Supplement
Effect of charcoal-containing cigarette filters on gas
Emergency Response and Air
Toxicants Branch, Division of
Laboratory Sciences, National
Center for Environmental Health,
Centers for Disease Control and
Prevention, Atlanta, Georgia,
USA
Correspondence to:
Gregory M Polzin, 4770 Buford
Highway, NE, Atlanta, GA
30341, USA; GPolzin@cdc.gov
Received 25 June 2007
Accepted 1 December 2007
i10
phase volatile organic compounds in mainstream
cigarette smoke
G M Polzin, L Zhang, B A Hearn, A D Tavakoli, C Vaughan, Y S Ding, D L Ashley,
C H Watson
ABSTRACT
Of the chemicals identified to date in mainstream
cigarette smoke with known toxicological properties, the
volatile organic compounds (VOCs) are considered the
most hazardous group owing to their high abundance and
toxicity. In this research we evaluate a recently introduced
line of cigarettes that contain charcoal in their filters. The
amount of charcoal in these filters ranged from 45 mg to
180 mg and were either dispersed among the filter
material or contained in a small cavity in the filter
segment. Charcoal has long been used for removing VOCs
from both water and air. Our findings indicate that these
cigarettes reduce machine generated mainstream smoke
deliveries of a wide range of VOCs compared to a similar,
non-charcoal filtered, cigarette. However, this reduction is
dependent not only on the amount of charcoal present but
also on the volume of smoke being drawn through the
filter. While a brand with 45 mg charcoal reduces VOC
delivery under ISO smoking conditions, charcoal saturation
and breakthrough occur under more intense smoking
conditions. Breakthrough is minimised for brands with the
most charcoal. Overall, the brands with the most charcoal
are effective at reducing VOC deliveries under even
intense smoking conditions.
Despite decades of public and private anti-smoking
campaigns, cigarette smoking remains the leading
cause of preventable deaths in the United States.
Each year an estimated 440 000 people in the
United States1 and over five million throughout the
world2 die from smoking-related illnesses.
Although nicotine is the primary source of tobacco
dependence,3 many other chemical constituents in
mainstream cigarette smoke are causative agents
for cancer, heart disease and birth defects.4
Mainstream cigarette smoke comprises more than
4000 compounds,5 6 many of which are classes of
toxic chemicals such as harmful gases, volatile
organic compounds (VOCs), polycyclic aromatic
hydrocarbons (PAHs), metals and metalloids and
tobacco-specific nitrosamines (TSNAs). The pre-
sence of these and other toxic chemicals has
resulted in the International Agency for Research
on Cancer (IARC) classifying cigarette smoke as a
group 1 carcinogen.2
Exposure to certain toxic chemicals and chemical
classes in tobacco smoke has been linked to adverse
health outcomes in both people and laboratory
animals. For example, cigarette smoke includes
large quantities of gaseous species such as carbon
monoxide (CO)—a cardiovascular toxin7—and
many other highly volatile compounds including
Tobacco Control 2008;17(Suppl I):i10–i16. doi:10.1136/tc.2007.022517
benzene, a known human carcinogen.8 Many of
the less volatile components, including the PAHs,
metals and TSNAs, are associated with cigarette
smoke particulate matter and contain potent
carcinogens such as benzo[a]pyrene, cadmium,
N’-nitrosonornicotine (NNN) and 4-(methylnitro-
samino)-1-(3-pyridyl)-1-butanone (NNK). Chronic
exposure to these and other compounds is what
constitutes the inherent health risks associated
with smoking.
But which chemical species are most responsible
for smoking’s adverse health consequences is a
matter of some debate. Much work has focused on
the potent carcinogenic TSNAs.4 These compounds
merit concern because NNN contributes to cancers
of the upper aero-digestive tract, while NNK has
been linked to adenocarcinoma much deeper in the
lungs.9 Increased exposure to such smoke consti-
tuents is associated with higher disease prevalence,
prompting researchers to examine both the toxi-
city of individual chemicals and their relative
abundance in smoke. Fowles and Dybing10 used
available toxicity indices in combination with
standardised machine-smoked constituent levels
to rank chemicals in mainstream smoke by their
toxicity. They concluded that many toxic chemi-
cals present in the volatile phase of cigarette smoke
dominate the hazard risk indices. Such risk
calculations are, however, typically based on
machine smoke yields—known to underestimate
deliveries of modern cigarettes—as a predictor for
human exposure.11 Moreover, these relative rank-
ings could shift somewhat depending on the
mainstream smoke concentrations of specific che-
micals from different machine smoking regimens10
or from an individual’s smoking behaviour.
Removal of the chemicals produced by the
combustion of the tobacco during cigarette smok-
ing has been addressed over the years by the
inclusion of an activated charcoal filter. Activated
charcoal is formed from carbonaceous materials
and, having an extremely high surface area, allows
adsorption and entrapment of chemicals within
the porous structure of the carbon matrix.
Activated charcoal has a wide variety of uses,
ranging from the treatment of orally ingested
poisons,12 removal of contaminants from water13
and removal of chemicals from industrial gas
streams.14 In mainstream smoke deliveries, char-
coal-filtered cigarettes have been shown to sig-
nificantly reduce many volatile components.15–20
Filtered cigarettes containing activated charcoal
have been commercially available in the United

