive Request 253
- Wallace Court Order 20010629
- Longden Int3
- Longden Int2
- Longden Int7
- Willard Brown 1rfp15
- Willard Brown 1rfp22
- Willard Brown 1rfp19
- Willard Brown 1rfp54
- Willard Brown 1rfp56
- Willard Brown 1rfp58
- Wallace 8rfp40
- Named Person
- Tcrc
- Parmele, H.B.
- Jeffrey, R.N.
- Rjr
- Kosak
- American
- Ftc
- Philip Morris
- Iso
- Coresta
- Usda
- Lindsey
- Hicks, G.W.
- Atf Natl Laboratory
- American Chemical Society
- Grob, K.
- Ishiguro
- Sugawara
- Guerin
- Fieser, L.F.
- Newman, M.S.
- Cook, J.W.
- Lacassagne, A.
- Buuhoi, N.P.
- Clar, E.
- Surgeon General
- Snook
- Willard, R.L.
- Furman
- Kennaway
- Lamphier, D.R.
- Roffo
- Groffith
- Dalhamn
- Hiller
- Rodgman, A.
- Keith
- Procter
- Epa
- Harke
- Maskarinec
- Neurath
- Esquier
- Hee
- Risner
- Jenkins
- Mait
- Fischer, K.
- Deluca
- Demain
- Brahms
- Beethoven
- Bach
- Benner
- Burdick
- Burton
- Touey
- Mumpower
- Information Resources
- Brice
- Willits
- Analytical Chemistry
- Crowder
- Bates
- Shell Chemical
- Martin
- Williams
- Schumacher
- Wolfrom, M.L.
- Oh, S.T. Univ
- Severson
- Adams
- Rayburn
- Davis
- George
- Kensler
- Resnik, F.
- Spears, A.
- Lorillard
- Carpenter
- Holmes
- Seligman
- Wolff
- Bgsm
- Hoffmann
- Newsome
- Liggett
- Stedman
- Schlotzhauer
- Schmeltz
- Cook, L.C.
- Mckennis
- Bowman
- Cundiff
- Glock
- Griffith
- Celanese
- B.&W.
- Nc, S.T. Univ
- Ornl
- Chung, H.
- Kelly, S.
- Oldaker, G.B. III
- Tabakforsch Intl
- Hhs
- Chem Comm
- Japan Monopoly
- Wake Forest Univ
- Characteristic
- Marginalia
- Box
- Rjr7010
- Author
- Green, C.R.
- Rodgman, A.
- Rjr
- Brand
- Premier
- UCSF Legacy ID
- zcc63a00
Document Images
I
1
I
I
I
1
I
a
I
first question than to answer the second. For instance, the 2nd and 14t1i volumes of
Recent Advances in Tobacco Science are dedicated to discussions of "Leaf composition
and physical properties in relation to smoking quality and aroma" and "Chemical and
sensory aspects of tobacco flavor," respectively. For example, scientists from American
Tobacco presented papers documenting the detrimental character of nornicotine towards
acceptable smoke flavor (11/03,16/22). Much of the analytical work relating chemical
composition to consumer preference may be proprietary and not available in publication.
The second major goal in tobacco science analytical methodology has been the
search for components of tobacco and its smoke that may be responsible for the human
health risks associated with tobacco usage. TCRC's record is filled with papers
attempting to isolate the "culprit du jour," e.g., benzo[a]pyrene (BaP), other polycyclic
aromatic hydrocarbons, aza-arenes, aldehydes and ketones, nitrogen oxides, N-
nitrosamines, aromatic amines, chlorinated hydrocarbons, free radicals, etc. We have the
so called "List of 43" tumorigenic agents in tobacco and smoke prepared by Hoffmann
and Hecht (47). In terms of human lung cancer, this list has its limitations because most of
the components have not been shown to be 1) tumorigenic to any human tissue or 2)
tumorigenic to lung tissue in laboratory animals (48). Only 5 of the 43 compounds,
benzo[a]pyrene, N-nitrosodimethylamine, N-nitrosodiethylamine, cadmium, and polonium-
21Q, have produced lung tumors in animal experiments but these studies were conducted
at massive dose levels relative to smoking or ETS exposure. Results of the BaP, N-
nitrosodimethylamine, and N-nitrosodiethylamine inhalation studies are rated as
"equivocal" by RTECS (49).
