Content uploaded by David Ashley
Author content
All content in this area was uploaded by David Ashley on Mar 17, 2020
Content may be subject to copyright.
Analysis of Volatile Organic
Compounds in Mainstream Cigarette
Smoke
GREGORY M. POLZIN,*
RACHEL E. KOSA-MAINES,
DAVID L. ASHLEY, AND
CLIFFORD H. WATSON
Emergency Response and Air Toxicants Branch, Division of
Laboratory Sciences, National Center for Environmental
Health, Centers for Disease Control and Prevention
Mainstream cigarette smoke is a complex aerosol
containing more than 4400 chemicals. The proliferation of
new brands has necessitated development of faster
and more reliable methods capable of analyzing a wide
range of compounds in cigarette smoke. Although the
International Agency for Research on Cancer has classified
whole cigarette smoke as a human carcinogen, many of
the individual chemicals are themselves highly biologically
active as carcinogens, teratogens, or have implications
for cardiovascular disease. Among these chemicals are
many volatile organic compounds (VOCs), e.g., benzene,
ethylbenzene, and styrene. To analyze VOCs in mainstream
cigarette smoke, we developed a novel headspace
collection technique using polyvinylfluoride bags for sample
collection followed by cannula transfer to evacuated
standard 20-mL auto sampler vials. Coupling collection of
the vapor-phase cigarette smoke with automated analysis
by solid-phase microextraction and gas chromatography/
mass spectrometry enabled us to routinely quantify selected
VOCs in mainstream cigarette smoke. This technique has
similar reproducibility to previous cold trap and impinger
collection methods with significantly higher sample
throughput and virtually no solvent waste. In this report
we demonstrate the method’s analytical capabilities
by quantitatively analyzing 13 selected VOCs in mainstream
cigarette smoke from top-selling domestic brands.
Introduction
Cigarette smoking remains the leading preventable cause of
premature death in the United States (1). From 1995 to 1999,
an estimated 440 000 deaths were attributed each year to
cigarette smoking (1). Although nicotine is the main chemical
component of tobacco smoke that keeps people using the
product (2), many of the other compounds in mainstream
cigarette smoke are associated with cancer, birth defects, or
heart disease (3). The combustion of cigarette tobacco filler
during cigarette smoking creates an aerosol containing
numerous chemical compounds (4,5). Among the thousands
of reported chemicals generated during smoking are several
specific classes of hazardous compounds that merit concern
(3,4,6-10). The International Agency for Research on Cancer
(IARC) groups individual chemicals and chemical mixtures
according to their carcinogenicity toward humans. Cigarette
smoke contains numerous compounds classified by IARC as
known, probable, and possibly carcinogenic. Although whole
cigarette smoke as a mixture has been classified as a human
carcinogen (11), many of the individual volatile organic
compounds (VOCs) present in whole smoke, such as benzene
(12), ethylbenzene (13), and styrene (14), are known or
potential human carcinogens. Although VOCs comprise only
a small fraction (by weight) of mainstream cigarette smoke
(3), smoking is a primary exposure source for many toxic
volatile compounds, and this fraction has been proposed as
the most hazardous fraction of mainstream smoke (15). For
example, cigarette smoking accounts for nearly half of all
Americans’ exposure to benzene, a known human carcinogen
(16).
Several cigarette design factors influence the delivery and
composition of mainstream cigarette smoke. Because ciga-
rette yields are determined by machine smoking, design
differences that reduce machine yields may not reduce a
smoker’s exposure. Filter ventilation reduces machine gen-
erated mainstream smoke yields but smokers often com-
pensate by blocking the vent holes, taking larger puffs, or
puffing more frequently (17). Similarly, changes to static burn
rates, tobacco filler composition and weight, and different
filter materials can result in a reduction in machine yields.
Smokers who compensate in an effort to obtain the desired
level of nicotine defeat these design features and obtain a
much different exposure than estimated by standardized
machine smoking measurements. An accurate assessment
of cigarette design changes and their impact on mainstream
smoke yields are required to help accurately estimate the
health impact these changes may have.
Earlier studies of VOCs in mainstream cigarette smoke
were accomplished by using a variety of techniques for both
collection and analysis. Previous researchers employed
collection techniques such as solvent-filled impinger trains
(18,19), adsorbent materials (20-22), cold traps (23,24),
and direct injection of the gas sample (25,26) to isolate the
wide range of chemicals in cigarette smoke. Although these
methods exhibited good reproducibility, the possibility of
sample breakthrough, tedious cleanup steps, generation of
solvent waste, and long sample-preparation times often
limited their utility. Factors such as impinger solvent volumes,
evaporation, reactivity, breakthrough, and perturbation of
the machine puff also potentially contribute to increased
variability in any analytical technique. Minimizing the
influence of such variables improves accuracy and precision.
We present a method for analyzing VOCs in mainstream
cigarette smoke that allows for high-throughput analysis at
a relatively low cost using commercially available materials
and equipment. Key features of this method include the
appropriate selection of internal standards, inert poly-
vinylfluoride (PVF) bags for vapor-phase collection, cannula
transfer of smoke to an evacuated headspace vial, and
subsequent automated solid-phase microextraction (SPME)
and gas chromatography/mass spectrometry (GC/MS). Other
advantages include minimal use of solvents, virtually no
sample preparation or cleanup required, and no sample
carryover. Method validation is discussed along with the
analysis of ketones and arenes in 41 cigarette varieties that
provide a representative sampling of domestic cigarettes
spanning the current range of nicotine deliveries. We also
examine the delivery of VOCs as a function of selected
cigarette design features and their delivery relative to nicotine
delivery.