States since the RJ Reynolds Tobacco Company’s 1954
introduction of its Tareyton brand. Charcoal-filtered cigarettes
held a US market share as high as 6% in the 1960s. A steady
decline over the years has, however, resulted in charcoal filters
capturing less than 1% of today’s US cigarette market. While
charcoal-filtered cigarettes are today less common in the United
States, smokers in several countries, including Japan, Korea,
Venezuela and Hungary, predominantly favour charcoal-filtered
cigarettes; in these markets, shares approach 90%.21 22
Yet the question of whether charcoal filters actually reduce
exposure to harmful chemicals in cigarette smoke remains open.
Indeed, there is a lack of conclusive data that charcoal filtration
reduces risk of smoking-related diseases.23 24 Still, several
researchers have speculated that such reductions would be
prudent as a means of reducing exposure to many toxic
chemicals in smoke.25 26 Because so many factors contribute to
cancer and heart disease (for example, smoking, diet, environ-
ment and genetics) none of these studies have conclusively
demonstrated that charcoal-filtered cigarettes reduce the inci-
dence of smoking-related illness. Although no definitive
population data have shown a clear decrease in cancer rates,
researchers have demonstrated that smokers of charcoal-filtered
cigarettes do demonstrate significant reductions in many
exposure biomarkers27 although their exposure to nicotine and
CO was not reduced. Nevertheless, given that charcoal-filtered
cigarettes have shown promising reductions in associated
exposure biomarkers linked to exposure to VOCs in mainstream
cigarette smoke, there is great interest in exploring this
technology for emission reductions.28
We noted that one major US cigarette manufacturer recently
began test marketing four different dual-component cigarette
filters with activated charcoal in one of the components and
cellulose acetate in the other. The manufacturer markets these
prototype products under its flagship brand name and many
design features are comparable to its popular, similarly named,
cellulose-acetate filtered, ‘‘ultralight’’ cigarette. All of these
prototypes have filter ventilation levels, ‘‘tar’’ and nicotine
deliveries that are comparable to the non-charcoal-filtered
ultralight variety. The prototype brands make no external
package reference to charcoal-containing filters, nor do they
make any claims regarding delivery reductions or health
benefits. Thus to examine the effectiveness of charcoal filtration
for reducing vapour phase VOC levels in mainstream cigarette
smoke, we measured the four prototypes’ key physical design
properties and mainstream smoke deliveries. The similarity in
these designs gave a basis by which to evaluate how changes in
charcoal filter designs may impact the delivery of vapour phase
VOCs.
METHODS
Materials
We purchased all target compounds from Aldrich Chemical Co
(Milwaukee, Wisconsin, USA). Methanol was purchased from
Tedia Company Inc (Fairfield, Ohio, USA). The ultra-high
purity gas helium was obtained from AirGas, Inc (Atlanta,
Georgia, USA). Glass fibre filters, 44 mm diameter, commonly
referred to as Cambridge filter pads (CFP) were purchased from
Whatman (Maidstone, UK). Tedlar brand PVF gas sampling
bags, carboxen-polydimethylsiloxane (Carboxen-PDMS) solid-
phase microextraction (SPME) fibres and Mininert septum caps
were purchased from Supelco (Bellefonte, Pennsylvania, USA).
All chemicals and solvents were used without further purifica-
tion.
Tobacco Control 2008;17(Suppl I):i10–i16. doi:10.1136/tc.2007.022517
Supplement
Cigarettes for testing
Charcoal-filtered and traditional cigarettes were obtained from
the test markets. Brands A and C were purchased in Atlanta,
GA. Brands B, D and E were obtained in Bismarck, ND, Tampa,
Florida, and Salt Lake City, Utah, respectively. All cigarettes
were stored in their original packaging at 270uC until needed.
For at least 24 hours before smoking, the cigarettes were
conditioned in an environmental chamber (Parameter
Generation Control, Inc, Black Mountain, North Carolina,
USA) maintained as specified in ISO 3402:1999 at a temperature
of 22uC, with 60% relative humidity.
Smoking conditions
After conditioning the cigarette, filter ventilation levels were
measured using a QTM-5 filter ventilation and pressure-drop
apparatus (Cerulean, Milton Keynes, UK). Cigarettes were
smoked using three different smoking regimens; the ISO
3308:2000 regimen (35 ml puff of 2-second duration every
60 seconds), the Massachusetts regimen (45 ml puff of 2-second
duration every 30 seconds with the filter ventilation holes 50%
obstructed) and the Health Canada Intense regimen (55-ml puff
of 2-second duration every 30 seconds with the filter ventilation
holes 100% obstructed) 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 paper overwrap +3 mm,
whichever was greater.
Nicotine determination
Mainstream smoke nicotine deliveries were determined from
extraction of the CFP in 2-propanol. Nicotine levels in the
extracts were analysed and quantified using gas chromatogra-
phy with flame ionisation detection as previously reported.29
VOC determination
The vapour-phase portion of mainstream cigarette smoke was
collected in individual 1-litre PVF bags attached directly ASM
500 puffing engines’ exhaust ports as previously described.30 To
reduce background and sample carryover, 30 blank puffs per
port were taken before attaching the PVF bags. Before smoking,
an internal standard solution was added to each 1-litre PVF bag
using a gas-tight syringe. After smoking, portions of the smoke
samples were transferred from the PVF bags via cannula to
evacuated 20-ml headspace vials (Microliter Corporation,
Atlanta, Georgia, USA). The headspace vials were subsequently
loaded on a Combi-Pal auto sampler (Leap Technologies,
Carrboro, North Carolina, USA) equipped with a 75-mm
Carboxen-PDMS SPME fibre for sample preconcentration and
transferred to the heated inlet of a 6890 gas chromatograph
coupled to a 5973 mass spectrometer (Agilent Technologies,
Palo Alto, California, USA). The smoke samples were incubated
at 30uC for 2 minutes then sampled with the SPME fibre for
30 seconds. Analytes desorbed from the fibre at 260uC in the
GC’s heated sample inlet were cryo-focused on an Agilent DB-
624 capillary column (30.0 m 6 320 mm 6 1.80 mm) head.
Constant helium flow was maintained at an average linear
velocity of 46 cm/s. The GC oven programme started at 220uC,
held for 2 minutes and ramped to 200uC at 8uC/min, for a total
run time of 29.50 minutes. The GC transfer line was
maintained at a temperature of 255uC and the MS source and
quadrupole temperatures were 230uC and 150uC, respectively.
Data were acquired in full-scan mode over a detection range
spanning 30–200 mass-to-charge units.
i11
 Supplement