At least two points need to be made about this search for toxic agents: 1) the
levels of reported toxic agents and 2) the presence of constituents with anti-toxic
properties. About the first, Wakeham (46) remarked,
...Some analysts with sophisticated methods are no, w finding chemicals in
fraetional nanogram quantities, o, ften ignoring the fact that toxicity is a
fanction of concentration. They are looking for the needle in the
haystack, as it were, to prevent the cow Jrom eating it.
14
52118 5214
l
I
I
I
I
.
..~-
Tobaceo
. -Smoks
A B
SO
55 60 65 70 75 80 85
90 95
Legaro Year
eDM)&Mte Qsttt eoehticaLietboft
x
.. . A pria to 19Si
B 19631960
C 19641970
l D 1970 b mid-
E 1970s
mid-1970s to dats
I
I
1
The tremendous escalation in the number of identified tobacco smoke (and
tobacco) components from 1954 to-date was achieved by successive advances in analytical
methodology, particularly the technologies concerned with the separation and
identification of individual components in complex mixtures. In Fig. I is shown the
increase in the number of identified tobacco smoke components from 1950 to date.
Also noted in Fig. I are the approximate dates when the major analytical advances were
implemented in the study of tobacco smoke composition.
Figure 1
Number of Identified Tobacco Smoke Components Reported since Kosak (40):
Accumulative by Year
'alsssleal' dwnical tactniques
adumn chramatopraphy
gas chromampraphy
plass cap5lary pa eMomatopaphy
aapbd with maea speckametry
derkatives ta hiph readuion pas
chromatopraphy (HR6C), HPLC,
- masa spedrom.try
Prior to 1950, the major part of the isolation and characterization of individual
components from tobacco smoke involved so-called "classical" chemical techniques, i.e.,
the fractionation of tobacco extracts or cigarette smoke condensate (CSC) into neutral
(aliphatic hydrocarbons, PAHs (polycyclic aromatic hydrocarbons), and esters), acidic
(acids and phenols), and basic fractions (amines, nicotine-related alkaloids), followed by
crystallization and/or distillation of these subfractions.
In the early 1950s, liquid column chromatography on alumina, silicic acid, or
Florosil® of the neutral or acidic fractions permitted further separation of components
10
1 52118 5210
I
I
I
l
.l
I
I
I
I
I
Examination of, Fig. I indicates that, as each analytical, separation, and
identification technology was developed and utilized, the number of smoke components
identified per year increased.
It is realized that various investigators who pioneered an emerging analytical
technology were often involved with the development and/or use of the technology prior
to the period indicated in Fig. 1. No slight of their noteworthy contributions is intended.
The periods indicated in Fig. I are those when the analytical technology in question was
sufficiently advanced and used by almost all investigators involved, in the analysis of
tobacco, tobacco smoke and/or definition of their composition.
The next dramatic break in the number vs. year plot occurred in the late-1970s
when the glass capillary' gas chromatograph was coupled directly to the mass
spectrometer. We recall that at this time there was a good-hearted argument between
mass spectroscopists and chromatographers as to whether the mass spectrometer was just
another GC detector or the gas chromatograph was a sophisticated inlet device for the
mass spectrometer. Present day mass selective detectors appear to support the
cluomatographers' position. Regardless, this mating of technologies permitted separation
of the components of a particular tobacco or smoke fraction or subffaction and
determination of the molecular weight and/or fracture pattern of each component or
mixture of closely related components immediately after they exited the chromatograph.
Interpretation of the data thus obtained, occasionally in concert with findings from W,
IIt, and/or NMR studies, permitted very rapid and unequivocal identification of smoke
components. Even with extremely low amounts of a particular smoke fraction or
subfraction, definitive chromatograms and MS data may be generated on several hundred
chromatographic peaks, each of which may represent an individual compound or a mixture
of several compounds.
An outstanding example of the use of GC-mass spectrometry in the definition of
tobacco smoke composition is the study by Snook et al. (28/40, 45) on the PAHs in
cigarette smoke. Nearly 1,000 individual PAHs, their homologs, and oxygen analogs were
' The term glass capillary is used in tlx generic sense to include Posed silica columns.
12
52118 5212
I
.I
I
I
I
.1
1
f
J
J
reported (Table II), but many of the identifications were subsequently shown to be
incorrect.