* Corresponding author phone: (770) 488-7292; fax: (770) 488-
0181; e-mail: GPolzin@cdc.gov.
Environ. Sci. Technol.
2007,
41,
1297-1302
10.1021/es060609l Not subject to U.S. Copyright. Publ. 2007 Am. Chem. Soc. VOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 91297
Published on Web 01/11/2007
Materials and Methods
Materials. All target compounds were purchased from Aldrich
Chemical Co. (Milwaukee, WI). Methanol was purchased from
Tedia Company Inc. (Fairfield, OH). The ultrahigh purity gas
helium was obtained from AirGas, Inc. (Atlanta, GA). Glass
fiber Cambridge filter pads (CFP), 44 mm, were purchased
from Whatman (Maidstone, England). Tedlar brand PVF bags,
carboxen-polydimethylsiloxane (carboxen-PDMS) SPME
fibers, and Mininert septum caps were purchased from
Supelco (Bellefonte, PA). All chemicals and solvents were
used without further purification.
Internal Standards. An internal standard (IS) stock
solution was prepared by adding 1 mL each of the perdeutero
compounds acetone-d6, benzene-d6, styrene-d8, tetrahydro-
furan-d8, toluene-d8, and p-xylene-d10 and diluting to 10 mL
with methanol. Working solutions of 1:100 and 1:1000
dilutions, in methanol, were prepared monthly as ISs for the
vapor and particulate phases respectively. These ISs were
stored in airtight amber vials with Mininert septum caps and
removed with a Hamilton (gas-tight) syringe as needed. A
10-µL aliquot of IS was added to each sample before analysis.
Preparation of Standard Curves. We made a calibration
stock solution by the addition of neat standards to methanol.
The amount of each analyte added to the stock calibration
mixture was recorded to the nearest 0.01 mg and diluted
with methanol to a final volume of 10 mL. Serial dilutions
of the calibration stock solution generated appropriate
concentration ranges for each analyte (Table 1). These
solutions were analyzed to prepare calibration curves for all
13 analytes in both the vapor and particulate phases. For the
vapor phase, for which PVF bags were used to collect the
smoke sample, calibration curves were generated by injection
of a standard solution and IS into an empty bag and filled
with 11 puffs of 35 mL from the smoking machine. The bag
was then removed from the smoking machine and treated
as a sample. Similarly, calibration curves for the particulate
phase collected on the CFP were prepared by the addition
of the standard solution and IS to a blank 20-mL headspace
vial and treated as samples. All calibration curves were linear
with R2values exceeding 0.99. Calibration curves were
prepared weekly and used for all analytical runs using the
associated IS.
Smoking Conditions. Prior to smoking, cigarette filter
ventilation levels were measured with a QTM-5 (Cerulean,
Milton Keynes, UK). For 24 h before smoking, cigarettes were
conditioned in an environmental chamber (Parameter
Generation Control, Inc, Black Mountain, NC) that was
maintained, as specified in ISO 3402:1999, at a temperature
of 22 °C and 60% relative humidity. Cigarettes were smoked
according to the conditions detailed in ISO 3308:2000 (35
mL puff of 2 s duration every 60 s) on an automated linear
16-port ASM 500 smoking machine (Cerulean, Milton Keynes,
UK) to a butt length of either filter +8 mm or filter overwrap
+3 mm, whichever was greater.
VOC Collection-Vapor Phase. The vapor-phase portion
of mainstream cigarette smoke was collected in individual
1-L PVF bags attached directly to the exhaust ports of the
ASM 500 puffing engines. To reduce background and sample
carryover, the smoking machine was programmed to com-
plete 30 blank puffs per port before attachment of the PVF
bags. Background levels of all analytes were measured on a
per port basis and corrections, if needed, were applied to all
results for that day’s analytical runs. Prior to smoking, 10 µL
of the 1:100 IS solution was added to each 1-L PVF bag. After
smoking, the PVF bags were closed and immediately removed
from the smoking machine. A portion of the smoke sample,
collected in the PVF bag, was subsequently transferred to an
evacuated 20-mL headspace vial (Microliter Corporation,
Atlanta, GA) by a 27 gauge cannula. Vials, containing the
analytical samples were subsequently loaded onto a LEAP
Combi-Pal auto sampler (LEAP Technologies, Carrboro, NC)
equipped with SPME sampling for analysis by GC/MS.
VOC Collection-Particulate Phase. The particulate phase
of the mainstream cigarette smoke was collected on a
standard 44-mm CFP. After smoking, the CFP was removed
from the holder and placed in a 20-mL headspace vial and
spiked with 10 µL of the 1:1000 IS solution. The sample was
then loaded on a LEAP Combi-Pal autosampler and analyzed
by automated SPME/GC/MS.
VOC Analysis. VOCs were quantitatively analyzed using
an Agilent 6890 gas chromatograph coupled to an Agilent
5973 mass spectrometer. The sample was incubated at 30 °C
for 2 min before the vial was sampled for 30 s with a 75-µm
carboxen-PDMS SPME fiber. Analytes were then desorbed
from the fiber at 260 °C in a heated inlet and focused onto
an Agilent DB-624 capillary column (30.0 m ×320 µm×1.80
µm). Helium flow was maintained in constant flow mode at
an average linear velocity of 46 cm/sec. For the analysis of
vapor-phase samples, the GC oven, equipped with liquid
nitrogen cryo cooling, was programmed to start at -20 °C,
hold for 2 min, and ramp to 200 °Cat8°C/min, for a total
run time of 29.50 min. The GC analysis of the particulate
phase samples differed in that the oven started at 40 °C, was
held constant for 2 min, and then ramped to 230 °Cat6
°C/min, for a total run time of 35 min. Both methods used
a transfer line temperature of 255 °C and source and
quadrupole temperatures of 230 °C and 150 °C, respectively.