similar to brand C. Brand D had a much longer charcoal-on-tow

Figure 1 Cutout views of the filters for the five brands in this study.
Quality control materials
For this study, the 2R4F research cigarette (University of
Kentucky, Lexington, Kentucky, USA), was used as the quality
control (QC) material. Each analytical run included QC
cigarettes. For each analyte, levels from the gas phase of
mainstream cigarette smoke for at least 20 2R4F cigarettes,
smoked over at least four weeks, were characterised to
determine the mean and the 95th and 99th confidence intervals.
Acceptance criteria for QC and samples followed the criteria
prescribed by Taylor.31
Statistical analysis
Statistical analysis was performed using the Statistical Analysis
Software v8.0 (SAS Institute, Cary, North Carolina, USA).
RESULTS
The five cigarette brands (table 1) were all standard king-sized
length (nominal 85 mm):
c Brand A was the standard cellulose acetate filtered ultralight
cigarette, while the four charcoal-filtered varieties (B-E) had
differing filter construction (fig 1).
c In brands B–D, the charcoal was dispersed in cellulose
acetate fibres, commonly referred to as ‘‘charcoal on tow’’ or
as a ‘‘Dalmatian’’ style filter. These dispersed charcoal
designs were of differing lengths and charcoal content
(table 1).
c In cigarette brand E, the charcoal was localised to a fixed
compartment or cavity.
Except for the filters, brands A and B appeared to be identical
ultra-light variants having no differences in pack labelling.
Brands C, D and E, from three different test market cities, were
identically labelled with respect to each other, but labelled
differently from brands A and B. Brand B, the charcoal-
containing, labelling-equivalent of brand A, had a charcoal filter
style charcoal filter and significantly more charcoal than brands
B and C. Brand E, containing the cavity style charcoal filter, had
the highest charcoal weight (table 1). Charcoal weights were
determined by hand-separating then weighing the charcoal from
the filter material of five cigarettes. Under low-power magni-
fication (106), the charcoal from all five brands appeared
similar; all five brands also had similar ventilation levels of
approximately 45% (table 1).
Nicotine-free dry particulate matter (‘‘tar’’), nicotine and
carbon monoxide (CO) were determined for all five brands
(table 2). Brands A and B, identical except that brand B was
charcoal-filtered, had comparable deliveries of ‘‘tar,’’ nicotine
and CO under all three machine smoking conditions. Brands D
and E, with the longer filters and lower tobacco filler amounts
(table 1), delivered 10% to 15% less ‘‘tar,’’ nicotine and CO
under all three smoking regimens. Brand C had intermediate tar,
nicotine and CO deliveries compared to the others. The
variation in deliveries for ‘‘tar,’’ nicotine and CO were relatively
small compared to the variations observed for the vapour-phase
VOC analytes (table 2).
Under the ISO, Massachusetts and Health Canada smoking
protocols, variations in the five brands’ cigarette filter construc-
tion yielded dramatically different total vapour phase VOC
deliveries. We collected quantitative data on 22 VOCs in the gas
phase (2,3-butadione, 2,5-dimethylfuran, 2-butanone, 2-methyl-
furan, 2-pentanone, 3-buten-2-one, 3-pentanone, acetaldehyde,
acetone, acetonitrile, acrolein, acrylonitrile, benzene, ethylben-
zene, isobutyronitrile, isoprene, o-xylene, propionaldehyde,
propionitrile, m-/p-xylene, styrene and toluene). For simplicity,
however, only acetaldehyde, acrolein, benzene, styrene and the
total of all 22 VOCs measured in the gas phase are reported
(table 2). Additional data for other analytes are available on
request. Under all three smoking regimens, cigarette brands D
and E, containing the most activated charcoal, delivered
significantly lower quantities of total VOCs analysed compared
to brands A–C.
When examining individual volatile constituents, however,
some differences appeared (table 2). For acetaldehyde, acrolein
and benzene, the levels of these constituents in the vapour
phase of mainstream smoke using the ISO regimen were highest
for A (the brand containing a cellulose acetate filter),
intermediate for brands B and C (which contained the lowest
amount of charcoal) and lowest for brands D and E (which
contained the most activated charcoal). These differences in
mainstream smoke levels are also found for acetaldehyde,
acrolein and benzene, but were found to a lesser degree when
using the Massachusetts smoking regimen. For the Health
Canada regimen, the differences between the smoke emission
levels for acetaldehyde, acrolein and benzene between brands B
and C and brand A are no longer statistically significant in most
cases, while the differences between brands D and E and brand
A are still statistically significant though reduced. For styrene,