By use of various analytical methodologies, the identities of many components
cataloged by Kosak in 1954 (Table II) were resolved. Of the approximately 80 entries,
the identities of 33 (over 40°/a) of the components were questioned by Kosak because he
did not "consider the evidence cited in the literature to be definitive proof' of their
identities. Two of the listed items (benzopyrene, "condensed aromatics") were reported
by Roffo (36) who did not study tobacco smoke generated by a process approximating
human smoking but studied a"desttvctive distillate" from tobacco. Several of the
components (a-, (i-, and y-socratine,;obelin, lohitam, anodmin, Iathraein, poikiline, and
gudham) first reported by Wenusch and Schbller (37) and listed under "Alkaloids" were
subsequently demonstrated to be mixtures or a component Gsted elsewhere in Table 11.
Foe;example, Kuffher et al. (38) demonstrated that obelin is a salt of ammonia, a- and Q-
socratine are mixtures of nicotyrine and 2,3'-bipyridine, and y-socratine is !-nornicotine.
Poikiline is identified as 4-amino-l-(3-pyridyl)-butanone.
And finally, the complexity of either tobacco or its smoke presents a professional
challenge to the analytical chemist. Analytical chemists who can master the difficulties of
separating, identifying, and quantifying components of tobacco and smoke are masters of
analytical technology.
Whatever the reason for our intense scrutiny of tobacco and smoke, there appears
to be a relationship between the techniques applied and the numbers of compounds either
quantified or merely identified. In 1958, RJRT R&D personnel cataloged from in-house
research and literature reports 230 identified compounds in tobacco and 202 identified
components in cigarette mainstream smoke; 54 components were common to both 6sts.
The RJRT R&D compilation of tobacco smoke components has been revised periodically,
more recently in a computerized catalog: The number of identified and cataloged tobacco
smoke components is about 3,900; the number of identified tobacco components is about
3,800. Relative to tobacco composition, about 2,800 components are unique to tobacco
smoke. For the purposes of Fig. 1, we estimate that there are about 4,800 known
components in tobacco smoke (39).
8
52118 5208
I
I
deteated; and many were identified unequivocally. These results should be compared with
the ridiculous controversy that raged in the early 1950s as to whether PAHs were indeed
present in cigarette mainstream smoke.
The early- to mid1970s also saw the emergence of high performance liquid
chromatography (HPLC), a highly efficient and effective variation of liquid column
chromatography. See reviews by Bell (4), Haeberer and Chortyk (5), Green et al. (8), and
Court (14). HPLC made possible the identification and quantification of many
compounds, e.g., heat sensitive and ionic substances, that are difficult to analyze by other
means.
From the existing research on tobacco and smoke. composition, one can only
conclude that both are very complex mixtures requiring the most advanced analytical
technology to completely characterize their makeup. There are currently about 6,700
chemical substances known to be present in tobacco and smoke. Wakeham (46) as early
ac"1971 opined that as many as a hundred thousand chemical components may exist in
tobacco smoke alone. Numerous authors have implied that this is a unique property of the
tobacco plant and its smoke. However, this known complexity is more likely not unique
but due to the intense analytieal scrutiny to which this valuable and controversial
commodity has been subjected.
"HOLY GRAII S" OF TOBACCO RESEARCH
When 50 years of TCRC analytical research on tobacco and smoke are viewed, it
is obvious that the objective of this work is an attempt to answer one or both of the
following questions:
1. What components of tobacco and its smoke are important to the flavor,
sensory experience, and consumer acceptance of tobacco products?
2. What components of tobacco and its smoke may be responsible for the
human health risks ass6ciated with tobacco usage?
Unfortunately, the brief answer to both of these questions is that we still do not know.
. J.
More has been written and discussed within the TCRC forum in an attempt to answer the t+
S.-.
13 . ~
52118 5213
a
I
~
I
_1
.1
I
I
~
I
J
"mellowing" effect on the smokability of the tobacco. TCRC audiences have heard many
reports on the chemical changes that occur during aging.
Filling power of purchased or processed tobaccos has great impact on the
economics of cigarette production because cigarettes are sold by unit rather than by
weight. The seemingly simple analytical methodology of determining the volume occupied
by a known mass of tobacco under a defined pressure has led to numerous TCRC
presentations, cf. 12/04, 25/04, 30/3 t.