Data were acquired in full-scan mode over 30-200 amu.
Cigarettes for Testing. We selected premium and value
cigarette brands, ranging from full-flavor to ultralight,
including both mentholated and non-mentholated varieties,
from the top four domestic cigarette producers for analysis.
The cigarettes were purchased at commercial retail outlets
in the Atlanta metropolitan area or were provided by the
Massachusetts Department of Public Health. The 1R4F and
2R4F research cigarettes were purchased from the University
of Kentucky Tobacco and Health Research Institute (Lex-
ington, KY). All cigarettes were stored in their original
packaging at -70 °C until being conditioned prior to analysis.
Quality Control Materials. The 2R4F research cigarette
was used as a quality control (QC) material. QC cigarettes
were smoked daily and analyzed with all runs performed
that day. All 13 analytes measured in the smoke of the 2R4F
cigarettes were characterized to determine the mean and
the 95th and 99th confidence intervals for each analyte
studied. Acceptance criteria for QC and blank samples
followed the criteria prescribed by Taylor (27).
TABLE 1. Linear Calibration Range, Calculated Limits of
Detection (LOD), and Correlation Coefficients (R2) for the
Vapor-Phase Portion of Mainstream Cigarette Smoke
analyte
linear range
(µg)
LOD
(µg/cigarette)
R
2
benzene 0.86-344 0.09 0.9999
p
-xylene 0.086-34.4 0.01 0.9991
acetone 2.78-1110 2.16 0.9986
o
-xylene 0.091-36.4 0.02 0.9998
2,3-butanedione 0.98-394 0.76 0.9999
2-pentanone 0.080-32.2 0.22 0.9994
toluene 0.84-338 0.21 1.0000
styrene 0.091-36.2 0.28 0.9995
ethylbenzene 0.089-35.4 0.10 0.9996
3-ethyltoluene 0.045-18.2 0.09 1.0000
3-euten-2-one 0.85-340 0.15 0.9999
3-pentanone 0.081-32.2 0.24 0.9989
2-butanone 0.79-316 0.81 0.9994
1298 9ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 4, 2007
Results and Discussion
Calibration, detection, and quantification of the 13 analytes
were straightforward. The linear quantification range, LOD,
and linear regression coefficients were more than adequate
to provide quantitative data for the analytes studied (Table
1). The LODs for these analytes, calculated as three times the
standard deviation extrapolated to zero concentration (27),
ranged from 0.01 µg to 2.12 µg per cigarette. Although not
needed for this study, switching to single ion monitoring
rather than a full spectral scan should improve LODs. The
results for each analyte in this study are presented on a per-
cigarette basis which simplifies data analysis and provides
for direct cigarette-to-cigarette comparisons.
Accurate quantification of the chemicals in mainstream
cigarette smoke dictates the analysis of both the particulate
and vapor phases. We accomplished this by parallel analysis
of the particulate matter located on the CFP and the vapor-
phase sample. Because of the highly volatile nature of the
analytes selected for this study, we found the fraction in the
particulate phase to be extremely low. Quantification of the
13 analytes in the particulate- and vapor-phase samples for
the 2R4F research cigarettes confirmed that these analytes
resided mainly in the vapor phase. In fact, the levels of 12
of the 13 analytes on the CFP had values below the method
limit of detection (LOD). The only measurable analyte in the
particulate phase, 2,3-butanedione, represented less than
2% of the measured vapor sample. Because of the nearly
complete gas-phase partitioning, we only report levels
measured in the vapor phase for the remainder of the
analyses.
Because of the potential for sample loss and the reactivity
of selected analytes in this study, the choice of an appropriate
IS was important. Isotopically labeled ISs for each analyte
was not practical because of the large number and wide
concentration ranges of the chemicals in cigarette smoke.
Limiting the number of ISs simplifies sample preparation
times and increases SPME sensitivity. Also, the use of a limited
number of ISs can greatly reduce the cost per analysis.
Therefore, we selected six perdeutero ISs for quantification
of VOCs in mainstream cigarette smoke. Carbonyl com-
pounds were quantified using either acetone-d6or THF-d8
and aromatic compounds using benzene-d6, toluene-d8,
FIGURE 1. Vapor-phase sample stability as a function of time.
FIGURE 2. A typical vapor-phase full scan chromatogram for the 2R4F research cigarette. Compounds quantified in this study are labeled.
For clarity, signal before 7 min and after 21 min are not shown.