Table 1 Physical parameters for the five brands in this study

Filter length Filter ventilation Tobacco Charcoal
(mm) (%) weight (g) weight (mg)

Brand A 27 45 0.58 0
Brand B 27 45 0.59 45
Brand C 27 44 0.56 45
Brand D 33 43 0.50 120
Brand E 33 46 0.54 180

i12 Tobacco Control 2008;17(Suppl I):i10–i16. doi:10.1136/tc.2007.022517

 Supplement

Table 2 Mainstream smoke constituent deliveries for brands in this study measured under ISO (35 ml puff,
2-second duration, 60 seconds between puffs), Massachusetts (45 ml puff, 2-second duration, 30 seconds
between puffs, ventilation holes 50% obstructed), or Health Canada (55 ml puff, 2-second duration,
30 seconds between puffs, ventilation holes 100% obstructed) regimens
Analyte (units) Brand ISO Massachusetts Health Canada

Tar (mg/cig) A 5.5 (0.3) 14.4 (0.9) 21.5 (1.1)

B 5.1 (0.4) 13.1 (0.5) 20.5 (0.3)
C 3.5 (0.9) 12.7 (0.1) 19.6 (1.1)
D 3.8 (0.3) 10.9 (0.1) 17.3 (0.2)
E 4.5 (1.0) 12.3 (0.6) 16.2 (1.7)
Nicotine (mg/cig) A 0.52 (0.03) 1.16 (0.10) 1.67 (0.07)
B 0.52 (0.00) 1.16 (0.05) 1.60 (0.00)
C 0.47 (0.02) 1.20 (0.10) 1.69 (0.31)
D 0.38 (0.04) 0.94 (0.04) 1.30 (0.03)
E 0.49 (0.06) 1.04 (0.09) 1.52 (0.27)
CO (mg/cig) A 7.0 (0.3) 17.9 (1.4) 25.7 (0.5)
B 7.1 (0.4) 18.5 (0.8) 28.4 (3.4)
C 6.5 (0.3) 17.6 (0.1) 24.9 (0.5)
D 5.6 (0.6) 15.5 (1.1) 23.7 (0.2)
E 6.0 (1.2) 16.2 (0.8) 21.3 (1.0)
Acetaldehyde (mg/cig) A 482.0 (49.6) 1428 (499) 1221 (88)
B 255.0 (33.2) 954 (293) 1034 (106)
C 325.3 (95.3) 1210 (395) 1165 (162)
D 52.0 (24.6) 485.2 (141.6) 741.6 (83.2)
E 133.0 (30.3) 622.9 (190.2) 574.4 (32.2)
Acrolein (mg/cig) A 77.0 (21.1) 243.9 (104.0) 235.3 (49.1)
B 22.3 (3.6) 134.3 (19.5) 184.6 (33.1)
C 34.2 (8.5) 172.5 (59.7) 218.6 (46.6)
D 2.1 (1.1) 23.9 (10.4) 64.6 (18.2)
E 13.4 (4.6) 57.0 (20.4) 64.3 (7.4)
Benzene (mg/cig) A 38.0 (7.1) 110.7 (34.6) 106.4 (13.5)
B 6.0 (2.7) 49.5 (8.5) 68.6 (12.5)
C 10.4 (4.2) 57.7 (12.9) 84.5 (16.9)
D 0.4 (0.0) 2.6 (1.4) 12.7 (2.4)
E 4.0 (1.0) 15.8 (3.1) 16.9 (1.8)
Styrene (mg/cig) A 3.8 (1.3) 21.9 (10.1) 20.6 (6.2)
B 0.5 (0.4) 3.3 (0.7) 6.5 (2.3)
C 1.5 (0.7) 6.5 (4.0) 8.5 (3.1)
D 0.4 (0.5) 1.1 (0.4) 1.5 (0.8)
E 1.0 (0.6) 4.0 (2.1) 2.8 (0.8)
Sum of the 22 volatile organic A 1894 (337) 5852 (2138) 5587 (1029)
compounds measured (mg/cig) B 618.0 (164.8) 3399 (735) 4245 (696)
C 1004 (212) 4341 (1529) 5095.1 (1141)
D 135.5 (63.8) 787 (205) 1733 (244)
E 353.6 (136.6) 1415 (380) 1540 (143)
Results are reported as the average of three observations and the standard deviations are included in parentheses.