Manufacturers and suppliers to the tobacco industry use analytical technology to
ensure that they provide a consistent product to their consumers. Likewise they have a
product stewardship responsibility to know the ingredients in their products and that these
meet all regulatory standards. Manufacturers need to understand how their products work
so that new, improved products that are more economical or meet market-place demands
can; be developed. Learning based upon analytical methodology is the key to this
understanding. Additionally, manufacturers and suppliers use analytical results to defend
themselves i"n litigation and support advertising claims for their products.
Regulators utilize analytical methodology to assess the safety of products, inform
consumers and the public, and classify various tobacco products for taxation purposes.
The FTC "tar" and nicotine smoking methods are examples of analytical methodology
useAj)y regulators (22).
jjISTORICAL REVIEW
A review of the TCRC literature reveals that 2,008 titles are recorded for
presentation.a Aside from the reviews, nearly every paper given at the TCRC either
presents new or uses existing analytical methodology. Our count indicates that at least
779 papers describe analytical technology directly and 307 additional papers are concerned
with isolation and identification techniques. Practically everyone who has conducted
a Nthough 2,008 titles are recorded, an unknown, but few, number of papers have been withdrawn at
the
last minute. The record is unclear on the precise number but it is believed to range from 5 to 10
papers.
This is a remarkable record.
21 ,
52118 5221
I
I
I
J
I
I
I
.l
Sugar, total nitrogen, nicotine, total alkaloids, and protein levels are examples of targets
for the development of new tobacco cultivars. See Mendell et al. (54) for additional
factors.
Once tobaccos reach the auction floor, manufacturers want to know more about
the physical and chemical characteristics of their intended purchase to achieve product
consistency. However, it is rarely possible to conduct analyses before tobacco purchase.
For this decision, we must rely on the art of experienced buyers. Several attempts have
been made to relate the color (10/25, 12/01, 13/04), feel, and odor clues used by the buyer
to analytical results. Numerous studies presented at the TCRC have attempted to
correlate tobacco volatiles obtained by steam distillation or headspace analysis to tobacco
quality. These studies have met with only modest success.
Table V
Reasons for Application of Analytical Technology to the Study of Tobacco Products
.y;
Area Need
Agronomy - Chemical characterization and breeding to obtain the greatest profit
Agronomic chemical residues, e.g., fertilizers, pesticides, herbicides
Production research, e.g., curing methods, sucker control
Alternative uses of tobacco e. ., protein production 55
Buyers Purity, e.g., sand
Usabilit e. ., fines stems, fillin power
Maniifacturers Quality assurance
and Suppliers Product stewardship
Product development
Product understanding
Regulatory requirements
Pest control
Product defense
Regulators Safety
Advertising
Consumer information
Taxation
Once purchased, tobacco is subjected to chemical and physical tests. Among these
are moisture, sand, stems, and Iamina content measurements. Green tobaccos, i.e., new
crop tobaccos, are usually stored for months to years because this aging has a
20
,, 52118 5220
I
-1
I
I
I
I
1
adapted hot wire anemometer for measurements of the velocity and duration of an inspired
puff. For the regular length unfiltered cigarettes reported in this study, average
characteristics for the 120 subjects were a 1.9-sec. puff, taken twice a minute, with a 44-
cm3 puff volume. Adams (20/31) reported on U.K. studies with tipped cigarettes that
employed a flow measuring orifice and found values of 31-cm'volume, 1.4-sec. duration,
and 1.8 puffs per minute. Gottscho (27/16) presented results of an international study on
smoking parameters of cigar smokers. They took on average 20-cm' puffs. In 1982,
Griffith (36/35) presented details of methodology used to characterize human smoking
behavior at the University of Kentucky Tobacco and Health Research Institute. Jenkins et
al. (38/49) described a variable puff parameter smoking machine to study changes in
smoke composition as a function of puff shape and other parameters. McBride (39/35)
reported on the construction and performance of a "smoke duplicator" system that
aocyirately records and then reproduces human smoking behavior. Jenkins et al. (40/56)
presented results from an ORNL project that combined a pressure transducer and a small
IR backscgnering detector in a cigarette holder that measured not only human smoking
parameters but also particulate matter yields. And finally, Borgerding and Winkler (49/47)
studied the effect of alternate puffing regimes on cigarette yields. They concluded that
although smoking parameters, e.g., volume, duration, and interval, affect yields, the
rankings of products remain the same. Earlier work on the effects of changing smoking
parameters was conducted by Butler et al. (13/06) and Frisch and Spivey (22/26).