TABLE 2. Analyte Yields (in Micrograms Per Cigarette),
Standard Deviation (σ), and Percent Relative Standard
Deviation (%RSD) for the 2R4F Kentucky Research Cigarette
over a 6 Month Period (N)100)
analyte average σ%RSD
benzene 44.1 5.1 11.7
toluene 57.4 5.8 10.2
styrene 2.2 0.4 17.0
o
-xylene 1.7 0.3 14.8
m
/
p-
xylene 9.9 1.3 13.6
3-ethylbenzene 4.4 0.6 13.3
3-ethyltoluene 1.6 0.6 37.1
acetone 366.9 64.2 17.5
3-buten-2-one 45.7 7.5 16.3
2,3-butanedione 89.3 13.1 14.7
2-butanone 86.3 11.8 13.7
2-pentanone 12.8 2.4 18.6
3-pentanone 6.1 1.3 20.8
TABLE 3. Vapor-phase Mainstream Cigarette Smoke Yields of
Previously Reported Volatile Organic Compounds from a
Kentucky 1R4F Research Cigarette. Literature Reported Values
Included for Comparison. Values Are Shown (Standard
Deviation (n)5)
analyte
delivery
(µg/cig)
previously Reported
Results (µg/cig)
benzene 41 (445
,
a
51,
b
59,
d
41
e
toluene 60 (7 68,
a
73,
b
61
e
styrene 2.9 (0.6 2.1,
a
1.8
e
xylenes 12.9 (1.9 10.6
e
acetone 340 (41 284,
c
380
d
a
From ref
23
.
b
From ref
28
.
c
From ref
8
.
d
From ref
26
.
e
From ref
24
.
VOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 91299
p-xylene-d10, or styrene-d8. Appropriate pairing of IS to
analyte was determined by a similarity in vapor pressure,
structure, and as was the case for styrene, reactivity, to achieve
optimal linear response.
Sample stability was of great concern because of numerous
reactive compounds associated with mainstream cigarette
smoke. Previous reports have indicated problems with sample
stability over time (24,26). To examine sample stability, a
2R4F cigarette was smoked and aliquots of the vapor phase
were transferred by cannula into 10 individual headspace
vials. These vials were sequentially sampled and analyzed
over the following 20 h. During this time, a steady decrease
in the raw area counts was observed for all analytes. Although
both analyte and IS ion counts decreased over time, change
in the analyte relative response ratio was less than 15% for
all analytes over a 20-hr period (Figure 1). Therefore, to
minimize temporal variability, all samples were analyzed
within 20 h of collection.
Concentration of the sample on the SPME fiber, coupled
with the relatively high abundances of the chosen analytes
in mainstream cigarette smoke, provided sufficient sensitivity
to allow full-scan mass spectral detection. Using full-scan
data acquisition provided quantitative analysis of the analytes
while retaining spectral information for qualitative analysis
of additional smoke constituents. The total ion chromatogram
for a 2R4F research cigarette (Figure 2) demonstrates the
good separation, high signal response, and adequate run
time for all 13 analytes. Given the chemical complexity of
smoke and the large number of chemicals readily identified
in the vapor phase, this method could be readily expanded
to analyze additional, volatile, cigarette smoke constituents.
Method reproducibility was assessed by evaluating the
results generated over a 6 month period for all 13 analytes
when a 2R4F cigarette was machine smoked. During that
time more than 100 analytical determinations were made
for the mainstream smoke of 2R4F cigarettes. The average
deliveries and standard deviations for the analytes studied
were determined from these data (Table 2). The calculated
relative standard deviation for all 13 compounds was, on
average, 16.9% for the 2R4F cigarette. Day-to-day variability
in this method can be attributed both to the variability in the
analytical process and sample-to-sample variability among
the 2R4F cigarettes.
The recent replacement of the 1R4F reference cigarette
with the 2R4F research cigarette made delivery comparisons