the levels in the vapour phase of mainstream smoke remain
significantly different between brand A and brands B-E at all
smoking regimens.
DISCUSSION
The non-charcoal-filtered brand A served as a reference for
comparison to brands B–E because, excluding filter construc-
tion, A has similar physical properties (for example, length,
diameter and filter ventilation). Visual comparison of brand A
and B packs or individual intact cigarettes showed no
differences. Brand C cigarettes contained a similar charcoal
filter to brand B (fig 1) and were visually similar to brand A and
B cigarettes. The filters of brands D and E had distinct design
differences. Brand D had a longer charcoal-containing segment
and brand E contained a pocket or cavity containing charcoal
granules. The filters of cigarette brands B, C and D were visually
similar and resembled the charcoal-on-tow style filters in other
US charcoal-filtered cigarettes and popular Japanese filtered
cigarettes. Brand E’s charcoal cavity design has appeared in
other commercial cigarette brands and is available from several
manufacturers.
With the similarities among all five brands, it was not
surprising that the ISO machine- smoked deliveries of tar,
nicotine and carbon monoxide (TNCO) differed only slightly.
Differences in TNCO values were small compared to the
dramatic differences observed for the vapour phase VOCs.
Statistical analysis (general linear model) established a correla-
tion between carbon monoxide (p,0.003) and nicotine
(p,0.02) to tobacco filler weight. Tar, however, correlated
poorly (p,0.3) with tobacco weight. The average tar and CO
deliveries were slightly lower in the smoke of cigarettes with
longer filters and higher charcoal content (brands D and E).
Their longer filter lengths displaced several millimetres of
tobacco rod length, resulting in an average of 1.3 fewer puffs
per cigarette (,15% less smoke volume) under all smoking