The second broad group of scientists participating in TCRCs has been those
involved with increasing our knowledge about tobacco and its products. Often they
conduct studies with unique equipment and difficult to reproduce methodologies.
Sometimes the value of this exploratory work is so significant that more routine analytical
techniques arise from their work. However, most of the knowledge-seeking studies
reported at TCRCs are so unique that they are never repeated.
26
52118 5226
I
I
_l
J
-1
I
_J
giVALYTICP t. TECHNOLOGIES DISCUSSED AT TCRCe
The rich 50-year history of the TCRC includes the discussion of numerous
analytical methodologies. We must remember that in 1947 the state of analysis was much
different than today. Major techniques for isolation and identification of constituents in
mixtures included crystallization, distillation, and gravimetry. The use of chemical
reactions to form crystalline substances that could be purified and weighed or colored
derivatives quantifiable by spectroscopy was popular. Instrumentation depended upon
vacuum tubes that were both bulky and notoriously unstable. Digital computers and
microprocessors were not yet invented. Today's simple tasks of multiplying and dividing
on a hand-held calculator available at any local shopping center for less than five dollars
were laboriously accomplished by manual calculation, slide rules, or with the aid of log
tables.
Discussion about the principles and practice of each of these techniques as related
to tobacco'science or even to TCRC presentations is beyond the scope of this review. We
intend to provide commentary on the most important techniques, direct the reader to
existing reviews, supplement earlier reviews with more recent presentations, and provide
top-line information where none exists.
Chiomatoeraohv
Foremost among the analytical technologies applied to tobacco science is
chromatography. Although the invention of chromatography, i.e., column adsorption
chromatography, is attributed to M. S. Tswett in 1903 (56), the sophisticated practice of
the technique as we know it today had not progressed greatly by the Ist TCRC. Because
chromatography is a branch of analytical chemistry devoted to the separation of mixtures
and also because of the complexity of tobacco and smoke composition, tobacco scientists
have been leaders in the practice and advancement of this technique. TCRC presentations
have been made on the following variations of chromatography: affinity, anion-exchange,
cation-exchange, carbon column, counter current, flash, gas-liquid, gas-solid, gel
27
52118 5227
I
I
.l
1
I
I
I
I
I
J
Although much has been made over the controversial compounds found in tobacco
and smoke during the past 50 years of the TCRC, the inhibitors, anticarcinogens, and
antimutagens present in tobacco and its smoke have received scant attention. Comparison
of the Gst of identified components in tobacco smoke with 6sts of compounds shown to
possess inhibitory or antitumorigenic action in tumorigenesis-type experiments in laboratory
animals (52) reveals not only that tobacco smoke contains numerous antitumorigens but also
that the levels in smoke of many of them far exceed those of the smoke components
categorized by some investigators as "tumorigens." Tobacco smoke components
demonstrated to be inhibitors and antitumorigens to these categorized components are listed in
Table IV. Furthermore, Lee and co-workers (48/44, 53) have presented in vitro
methodologies at the TCRC, demonstrating that both nicotine and cotinine inhibit the
mutagenicity of several N-nitrosamine components of tobacco or tobacco smoke.
USES OF ANALYTICAL TECHNOLOGY
Tobacco and its products are among the most studied substances on record. There
are numerous reasons for this intense scrutiny but basically it reflects the monetary value
of.tobacco commerce. Tobacco and its products have been studied by various groups,
including those involved with growing and harvesting the crop, the buyers and processors,
manufacturers, suppliers, research institutes, universities, and government agencies. Why
do all these folks use analytical methodology to study our crop and products? We have
attempted to provide some answers in Table V.
At the farm level, growers need to know the minimum amount of agronomic
chemicals, e.g., fertilizers, herbicides, pesticides, sucker control agents, that are needed to
grow a healthy, good-smoking crop. Farmers, extension agents, university scientists,
government scientists, and manufacturers are also interested in the effects that the many
tobacco production factors such as topping, harvesting, and curing have on tobacco's
physical and chemical characteristics. Breeders are interested in knowing the constituents
of tobacco that impart desirable smoking characteristics to the crop.
! J
,L-A
r.1
18 (~J
~.
52118 5218