with other researchers tenuous. To make meaningful com-
parisons, we also analyzed some of the older 1R4F cigarettes.
However, of the 13 analytes reported here, only values for
TABLE 4. Delivery of Aromatic Volatile Organic Compounds in Mainstream Cigarette Smoke of 41 U.S. Brandsa
full flavored brands % vent benzene toluene styrene
o
-xylene
m
/
p
-xylene Et-benzene 3-Et-toluene
Basic 0 49.8 (5.9 69.0 (9.6 4.3 (0.4 2.4 (0.1 13.1 (1.2 6.1 (0.7 2.3 (0.3
Camel (M) 0 52.9 (3.6 74.6 (5.8 5.0 (0.6 2.7 (0.3 15.0 (1.3 6.9 (0.6 2.8 (0.7
Newport (M) 0 50.7 (2.5 72.3 (3.7 4.3 (0.0 2.4 (0.0 13.2 (0.4 6.3 (0.3 2.2 (0.6
Salem (M) 0 56.8 (6.9 81.2 (12.7 5.3 (1.1 2.9 (0.5 16.0 (2.8 7.5 (1.2 2.9 (0.9
Camel Jade (M) 1 57.1 (3.5 82.4 (3.7 5.4 (0.4 3.0 (0.3 16.4 (1.4 7.8 (0.7 2.7 (1.1
Kool (M) 1 50.5 (4.5 73.1 (3.1 4.2 (0.2 2.3 (0.2 13.4 (0.9 6.2 (0.3 1.8 (0.4
GPC 5 56.4 (0.7 75.0 (2.8 4.7 (0.2 2.6 (0.1 14.4 (0.8 6.9 (0.3 2.4 (0.0
Marlboro 12 46.8 (0.5 67.0 (3.6 3.9 (0.3 2.2 (0.1 12.8 (0.4 5.6 (0.2 2.3 (0.3
Benson & Hedges 13 50.7 (2.4 73.8 (6.4 3.8 (0.9 2.4 (0.4 14.1 (2.2 6.1 (0.9 2.4 (0.8
Camel 15 56.6 (0.9 76.1 (2.0 4.8 (0.1 2.7 (0.1 14.7 (0.6 6.7 (0.2 2.6 (0.2
Marlboro (M) 16 42.8 (3.3 62.4 (6.2 3.2 (0.2 1.9 (0.2 11.1 (1.0 5.1 (0.4 1.6 (0.3
Kent 20 38.3 (2.1 57.2 (5.8 3.0 (0.4 1.7 (0.1 9.8 (0.3 4.6 (0.7 1.3 (0.9
Doral 22 50.0 (7.4 66.0 (12.0 4.6 (0.7 2.7 (0.4 14.6 (1.5 6.3 (0.6 2.9 (0.7
Winston 23 51.5 (1.5 68.9 (1.7 4.1 (0.3 2.4 (0.1 13.7 (0.1 6.1 (0.2 2.2 (0.1
medium or mild brands
GPC 12 43.8 (3.6 57.9 (8.5 3.1 (0.2 1.7 (0.1 9.7 (0.9 4.8 (0.3 1.3 (0.2
Marlboro (M) 13 43.9 (3.1 62.0 (5.2 3.0 (0.4 1.9 (0.2 10.8 (0.9 5.2 (0.5 1.5 (0.2
Newport (M) 20 42.2 (4.2 58.5 (6.1 3.2 (0.5 1.8 (0.2 10.0 (1.3 4.7 (0.6 1.7 (0.3
Marlboro 22 41.9 (4.3 57.6 (7.5 3.0 (0.3 1.9 (0.1 11.0 (0.9 4.9 (0.5 1.6 (0.2
light brands
Basic 15 42.0 (0.7 55.1 (1.4 2.9 (0.2 1.8 (0.1 10.1 (0.3 4.7 (0.1 1.6 (0.4
GPC 21 38.2 (3.5 50.3 (4.3 2.5 (0.4 1.5 (0.2 8.5 (1.0 4.1 (0.5 1.4 (0.3
Marlboro 22 40.4 (3.0 55.4 (6.5 2.7 (0.1 1.8 (0.2 10.4 (1.2 4.6 (0.5 1.8 (0.3
Newport (M) 23 31.6 (3.1 43.2 (5.4 1.8 (0.4 1.2 (0.1 6.8 (0.9 3.3 (0.5 0.9 (0.2
Camel Jade (M) 24 42.8 (3.8 60.3 (4.7 3.4 (0.3 1.9 (0.2 10.9 (1.1 5.3 (0.5 1.6 (0.4
Camel (M) 25 39.6 (3.9 57.1 (9.3 2.8 (0.6 1.6 (0.3 9.7 (1.8 4.8 (1.0 1.1 (0.6
Marlboro (M) 25 37.0 (5.9 51.3 (9.4 2.3 (0.3 1.4 (0.3 8.6 (1.7 4.1 (0.6 0.9 (0.5
Doral 26 38.2 (5.8 53.4 (6.9 2.9 (0.6 1.8 (0.3 10.0 (1.4 4.4 (0.7 1.8 (0.2
Camel 30 39.4 (3.5 52.6 (6.6 2.6 (0.3 1.6 (0.2 9.0 (1.1 4.2 (0.5 1.4 (0.3
Winston 30 44.5 (3.4 57.2 (5.7 2.9 (0.4 1.9 (0.2 10.4 (1.4 4.8 (0.5 1.5 (0.2
Misty 49 30.7 (2.6 39.9 (3.3 1.8 (0.3 1.2 (0.1 6.9 (0.8 3.2 (0.3 0.9 (0.1
Misty (M) 50 25.4 (1.5 32.3 (1.9 1.2 (0.3 0.8 (0.1 4.8 (0.5 2.2 (0.2 0.5 (0.0
ultralight brands
Basic 32 25.6 (5.0 32.0 (6.7 1.4 (0.5 1.0 (0.3 5.6 (1.7 2.7 (0.7 0.7 (0.4
Marlboro 47 25.5 (2.8 33.6 (4.8 1.4 (0.3 1.0 (0.2 5.9 (1.1 2.7 (0.3 0.8 (0.4
GPC 49 26.7 (3.5 34.5 (4.0 1.9 (0.3 1.2 (0.2 6.5 (1.0 3.0 (0.4 1.1 (0.4
Camel 52 29.3 (1.1 39.7 (2.5 2.1 (0.3 1.4 (0.1 7.6 (0.6 3.4 (0.3 1.3 (0.4
Marlboro (M) 53 27.3 (2.6 33.3 (3.3 1.2 (0.3 1.0 (0.1 5.8 (0.5 2.7 (0.3 0.7 (0.1
Winston 55 26.6 (0.2 31.7 (1.8 1.5 (0.3 1.1 (0.1 6.2 (0.4 2.8 (0.3 0.9 (0.1
Doral 61 19.7 (1.9 27.7 (2.9 1.2 (0.2 1.0 (0.0 5.7 (0.3 2.5 (0.2 0.8 (0.2
Misty (M) 67 15.7 (1.3 21.0 (3.1 0.7 (0.2 0.7 (0.1 4.0 (0.8 1.7 (0.2 0.4 (0.3
True 68 15.0 (1.1 18.8 (2.1 0.9 (0.3 0.6 (0.1 3.7 (0.3 1.7 (0.2 0.5 (0.0
Carlton 77 6.3 (0.6 7.3 (0.7 0.3 (0.1 0.3 (0.1 1.9 (0.3 1.0 (0.1 0.2 (0.1
Carlton (M) 78 3.7 (0.9 4.5 (1.3 0.3 (0.3 0.2 (0.1 1.5 (0.4 0.8 (0.2 0.0 (0.0
a
All analytes are expressed as micrograms per cigarette along with the corresponding standard deviation, (M) denotes a mentholated brand.