Tobacco Control 2008;17(Suppl I):i10–i16. doi:10.1136/tc.2007.022517 i13

 Supplement
Figure 2 Total volatile organic compounds (VOC) deliveries for all five
brands under ISO (35 ml puff, 2-second duration, 60 seconds between
puffs), Massachusetts (45 ml puff, 2-second duration, 30 seconds
between puffs, ventilation holes 50% obstructed), or Health Canada
(55 ml puff, 2-second duration, 30 seconds between puffs, ventilation
holes 100% obstructed) regimens. Results are reported as the average of
three observations.
regimens. When comparing tar, nicotine or carbon monoxide
deliveries to filter length or charcoal content among the five
designs, however, no statistically significant differences
(p,0.05) were observed. Thus using the ISO smoking regimen,
it would appear that these charcoal filtration designs have little
effect on the delivery of tar, nicotine and carbon monoxide.
Under the more intense (Massachusetts or Health Canada)
smoke regimens, for all five brands, higher mainstream smoke
deliveries of tar, nicotine and CO were obtained (table 2).
Nicotine delivery increased ,200% for all five brands when
smoked according to the Health Canada conditions relative to
the ISO regimen. Similarly, increases for the delivery of tar and
CO (,300% and ,200%) were observed for all five brands.
These results are not surprising, given that the more intense
smoking method takes more frequent puffs, larger puffs and
because of the 50% blocked (Massachusetts) or 100% blocked
(Health Canada) filter ventilation holes, has reduced or no air
dilution of the mainstream smoke. Comparing mainstream
smoke deliveries between the non-charcoal-filtered brand A and
the charcoal-filtered brands B-E suggests that under the intense
smoking regimens the filter differences have little effect on the
deliveries of tar, nicotine and CO.
Significant reductions in vapour phase VOC deliveries under
the ISO smoking regimen were observed for all charcoal-filtered
brands (table 2). For example, the mainstream smoke total of
the 22 VOCs analysed from brands B–E (charcoal-containing)
compared to brand A (non-charcoal-filtered) were, depending on
brand, reduced 9–93% (fig 2). Although direct comparisons
between theses five brands are complicated by differences in
filter construction, some trends are evident. By comparing
brands A–C, which have similar filter lengths, it is clear that
under ISO smoking conditions vapour phase VOC deliveries
were lower for the charcoal-containing brands B and C
compared to brand A. Not surprisingly, brands D and E, which
contain higher charcoal content and longer filter lengths, had
even lower VOC deliveries. Interestingly, the charcoal cavity-
filtered brand E (180 mg charcoal) contained the most charcoal
but was less effective at VOC removal under the ISO regimen
than brand D (120 mg charcoal). These data support the finding
that increasing charcoal content can reduce vapour phase
i14
Figure 3 Nicotine corrected mainstream smoke constituent deliveries
for (A) acrolein and (B) benzene for brands in this study measured under
ISO (35 ml puff, 2-second duration, 60 seconds between puffs),
Massachusetts (45 ml puff, 2-second duration, 30 seconds between
puffs, ventilation holes 50% obstructed) or Health Canada (55 ml puff, 2-
second duration, 30 seconds between puffs, ventilation holes 100%
obstructed) regimens. Results are reported as the average of three
observations.
mainstream smoke VOC levels, but the configuration of the
charcoal filter can affect reduction efficiency.
The original FTC or ISO smoking regimen typically under-
estimates smoker’s intake from lower delivery ‘‘light’’ and
‘‘ultralight’’ brands. Because these cigarettes are typically
smoked more intensely, various smoking regimens have been
proposed to better characterise the delivery of smoke constitu-
ents under such intense smoking conditions.11 When machine-
smoked under the ISO regimen, substantial reductions are
observed in the total vapour phase VOC levels for all charcoal-
containing brands compared to brands with similar filter length
and no charcoal (fig 2). When smoked using the more intense
Massachusetts or Health Canada regimens, however, brands
with the lowest charcoal amounts—brands B and C—yielded
total VOC deliveries not statistically different from the non-
charcoal-filtered brand A (table 2 and fig 2). This suggests that
the increased mainstream smoke volume generated by the
more intense regimens saturates the charcoal in the filters of
brands B and C, allowing VOC breakthrough. Brands D and E,
with higher charcoal content, under all smoking regimens
provided statistically significant VOC reductions compared to
the other brands (table 2 and fig 2). Thus, measuring VOC
deliveries under the ISO regimen alone could yield potentially
Tobacco Control 2008;17(Suppl I):i10–i16. doi:10.1136/tc.2007.022517