1300 9ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 4, 2007
acetone (8,26), benzene (22-24,26,28), styrene (23,24),
toluene (23,24,28), and total xylenes (24) were reported in
previous studies for the 1R4F research cigarette. Quantified
values for the deliveries of these five analytes in 1R4F cigarette
smoke were comparable to previously reported values (Table
3). These previous methods used a variety of smoke collection
techniques, including the use of solvent filled cold traps as
well as directly sampling the cigarette vapor phase or smoke,
which differed from the PVF bag technique used in the current
study. The agreement with previous work helps to support
the validity of our reported results. Additionally, the measured
values of all 13 analytes for the 1R4F research cigarette are
comparable to the newer 2R4F research cigarette (Tables 4
and 5) and should provide a useful future reference.
Brand and manufacturer differences among top-selling
domestic cigarettes were investigated for the analytes in this
study. Delivery data, along with filter ventilation are reported
in Tables 4 and 5. When separated into groups on the basis
of their FTC mainstream smoke delivery designations (full-
flavor, medium, light, or ultralight), no statistical differences
(p<0.05) existed between the individual VOC smoke
concentrations in the 41 commercial brands examined from
the four major domestic manufacturers for brands grouped
together according to comparable delivery. Further statistical
analysis of analyte delivery for all 41 cigarette brands showed
that filter ventilation was the main cigarette design parameter
influencing delivery of the 13 analytes. For example, benzene
delivery, when analyzed as a function of filter ventilation
(Figure 3), showed a good linear relation for all 41 brands,
as did the other 12 VOCs analyzed in this study. In
comparison, physical characteristics such as filler mass and
filter length were much poorer predictors of the overall
mainstream smoke delivery of these analytes.
One limitation of the current study is the smoking
parameters were based on standardized machine smoking
methods, which do not accurately estimate smoke deliveries
for modern cigarettes (29). Kozlowski et al. (17) have shown
that smokers of low delivery cigarettes may compensate to
obtain sufficient nicotine by blocking the vent holes or by
taking more frequent, larger, and deeper puffs. Such com-
pensation behavior would result in much higher deliveries
of all constituents in cigarette smoke. If the present results
are normalized to the reported nicotine delivery of the
cigarette, all cigarettes studied had, within standard deviation,
nearly identical deliveries of all 13 analytes (Figure 4). Work
is under way in our laboratories to define the relationship
between the amount of nicotine delivered under a wide range
of smoking conditions and the resultant potential exposure
to these and other VOCs in mainstream cigarette smoke.
TABLE 5. Delivery of Volatile Organic Ketones in Mainstream Cigarette Smoke of 41 U.S. Brandsa
full flavored brands % vent acetone 2,3-butanedione 2-pentanone 2-butanone 3-pentanone 3-buten-2-one
Basic 0 401.4 (43.5 105.4 (3.5 15.1 (0.4 95.5 (6.4 6.6 (0.4 60.7 (2.7
Camel (M) 0 511.7 (34.7 128.1 (9.0 21.4 (2.2 126.2 (9.5 9.7 (0.9 78.2 (8.6
Newport (M) 0 488.7 (6.0 115.1 (8.7 18.8 (2.6 116.8 (7.5 8.5 (1.0 74.4 (3.5
Salem (M) 0 534.2 (47.3 128.0 (15.7 21.2 (2.9 127.9 (13.8 9.8 (1.2 81.5 (8.7
Camel Jade (M) 1 541.4 (39.6 124.9 (12.5 21.2 (3.2 129.5 (13.5 10.3 (1.8 80.1 (8.7
Kool (M) 1 494.0 (81.9 107.2 (30.6 16.8 (4.9 107.7 (24.3 8.3 (2.3 79.1 (18.8
GPC 5 477.8 (25.8 115.5 (3.9 18.2 (0.5 114.7 (4.4 8.1 (0.4 79.3 (6.0
Marlboro 12 402.7 (11.7 101.3 (15.0 15.1 (2.5 94.8 (9.1 6.7 (0.9 64.0 (9.8
Benson & Hedges 13 438.1 (27.0 100.4 (10.6 15.9 (2.3 101.2 (8.8 7.4 (1.5 64.6 (5.8
Camel 15 470.7 (10.0 122.2 (5.0 18.6 (1.6 115.4 (5.3 8.5 (1.5 73.5 (5.5
Marlboro (M) 16 376.9 (31.4 106.3 (9.5 14.8 (1.4 92.1 (7.0 6.7 (0.7 53.5 (5.8
Kent 20 376.0 (51.3 95.0 (6.2 12.0 (3.0 83.3 (2.1 5.3 (1.6 49.2 (2.9
Doral 22 472.5 (97.0 145.0 (71.0 22.5 (8.5 129.6 (46.3 9.7 (3.0 94.8 (47.5
Winston 23 433.3 (36.2 119.6 (10.3 16.5 (2.9 104.3 (12.7 6.9 (1.4 68.7 (6.4
medium or mild brands
GPC 12 356.1 (21.4 82.4 (9.9 10.9 (0.8 77.6 (1.6 4.7 (0.3 50.5 (6.4
Marlboro (M) 13 358.6 (31.3 88.6 (13.6 12.8 (1.8 83.9 (9.4 5.6 (0.4 47.7 (9.3
Newport (M) 20 375.3 (31.0 94.1 (12.9 14.5 (3.2 91.2 (13.8 6.8 (1.3 55.5 (7.7
Marlboro 22 332.2 (35.5 79.8 (2.3 12.2 (0.7 78.5 (6.2 5.3 (0.4 53.2 (3.8
light brands
Basic 15 336.9 (3.2 84.0 (5.7 12.3 (0.9 79.4 (2.2 5.4 (0.4 50.6 (1.5
GPC 21 304.7 (17.7 75.8 (8.6 10.7 (1.9 70.5 (8.4 4.7 (0.7 44.8 (6.2
Marlboro 22 321.2 (20.9 75.0 (6.7 12.0 (1.4 76.1 (5.7 5.2 (0.7 50.1 (2.3
Newport (M) 23 301.6 (28.8 73.5 (4.1 10.2 (0.9 68.0 (4.9 4.6 (0.3 41.2 (1.2
Camel Jade (M) 24 353.2 (27.8 88.9 (5.1 13.1 (0.4 84.8 (4.1 6.4 (0.3 50.1 (3.2
Camel (M) 25 353.4 (59.8 78.7 (3.9 10.7 (0.5 76.9 (8.5 4.6 (0.9 44.9 (4.9
Marlboro (M) 25 313.0 (20.3 76.8 (2.8 9.9 (1.8 69.6 (7.0 4.3 (0.8 40.5 (2.7
Doral 26 294.5 (34.7 73.7 (6.8 10.2 (0.6 67.5 (6.3 4.5 (0.5 42.4 (3.9
Camel 30 299.6 (28.6 72.6 (2.5 9.8 (1.0 68.0 (3.1 4.7 (0.8 42.3 (3.4
Winston 30 350.1 (21.0 88.3 (3.0 12.1 (0.6 81.7 (4.1 5.2 (0.6 52.2 (2.5
Misty 49 255.5 (17.9 68.0 (1.2 8.9 (0.4 59.3 (3.0 3.7 (0.3 38.8 (1.4
Misty (M) 50 232.6 (21.3 64.7 (18.8 7.5 (2.4 52.1 (9.8 3.6 (1.2 38.3 (14.8
ultralight brands
Basic 32 194.0 (21.4 48.8 (6.8 6.0 (1.4 43.1 (6.6 2.7 (0.5 25.4 (4.3
Marlboro 47 202.3 (25.1 48.0 (5.3 6.8 (1.3 45.5 (6.6 3.0 (0.5 30.5 (5.3
GPC 49 213.6 (24.2 56.7 (8.4 7.1 (1.4 50.0 (7.3 3.3 (0.7 29.2 (5.5
Camel 52 228.8 (12.7 57.3 (2.0 7.6 (0.8 52.3 (1.7 3.6 (0.5 31.1 (2.4
Marlboro (M) 53 197.4 (15.9 44.3 (3.7 6.1 (0.4 43.7 (2.6 2.6 (0.3 27.9 (3.6
Winston 55 195.2 (5.8 49.5 (2.1 5.7 (0.2 42.7 (1.4 2.5 (0.1 26.2 (0.2
Doral 61 156.1 (13.8 36.5 (0.7 4.8 (0.4 34.0 (1.6 2.1 (0.2 20.1 (2.4
Misty (M) 67 139.9 (6.2 36.4 (3.4 3.9 (0.6 29.7 (2.4 1.7 (0.3 17.7 (1.0
True 68 122.2 (9.0 32.6 (2.4 3.3 (0.2 26.3 (1.6 1.5 (0.1 15.2 (0.6
Carlton 77 55.2 (4.7 15.1 (1.9 1.4 (0.1 11.7 (1.0 0.6 (0.0 7.5 (1.0
Carlton (M) 78 45.6 (10.6 12.7 (2.0 1.1 (0.2 10.0 (1.8 0.5 (0.0 5.5 (0.9
a
All analytes are expressed as micrograms per cigarette along with the corresponding standard deviation, (M) denotes a mentholated brand.
VOL. 41, NO. 4, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 91301
Disclaimers
Use of trade names is for identification only and does not
imply endorsement by the Centers for Disease Control and
Prevention or by the U.S. Department of Health and Human
Services. The findings and conclusions in this report are those
of the author(s) and do not necessarily represent the views
of the Centers for Disease Control and Prevention.
Literature Cited
(1) Centers for Disease Control and Prevention. Annual smokings
attributable mortality, years of potential life lost and economic
costssUnited States, 1995-1999. Morbidity and Mortality
Weekly Report 2002,51, 300.
(2) U.S. Department of Health and Human Services. The Health
Consequences of Smoking: Nicotine Addiction: A Report of the
Surgeon General; DHHS Publication no. (CDC) 88-8406; Center
for Health Promotion and Education, Office on Smoking and
Health: Rockville, MD, 1988.
(3) Hoffmann, D.; Hoffmann, I. The changing cigarette, 1950-1995.
J. Toxicol. Environ. Health 1997,50, 307.
(4) Stedman, R. L. The chemical composition of tobacco and
tobacco smoke. Chem. Rev. 1968,68, 153.
(5) Roberts, D. L. Natural tobacco flavor. Recent Adv. Tob. Sci. 1988,
14, 45.
(6) Matsushima, S.; Ishiguro, S.; Sugawara, S. Composition studies
on some varieties of tobacco and their smoke 1. Major
components in smoke condensate. Beitrage Zur Tabakforschung
Int. 1979,10, 31.
(7) Smith, C. J.; Livingston, S. D.; Doolittle, D. J. An international
literature survey of “IARC Group I carcinogens” reported in
mainstream cigarette smoke. Food Chem. Toxicol. 1997,35, 1107.
(8) Smith, C. J.; Perfetti, T. R.; Rumple, M. A.; Rodgman, A.; Doolittle,
D. J. “IARC Group 2A carcinogens” reported in cigarette
mainstream smoke. Food Chem. Toxicol. 2000,38, 371.