misleading data, suggesting to smokers that using these
products may reduce their exposure to VOCs no matter how
these cigarettes are smoked. Because brands A–E are all high-
ventilation cigarettes, smokers are likely to smoke these more
intensely to obtain the nicotine they crave and overcome any
benefit that could occur from reduced delivery of volatiles.10 11
Such observations underscore the need for testing tobacco
products under a broad range of usage patterns to understand
fully if particular cigarette design change can achieve mean-
ingful reductions for smokers under a range of smoking
conditions.
A more appropriate means for comparing brands is to
normalise vapour phase mainstream smoke VOC deliveries
with nicotine. After adjusting for nicotine delivery or when
smoked under a more intense regimen, the majority of standard
US domestic king-sized filtered cigarettes, regardless of filter
ventilation, deliver comparable levels of these VOCs.30 This is
important because, as smokers require sufficient dosages to
satisfy their nicotine cravings, users of highly ventilated brands
usually compensate by taking larger or more frequent puffs or
by blocking vent holes, resulting in higher inhalation of toxic
mainstream smoke constituents.32 Comparing nicotine-normal-
ised VOC levels from the charcoal-filtered brands B–E to the
non-charcoal-filtered brand A indicated that when smoked
under ISO conditions, the nicotine-normalised VOC levels from
harmful compounds such as acrolein and benzene were reduced
by charcoal filtration (fig 3). Under more intense smoking
regimens, brands B and C, which contained less charcoal, had
nicotine-normalised VOC deliveries similar to those of the non-
charcoal brand (A). Brands D and E, however, with higher
charcoal levels, maintained significant reductions. Such beha-
viour demonstrates that enough charcoal needs to be present in
the cigarette filters to prevent breakthrough and to maintain
reductions of VOCs in mainstream smoke.
Brands D and E, with their higher charcoal content, were in
fact superior at reducing vapour phase mainstream smoke
VOC deliveries. When these brands were test-marketed, no
indication was revealed of which variant was available in each
location. The prototype brands were identically labelled and,
presumably, consumers had no knowledge of whether a
recently introduced brand contained high or low charcoal
amounts. Even a consumer knowingly smoking a charcoal-
filtered cigarette may not realise that their smoking behaviour
can substantially reduce the efficiency of the filter. This study
also highlights the importance of consumers and public health
decision-makers knowing which of the various test market
cigarette configurations is selected to be sold nationally and
what compromises are made to produce a commercially
acceptable product. For example, if the results for brand D
using the ISO method were used to promote a total VOC
reduction for this product, but brand C was the one that was
actually marketed, this would be deceptive to both smokers
and to public health officials.
Activated charcoal filtration is a well accepted means for
removing VOC contamination from air and water. When
present in sufficient quantities it has great utility for reducing
VOCs in mainstream cigarette smoke under the most intense
smoking conditions. The use of activated charcoal for filtration
and purification of air streams encompasses a wide range of
applications—from large industrial scrubbers to personal
respirator masks. Additionally, water treatment plants use
activated charcoal for filtration and charcoal filters are sold for
home water purification. The effectiveness of these products
has routinely been demonstrated as an effective means to reduce
Tobacco Control 2008;17(Suppl I):i10–i16. doi:10.1136/tc.2007.022517
Supplement
VOCs in both air and water. Yet in all of these commercial and
domestic applications, because charcoal actively absorbs VOCs
over time it becomes saturated, thus necessitating filter
replacement or regeneration. One concern arising from use of
an active filtration agent such as charcoal in cigarette filters is a
reduction in efficiency from the absorption of volatile chemical
species such as those associated with pack aroma over time.
Ceschini examined this phenomenon in detail and found that
after an initial loss (,20–30%) in the capacity of activated
charcoal during the first two weeks, the filters’ effectiveness
remained consistent up to 6 months.33 Consistent with
Ceschini’s findings, in our own work, over a 12-week period
between initial receipt and the completed analysis of these
cigarette packs, we noticed no change in the efficiency of the
charcoal filters.
In accordance with statements by the US Surgeon General,34
toxic compounds in cigarette smoke should be removed or their
levels should be reduced to the lowest possible values. We do
not know whether a reduction in mainstream smoke VOC
levels would influence the adverse health effects associated with
smoking. If sufficient charcoal is used, charcoal filters have the
potential to reduce VOC levels in mainstream smoke under
some machine smoking conditions, but little or no reduction
was seen for other, less volatile harmful constituents. Charcoal-
filtered cigarettes with reduced mainstream smoke VOC
deliveries cannot be considered ‘‘safer’’ and there is a significant
risk of the public making such conclusions as happened
previously with ‘‘light’’ cigarettes.32 Although a cigarette filter
with sufficient charcoal achieves reduced VOC levels in main-
stream smoke, there would not be a corresponding reduction in
sidestream cigarette smoke. Therefore, the VOC exposure from
sidestream smoke for the active or any passive smokers would
not be reduced.
Preliminary results indicate that levels of other important
but less volatile toxic constituents (for example, TSNAs,
PAHs) are not affected or are reduced to a lesser extent than
the VOCs examined here. Considerations involving the net
effects of all toxic smoke constituents are important when
drawing conclusions about the potential for reduced main-
stream smoke deliveries and any resulting change in toxicity.
In fact, a recent article indicated that a charcoal filtered
cigarette may have no less of a ‘‘harm index’’ when compared
in a nicotine normalised manner.20 Any efforts aimed at
reducing risks must consider how people smoke—both in
terms of the toxicity levels delivered under such conditions
and in terms of whether any design changes actually produce a
reduced exposure or only a false sense of safety. To the best of
our knowledge, currently no data show that selectively
reducing specific chemicals from smoke reduces a smoker’s
risk. Scherer et al27 found reduced body burden of mainstream
smoke volatiles among smokers of charcoal-filtered cigarettes.
Such findings are encouraging and they suggest possibilities
may exist for reduced exposure through cigarette design
improvements. When it is technically feasible and relatively
easy to do so, manufacturers should be encouraged to remove
harmful chemicals from smoke. In the meantime, cessation
remains the only proved means of reducing the risks associated
with cigarette smoking.
Competing interests: None.
Disclaimer: Use of trade names is for identification only and does not imply
endorsement by the Centers for Disease Control and Prevention or by the US
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.
i15
 Supplement

REFERENCES
1. Centers for Disease Control and Prevention. Annual smoking-attributable
mortality, years of potential life lost and economic costs—United States, 1995–1999.
Morb Mort Wkly Rep 2002;51:300.
2. International Agency for Research on Cancer. IARC monographs on the
evaluation of the carcinogenic risk of chemicals to humans. IARC: Lyon, France,
2002;83.
3. US 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.
4. Hoffmann D, Hoffmann I. The changing cigarette, 1950–1995. J Toxicol Environ
Health 1997;50:307.
5. Stedman RL. The chemical composition of tobacco and tobacco smoke. Chem Rev
1968;68:153.
6. Roberts DL. Natural tobacco flavor. Recent Adv Tobacco Sci 1988;14:45.
7. US Department of Health and Human Services. The health consequences of
smoking: cardiovascular disease. A report of the Surgeon General, 1983. Rockville,
MD: Public Health Service, Office on Smoking and Health, 1983 (DHHS Publication No
(PHS) 84–50204)
8. National Toxicology Program. Report on carcinogens, 10th ed. Research Triangle
Park, NC: US Department of Health and Human Services, Public Health Service,
National Institutes for Health, 2002.
9. Hecht SS. Biochemistry, biology, and carcinogenicity of tobacco-specific N-
nitrosamines. Chem Res Toxicol 1998;11:559–603.
10. Fowles J, Dybing E. Application of toxicological risk assessment principles to the
chemical constituents of cigarette smoke. Tob Control 2003;12:424.
11. Kozlowski LT, Rickert WS, Pope MA, et al. Estimating the yield to smokers of tar,
nicotine, and carbon-monoxide from the lowest yield ventilated filter-cigarettes.
Br J Addict 1982;77:159.
12. Bond GR. The role of activated charcoal and gastric emptying in gastrointestinal
decontamination: a state-of-the-art review. Ann Emerg Med 2002;39:273–86.
13. Yu JJ, Chou SY. Contaminated site remedial investigation and feasibility removal of
chlorinated volatile organic compounds from groundwater by activated carbon fiber
adsorption. Chemosphere 2000;41:371–8.
14. Ruddy EN, Carroll LA. Select the best VOC control strategy. Chem Eng Prog
1993;89:28–35.
15. Steinik M, Smith M. The path to a safer cigarette. Advance Lights. Trionic filter.
Global Equity Research, JP Morgan, 28 July 2004. Packet insert.
16. Lorillard L, Ihrig AM. ‘‘The selective removal of aldehydes from cigarette smoke’’. 20
Feb 1980. Bates: 80007395-80007421.http://tobaccodocuments.org/lor/80007395-
7421.html. Accessed Nov 2006.
17. Cundiff RH. Analytical research; Markunas, PC; Unk. ‘‘the Spectrophotometric
Determination of Hydrogen Cyanide in Cigarette Smoke.’’. 11 May 1964. Bates:
500963767–500963782.http://tobaccodocuments.org/rjr/500963767-3782.html
Accessed 11/06.
18. British American Tobacco Company. 54 page document. No date. Bates:
402368992–402369045.http://tobaccodocuments.org/guildford_misc/402368992-
9045.html Accessed 11/06.
19. Xue L, Thomas CE, Koller KB. Mainstream smoke gas phase filtration performance of
adsorption materials evaluated with a puff-by-puff multiplex GC-MS method. Beitr
Tabakforsch Int 2002;20:251–6.
20. Laugesen M, Fowles J. Marlboro UltraSmooth: a potentially reduced exposure
cigarette? Tob Control 2006;15:430–5.
21. John AL. Japan. Always something new. Tob Int 1996 August:30–35.
22. Fisher B. Filtering new technology. Tob Reporter 2000;127:46–7.
23. Marugame T, Sobue T, Nakayama T, et al. Filter cigarette smoking and lung cancer
risk; a hospital-based case-control study in Japan. Br J Cancer 2004;90:646–51.
24. Muscat JE, Takezaki T, Tajima K, et al. Charcoal cigarette filters and lung cancer risk
in Aichi Prefecture, Japan Cancer Sci 2005;96:283–7.
25. Djordjevic MV, Stellman SD, Zang E. Doses of nicotine and lung carcinogens
delivered to cigarette smokers, J Natl Cancer I 2000;92:106–11.
26. Stellman SD, Takezaki T, Wang L, et al. Smoking and lung cancer risk in American
and Japanese men: an international case-control study. Cancer Epidemiol Prev
2001;10:1193–9.
27. Scherer G, Urban M, Engl J, et al. Influence of smoking charcoal filter tipped
cigarettes on various biomarkers of exposure, Inhal Toxicol 2006;18:821–9.
28. Laugesen M, Fowles J. Scope for regulation of cigarette smoke toxicity: the case for
including charcoal filters. N Z Med J 2005;118:U1402.
29. Calafat AM, Polzin GM, Saylor J, et al. Determination of tar, nicotine, and carbon
monoxide yields in the mainstream smoke of selected international cigarettes, Tob
Control 2004;13:45–51.
30. Polzin GM, Kosa-Maines RE, Ashely DL, et al. Analysis of volatile organic compounds
in mainstream cigarette smoke. Environ Sci Technol 2007;4:1297–302.
31. Taylor JK. Quality assurance of chemical measurements. Boca Raton, FL: CRC Press,
1987.
32. Burns DM, Major JM, Shanks TG, et al. Risks associated with smoking cigarettes
with low machine-measured yields of tar and nicotine; Smoking and Tobacco Control
Monograph 13; NIH Publication 02–5074. Bethesda, MD: US Department of Health
and Human Services, National Institutes of Health, National Cancer Institute,
2000:65–158.
33. Ceschini P. Decrease of the carbon retention capacity in cigarette filters. Beitr
Tabakforsch Int 1972;6:220–2.
34. Centers for Disease Control and Prevention. Reducing tobacco use: a report of
the surgeon general. Atlanta, GA: Office on Smoking and Health, 2000.

i16 Tobacco Control 2008;17(Suppl I):i10–i16. doi:10.1136/tc.2007.022517