(9) Smith, C. J.; Perfetti, T. A.; Rumple, M. A.; Rodgman, A.; Doolittle,
D. J. “IARC Group 2B carcinogens” reported in cigarette
mainstream smoke. Food Chem. Toxicol. 2001,39, 181.
(10) Rustemeier, K.; Stabbert, R.; Haussmann, H. J.; Roemer, E.;
Carmines, E. L. Evaluation of the potential effects of ingredients
added to cigarettes. Part 2: Chemical composition of main-
stream smoke. Food Chem. Toxicol. 2002,40, 93.
(11) IARC. IARC Monographs on the Evaluation of the Carcinogenic
Risk of Chemicals to Humans. IARC: Lyon, France, 1986; Vol.
38.
(12) IARC. IARC Monographs on the Evaluation of the Carcinogenic
Risk of Chemicals to Humans. IARC: Lyon, France, 1982; Vol.
29.
(13) IARC. IARC Monographs on the Evaluation of the Carcinogenic
Risk of Chemicals to Humans. IARC: Lyon, France, 2000; Vol.
77.
(14) IARC. IARC Monographs on the Evaluation of the Carcinogenic
Risk of Chemicals to Humans. IARC: Lyon, France, 2002; Vol.
82.
(15) Fowles, J.; Dybing, E. Application of toxicological risk assessment
principles to the chemical constituents of cigarette smoke. Tob.
Control 2003,12, 424.
(16) National Toxicology Program. Report on Carcinogens, 10th Ed.;
U.S. Department of Health and Human Services, Public Health
Service, National Institutes for Health: Research Triangle Park,
NC, 2002.
(17) Kozlowski, L. T.; Rickert, W. S.; Pope, M. A.; Robinson, J. C.;
Frecker, R. C. Estimating the yield to smokers of tar, nicotine,
and carbon-monoxide from the lowest yield ventilated filter-
cigarettes. Br. J. Addict. 1982,77, 159.
(18) Houlgate, P. R.; Dhingra, K. S.; Nash, S. J.; Evans, W. H.
Determination of formaldehyde and acetaldehyde in main-
stream cigarette-smoke by high-performance liquid-chroma-
tography. Analyst 1989,114, 355.
(19) Miyake, T.; Shibamoto, T. Quantitative-analysis by gas-chro-
matography of volatile carbonyl-compounds in cigarette-smoke.
J. Chromatogr. A 1995,693, 376.
(20) Higgins, C. E.; Griest, W. H.; Olerich, G. Application of Tenax
trapping to analysis of gas-phase organic-compounds in ultralow
tar cigarette-smoke. J. AOAC 1983,66, 1074.
(21) Nunez, A. J.; Gonzalez, L. F.; Janak, J. Pre-concentration of
headspace volatiles for trace organic-analysis by gas-chroma-
tography. J. Chromatogr. A 1984,300, 127.
(22) Takanami, Y.; Chida, M.; Hasebe, H.; Sone, Y.; Suhara, S. Analysis
of cigarette smoke by an online thermal desorption system and
multidimensional GC-MS. J. Chromatogr. Sci. 2003,41, 317.
(23) Byrd, G. D.; Fowler, K. W.; Hicks, R. D.; Lovette, M. E.; Borgerding,
M. F. Isotope-dilution gas-chromatography mass-spectrometry
in the determination of benzene, toluene, styrene and acrylo-
nitrile in mainstream cigarette-smoke. J. Chromatogr. A 1990,
503, 359.
(24) Darrall, K. G.; Figgins, J. A.; Brown, R. D.; Phillips, G. F.
Determination of benzene and associated volatile compounds
in mainstream cigarette smoke. Analyst 1998,123, 1095.
(25) Brunnemann, K. D.; Kagan, M. R.; Cox, J. E.; Hoffmann, D.
Analysis of 1,3-butadiene and other selected gas-phase com-
ponents in cigarette mainstream and sidestream smoke by gas-
chromatography mass selective detection. Carcinogenesis 1990,
11, 1863.
(26) Dong, J. Z.; Glass, J. N.; Moldoveanu, S. C. A simple GC-MS
technique for the analysis of vapor phase mainstream cigarette
smoke. J. Microcolumn Sep. 2000,12, 142.
(27) Taylor, J. K. Quality Assurance of Chemical Measurements; CRC
Press: Boca Raton, FL, 1987.
(28) Brunnemann, K. D.; Kagan, M. R.; Cox, J. E.; Hoffmann, D.
Determination of benzene, toluene and 1,3-butadiene in
cigarette-smoke by GC-MSD. Exp. Pathol. 1989,37, 108.
(29) Burns, D. M.; Major, J. M.; Shanks, T. G.; Thun, M. J.; Samet,
J. M. Risks associated with smoking cigarettes with low machine-
measured yields of tar and nicotine; Smoking and Tobacco
Control Monograph 13; NIH Publication 02-5074.; U.S. Depart-
ment of Health and Human Services, National Institutes of
Health, National Cancer Institute: Bethesda, MD, 2000; pp 65-
158.
Received for review March 15, 2006. Revised manuscript
received October 18, 2006. Accepted October 18, 2006.
ES060609L
FIGURE 3. The relation between filter ventilation and the level of
benzene in the vapor phase of mainstream cigarette smoke for top
selling domestic cigarettes. Each data point is the result of at least
three measurements and error bars representing the standard
deviation are included. A linear regression of the points in provided
for reference.
FIGURE 4. Average selected analyte deliveries normalized to Federal
Trade Commission (FTC) nicotine values. Errors bars are at one
standard deviation.
1302 9ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 4, 2007
